![ehicle platoon based on vehicle-infrastructure cooperation controller design based on cost specifications, and introduced a synchronization mechanism to compensate for the time delay caused by wireless communication accurately. Harfouch et al. [82] designed a switching control strategy based on communication delay. When the vehicle network is in good condition, it starts a Coordinated Automatic Cruise Control (CACC) model. Otherwise, Adaptive Cruise Control (ACC) is enabled and analyzes the stability of this switch control strategy. At present, the main problem in the vehicle platoon research is that most of the teams consider the grouping strategy and the following model in the case of normal traffic flow and rarely mention the fleet motion control under emergency or network delay conditions, which is precisely the key whether platoon technology can be put into practical use. In the future, in the IR system, the reliability of V2I will be fully utilized to make up for the shortcomings of V2V randomness, and real-time high-precision maps and highprecision wireless positioning services will be provided to ensure the reliability and safety of vehicle platoon operation. At present, there are still many problems to be solved before the commercial vehicle platoon is commercialized. For example, in complex traffic environments and severe weather conditions, the reliability of some technical means (wireless communication, sensors) cannot be completely guaranteed, and the vehicle security assurance strategy has not yet been subjected to systematic verification and testing. How to prevent the vehicle queue information system from being attacked by hackers is also an urgent problem to be solved.](https://www.researchgate.net/profile/Zhigang-Xu-8/publication/349477084/figure/fig3/AS:993587684708353@1613901172940/ehicle-platoon-based-on-vehicle-infrastructure-cooperation-controller-design-based-on_Q320.jpg)
Content uploaded by Zhigang Xu
Author content
All content in this area was uploaded by Zhigang Xu on Feb 21, 2021
Content may be subject to copyright.
China Journal of Highway and Transport No. 08
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 1
______________________________________
Received: 2018-10-04
Supported by: National Key Research and Development Program of China (2018YFB0105104); Open Project of State Key Laboratory of
Software Architecture of Neusoft Corporation (211924180084); National Natural Science Foundation of China (61973045)
First author: XU Zhi-gang (1979–), male, from Ezhou City, Hubei Province, professor, PhD, E-mail: xuzhigang@chd.edu.cn
Corresponding author: ZHAO Xiang-mo (1966–), male, from Chongqing City, professor, doctoral supervisor, PhD, E-mail:
xmzhao@chd.edu.cn
Citation: XU Zhi-gang, LI Jin-long, ZHAO Xiang-mo, LI Li, WANG Zhong-ren, TONG Xing, TIAN Bin, HOU Jun, WANG Gui-ping, ZHANG Qian. A
Review on Intelligent Road and Its Related Key Technologies [J]. China Journal of Highway and Transport, 2019 (08): 1–24.
A Review on Intelligent Road and Its Related Key Technologies
XU Zhi-gang1, LI Jin-long1, ZHAO Xiang-mo1, LI Li1, WANG Zhong-ren2, TONG Xing3, TIAN Bin1, HOU Jun1, WANG
Gui-ping4, ZHANG Qian5
1. The Joint Laboratory for Internet of Vehicles, Ministry of Education-China Mobile Communications Corporation, Chang’an Uni-
versity, Xi’an 710064, Shaanxi, China;
2. California Department of Transportation, Sacramento CA95833, California, USA;
3. Qilu Transportation Information Group Co., Ltd., Jinan 250102, Shandong, China;
4. School of Information Engineering, Chang'an University, Xi’an 710064, Shaanxi, China;
5. State Key Laboratory of Software Architecture State Key New Technology, Neusoft Corporation, Shenyang 110179, Liaoning, China
Abstract: Based on the technical framework of automated highway systems (AHSs), the influence of different
driving factors such as primary application, communications technology, green energy technology, and automated
driving technology, the evolution of concepts, development of technology, and future changes in intelligent roads
(IRs) were reviewed in this paper. According to the current development trend in information technology, the con-
cept and technical framework of IRs have been extended and expanded on the basis of studies on AHSs. The direc-
tion of evolution of the IR system in the future, and its system architecture, which consists of a management layer,
network layer, and application layer, were presented. Meanwhile, focusing on the current popular technologies and
the development direction of science and technology in the future, the status of research on emerging technologies
driving the rapid development of IRs was summarized, such as ubiquitous wireless communication, high-precision
positioning and navigation, vehicle platoon control, wireless charging, intelligent road materials, road active safety
control technology, vehicle-to-road information interaction for Mobility-as-a-Service, intelligent decision planning
technology combined with infrastructure, etc. Based on the development characteristics of these eight key technolo-
gies, some recommendations on the application and promotion of IR technology in the future are presented. Mean-
while, the influence and impact of emerging technologies such as vehicle-integrated integration, intelligence
parallel systems, artificial intelligence, and traffic information security on the future development of IR were ana-
lyzed. Finally, this paper systematically predicted the commercialization promotion route of IR technologies, and
that the application of IR will further reduce the cost of technologies and equipment in autonomous driving,
providing a safer, more stable and efficient traffic environment for autonomous driving in the future. The research
results of this paper have significance as reference for current and future technology research and development of
IR and engineering applications of IR. DOI: 10.19721/j.cnki.1001-7372.2019.08.001-en
Keywords: traffic engineering; intelligent road; review; intelligent transportation system; road infrastructure; par-
allel intelligent road; vehicle-infrastructure cooperation
0 Introduction
Intelligent road (IR) is a multi-functional integrated road
infrastructure system that can provide a large amount of
global, real-time and a priori information to assist Intelligent
and Connected Vehicle (ICV) [1] environment perception and
instant communication, which can eliminate driving safety
and traffic congestion and make future road traffic systems
safer, more efficient, more environmentally friendly and
more comfortable.
IR is an important part of the Intelligent Transportation
System (ITS) [2]. In the early 1960s, scholars and institutions
proposed the concept of automated highways [3–5]. In the
1980s, the Federal Highway Administration (FHWA) con-
ducted research on Automated Highway Systems (AHS) [6–7].
Related studies indicate that AHS is a system that collects
information through sensors installed on roads and vehicles
and uses that information to drive the vehicle with little or no
intervention [8]. The technology that it studies is designed to
guide vehicles to improve traffic flow and road safety by
reducing accidents while reducing fuel consumption and
pollution [9]. The focus of AHS is to provide infrastructure
support for automated vehicles, and the roadside control
system and the vehicle itself can be responsible for the safety
and efficiency of traffic flow [10]. AHS is a solution that can
solve road congestion, driving safety problems and improve
road utilization. It can increase road vehicle capacity [11], and
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 2
its roadside control system can optimize the entire road ca-
pacity and traffic flow [12]. Hall et al. [13] proposed that road
automation with control, sensing, and communication tech-
nologies applied to road vehicles could improve road per-
formance and increase road capacity by about three times.
AHS introduces the safety and efficiency of road traffic
through automation [14]. Besides, there is the Intelligent Ve-
hicle and Highway System (IVHS), a comprehensive pro-
gram initiated by the US government under the Multimodal
Transport Ground Transportation Efficiency Act of 1991 to
improve safety, reduce congestion, and improve mobility,
reduce environmental impact, save energy, and increase
economic productivity in transportation [15].
IR is a way of the ICV to improve the driving environ-
mental perception and environmental adaptability, to help it
achieve a reliable and secure autonomous navigation, and to
help ICV provide a safer and more convenient way of travel.
Studies have shown [16] that in the fully automated driving
phase, autonomous vehicles can completely avoid traffic
accidents and improve traffic efficiency by more than 30%.
However, there are still many hidden dangers in the current
self-driving environmental perception and ability to respond
to various emergencies [17]. By installing sensing devices,
communication devices, control devices, etc. [18] on the road
infrastructure, we can make the roads “smart” and provide
environmental perception and communication support for the
vehicles, which can break through limitations of the single-
vehicle intelligent system on the perception of the sur-
rounding environment of the vehicle from the time and space
dimensions [19]. However, there are not many studies on IR
infrastructure. In the early AHS research, Carbaugh et al. [20]
studied four different AHSs and quantified the impact of
AHS on vehicle cooperation and operating speed. Li[21] pro-
vided a service-oriented architecture (SOA) for AHS and
established a highway information system that expanded the
annotations of sensor resources, data resources, and pro-
cessing resources into services. Baskar et al. [22] proposed an
integrated traffic management and control method based on
the AHS-based hierarchical traffic control architecture. In
IVHS, AHS is considered an advanced vehicle control sys-
tem. Therefore, AHS can be defined as follows: “AHS con-
sists of three parts: self-driving vehicle, vehicle
communication and coordination, and IR infrastructure,
whose goal is to build a vehicle-road integrated system
through communication, electronics and automation tech-
nology to achieve the maximum road capacity in an envi-
ronment of real-time communication [23]”.
With the continuous advancement of science and tech-
nology, AHS is gradually moving towards an intelligent di-
rection. IR will be a necessary facility for providing users
with a ubiquitous mobile service supply environment [24].
According to relevant literature research, it can be seen that
the concept and application of IR are still in the stage of
evolution, and its connotation and extension will continue to
expand. However, its essence is still positioned in
intelligence. The difference from the traditional road is that
its acquisition of the state information of the road is real-time;
the information in its internal elements is shared; its man-
agement and decision support system is intelligent to a cer-
tain extent, and its goal is to make the road safer, efficient,
energy-efficient, environmentally friendly and sustainable.
In order to clarify the development status of IR and their
key technologies all over the world, this paper takes AHS as
the prototype. It systematically sorts out the core concepts
and technical architecture of IR, reviews and summarizes the
development process and latest development of IR in stages.
The development of the key technologies of IR construction
and development has been analyzed and reconstructed, and
the future development trend of IR has been prospected and
considered.
1 Development of IR
1.1 American AHS system architecture and
evolution
The IR technology system integrates technologies such as
computing, sensing, communication, network, automatic
control, and intelligent materials. It is a multi-functional
technology development framework for environment per-
ception, network interconnection, and system integration.
The National Automated Highway System Consortium
(NAHSC) defines several alternative AHS concepts, from
collaborative to fully automated, based on the degree of col-
laboration between vehicles and infrastructure, including
autonomous driving, cooperative collaboration, facility
support and management control, and introduces the evolu-
tionary concepts of AHS’s position retention, lane change,
congestion mitigation and flow control functions at different
stages. Figure 1 shows these alternative concepts and their
four functions [25]. However, with the rapid development of
science and technology and the tremendous changes in the
transportation environment, the concept of AHS has become
more extensive and innovative.
On the basis of AHS, this paper adds new technology to
the conceptual content of IR and expands its function. It adds
two functions of energy supply and road maintenance on the
basis of AHS, as well as the evolution of functions in dif-
ferent stages. It establishes a technical framework for IR
systems that meets social development and higher public
travel requirements. The evolution direction of the future IR
system is predicted, namely that there may be a stage of
coexistence of artificial driving and semi-autonomous driv-
ing in the early stage of autonomous driving. With the con-
tinuous improvement of IR information sensing, processing,
transmission, computing, and fusion decision systems, IR
will eventually reach the intersection of physical highways
and digital highways. The two will realize the interaction
of virtual and real information, jointly complete the
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 3
Figure 1 Framework of IR based on AHS concept [25]
Figure 2 Concept and structure of IR
decision-control-management function, and become an ad-
vanced stage with parallel intelligence, as shown in Figure 1.
To realize the function of the IR and integrate the current
technical conditions, Figure 2 shows the level of the IR sys-
tem and sorts out the structural level and related concepts of
the current IR. In this framework, the IR will sense and col-
lect the driving state and road condition of the vehicle in real
time through the roadside equipment, and then realize the
interconnection between the entities of the IR through the
ubiquitous network. Then it uses big data and cloud platform
technology for dynamic interaction of data, information
mining, intelligent decision-making, etc., to provide com-
prehensive and efficient information services for vehicles,
drivers, managers, and other participants.
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 4
The future IR system architecture will be a high-tech road
complex integrating environmental perception, planning
decision-making, information interaction, and automatic
repair. The structure level concept of the IR system in Figure
2 is simplified and integrated. This paper mainly divides it
into three structural levels: perception control layer, network
communication layer, and information management layer.
a) Perception control layer
The perception control layer corresponds to the sensing
and collecting module in Figure 2 and is mainly composed of
a roadside device and a smart vehicle, thereby realizing an
effective collection of road and vehicle information and ef-
fective control of the executing device. It is also responsible
for the automatic monitoring and dynamic management of
the underlying sensor’s operating status. The detection and
recognition accuracy cannot meet the development needs of
autonomous driving. IR can comprehensively acquire rich
surrounding environment information by combining roadside
equipment and vehicle sensors, and provide rich and com-
prehensive road condition information and decision-making
basis for autonomous driving. At the same time, deep learn-
ing is also applied in the aspect of environmental perception,
which has significant advantages in the identification and
classification of road objects and images [26–28].
b) Network communication layer
The network communication layer corresponds to the
network communication module in Figure 2. It is mainly
composed of various communication technologies such as
V2X (Vehicle to Everything) communication technology and
network protocol, mostly implementing sensor networks,
vehicle networking, optical networks, and various wireless
networks. Inter-connectivity ensures efficient transmission of
all types of information while meeting network stability and
information security under extreme conditions. Through the
integration of interoperable inter-vehicle communication
networks, inter-vehicle communication networks, and
wide-area communication networks, the network communi-
cation layer can more effectively obtain driver information,
vehicle attitude information and vehicle-surrounding envi-
ronment information data, which provides rich data integra-
tion and analysis channels. In terms of matching different
levels of intelligent carriers and IR connections, IR provides a
standardized interface for data transmission, such as vehicle-
road communication and perception information for carriers.
c) Information management layer
The information management layer corresponds to the
decision processing and service providing module in Figure
2, which is mainly composed of cloud platform technology
and big data technology and realizes the storage of human-
vehicle-road data, information mining, and decision support.
Currently, commonly used decision-making methods include
state machines, decision trees, deep learning, and reinforce-
ment learning [29]. Through the real-time state analysis of the
vehicle road data, the guidance and diversion of the traffic
flow in time and space can be realized to avoid road conges-
tion, improve lane management, and publish traffic guidance
information through variable traffic signs or digital traffic
broadcasts.
1.2 The development stage of IR
To better understand the development and current situation
of IR, according to the technological development path of IR,
this paper summarizes the development of IR into four stag-
es: concept primary application stage, communication tech-
nology driven development stage, green energy technology
driven development stage, and automated driving technology
driving development stage. The application of each stage of
technology enriches and expands the concept and content of
IR, moving from the original conceptual model to the multi-
functional and diversified technology integration.
1.2.1 Concept primary application stage
From the 1970s to the 1990s, the concept of IR began to
enter the public eye. Many countries started relevant research
and system development, and many IR research cooperation
projects began to take off. The California PATH project,
established in 1986, is the first research project in North
America focusing on intelligent transportation systems. It has
conducted comprehensive systematic research and long-term
experimental exploration of AHS. PATH plans to divide the
AHS architecture into five levels: Network Layer, Link
Layer, Coordination Layer, Management Layer, and Physical
Layer [30]. As shown in Figure 3, the function of the network
layer is to estimate the network state of the highway based on
information collected by the roadside or vehicle sensors. The
link layer obtains the acceleration and queuing length target
values of the vehicle through the broadcast link and controls
the traffic flow in the road segment or link of the highway
network [31]. The coordination layer is mainly to coordinate
the execution of the manipulation between the vehicle groups
while assisting the management in performing the manipu-
lation. A closed-loop control system consisting of manage-
ment and physical layers controls each vehicle in both the
vertical and horizontal directions. In September 1988, the
first use of permanent magnets embedded in the road for
vehicle guidance experiments demonstrated the ability to
accurately detect vehicle position based on magnetic field
measurements [32]. In February 1991, a 300 m long test
track was constructed, which provided a venue for testing
automatic vehicle steering control based on the magnetic
guidance concept and implemented the first automatic
steering control on the Toyota Celica. Besides, the Virginia
Smart Road Project, which began in the 1980s, is a 9.17 km
long restricted access highway that has been built as the only
road in the northern United States that can test autonomous
driving, network technology, and smart transportation
systems [33].
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 5
Figure 3 Hierarchical architecture of the AHS in the PATH
program [31]
In 1973, Japan began the first Intelligent Transportation
Control System for the Comprehensive Automobile Control
System (CACS), which studied the world’s first road auto-
matic navigation system [34]. From the mid-1980s to the
mid-1990s, Japan successfully developed projects such as the
Road Automobile Communication System (RACS) and the
Advanced Mobile Traffic Information and Communication
System (AMTICS) [34–35]. In 1987, Europe began the Program
for a European Traffic System with Higher Efficiency and
Unpreceden ted Safety (PROMETHEUS), whose primary
purpose is to research in the field of transportation in-
formatization [36]. In 1990, the United States established an
intelligent vehicle system organization and began to use
roads and communications to study road traffic to improve
road traffic efficiency. In January 1994, Japan established the
Vehicle, Road and Traffic Intelligence Society (VETIS), and
in July 1995, it established the Vehicle Information and
Communication System (VICS) Center. Since 1999, National
Intelligent Transport Systems Center of Engineering and
Technology has carried out in-depth research and develop-
ment of IR systems with vehicle magnetic induction control
technology as the starting point [37].
1.2.2 Communication technology driven develop-
ment stage
In September 2003, the European ITS organization, the
European Road Transport Telematics Implementation Coor-
dination Organization (ERTICO), proposed the concept of
Electronic Safety (ESafety) and began research on the Eu-
ropean vehicle-infrastructure cooperation system. At the
same time, the European Union began to lead the research
and application of European ITS. On January 1st, 2011, the
EU officially launched the Drive C2X, a vehicle networking
project, which was successfully tested in 2014. In 2010, the
United States launched the “Intelligent Transportation Sys-
tem Strategic Plan (2010–2014)”. Then in August 2012, the
“Intelligent Transportation System Strategic Plan
(2010–2014): Progress in 2012” was released. The
Connected Vehicle was established as a general term for
research related to vehicle routing, and the security-based
Dedicated Short Range Communications (DSRC) Vehicle to
Vehicle (V2V) technology has achieved initial results and
completed typical applications. In 2006, the Japanese gov-
ernment and 23 well-known companies launched the
SmartWay program, which is the basis for the development of
the Japanese vehicle road system. It designed the system
framework and open platform structure. The SmartWay sys-
tem was popularized throughout Japan in 2010. China’s re-
search on vehicle-infrastructure cooperation technology
started late. In October 2010, the State Council proposed two
projects involving critical technologies for the Internet of
Vehicles in the “836” plan. That is to say, the key technology
research of smart vehicle-infrastructure cooperation and the
key technology research of synergetic linkage control of
regional traffic in large cities passed the acceptance test in
February 2014. The project completed the system framework
of the vehicle-infrastructure cooperation system and pro-
posed the integrated testing and demonstration scheme of the
vehicle-infrastructure cooperation system, breaking through
a number of key technologies, as shown in Figure 4 [38]. As
can be seen from Figure 4, Vehicle to Infrastructure (V2I)
communication and V2V communication and applications
provide an effective platform for improving active traffic
safety applications.
Figure 4 Vehicle-infrastructure cooperation to achieve multi-
modal traffic interconnection [38]
1.2.3 Green energy technology driven development
stage
In the green energy technology driven development stage,
people pay more attention to environmental protection con-
struction and green travel modes. At the 16th Intelligent
Transportation World Congress, Congress focused on the
future of transportation, climate change and the development
of intelligent transportation, with the aim of enabling many
technologies in the field of intelligent transportation to con-
tribute to energy conservation and emission reduction. At this
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 6
stage, energy-saving emission reduction and environmental
protection are achieved by installing power generation
equipment such as photovoltaic panels on the road and using
green clean energy. In terms of energy conservation and
emission reduction, a new type of IR with clean energy using
solar energy has emerged. The European Highway Research
Laboratory has proposed the concept of Forever Open Road [39],
which includes applying directly to the road itself rather than
at the edge of the road. The reason is that the construction of
asphalt pavements creates costs that can be deducted from the
cost of more expensive solar panels. Besides, solar roads can
also provide lane markers with integrated LEDs and heating
elements that prevent snow and ice from accumulating in
winter. In theory, the power generation of road panels will
cover construction costs over time, and if widely deployed,
they will provide a decentralized grid system. In 2010, the
$100,000 investment in a solar panel prototype by the US
Federal Highway Administration was a multi-layer prefab-
ricated panel with a transparent glass surface and solar panels
that formed the surface layer and was equipped with em-
bedded LED lighting. Below it is an electronic layer con-
taining microprocessors that control lighting, heating,
communication, and monitoring, and finally, the bottom is a
substrate layer that distributes power and data signals. In the
first half of 2014, the Netherlands remodeled the first IR that
could shine on the N329 road in its capital, Amsterdam.
Subsequently, the United States, France, the United King-
dom, and others have launched their new energy roads. In
December 2017, the world’s first photovoltaic highway was
put into operation in Jinan, China. This 2 km road has been
connected to the grid for power generation and will be mobile
charging in the future. In addition, the United States, France,
the United Kingdom, South Korea, etc. have also built their
photovoltaic power generation roads to explore the applica-
tion of solar energy in the road. This green road, represented
by photovoltaic pavement, can provide effective clean energy
protection for automatic de-icing of roads and wireless
charging of electric vehicles.
1.2.4 Autonomous driving technology driven devel-
opment stage
In the autonomous driving technology driven development
stage, the ICV technology aiming at autonomous driving has
been continuously developed. The optimization goal of
safety, comfort, energy-saving, and efficient driving has been
realized soon, and the goal of unmanned driving is realized in
the long term [40]. Developed countries such as the United
States, Japan, and Europe have begun to intelligently trans-
form and upgrade the communication equipment and envi-
ronmental sensing equipment of some experimental roads to
assist ICVs in completing highly automated driving. More
than 1 800 research projects have been conducted at the
Smart Road test site in Virginia to test autonomous driving,
network technology, and smart transportation systems. The
most distinctive feature is the establishment of a road system
with a weather simulation system [33].
At present, the controlled and closed intelligent network
test site bears many functions such as ICV technology re-
search, module development, and performance verification.
The test field simulates a variety of road scenes and is
equipped with relevant communication and positioning
equipment, and it is also a necessary facility for transforming
IR. In the intelligent networked automobile test site, foreign
construction time is earlier, and many can provide a reference
for the construction of IR. The famous Mcity test site [41–42] in
Ann Arbor, Michigan, USA, is a $10 million joint venture
between the University of Michigan and the Michigan De-
partment of Transportation for research and development,
test evaluation of autonomous driving, vehicle networking
technology, and electrical safety systems. The simulated
town is the world’s first test site designed for ICVs. Founded
in 2014, the Swedish Astazero Safety Technology Compre-
hensive Test Site is the largest intelligent vehicle test site in
Europe. The tests include communication technology, V2V
and V2I functions, vehicle dynamics, etc. [43]. The City Cir-
cuit test site, built by the famous British automotive test
service provider Mira, is located in Midland, England. It
covers an area of about 3.04 million square meters and pro-
vides a completely reproducible security real environment for
traditional vehicles and ICVs. South Korea, Japan, etc. also
owned and built their own smart car test sites.
In recent years, the construction and development of
China’s smart car test site have been rapidly developed. On
June 7, 2016, the closed test area of China’s first national
intelligent networked car (Shanghai) pilot demonstration
zone approved by the Ministry of Industry and Information
Technology was officially opened. The closed test zone of
this pilot demonstration zone is located in Jiading, which is
the most abundant closed test area for V2X communication
technologies such as DSRC and LTE-V and the most func-
tional scenes in the world. At present, the demonstration
zones that have been approved in China include five
demonstration zones in Beijing, Jilin, Chongqing, Wuhan,
and Zhejiang. In addition to the above demonstration areas,
the Xiangjiang New District Intelligent Driving Test Zone in
Hunan and the M-CITY project with a total investment of
CNY 10 billion in Shenzhen are also under construction. In
terms of scientific research and construction in universities,
the car networking and smart car testing grounds of Chang’an
University are at the forefront of China. The test site covers
an area of 280 000 m2, with a 2.4 km high-speed circular
runway, a 1.1 km straight test track, a 13 000 m2 steering
stability test plaza, an F3 lane, a car driving training ground,
and five reliability-enhanced typical test roads. Three kinds
of low-adhesion-coefficient pavements and other special
vehicle test road facilities integrated 4G-LTE, LTE-V, Wi-Fi,
802.11p, and EUHT wireless networks, and built a relatively
complete vehicle network communication system, as shown
in Figure 5. The test site road system can be used as an in-
telligent highway prototype system. On July 10, 2018, the
Ministry of Transport has made autonomous-drving closed
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 7
field test base qualification confirmation and awarding based
on “Technical Guide for Autonomous-driving Closed Test
Site Construction (Interim)” for Chang’an University, Re-
search Institute of Highway Ministry of Transport, and China
Merchants Chongqing Vehicle Inspection Institute. With the
opening of the self-driving car road test and the autonomous
driving test site, the intelligent development and commer-
cialization of the road infrastructure will also be accelerated.
Figure 5 IR prototype system in vehicle-infrastructure coopera-
tion test site of Chang’an University
2 The key technologies of IR
IR is a complicated and huge systematic project integrat-
ing various aspects of professional technology, involving
information collection and processing, information encryp-
tion and transmission, road materials, intelligent management
and detection and other aspects of technology. Based on the
current mainstream technology and promising development
direction, this paper summarizes the eight key technologies
applied in the future on IR, including ubiquitous wireless
communication technology, vehicle full-time continuous
high-precision positioning technology, vehicle platoon based
on vehicle coordination technology, intelligent road material
technology, wireless charging technology, road control
technology for active safety, vehicle-to-road information
interaction for Mobility as a Service (MaaS) and intelligent
decision-making and planning technology combined with
infrastructure.
2.1 Ubiquitous wireless communication technology
The ubiquitous network is a ubiquitous network coverage
based on heterogeneous network convergence and spectrum
resource sharing. It can leverage existing and new network
technologies to achieve ubiquitous and on-demand infor-
mation acquisition, delivery, storage, awareness, decision
making, and use of integrated network systems such as inte-
grated services. In the IR, the complementarity of various
technologies of wireless communication becomes more and
more obvious and significant. Different access technologies
for different terminals of people, vehicles, and roads have
different coverage areas, different applicable areas, different
technical features and different access rates, the integration
process and the diversification of the combined network are
significant, as shown in Figure 6. Ubiquitous wireless
communication technology is an important guarantee for the
normal operation of IR facilities. The main task now is how
to integrate wireless communication technology with road
facilities better and solve network performance problems of
vehicle-road communication.
Figure 6 Application of ubiquitous wireless technology in IR
There is a large amount of research to test and apply
wireless communication technology on IR. Among them,
Radio-over-Fiber (ROF) technology [44] is a communication
system that can be used to construct new V2I information
interaction in the future, which can realize high-rate and
reliable transmission and interaction of various information.
The Carlink platform [45] based on roadside wireless base
stations supports inter-car communication. It provides a
wireless service based on hybrid intelligent transportation for
real two-way communication entities for various traffic and
security services. Visible Light Communication (VLC)
technology [46] can be applied to autonomous driving queues.
Noh et al. [47] proposed a V2I collaboration system that ex-
tends the scope of context perception and improves the highly
automated driving situational perception. The deployment
conditions and communication settings of the Road Side Unit
(RSU) on the IR have an impact on the communication
quality of the V2I supporting the IEEE 802.11p protocol [48],
mainly affecting the coordinated driving of the vehicle, es-
pecially in the traffic interference scenario. To address the
effects of environmental disturbances, Jia et al. [49] developed
an enhanced cooperative micro-follow-up traffic model that
considers V2X, while applying WAVE technology for inter-
active communication [50]; vehicles can communicate using
infrastructure. Even though the V2V communication service
is suddenly interrupted, the infrastructure can provide reliable
service.
In terms of network performance services for vehicle
communication, Dey et al. [51] evaluated the performance of
Het-Net, including Wi-Fi, DSRC, and LTE technologies for
V2V and V2I communications. Based on this, an application
layer switching method was developed to enable Het-Net
communication for two Connected Vehicle Technology
(CVT) applications: traffic data collection and forward col-
lision warning. It has also successfully used the application
layer switching technology to maintain a seamless
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 8
connection of CVT applications, which can be adopted in
future connected vehicles supporting Het-Net. Chennikara-
Varghese et al. [52] proposed a conceptual framework for
integrating RSU with Local Peer Groups (LPG) organization
and routing protocols. Still, in LPG, it is also necessary to
develop efficient and reliable unicast and Multicast protocols,
as well as multi-channel MAC protocols that support such
LPG-to-RSU protocol extensions. In interactive communi-
cation, V2V and V2I communications have a significant
impact on connection performance. A car-to-vehicle ap-
proach can be considered to assist in the transmission of
safety-related information between the vehicle and the infra-
structure [53]. In terms of cooperation between multiple RSUs
in a two-way road scenario, Ko et al. [54] proposed a maximum
service algorithm to improve the channel efficiency of V2I
and V2V communications to maximize system performance.
Reference [55] studied beam design to maximize the data rate
of a millimeter-wave V2I system based on beam switching.
Ubiergo et al. [56] proposed a control strategy of consulting
speed limit based on V2I communication. Reference [57]
proposed a new analysis model for performance analysis and
clustering design in self-organizing vehicular networks to
achieve the required system reliability and network
throughput.
Ubiquitous wireless communication technology is the ba-
sis for highly intelligent vehicle-to-road communication and
collaboration. Although the relevant industries have already
achieved many important achievements, there are still many
practical problems in China that need to be solved and
overcome, for example, channel partitioning and frequency
resource allocation, self-organizing network architecture
design, link multiplexing technology, heterogeneous-network
vertical switching method, antenna gain and sensitivity, and
DSRC new standard formulation. At present, the more ma-
ture commercial technology is the Long Term Evolution-
Vehicle (LTE-V) technology promoted by Huawei, Datang
and other enterprises in China. This technology has a com-
patible cellular network and multiple advantages such as
system compatibility with DSRC technology and smooth
transition to 5G.
2.2 Vehicle full-time continuous high-precision
positioning and navigation technology
High-precision positioning technology is the basis for ve-
hicle safety applications and personalized traffic information
services. With the gradual maturity of the Beidou system’s
sub-meter precision positioning technology, and the combi-
nation of wireless positioning technology such as WLAN
signal positioning and RF wireless tag positioning installed
on the road infrastructure, high-precision positioning tech-
nology is directly applied to urban traffic planning and the
management, intelligent public transport, vehicle safety and
assisted driving, intelligent travel and other fields to promote
the technological upgrading of IR. However, the existing
GPS, inertial navigation and other devices have insufficient
positioning accuracy and are easily affected by the environ-
ment, which cannot meet the requirements of IR applications.
Therefore, exploring new, high-precision and economic po-
sitioning technology is the key to realizing IR.
In the research of high-precision positioning technology,
wide-area precise positioning technology [58] can be applied
to vehicle network systems, which can provide lane-level
positioning services in urban environments. O’Keefe et al. [59]
proposed a method of tightly coupling carrier-phase differ-
ential GPS with ultra-wide bandwidth (UWB) ranging for
V2I relative navigation. Haak et al. [60] proposed a lane-level
high-precision vehicle positioning method that solves the
problem that vehicle positioning accuracy is susceptible to
digital map quality. In Reference [61], LED street lights with
different color temperatures and optical focus are installed on
the lane, and the position of the self-driving vehicle is de-
termined by analyzing the coordinates of the chromaticity
points [61]. Kawamura et al. [62] developed a vehicle naviga-
tion system with a Radio Frequency Identification Devices
(RF-ID) system that uses information embedded in the RF-ID
tag under the road to find the location of the vehicle. The
RF-ID tag has the ability to intelligent the driving lane
through driving support information. Pashaian et al. [63]
proposed a map matching method based on fuzzy logic and
neural network and developed a vehicle navigation system
that combines low-cost GPS receivers and microcontroller
chips. Dai et al. [64] developed a Location Based Service
(LBS) based on self-organizing vehicular network, using
RSU as a road infrastructure support for processing K-
Nearest Neighbor (KNN) queries. Kim [65] proposed an in-
frastructure-based path tracking control system that includes
an infrastructure sensor module, a vehicle controller, and an
actuator module. Combined with the above high-precision
positioning technology, in the future construction of intelli-
gent highways, LED sensors, radio frequency identification
and other sensors can be interconnected to achieve compre-
hensive and timely vehicle information positioning.
In improving vehicle navigation accuracy, Pashaian et al. [66]
proposed two methods based on fuzzy logic and neural net-
work to solve the matching problem in car navigation sys-
tems. Zhang et al. [67] proposed a solution for co-localization
using symmetric measurement equation filters to solve data
correlation problems. Ponte et al. [68] proposed a solution that
uses radar sensors to detect surrounding road infrastructure
and locate associated errors based on the Global Navigation
Satellite System (GNSS). To solve the carrier phase integer
fuzzy, Vu et al. [69] proposed a smooth frame for estimating
the sensor platform trajectory using inertial measurement
unit, dual-frequency GPS pseudorange and carrier phase
receiver, and its positional accuracy is centimeter. Liu et al.[70]
proposed a low-cost coordinated vehicle positioning solution
integrated through dedicated short-range communication and
dead reckoning to facilitate DSRC/DR-based integration
solutions to compensate for GNSS and achieve desired ac-
curacy and availability.
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 9
In the future of IR, real-time high-precision positioning of
large-scale vehicles in the road will help to fine-tune the
characteristics of traffic flow and precise control, which is
conducive to real-time monitoring of urban traffic and traffic
flow guidance control, so that vehicles can realize automatic
identification and collision warning of abnormal behavior
according to the current position information and micro-
scopic motion recognition such as lane change, overtaking,
and motion direction correction. Figure 7 shows the overall
solution for real-time precise positioning of IR large-scale
vehicles through intelligent wireless positioning technology,
an inertial navigation system, satellite positioning, vehicle
autonomous positioning, and roadside facility assisted posi-
tioning and high-precision map. The plan will make full use
of the positioning information provided by the roadside fa-
cilities to make up for the lack of satellite positioning, to
ensure the real-time acquisition of high-precision positions of
all vehicles on the IR, realize the visibility, storage and con-
trol of the vehicle’s motion trajectory, and contribute to the
early warning, emergency response, and post-mortem analy-
sis of traffic accidents.
Figure 7 Several techniques used in high precision positioning
2.3 Vehicle platoon technology based on vehicle-
infrastructure cooperation
As shown in Figure 8, the vehicle platoon uses the distance
detecting sensor and the vehicle-vehicle/vehicle-road com-
munication technology to make all the vehicles automatically
follow the head vehicle, so that the entire fleet forms a uni-
form whole. Since the control of the speed and the relative
distance between the vehicles in the platoon is synchronized,
the driving mode of the vehicle can significantly shorten the
vehicle spacing and reduce the wind resistance, thereby sig-
nificantly reducing the overall fuel consumption of the pla-
toon. The platoon allows the vehicles in the queue to travel at
a higher average speed, thus improving road traffic efficiency
and driving safety. Reliable communication links in the
vehicle queue are key, but in the harsh environment of the
road environment, tunnels, bridges, mountains, high-rise
buildings, vegetation, and other roads obstruct the objects,
and electromagnetic interference will affect the quality of
wireless communication between vehicles, thus affecting the
platoon safety and control performance. At present, research
on the vehicle platoon mainly focuses on network commu-
nication and motion control optimization.
In the network communication technology of vehicle
platoon, Bernardo et al.[71] regarded the vehicle platoon con-
trol and management process as a time-delay system, using
the Lyapunov–Razumikhin theorem to analyze its time-delay
and explore the tolerable maximum communication delay of
vehicle platoon system. Lei et al. [72] built a simulation plat-
form to study the impact of packet loss rate of wireless
communication on the performance of the vehicle platoon
system through simulation method and to obtain the mini-
mum packet loss rate required to ensure the stability of the
fleet. At present, IEEE 802.11p-based DSRC has become the
first network studied by vehicle platoon researchers [73]. IEEE
802.11p has the characteristics of self-organization and low
latency, which can meet the needs of workshop communica-
tion in vehicle platoon under most conditions. However,
vehicle platoon communication technology still has its
shortcomings: In the environment with high vehicle density,
it is easy to lose packets [74]; the information transmission
distance is short, about 300 m [75]; and it is easy to be blocked
and covered. These problems have not been well resolved. IR
can compensate and optimize the instant communication and
network stability of the vehicle platoon through V2X com-
munication and environment-aware devices.
In terms of motion control and optimization of the vehicle
platoon, Saeednia et al. [76] divided the truck group into three
stages: fleet formation, fleet maintenance, and fleet dissipa-
tion. A hybrid strategy of hybrid overtaking and deceleration
modes were proposed to optimize the expected speed of the
vehicle platoon formation process. Hall et al.[77] optimized
some static parameters such as optimal captain and platoon
order during the formation of the team. Amoozadeh et al. [78]
developed a fleet marshaling protocol based on Vehicular
Ad-hoc Network (VANET) wireless communication, which
decomposed all vehicle grouping strategies into three basic
types: consolidation, splitting and lane change. Zheng et al. [79]
proposed a Distributed Model Predictive Control (DMPC)
for heterogeneous vehicle groupings with one-way topolo-
gies and previously unknown expected set points, using the
sum of local cost functions. As a Lyapunov function candi-
date, it is proved that the asymptotic stability of DMPC can
be achieved by clarifying the sufficient conditions of the
weight of the cost function. At the same time, in Reference [80],
the influence of information flow topology on the internal
stability and scalability of heterogeneous vehicle group
movement in rigid structures was studied. Sabau et al. [81]
proposed a new distributed control architecture for hetero-
geneous grouping of autonomous vehicles, optimized the
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 10
Figure 8 Vehicle platoon based on vehicle-infrastructure cooperation
controller design based on cost specifications, and introduced
a synchronization mechanism to compensate for the time
delay caused by wireless communication accurately. Har-
fouch et al. [82] designed a switching control strategy based on
communication delay. When the vehicle network is in good
condition, it starts a Coordinated Automatic Cruise Control
(CACC) model. Otherwise, Adaptive Cruise Control (ACC) is
enabled and analyzes the stability of this switch control strat-
egy. At present, the main problem in the vehicle platoon re-
search is that most of the teams consider the grouping strategy
and the following model in the case of normal traffic flow and
rarely mention the fleet motion control under emergency or
network delay conditions, which is precisely the key whether
platoon technology can be put into practical use.
In the future, in the IR system, the reliability of V2I will be
fully utilized to make up for the shortcomings of V2V ran-
domness, and real-time high-precision maps and high-
precision wireless positioning services will be provided to
ensure the reliability and safety of vehicle platoon operation.
At present, there are still many problems to be solved before
the commercial vehicle platoon is commercialized. For ex-
ample, in complex traffic environments and severe weather
conditions, the reliability of some technical means (wireless
communication, sensors) cannot be completely guaranteed,
and the vehicle security assurance strategy has not yet been
subjected to systematic verification and testing. How to
prevent the vehicle queue information system from being
attacked by hackers is also an urgent problem to be solved.
2.4 Wireless charging technology
Wireless Charging Technology (WCT) refers to the
charging of electrical equipment using electromagnetic fields
or electromagnetic waves without connecting through phys-
ical wires. At kilowatt power levels, the transmission distance
increases from a few millimeters to hundreds of millimeters,
and grid load efficiency is higher than 90%. These advances
have made the WCT very attractive for electric vehicle (EV)
charging applications in stationary and dynamic charging
scenarios [83]. At present, there are three kinds of wireless
charging methods: electromagnetic induction type, electro-
magnetic resonance type, and radio wave type. The electric
vehicle wireless charging system is a complex nonlinear
magnetoelectric coupling system. There are three mainstream
wireless charging standards: PMA (Power Matters Alliance),
Qi (Wireless Charging Alliance’s wireless charging standard)
and A4WP (Alliance for Wireless Power). At present, Qual-
comm Halo equipment has an energy conversion efficiency
of more than 90%, which can charge more power in a shorter
time. Compared with the wired charging of electric vehicles,
wireless charging has the advantages of convenient use,
safety, reliability, no spark and electric shock, no dust and
contact loss, no mechanical wear, no corresponding mainte-
nance problems. It can adapt to rain, snow, and other bad
weathers and environments. The schematic diagram of
wireless vehicle charging is shown in Figure 9.
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 11
Figure 9 Application of wireless charging technology
In research on road wireless charging applications, De-
florio et al. [84] proposed a method for assessing wireless
inductive power transfer performance for charging electric
vehicles while driving. Xiao et al. [85] proposed an inductive
wireless charging channel composed of multiple spiral coils
for electric vehicles. Chen et al. [86] studied a WCT charging
system using a series of segmented primary coils, discussed
the relationship between primary coil size and transmission
efficiency, and proposed the promotion design of primary
coils for simultaneous power supply on two-lane roads. Wang
et al. [87] proposed a hybrid framework that combines wireless
charging and solar technology. The cluster heads in the net-
work are equipped with solar panels to remove solar energy,
and the remaining nodes are powered by wireless charging.
The hybrid frame can reduce the energy consumed by the
battery by 20% and save the vehicle’s moving cost by 25%.
By allowing partial recharging, one can further reduce the
energy consumption of the battery with a slight increase in
cost. In the future, the number of high-cost mobile chargers
can be reduced by deploying lower-cost solar sensors. Ref-
erence [88] proposed a follow-up model to study the motion
behavior of each EV near Wireless Charging Lane (WCL).
The paper points out that due to the high cost of WCL, WCL
is currently not applied across the highway. Applying Part
Wireless Charging Lane (PWCL) on some roads may be a
viable solution. Reference [89] studied the longitudinal safety
of electric vehicles based on PWCL on highways. When
PWCL is used, the distribution of EV significantly affects
longitudinal safety. Larger vehicle deceleration will result in
a higher longitudinal collision risk for EV, while the length of
PWCL has no significant effect. Jiang et al. [90] proposed an
inductively coupled wireless charging system for 48 V light
electric vehicles. They evaluated and designed the system’s
main power levels, including high-frequency inverters, res-
onant networks, full-bridge rectifiers, and load-matching
converters.
There are still many engineering problems in the current
wireless charging technology that need to be solved. In the
future, if the wireless charging of electric vehicles is realized,
the wireless charging module is placed on the IR, and the
power is supplied by using new energy such as solar energy,
the power battery capacity of the electric vehicle can be
greatly reduced, and the energy-saving and emission reduc-
tion are facilitated. The operating cost of the car provides a
new idea for expanding the driving range of electric vehicles.
The wireless charging technology of electric vehicles will be
in a hot spot in the industry for a long time to come.
2.5 Intelligent road material technology
In the IR system, intelligent material technology is the
basis of road intelligence: On the one hand, intelligent road
materials can sense environmental stimuli autonomously,
analyze, process and judge them, and take certain measures to
respond appropriately; on the other hand, the excellent
characteristics of smart materials enable the road surface to
have intelligent functions such as energy collection,
self-regulation, self-diagnosis, self-healing, and information
interaction, which enhances the road service capability.
In the research of intelligent material applications, smart
concrete adds intelligent components to the original com-
ponents, making concrete a multifunctional material with
self-perception and memory, self-adaptation and self-healing.
According to these characteristics, it is possible to effectively
predict the internal damage of concrete, meet the needs of its
own safety detection, prevent the brittle failure of concrete
structures, and automatically repair according to the test
results, significantly improving the safety and durability of
concrete. In recent years, a series of intelligent concrete, such
as self-destructive concrete, temperature self-adjusting con-
crete, and bionic self-healing concrete, have laid a solid
foundation for the study of intelligent concrete [91]. The
cement-based piezoelectric composite [92] can be used as the
sensing element of the new intelligent traffic monitoring
system, and its piezoelectricity enables powerful and accurate
real-time detection of the pressure caused by traffic flow.
Embedded cement-based piezoelectric sensors and related
measuring devices have excellent intelligent traffic moni-
toring capabilities, such as traffic flow detection, vehicle
speed detection, and dynamic weighing measurement. The
Intelligent Dust Sensor Network [93] can be used to monitor
road surface temperature and humidity conditions and to
detect icy road conditions. A new generation of asphalt
binders [94] is a smart material that adapts the mechanical
properties of a material to its actual changing load conditions
during the life of the material. The matrix has been modified
with magnetic particles when activated by magnetic fields
and can alter the mechanical properties of the adhesive,
providing a wide range of applications for implementing this
application in smart infrastructure (Figure 10), especially for
construction, repair and maintenance of asphalt pavements.
Jung et al. [95] proposed a piezoelectric energy harvester
module based on polyvinylidene fluoride polymer, which can
be used in IR applications. Alavi et al. [96] proposed a con-
tinuous health monitoring method for asphalt concrete
pavement based on piezoelectric self-sensing technology.
The signal sensed by the piezoelectric sensor from the traffic
load can be used to enhance the self-powered sensor capa-
bility and also for damage diagnosis. Nasir et al. [97] have
studied the road-surface solar collector system and found that
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 12
by absorbing heat from the road surface and using heat en-
ergy, one can reduce the heat of the road surface.
Figure 10 Sketch of phenomena appearing in the mechanomuta-
ble binders when activated by a magnetic Field [96]
In the future of IR, conductive materials such as conduc-
tive fibers and graphite are added to the pavement material to
make the road surface conductive, to achieve the function of
snow melting and ice melting through electric heat transfer.
The detection of micro-cracks on roads is challenging. The
self-healing technique of cracks can be used. Based on the
basic principles of bionics, the phenomenon of damage
healing and its principle can be simulated. The self-healing of
road damage can be stimulated by energy compensation or
material compensation. The addition of graphite, conductive
fibers, and the like to the pavement material enable the road
surface to have electrical conductivity, thereby achieving
energy supply by means of electric heating. Due to the
high-performance requirements of various aspects of road
materials and the input cost of materials themselves, there is
still a period of large-scale commercial application of intel-
ligent road materials. However, the application of intelligent
material technology in the future will significantly improve
the overall performance of IR.
2.6 Road control technology for active safety
In the IR system, the road control technology for active
safety is an important goal of road traffic design and intelli-
gent integration. Safety is integrated into the whole process
of road traffic design and operation management, and road
traffic information can be predicted in advance. As shown in
Figure 11, through the RSU, On-Board Unit (OBU), elec-
tronic information card, smartphone, etc., road situation
perception, construction area warning, weather warning of
specific locations, accident notification, roadside alarm,
parking notices, and pedestrian collision warnings can be
realized to sense, remind and plan the driving route and road
rights level of special operating vehicles in advance. The
electronic equipment of road infrastructure can assist the
special operation vehicle in carrying out high-efficiency
monitoring and tracking of the entire process and realize the
pre-diagnosis, accident warning and active intervention of the
road traffic safety situation; at the same time, it can carry out
active safety protection on the road surface itself and improve
the road tires. The friction, rain and snow, anti-slip warning
system and other functions can improve the safe passage
capacity of the road itself.
Figure 11 Application of IR safety technology
In the field of road control technology for active safety, the
road condition monitoring system [98] can detect and locate
road abnormalities by using low-cost MEMS accelerometers
and GPS receivers in the tablet. Laser ranging technology [99]
can establish an early warning system for automatic moni-
toring of snow depth, real-time, continuous and automatic
monitoring of snowy road conditions. Based on the road
surface temperature monitoring system of the General Packet
Radio Service (GPRS) technology and the Rich Internet
Application (RIA) model [100], it can automatically heat the
melting surface of the ice when the temperature data exceed
the warning range and send the alarm information to the
monitoring manager simultaneously. Accident risk can also
be assessed through intelligent infrastructure equipment, and
neural networks and fuzzy logic can be applied to transport
control to improve safety [101]. Zhou et al. [102] established a
road solar system to analyze the heat transfer mechanism and
effect of the system based on the monitored solar radiant heat,
the solar energy absorbed by the road and the heat stored in
the soil. Ito et al. [103] proposed a road condition monitoring
system that uses various sensor data in a severe communica-
tion environment, to collect various sensor data, select sensor
data for the purpose, and use the selected sensor data to un-
derstand the road conditions. At the intersection of cities,
sensor data can be used for real-time traffic signal control [104]
and cycle length prediction [105] for different levels of traffic
demand and signal interval, helping to reduce travel delays
and improve intersection safety factor. At the same time,
combined with automated vehicle trajectory control and
intersection optimization control method [106], it can achieve
huge benefits of reducing congestion, reducing collision risk
and reducing fuel consumption and intersection emissions.
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 13
At present, a lot of work has been done on the active safety
protection of the vehicle itself and the safety accident pre-
diction of the traffic flow information. In the future, the IR
will start from the road infrastructure itself and improve the
road safety and stability through real-time monitoring, ad-
vance prediction, automatic sensing, and automatic repair.
2.7 Vehicle-to-road information interaction tech-
nology for MaaS
Travel-as-a-service (MaaS) aims to establish a seamless
networked transportation system based on shared traffic
patterns and intelligent information technology [107]. Users of
MaaS can use the service as pay-as-you-go, or they can
purchase a mobile package based on their travel needs [108]. At
present, Germany, France, the Netherlands, and the United
States have begun to design and implement the MaaS solu-
tion. Through the application and smart card access system,
all the transportation modes of operators or cities can be
accessed, so that travelers can enjoy travel plans, travel res-
ervations, real-time information, personalized travel advice
and other services. The MaaS model is based on an integrated
technology that eliminates the barriers between different
modes of transportation and enables seamless travel [109]. As
shown in Figure 12, the technology includes six key tech-
nologies [110]. The traffic travel service platform is one of the
six, which is supported by the vehicle-road information in-
teraction technology and is the critical foundation for the
seamless travel of MaaS.
Figure 12 Multi-mode traffic seamless travel technology based
on IR
In the future city passenger travel, on-demand automated
mobile services are foreseeable [111], Autonomous Vehicle
Sharing and Reservation (AVSR) can effectively solve the
automatic vehicle travel chain by constructing a linear pro-
gramming model. The optimal solution and the required fleet
size can further increase the efficiency of the shared system [112].
The MaaS model supports ride-sharing through a variety of
modes of transport, but currently, it relies heavily on the
important role of the car [113]. How to combine the au-
to-driving and intelligent roads to solve the road congestion
problem based on this will become a key factor of a consumer
to weigh the time cost. Ravindran et al. [114] proposed an
information-centric 5G Information-centric Networking
(5G-ICN), which converts MaaS functions into 5G-ICN
slices. It outlines the functional configuration and functions
of interdependence and coordination between 5G-ICN slices
to meet MaaS goals. On the security issue, Thai et al. [115]
proposed a quantitative analysis framework for the vulnera-
bility of the Denial of Service (DoS) to MaaS system attacks.
At present, the research on MaaS’s solutions and technolo-
gies is still in progress. In the future, the technical solutions
combining autonomous driving and intelligent roadside
equipment will become the key development direction of the
MaaS model.
In the future of MaaS mode, the vehicle-road information
interaction technology transmits data information between
the road and the vehicle through information interaction. It
establishes a seamlessly connected road-vehicle information
network system. The road facilities provide scene-adaptive
traffic information for the vehicles. The vehicles provide
holographic location and status information for the roads,
enabling the entire road network and transportation network
to develop in the direction of information flow integration
system, providing flexible, efficient and people-oriented
travel services for travelers and realizing the transition from
private transportation to shared transportation, from discrete
traffic subsystems to integrated transportation systems. As a
high-quality data platform for real-time information on traffic
travel, IR can share part of the data collection cost and pro-
vide a unified data standard interface. IR optimizes the uti-
lization and optimal traffic for travel service providers
through road traffic perception and real-time positioning and
provides a data and service interface to improve the overall
operational efficiency of the system. For traffic management
agencies, IR, as an important information platform carrier, is
conducive to real-time effective supervision of traffic service
safety behaviors and helps to supplement and improve rele-
vant security measures. The construction and development of
IR in the future will be an important basis for the integration,
humanization and low-carbon construction of the MaaS
platform. It is important to support ensuring the development
of transportation towards being green and low-carbon, effi-
cient and convenient, economical and comfortable, safe and
reliable.
2.8 Intelligent decision-making and planning
techniques combined with infrastructure
IR provides a more efficient and stable way of sensing and
communication for autonomous driving and multi-vehicle
collaborative planning. The planning decision module of
autonomous driving is responsible for generating the driving
behavior of the vehicle, which is a key factor reflecting the
intelligent level of the autonomous vehicle [116]. However, in
the actual traffic scene, there are often multi-vehicle collab-
orative planning and decision-making problems. It is
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 14
challenging to solve practical operation problems only by
single-vehicle intelligent decision-making and planning. IR
serves as road infrastructure and a multi-vehicle sensing
communication platform to provide better ideasfor multi-
vehicle collaborative decision making and planning. How-
ever, many studies at present focus only on the decision-
making and planning of single-vehicle intelligence and less
on the impact of intelligent road infrastructure.
The performance of single-vehicle intelligent decision-
making behavior is an important indicator to measure the
intelligent level of intelligent vehicles, including vehicle
heel, lane change, overtaking, and inbound traffic. At present,
the decision-making methods of intelligent vehicle driving
behavior mainly include rule-based decision-making meth-
ods, Markov-based decision-making methods, and neural
network-based decision-making methods. The rule-based
decision-making method is primarily to manually build a
complex structure composed of a large number of rule mod-
ules. This method relies on the experience of rule-makers and
has high interpretability, but it lacks the adaptability to com-
plex and variable traffic environments. Liu et al. [117] proposed
a Markov decision algorithm that integrates the road envi-
ronment and the intention of a vehicle in urban road scenar-
ios, using road scenarios to refer to vehicle behavior changes
and then observing the deviations from reference behaviors to
infer the response of other vehicles. Naranjo et al. [118] estab-
lished a lane change decision model based on fuzzy logic to
improve computational efficiency. Perez et al. [119] proposed a
decision system based on fuzzy logic for autonomous over-
taking on two lanes. The limitation of fuzzy logic is that its
design relies on relevant empirical knowledge, which is often
difficult to be obtained. Probability-based driving decision-
making methods take into account the uncertainty of the
environment and can improve the robustness of intelligent
vehicle driving decisions. Ardelt et al. [120] proposed a prob-
abilistic method for vehicle lane change decision on the
highway and carried out a real vehicle test on the expressway
to verify the effectiveness of the method. Wolf et al. [121] pro-
posed a method of learning to drive a vehicle in a simulated
environment using Deep Q-Network. By benchmarking the
center distance of the lane, one can add other variables (such as
the angular deviation of the center line and vehicles) to im-
prove the learning and driving behavior of the vehicle. Lillic-
rap et al. [122] proposed a behavioral judgment and model-free
algorithm based on deterministic strategy gradient and
depth-dependent reinforcement learning. Ngai et al. [123] pro-
posed a multi-objective reinforcement learning method based
on the Q-learning algorithm for vehicle overtaking decision
and control. Besides, the autonomous driving terminal-to-
terminal solution and the direct acquisition strategy by gener-
ating an anti-network extraction framework can reproduce the
driving behavior of the driver in the face of an unexpected
situation [124]. Considering the macro and micro levels, a
comprehensive model is needed to capture the lane change
decision process and its impact on surrounding traffic [125].
Especially when a new lane-change model is developed, a
multi-level evaluation strategy should be preferred. However,
the current decision-making methods still lack versatility and
flexibility to cope with new complex scenarios. How to inte-
grate with the information and rules determined in the modern
and digital roads also presents challenges.
Multi-Vehicle Motion Planning (MVMP) refers to the
process of solving the path or trajectory of multiple vehicles
based on the initial position, driving purpose and constraints
of multiple vehicles [126]. How to effectively deal with
large-scale complex constraints in a centralized MVMP
problem and guarantee the solution efficiency of the problem
is the key factor affecting the development prospect of a
dynamic optimization method in the field of multi-vehicle
coordinated motion planning [127]. Desaraju et al. [128] pro-
posed a Decentralized Multi-agent Rapidly-exploring Ran-
dom Tree (DMA-RRT) and a cooperative DMA-RRT method
to solve the multi-agent planning path problem affected by
complex constraints. While considering vehicle dynamics,
time and space specifications, and energy-related require-
ments, Häusler et al. [129] proposed a numerical algorithm for
solving multi-vehicle motion planning problems. The cur-
rently distributed MVMP algorithm can be subdivided into
three categories according to the distributed mechanism:
priority allocation method, motion coordination method and
task decomposition method [130]. The priority allocation
method refers to assigning priorities to each vehicle accord-
ing to certain rules, and then performing motion planning for
all vehicles in order of priority from high to low. High-
priority vehicles have been solved when trajectory planning
is carried out for low-priority vehicles [131]. In the lane change
mission, Plessen et al. [132] aimed to maximize the capacity
and adjust the priorities through optimization calculations. In
the intersection task, Dresner et al. [133] proposed a “first
come, first served” priority allocation protocol. In a net-
worked vehicle environment, a cooperative vehicle intersec-
tion control system[134] enables collaboration between the
vehicle and the infrastructure to perform efficient intersection
operations and management when all vehicles are fully au-
tomated. The motion coordination method, also called the
path-speed decomposition method, transforms spatially con-
flicting paths into conflict-free security trajectories in time
and space by configuring speed variables on conflicting travel
paths in multiple workshops [135]. Ahmadzadeh et al. [136] pro-
posed a path planning method for a non-complete multi-
vehicle system with moving obstacles. The task decomposi-
tion method decomposes the complete multi-vehicle coor-
dinated motion process into a series of tasks or processes,
solves them one by one, and finally realizes the complete
coordinated motion process by combining the results of the
tasks. In solving the problem of the multi-vehicle coordinated
lane change, Desiraju et al. [137] divided the vehicle formation
into several groups and changed lanes according to small
batches. Atagoziyev et al. [138] completely decomposed the
collaborative lane changing task into a sequential lane
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 15
changing process in which the vehicles were completed one
by one. Abichandani et al. [139] proposed a general framework
based on mathematical programming that can adapt to all the
objectives and constraints of the MVMP problem, focusing
on the capabilities of various path primitives, spatial and/or
temporal discretization, and centralized/distributed decision
making.
Intelligent decision-making and planning techniques
combined with infrastructure can improve the intelligence
and networking capabilities of ICV decision-making and
planning. On the one hand, intelligent highways expand the
range of ICV perception through a variety of roadside sen-
sors, fully sense the driving environment, and optimize its
motion based on this mode to achieve autonomous intelligent
movement in the driving road view. On the other hand, ICV
uses vehicle-vehicle or a vehicle-road communication net-
work to share vehicle route planning and driving decision
information with other ICV and intelligent roadside facilities
and adopt a collaborative decision-making mechanism to
realize multi-vehicle collaborative planning operation, as
shown in Figure 13 [140].
Figure 13 Intelligent decision and planning techniques combined
with infrastructure [140]
3 Development prospects of IR
3.1 Development and application of IR in vehicle
road integration technology
In the construction of IR, the future will be more on the
existing infrastructure, using information technology, sensing
technology, network technology and other technologies for
system integration and transformation. The vehicle-
infrastructure cooperation sensing integration technology is
based on the roadside sensing device and supplemented by
vehicle perception, realizing the holographic and full-scale
integrated approach of vehicle and road, and improving the
sensing ability of the self-driving vehicle. IR provides an
effective technology realization carrier for the vehicle-
infrastructure cooperation integration technology and is a
realization platform for realizing the scale, industrialization
and integration application of the vehicle-infrastructure co-
operation technology; the vehicle-infrastructure cooperation
technology is also an important technical component of the
IR system and improves important technical support for the
integration of road traffic system information.V2X commu-
nication technology integrated with vehicle roads and
decision-making is based on the wireless communication
between vehicles and vehicles, vehicles and roadside infra-
structure, vehicles and pedestrians. The real-time perception
and timely warning of vehicle surrounding conditions will
become one of the current research hotspots of world coun-
tries to solve road safety problems. In the future, in the field
of unmanned driving, IR upgrades supported by dynamic
wireless charging will also receive more and more attention.
As shown in Figure 14, the driving control of smart vehicles
in the future will be completed by the roadside control system
of the intelligent highway and the onboard control system,
providing a safer, more stable and more efficient environment
for autonomous driving.
Figure 14 IR in the future
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 16
3.2 Development and application of parallel IR
based on virtuality and reality
A parallel system is a common system consisting of a
natural reality system and corresponding one or more virtual
or ideal artificial simulation systems [141–142]. The parallel IR
based on the parallel system uses the artificial simulation
road system as a modeling tool, driven by the information
data of the human-vehicle-road, using the big data calculation
test method for analysis and evaluation and realizing the
physical road system and the artificial simulation system. The
interaction, comparison, and analysis of the behavior be-
tween them complete the “learning” and “estimation” of their
respective future conditions and adjust their respective
management and control methods accordingly. The parallel
IR system framework is shown in Figure 15. The parallel IR
realizes the information flow between the physical highway
system and the virtual highway system through the interac-
tion of virtuality and reality. The physical road system senses
the “human-vehicle-environment” operation status on the
road in real time through sensors and communication devic-
es, preprocesses the collected information and transmits it to
the virtual highway system. The virtual road system consists
of a mathematical model and a specific algorithm. It will
output effective information according to the physical high-
way system for event detection, fault diagnosis, situation
assessment, and system optimization, and perform
closed-loop control to continuously optimize the traffic flow
conditions and vehicle motion conditions of the physical
highway system, so as to ensure the safety of physical roads,
traffic efficiency, environmental performance, and energy
utilization efficiency are always in an optimal state.
Figure 15 Framework of parallel intelligent road
3.3 Rapid development and application of artifi-
cial intelligence technology in IR
Artificial Intelligence (AI) technology, represented by
deep learning and enhanced learning, is rapidly being applied
in the field of environmental perception and deep data min-
ing. On the one hand, in the maintenance of IR, using big
data, artificial intelligence and other technologies, we can
compare the weather information, traffic flow, vehicle com-
position information, and road conditions across the country
to construct a library of road health testing and maintenance
management information based on big data. On the other
hand, deep learning and reinforcement learning can be ef-
fectively applied to in-vehicle systems and roadside systems,
providing better driving prediction and adaptive control,
supporting road-health automatic tracking and maintenance
decision-making, and continuously improving information
transmission and traffic decision-making accuracy. In terms
of cloud computing technology, distributed redundant storage
is generally adopted, which has the characteristics of pro-
cessing large-scale data and realizing data sharing. The mas-
sive data storage and computing requirements of
human-vehicle-road provide opportunities for cloud compu-
ting technology to move from the concept layer to the ap-
plication layer. In the future, combined with video
surveillance points built on the vehicle and roadside, wireless
broadband technology can be used to construct smart vision
vehicle networking, visual management of license plates,
accident scenes, events, traffic flows, etc., which will also
become a research hotspot.
3.4 Application and development of information
security technology in IR
As information and networking play an increasingly im-
portant role in various fields, information security has be-
come a key area of focus for various industries in various
countries. IR is a highly informative complex that integrates
sensing, control and human-vehicle-road data processing. In
the IR system, there are a large number of basic nodes; a large
number of sensing devices exchange information; and a uni-
fied communication protocol has not yet been formed. Once
maliciously attacked and controlled, it will have a devastating
impact on the safety of smart vehicles, road infrastructure,
drivers, passengers, and even pedestrians. Therefore, the
research on IR information security technology should not be
delayed. On the one hand, it should increase the input of
personnel and funds, determine data management objects and
implement hierarchical management, establish a data security
system for data storage security, transmission security and
application security, establish the data security technology
frameworkof IR and formulate technical standards for IR data
security in China. On the other hand, it is necessary to es-
tablish an emergency response system of IR information
security and a legal and ethical framework to ensure the
healthy development of the IR system.
3.5 Accelerated integration of autonomous driv-
ing technology and IR construction
The key series of autonomous driving technology is an
important determinant of the autonomous driving L4 and L5
levels, which can effectively improve driving safety, infra-
structure utilization and driving comfort. Although autono-
mous driving has made significant progress in applied
research, there are still many obstacles to achieve the com-
mercialization of autonomous driving. There are many
problems to be solved, mainly including the following: 1.
Advanced sensors such as laser radar are large in size, high in
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 17
cost, and susceptible to surrounding conditions such as
weather conditions. 2. Although the accuracy of onboard
sensor equipment is high, there are still limitations and blind
spots in the perception of environmental complexity and
diversity. 3. The autonomous single-vehicle intelligent ap-
plication is high in cost and difficult to be commercialized on
a large scale. 4. The safety of autonomous driving still needs
further research. 5. There are no test evaluation methods and
related standard regulations for autonomous driving vehicles.
The cross-integration of IR and autonomous driving tech-
nologies provides new technologies and new ways to solve
the above problems.
IR has the characteristics of perception and control and can
share the partial sensing control cost of the intelligent vehicle
by the large-scale IR. Through the combination of the road-
side sensing device of the IR and the self-driving vehicle with
only the necessary sensing device, the smart vehicle is further
improved in the low speed and miniaturization based on the
comprehensive sensing capability of the intelligent vehicle.
The V2X communication realizes the network-based intel-
ligent technology of the self-driving vehicle and IR, sup-
plementing the shortage of automatic intelligence. It can not
only help autonomous driving systems achieve comprehen-
sive and efficient decision-making and obstacle avoidance
planning capabilities but also provide conditions for vehicle-
group collaborative decision-making planning. The integra-
tion of IR and autonomous driving technology can greatly
improve the autonomy and intelligence of autonomous vehi-
cles, reduce the risks and hidden dangers of single-vehicle
intelligence, and further improve the driving safety of
self-driving vehicles. In the future, a test field combining IR
and self-driving vehicles will become a research hotspot.
3.6 The realization route of IR commercialization
promotion
Looking back on the development history of American
AHS and looking forward to the future, we can conclude that
the final realization and application of IR will take a long
time but will eventually be accepted by society. Based on the
literature analysis, this paper makes a preliminary prediction
of its development roadmap. As shown in Figure 16, the
roadmap includes the following six phases: 1. Conceptual
design and software simulation. 2. Construction of closed
campus prototype system. 3. Key technology breakthroughs
and core component development. 4. Intelligent highway
demonstration application. 5. Standard and norm formula-
tion. 6. Large-scale promotion and application. The above six
stages will overlap in time; there is no obvious dividing line;
and its development characteristics will be an evolving spiral
state in the IR. With the continuous iteration of driving
technology, the function and performance of the IR will
continue to improve, gradually approaching its ideal con-
ceptual design prototype. As shown in Figure 17, this paper
combs and integrates the development of IR. In the future,
with the continuous breakthrough of some emerging
technologies (5G communication, high-precision positioning
of vehicles, artificial intelligence, etc.), the large-scale pro-
motion and application of IR will be further accelerated. The
current popular autonomous driving technology route is
mainly based on the vehicle as the main body, equipped with
advanced in-vehicle sensors, controllers, actuators and other
equipment to make the vehicle an advanced mobile agent, but
the current high-sensitivity device also makes the cost high.
Therefore, part of the perception and control cost of the smart
vehicle can be shared by the large-scale IR. By combining the
roadside sensing device of the IR with the vehicle only in-
stalling the necessary sensing device, we can further reduce
the cost of the vehicle sensing device based on the compre-
hensive sensing capability of the intelligent vehicle. The
driving control of intelligent vehicles in the future will be
completed by the roadside control system of the IR and the
on-board control system, providing a safer, more stable and
more efficient environment for autonomous driving. There-
fore, the future technical route will gradually develop from
the “smart car, ordinary road” mode to the high-tech devel-
opment stage of “smart car, smart road”.
Figure16 Future development route of IR
Figure 17 Context arrangement of IR
4 Conclusions
(1) IR is a high-tech road complex combining sensing,
computing, communication, materials, artificial intelligence,
and system integration technology. Based on the AHS sys-
tem, this paper combs the core concepts, key technologies
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 18
and four historical development stages of IR and proposes a
three-tier architecture consisting of information management
layer, network communication layer and perception control
layer. Initially, the framework of parallel IR system com-
bining virtuality and reality is given, and it is pointed out that
the road will gradually evolve from the “smart car, ordinary
road” mode to the advanced stage of “smart car, smart road”.
(2) At present, the IR is still in the conceptual design stage,
and there are no mature industry norms and technical stand-
ards in the international arena. In this paper, the system ar-
chitecture and key technologies of IR are summarized in
detail, which has particular guiding significance for the future
development and engineering application of IR. With the
continuous updating and upgrading of technology, the con-
cept and content of IR will continue to be enriched and
diversified.
(3) At present, there are relatively few studies on IR. In
this paper, inevitably, some factors are not well considered
for the research and discussion of IR. With the advent of
autonomous driving bottlenecks and the rapid development
of intelligent transportation, there will be more attention and
study on IR in the future. In the future, if an international
association or organization related to IR is established, the
relevant concepts and definitions of IR will be standardized,
and the technical standard system, laws and regulations as-
sociated with IR will be formulated, which will help accel-
erate the development of IR from theoretical ideas to reality.
References
[1] LI Ke-qiang, DAI Yi-fan, LI Sheng-bo, et al. Stateof-the-art and Tech-
nical Trends of Intelligent and Connected Vehicles [J]. Journal of Auto-
motive Safety and Energy, 2017, 8(1): 1–14 (in Chinese).
[2] RAN Bin, TAN Hua-chun, ZHANG Jian, et al. Development Status and
Trend of Connected Automated Vehicle Highway System [J]. Journal of
Automotive Safety and Energy, 2018, 9(2): 119–130 (in Chinese).
[3] MORRISON H M, WELCH A F, HANYSZ E A. Automatic Highway
and Driver Aid Developments [J]. SAE Transactions, 1961, 69: 31–53.
[4] FENTON R E, COSGRIFF R L, OLSON K W, et al. One Approach to
Highway Automation [J]. Proceedings of the IEEE, 1968, 56(4):
556–566.
[5] IOANNOU P A. Automated Highway Systems [M]. New York: Springer
Science and Business Media, 2013.
[6] SINHA K C, COHN L F, HENDRICKSON C T, et al. Role of Advanced
Technologies in Transportation Engineering [J]. Journal of Transportation
Engineering, 1988, 114(4): 383–392.
[7] HALL R W, THAKKER V, HORAN T A, et al. Automated Highway
System Field Operational Tests for the State of California: Potential Sites,
Configurations and Characteristics[R]. Berkeley: University of California,
Berkeley, 1997.
[8] MANIVANNAN P V, RAMAKANTH P. Vision Based Intelligent Vehi-
cle Steering Control Using Single Camera for Automated Highway Sys-
tem [J]. Procedia Computer Science, 2018, 133: 839–846.
[9] HAMOUDA O, KA^ANICHE M, KANOUN K. Safety Modeling and
Evaluation of Automated Highway Systems [C]//IEEE. Proceedings of
the 2009International Conference on Dependable Systems and Networks.
New York: IEEE, 2009: 73–82.
[10] HOROWITZ R, VARAIYA P. Control Design of an Automated High-
way System [J]. Proceedings of the IEEE, 2000, 88(7): 913–925.
[11] TOMIZUKA M. Automated Highway Systems-An Intelligent Trans-
portation System for the Next Century [C]//IEEE. Proceedings of the
1997 International Symposium on Industrial Electronics. New York:
IEEE, 1997: 1–4.
[12] TSAO H S J. Traffic Control for Automated Highway Systems: A
Conceptual Framework [J]. Transportation Research Part C, 1995, 3(4):
227–246.
[13] HALL R W, CALISKAN C. Design and Evaluation of an Automated
Highway System with Optimized Lane Assignment [J]. Transportation
Research Part C, 1999, 7(1): 1–15.
[14] TSUGAWA S. An Overview on Control Algorithms for Automated
Highway Systems [C]//IEEE. Proceedings of the 1999International
Conference on Intelligent Transportation Systems. New York: IEEE, 1999:
234–239.
[15] GLLA, VARAIYA P. Smart AHS: A Simulation Framework for Auto-
mated Vehicles and Highway Systems [J]. Mathematical and Computer
Modelling, 1998, 27(9/10/11): 103–128.
[16] CHINA-SAE. Technology Roadmap for Energy Saving and New En-
ergy Vehicles[M]. Beijing: Mechanical Industry Press, 2016 (in Chinese).
[17] CHEN C, SEFF A, KORNHAUSER A, et al. Deepdriving: Learning
Affordance for Direct Perception in Autonomous Driving [C]//IEEE.
Proceedings of the 2015IEEE International Conference on Computer Vi-
sion. New York: IEEE, 2015: 2722–2730.
[18] VAN NUNEN E, KWAKKERNAAT M R J A E, PLOEG J, et al. Co-
operative Competition for Future Mobility [J]. IEEE Transactions on In-
telligent Transportation Systems, 2012, 13(3): 1018–1025.
[19] SHLADOVER S E. Automated Vehicles for Highway Opera-
tions(Automated Highway Systems) [J]. Journal of Systems and Control
Engineering, 2005, 219(1): 53–75.
[20] CARBAUGH J, GODBOLE D N, SENGUPTA R. Safety and Capacity
Analysis of Automated and Manual Highway Systems [J]. Transportation
Research Part C, 1998, 6(1/2): 69–99.
[21] LI H. A Service-oriented Architecture for Networked Highway [J].
Procedia Engineering, 2011, 15: 2093–2097.
[22] BASKAR L D, DE SCHUTTER B, HELLENDOORNH. Traffic
Management for Automated Highway Systems Using Model-based Pre-
dictive Control [J]. IEEE Transactions on Intelligent Transportation Sys-
tems, 2012, 13(2): 838–847.
[23] BENDER J G. An Overview of Systems Studies of Automated Highway
System [J]. IEEE Transactions on Vehicular Technology, 1991, 40(1):
82–99.
[24] TERZIYAN V, KAYKOVA O, ZHOVTOBRYUKH D. Ubiroad: Se-
mantic Middleware for Context-aware Smart Road Environments
[C]//IEEE. Proceedings of the 2010 5th International Conference on In-
ternet and Web Applications and Services. New York: IEEE, 2010:
295–302.
[25] CHEON S. An Overview of Automated Highway Systems(AHS)and
the Social and Institutional Challenges They Face[R]. Berkeley: Univer-
sity of California Transportation Center, 2003.
[26] VAN DRIEL C J G, VAN AREM B. The Impact of a Congestion As-
sistant on Traffic Flow Efficiency and Safety in Congested Traffic Caused
by a Lane Drop [J]. Journal of Intelligent Transportation Systems, 2010,
14(4): 197–208.
[27] ERHAN D, SZEGEDY C, TOSHEV A, et al. Scalable Object Detection
Using Deep Neural Networks [C]//IEEE. Proceedings of the 2014IEEE
Conference on Computer Vision and Pattern Recognition. New York:
IEEE, 2014: 2147–2154.
[28] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet Classifi-
cation with Deep Convolutional Neural Networks [J]. Advances in Neural
Information Processing Systems, 2012, 25(2): 1097–1105.
[29] ARULKUMARAN K, DEISENROTH M P, BRUND-AGE M, et al.
Deep Reinforcement Learning: A Brief Survey [J]. IEEE Signal Pro-
cessing Magazine, 2017, 34(6): 26–38.
[30] HSU A, SACHS S, ESKAFI F, et al. The Design of Platoon Maneuvers
for IVHS [C]//IEEE. Proceedings of the 1991 American Control Con-
ference. New York: IEEE, 1991: 2545–2550.
[31] ALVAREZ L, HOROWITZ R, LI P. Traffic Flow Control in Automated
Highway Systems [J]. Control Engineering Practice, 1999, 7(9):
1071–1078.
[32] SHLADOVER S E. PATH at 20-history and Major Milestones [J]. IEEE
Transactions on Intelligent Transportation Systems, 2007, 8(4): 584–592.
[33] GIBSON T. Virginia’s Smart Road: Where Researchers Make the
Extreme Weather [J]. Weatherwise, 2015, 68(4): 20–27.
[34] KAWASHIMA H. Japanese Perspective of Driver Information Systems
[J]. Transportation, 1990, 17(3): 263–284.
[35] OKAMOTO H. Advanced Mobile Information and Communication
System(AMTICS) [J]. SAE Paper881176.
[36] FESTAG A. Cooperative Intelligent Transport Systems Standards in
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 19
Europe [J]. IEEE Communications Magazine, 2014, 52(12): 166–172.
[37] LI Bin, WANG Chun-yan, WU Tao, et al. Research Review on Mag-
netic Guidance Technology of Intelligent Highway System in China [J].
Journal of Highway and Transportation Research and Development, 2004,
21(11): 66–69 (in Chinese).
[38] CHANG Cu-yu, XIANG Yong, SHI Mei-lin. Development and Status
of Vehicular Ad Hoc Network [J]. Journal of Communications, 2007,
28(11): 116–126 (in Chinese).
[39] LAMB M J, COLLIS R, DEIX S, et al. The Forever Open
Road-Defining the Next Generation Road [C]//AIPCR. Proceedings of
the 2011 AIPCR World Congress. Mexico City: AIPCR, 2011: 1–16.
[40] Editorial Department of China Journal of Highway and Transport.
Review on China’s Automotive Engineering Research Progress: 2017 [J].
China Journal of Highway and Transport, 2017, 30(6): 1–197 (in
Chinese).
[41] BRIEFS U. Mcity Grand Opening [J]. Research Review, 2015, 46(3):
1–2.
[42] EUSTICE R. University of Michigan’s Work Toward Autonomous Cars
[R]. Ann Arbor: University of Michigan, 2015.
[43] SKOGLUND M, PETIG T, VEDDER B, et al. Static and Dynamic
Performance Evaluation of Low-cost RTK GPS Receivers [C]//IEEE.
Proceedings of the 2016IEEE Intelligent Vehicles Symposium(IV). New
York: IEEE, 2016: 16–19.
[44] NAI W, DONG D, ZHENG W, et al. A Novel Vehicle-to-infrastructure
Communication System Providing Triple Play Services Based on
Radio-over-fiber Technique [C]//IEEE. Proceedings of the 2011IEEE
7th International Conference on Wireless Communications, Networking
and Mobile Computing(WiCOM). New York: IEEE, 2011: 1–4.
[45] SUKUVAARA T, NURMI P. Wireless Traffic Service Platform for
Combined Vehicle-to-vehicle and Vehicle-to-infrastructure Communica-
tions [J]. IEEEWireless Communications, 2009, 16(6): 54–61.
[46] ABUALHOUL M Y, SHAGDAR O, NASHASHIBIF. Visible Light
Inter-vehicle Communication for Platooning of Autonomous Vehicles
[C]//IEEE. Proceedings of the 2016IEEE Intelligent Vehicles Symposi-
um(IV). New York: IEEE, 2016: 508–513.
[47] NOH S, AN K, HAN W. Toward Highly Automated Driving by Vehicle-
to-infrastructure Communications [C]//IEEE. Proceedings of the
2015IEEE15th International Conference on Control, Automation and
Systems(ICCAS). New York: IEEE, 2015: 2016–2021.
[48] GOZ
Infrastructure Communications in Urban Environments [J]. IEEE Com-
munications Magazine, 2012, 50(5): 176–183.
[49] JIA D, NGODUY D. Enhanced Cooperative Car-following Traffic
Model with the Combination of V2Vand V2I Communication [J].
Transportation Research Part B, 2016, 90: 172–191.
Infrastructure Communication: Making a Case for Intelligent Roadside
Infrastructure [C]//ITS. 19th ITS World Congress. Washington DC: ITS,
2012: 1–11.
[51] DEY K C, RAYAMAJHI A, CHOWDHURY M, et al. Vehicle-
to-vehicle(V2V)and Vehicle-to-infrastructure(V2I)Communication in a
Heterogeneous Wireless Network-performance Evaluation [J]. Trans-
portation Research Part C, 2016, 68: 168–184.
[52] CHENNIKARA-VARGHESE J, CHEN W, HIKI-TA T, et al. Local Peer
Groups and Vehicle-to-infrastructure Communications [C]//IEEE. Pro-
ceedings of the 2007IEEE Globecom Workshops. New York: IEEE, 2007:
1–6.
[53] ZHAO J, CHEN Y, GONG Y. Study of Connectivity Probability of
Vehicle-to-vehicle and Vehicle-to-infrastructure Communication Systems
[C]//IEEE. Proceedings of the 2016IEEE 83rd Vehicular Technology
Conference(VTC Spring). New York: IEEE, 2016: 1–4.
[54] KO B, LIU K, SON S H. Towards Efficient Data Services in Vehicular
Networks Via Cooperative Infrastructure-to-vehicle and Vehicle-
to-vehicle Communications [C]//IEEE. Proceedings of the 2016 Inter-
national IEEE Conferences on Ubiquitous Intelligence and Computing,
Advanced and Trusted Computing, Scalable Computing and Communi-
cations, Cloud and Big Data Computing, Internet of People, and Smart
World Congress. New York: IEEE, 2016: 82–89.
[55] VA V, SHIMIZU T, BANSAL G, et al. Beam Design for Beam
Switching Based Millimeter Wave Vehicle-to-infrastructure Communi-
cations [C]//IEEE. Proceedings of the 2016IEEE International Confer-
ence on Communications(ICC). New York: IEEE, 2016: 1–6.
[56] UBIERGO G A, JIN W L. Mobility and Environment Improvement of
Signalized Networks Through Vehicle-to-infrastructure(V2I) Commu-
nications [J]. Transportation Research Part C, 2016, 68: 70–82.
[57] WANG H, LIU R P, NI W, et al. A New Analytical Model for Highway
Inter-vehicle Communication Systems [C]//IEEE. Proceedings of the
2014 IEEE International Conference on Communications. New York:
IEEE, 2014: 2581–2586.
[58] GUO C, GUO W, CAO G, et al. A Lane-level LBSSystem for Vehicle
Network with High-precision BDS/GPS Positioning [J]. Computational
Intelligence and Neuroscience, 2015, 2015: 1–13.
[59] O’KEEFE K, JIANG Y, PETOVELLO M. An Investigation of Tightly-
coupled UWB/Low-cost GPS for Vehicle-to-infrastructure Relative Posi-
tioning [C]//IEEE. 2014IEEE Radar Conference. New York: IEEE, 2014:
1295–1300.
[60] HAAK U, SASSE A, HECKER P. On the Definition of Lane Accuracy
for Vehicle Positioning Systems [J]. IFAC Proceedings, 2010, 43(16):
372–376.
[61] JAE-HOON J, MIN K, GI-SIG B. Position Recognition for an Au-
tonomous Vehicle Based on Vehicle-toled Infrastructure [C]//Springer.
2016 International Conference on Advanced Engineering Theory and
Applications. Berlin: Springer, 2016: 913–921.
[62] KAWAMURA T, KASHIWA T, TAGUCHI K, et al. Intelligent Driving
Lane with RF-ID for Vehicle Navigation System [C]//IEEE.
2014International Conference on Connected Vehicles and Expo. New
York: IEEE, 2014: 869–870.
[63] PASHAIAN M, MOSAVI M R, PASHAIAN M. Accurate Intelligent
Map Matching Algorithms for Vehicle Positioning System [J]. IJCSI In-
ternational Journal of Computer Science Issues, 2012, 9(2): 114–119.
[64] DAI C, CHOW C Y, ZHANG J. Utilizing Road-side Infrastructure for
Location-based Services in Vehicular Ad-hoc Networks [C]//IEEE. Pro-
ceedings of the 2013 IEEE 8th International Conference on Communica-
tions and Networking in China. New York: IEEE, 2013: 546–551.
[65] KIM W, KIM D, YI K, et al. Development of a Path-tracking Control
System Based on Model Predictive Control Using Infrastructure Sensors
[J]. Vehicle System Dynamics, 2012, 50(6): 1001–1023.
[66] PASHAIAN M, MOSAVI M R, PASHAIAN M. Accurate Intelligent
Map Matching Algorithms for Vehicle Positioning System [J]. Interna-
tional Journal of Computer Science Issues, 2012, 9(2): 114–119.
[67] ZHANG F, HINZ G, CLARKE D, et al. Vehicle-infrastructure Locali-
zation Based on the SME Filter [C]//IEEE. Proceedings of the 2015 IEEE
18th International Conference on Intelligent Transportation Systems. New
York: IEEE, 2015: 225–230.
[68] DE PONTE M F, DIAZ E M, RASHDAN I. Cooperative Infrastruc-
ture-based Vehicle Positioning [C]//IEEE. Proceedings of the 2016 IEEE
84th Vehicular Technology Conference. New York: IEEE, 2016: 1–6.
[69] VU A, FARRELL J A, BARTH M. Centimeter-accuracy Smoothed
Vehicle Trajectory Estimation [J]. IEEE Intelligent Transportation Sys-
tems Magazine, 2013, 5(4): 121–135.
[70] LIU J, CAI B, WEN Y, et al. Integrating DSRC and Dead-reckoning for
Cooperative Vehicle Positioning Under GNSS-challenged Vehicular En-
vironments [J]. International Journal of Ad Hoc and Ubiquitous Compu-
ting, 2015, 19(1/2): 111–129.
[71] DI BERNARDO M, SALVI A, SANTINI S. Distributed Consensus
Strategy for Platooning of Vehicles in the Presence of Time-varying
Heterogeneous Communication Delays [J]. IEEE Transactions on Intel-
ligent Transportation Systems, 2015, 16(1): 102–112.
[72] LEI C, VAN EENENNAAM E M, WOLTERINKW K, et al. Impact of
Packet Loss on CACC String Stability Performance [C]//IEEE. Pro-
ceedings of the 2011IEEE 11th International Conference on ITST
elecommunications. New York: IEEE, 2011: 381–386.
[73] KENNEY J B. Dedicated Short-range Communications (DSRC)
Standards in the United States [J]. Proceedings of the IEEE, 2011, 99(7):
1162–1182.
[74] CAMPOLO C, VINEL A, MOLINARO A, et al. Modeling Broadcast-
ing in IEEE 802. 11p/WAVE Vehicular Networks [J]. IEEE Communica-
tions Letters, 2011, 15(2): 199–201.
[75] EMMELMANN M, BOCHOW B, KELLUM C C. Vehicular Net-
working: Automotive Applications and Beyond[M]. Hoboken: John
Wiley & Sons, 2010.
[76] SAEEDNIA M, MENENDEZ M. Analysis of Strategies for Truck
Platooning: Hybrid Strategy [J]. Transportation Research Record,
2016(2547): 41–48.
[77] HALL R, CHIN C. Vehicle Sorting for Platoon Formation: Impacts on
Highway Entry and Throughput [J]. Transportation Research Part C, 2005,
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 20
13(5): 405–420.
[78] AMOOZADEH M, DENG H, CHUAH C N, et al. Platoon Management
with Cooperative Adaptive Cruise Control Enabled by VANET [J]. Ve-
hicular Communications, 2015, 2(2): 110–123.
[79] ZHENG Y, LI S E, LI K, et al. Distributed Model Predictive Control for
Heterogeneous Vehicle Platoons Under Unidirectional Topologies [J].
IEEE Transactions on Control Systems Technology, 2017, 25(3):
899–910.
[80] ZHENG Y, LI S E, WANG J, et al. Stability and Scalability of Homo-
geneous Vehicular Platoon: Study on the Influence of Information Flow
Topologies [J]. IEEE Transactions on Intelligent Transportation Systems,
2016, 17(1): 14–26.
[81] SABA(U), OARA(C, WARNICK S, et al. Optimal Distributed Control
for Platooning Via Sparse Coprime Factorizations [J]. IEEE Transactions
on Automatic Control, 2017, 62(1): 305–320.
[82] HARFOUCH Y A, YUAN S, BALDI S. An Adaptive Switched Control
Approach to Heterogeneous Platooning with Intervehicle Communication
Losses [J]. IEEE Transactions on Control of Network Systems, 2018, 5(3):
1434–1444.
[83] LI S, MI C C. Wireless Power Transfer for Electric Vehicle Applications
[J]. IEEE Journal of Emerging and Selected Topics in Power Electronics,
2015, 3(1): 4–17.
[84] DEFLORIO F P, CASTELLO L, PINNA I, et al. “Charge while Driving”
for Electric Vehicles: Road Traffic Modeling and Energy Assessment [J].
Journal of Modern Power Systems and Clean Energy, 2015, 3(2):
277–288.
[85] XIAO J, CHENG E, CHEUNG N, et al. Study of Wireless Charging
Lane for Electric Vehicles [C]//IEEE. Proceedings of the
2016International Symposium on Electrical Engineering(ISEE). New
York: IEEE, 2016: 1–4.
[86] CHEN Z, JING W, HUANG X, et al. A Promoted Design for Primary
Coil in Roadway-powered System [J]. IEEE Transactions on Magnetics,
2015, 51(11): 1–4.
[87] WANG C, LI J, YANG Y, et al. Combining Solar Energy Harvesting
with Wireless Charging for Hybrid Wireless Sensor Networks [J]. IEEE
Transactions on Mobile Computing, 2018, 17(3): 560–576.
[88] HE J, HUANG H J, YANG H, et al. An Electric Vehicle Driving Be-
havior Model in the Traffic System with a Wireless Charging Lane [J].
Physica A: Statistical Mechanics and Its Applications, 2017, 481:
119–126.
[89] LI Y, WANG W, XING L, et al. Longitudinal Safety Evaluation of
Electric Vehicles with the Partial Wireless Charging Lane on Freeways [J].
Accident Analysis & Prevention, 2018, 111: 133–141.
[90] JIANG W, XU S, LI N, et al. Wireless Power Charger for Light Electric
Vehicles [C]//IEEE. Proceedings of the 2015IEEE 11th International
Conference on Power Electronics and Drive Systems. New York: IEEE,
2015: 562–566.
[91] SUN J, DING W X. Research Status and Prospects of the Smart Con-
crete [C]//Trans Tech. Advanced Materials Research. Switzerland: Trans
Tech Publications, 2011, 194: 1118–1121.
[92] ZHANG J, LU Y, LU Z, et al. A New Smart Traffic Monitoring Method
Using Embedded Cementbased Piezoelectric Sensors [J]. Smart Materials
and Structures, 2015, 24(2): 1–8.
[93] FERZLI N A, IVEY R A, KING T, et al. An Application of “Smart
Dust”for Pavement Condition Monitoring [C]//International Society for
Optics and Photonics. Smart Structures and Materials 2006: Sensors and
Smart Structures Technologies for Civil, Mechanical, and Aerospace
Systems. New York: International Society for Optics and Photonics, 2006,
6174: 1–12.
[94] MORENO-NAVARRO F, IGLESIAS G R, RUBIO-G
Development of Mechanomutable Asphalt Binders for the Construction
of Smart Pavements [J]. Materials and Design, 2015, 84: 100–109.
[95] JUNG I, SHIN Y H, KIM S, et al. Flexible Piezoelectric Polymer-based
Energy Harvesting System for Roadway Applications [J]. Applied Energy,
2017, 197: 222–229.
[96] ALAVI A H, HASNI H, LAJNEF N, et al. Continuous Health Moni-
toring of Pavement Systems Using Smart Sensing Technology [J]. Con-
struction and Building Materials, 2016, 114: 719–736.
[97] NASIR D S N M, HUGHES B R, CALAUTIT J K. A CFD Analysis of
Several Design Parameters of a Road Pavement Solar Collector(RPSC)for
Urban Application [J]. Applied Energy, 2017, 186: 436–449.
[98] EL-WAKEEL A S, LI J, RAHMAN M T, et al. Monitoring Road Sur-
face Anomalies Towards Dynamic Road Mapping for Future Smart
Cities [C]//IEEE. Proceedings of the 2017IEEE Global Conference on
Signal and Information Processing. New York: IEEE, 2017: 828–832.
[99] WANG Y D, CHENG Y C, ZHANG P, et al. Research on Exploitation
Technology of Snow Disaster Forecast Warning System on Roads [J].
Applied Mechanics and Materials, 2014, 533: 350–353.
[100] TAN B. Road Surface Temperature Monitor System Realization Based
on Rich Internet Application Model and GPRS Technology [C]//IEEE.
Proceedings of the 2010IEEE 2nd International Workshop on Intelligent
Systems and Applications. New York: IEEE, 2010: 1–4.
[101] BEINAROVICA A, GOROBETZ M, LEVCHENK-OV A. Innovative
Neuro-fuzzy System of Smart Transport Infrastructure for Road Traffic
Safety [C]//Institute of Physics. IOP Conference Series: Materials Sci-
ence and Engineering. London: IOP Publishing, 2017, 236(1): 1–8.
[102] ZHOU Z, HU S, ZHANG X, et al. Characteristics and Application of
Road Absorbing Solar Energy [J]. Frontiers in Energy, 2013, 7(4):
525–534.
[103] ITO K, ARAI Y, HIRAKAWA G, et al. A Road Condition Sharing
System Using Vehicle-to-vehicle Communication in Various Communi-
cation Environment [C]//IEEE. Proceedings of the 2015IEEE9th Inter-
national Conference on Complex, Intelligent, and Software Intensive
Systems. New York: IEEE, 2015: 242–249.
[104] MOGHIMI B, SAFIKHANI A, KAMGA C, et al. Short-term Predic-
tion of Signal Cycle on an Arterial with Actuated-uncoordinated Control
Using Sparse Time Series Models [J]. IEEE Transactions on Intelligent
Transportation Systems, 2019, 20(8): 2976–2985.
[105] MOGHIMI B, SAFIKHANI A, KAMGA C, et al. Cycle-length Pre-
diction in Actuated Traffic-signal Control Using Arima Model [J]. Journal
of Computing in Civil Engineering, 2017, 32(2): 1–10.
[106] GUO Y, MA J, XIONG C, et al. Joint Optimization of Vehicle Tra-
jectories and Intersection Controllers with Connected Automated Vehicles:
Combined Dynamic Programming and Shooting Heuristic Approach [J].
Transportation Research Part C, 2019, 98: 54–72.
[107] JITTRAPIROM P, CAIATI V, FENERI A M, et al. Mobility as a
Service: A Critical Review of Definitions, Assessments of Schemes, and
Key Challenges [J]. Urban Planning, 2017, 2(2): 13–25.
[108] KAMARGIANNI M, MATYAS M, LI W, et al. ACritical Review of
New Mobility Services for Urban Transport [J]. Transportation Research
Procedia, 2016, 14: 3294–3303.
[109] KAMARGIANNI M, MATYAS M, LI W, et al. Feasibility Study
for“Mobility as a Service”Concept in London[R]. London: UCL Energy
Institute, 2015.
[110] DJAVADIAN S, CHOW J Y J. An Agent-based Day-to-day Adjust-
ment Process for Modeling‘Mobility as a Service’with a Two-sided
Flexible Transport Market [J]. Transportation Research Part B, 2017, 104:
36–57.
[111] LAMOTTE R, DE PALMA A, GEROLIMINIS N. On the Use of
Reservation-based Autonomous Vehicles for Demand Management [J].
Transportation Research Part B, 2017, 99: 205–227.
[112] MA J, LI X, ZHOU F, et al. Designing Optimal Autonomous Vehicle
Sharing and Reservation Systems: A Linear Programming Approach [J].
Transportation Research Part C, 2017, 84: 124–141.
[113] HENSHER D A. Future Bus Transport Contracts Under a Mobility as a
Service(MaaS)Regime in the Digital Age: Are They Likely to Change [J].
Transportation Research Part A, 2017, 98: 86–96.
[114] RAVINDRAN R, CHAKRABORTI A, AMIN S O, et al. 5G-ICN:
Delivering ICN Services over 5G Using Network Slicing [J]. IEEE
Communications Magazine, 2017, 55(5): 101–107.
[115] THAI J, YUAN C, BAYEN A M. Resiliency of Mobility-as-a-Service
Systems to Denial-of-service Attacks [J]. IEEE Transactions on Control
of Network Systems, 2018, 5(1): 370–382.
[116] OKUMURA B, JAMES M R, KANZAWA Y, et al. Challenges in
Perception and Decision Making for Intelligent Automotive Vehicles: A
Case Study [J]. IEEE Transactions on Intelligent Vehicles, 2016, 1(1):
20–32.
[117] LIU W, KIM S W, PENDLETON S, et al. Situation-aware Decision
Making for Autonomous Driving on Urban Road Using Online POMDP
[C]//IEEE. Proceedings of the 2015IEEE Intelligent Vehicles Symposium.
New York: IEEE, 2015: 1126–1133.
[118] NARANJO J E, GONZALEZ C, GARCIA R, et al. Lane-change
Fuzzy Control in Autonomous Vehicles for the Overtaking Maneuver [J].
IEEE Transactions on Intelligent Transportation Systems, 2008, 9(3):
438–450.
[119] PEREZ J, MILANES V, ONIEVA E, et al. Longitudinal Fuzzy Control
© 2019 China Academic Journals (CD Edition) Electronic Publishing House Co., Ltd. 21
for Autonomous Overtaking [C]//IEEE. Proceedings of the 2011 IEEE
International Conference on Mechatronics. New York: IEEE, 2011:
188–193.
[120] ARDELT M, COESTER C, KAEMPCHEN N. Highly Automated
Driving on Freeways in Real Traffic Using a Probabilistic Framework [J].
IEEETransactions on Intelligent Transportation Systems, 2012, 13(4):
1576–1585.
[121] WOLF P, HUBSCHNEIDER C, WEBER M, et al. Learning How to
Drive in a Real World Simulation with Deep Q-networks [C]//IEEE.
Proceedings of the 2017IEEE Intelligent Vehicles Symposium. New York:
IEEE, 2017: 244–250.
[122] LILLICRAP T P, HUNT J J, PRITZEL A, et al. Continuous Control
with Deep Reinforcement Learning[EB/OL]. [2015–07–09]. https: //arxiv.
org/pdf/1509. 02971. pdf.
[123] NGAI D C K, YUNG N H C. A Multiple-goal Reinforcement Learning
Method for Complex Vehicle Overtaking Maneuvers [J]. IEEE Transac-
tions on Intelligent Transportation Systems, 2011, 12(2): 509–522.
[124] SCHWARTING W, ALONSO-MORA J, RUS D. Planning and
Decision-making for Autonomous Vehicles [J]. Annual Review of Con-
trol, Robotics, and Autonomous Systems, 2018, 1: 187–210.
[125] ZHENG Z. Recent Developments and Research Needs in Modeling
Lane Changing [J]. Transportation Research Part B, 2014, 60: 16–32.
[126] LI B, SHAO Z, ZHANG Y, et al. Nonlinear Programming for Multi-
vehicle Motion Planning with Homotopy Initialization Strategies
[C]//IEEE. Proceedings of the 2017 13th IEEE Conference on Automa-
tion Science and Engineering. New York: IEEE, 2017: 118–123.
[127] LI Bai. Research on Computational Optimal Control Methods for
Autonomous Vehicle Motion Planning Problems with Complicated Con-
straints [D]. Hangzhou: Zhejiang University, 2018 (in Chinese).
[128] DESARAJU V R, HOW J P. Decentralized Path Planning for Multi-
agent Teams with Complex Constraints [J]. Autonomous Robots, 2012,
32(4): 385–403. -optimal
Motion Planning for Multiple Robotic Vehicles with Collision Avoidance
[J]. IEEE Transactions on Control Systems Technology, 2016, 24(3):
867–883.
[130] LI B, ZHANG Y, SHAO Z, et al. Simultaneous Versus Joint Compu-
ting: A Case Study of Multi-vehicle Parking Motion Planning [J]. Journal
of Computational Science, 2017, 20: 30–40.
[131] INDRANIL S, RATTANACHAI R, VIJAY K, et al. Implan: Scalable
Incremental Motion Planning for Multi-robot Systems [C]//IEEE. Pro-
ceedings of the 2016ACM/IEEE 7th International Conference on
Cyber-physical Systems. New York: IEEE, 2016: 1–10.
[132] PLESSEN M G, BERNARDINI D, ESEN H, et al. Multi-automated
Vehicle Coordination Using Decoupled Prioritized Path Planning for
Multi-lane One-and Bi-directional Traffic Flow Control [C]//IEEE. Pro-
ceedings of the 2016IEEE 55th Conference on Decision and Control. New
York: IEEE, 2016: 1582–1588.
[133] DRESNER K, STONE P. A Multiagent Approach to Autonomous
Intersection Management [J]. Journal of Artificial Intelligence Research,
2008, 31: 591–656.
[134] LEE J, PARK B. Development and Evaluation of a Cooperative Ve-
hicle Intersection Control Algorithm Under the Connected Vehicles En-
vironment [J]. IEEE Transactions on Intelligent Transportation Systems,
2012, 13(1): 81–90.
[135] LIAN F L, MURRAY R. Real-time Trajectory Generation for the
Cooperative Path Planning of Multi-vehicle Systems [C]//IEEE. Pro-
ceedings of the 200241st IEEE Conference on Decision and Control. New
York: IEEE, 2002, 4: 3766–3769.
[136] AHMADZADEH A, MOTEE N, JADBABAIE A, et al. Multi-vehicle
Path Planning in Dynamically Changing Environments [C]//IEEE.
2009IEEE International Conference on Robotics and Automation. New
York: IEEE, 2009: 2449–2454.
[137] DESIRAJU D, CHANTEM T, HEASLIP K. Minimizing the Disrup-
tion of Traffic Flow of Automated Vehicles During Lane Changes [J].
IEEE Transactions on Intelligent Transportation Systems, 2015, 16(3):
1249–1258.
[138] ATAGOZIYEV M, SCHMIDT K W, SCHMIDT EG. Lane Change
Scheduling for Autonomous Vehicles [J]. IFAC-Papers on Line, 2016,
49(3): 61–66.
[139] ABICHANDANI P, FORD G, BENSON H Y, et al. Mathematical
Programming for Multi-vehicle Motion Planning Problems [C]//IEEE.
Proceedings of the2012IEEE International Conference on Robotics and
Automation. New York: IEEE, 2012: 3315–3322.
[140] WARD E, EVESTEDT N, AXEHILL D, et al. Probabilistic Model for
Interaction Aware Planning in Merge Scenarios [J]. IEEE Transactions on
Intelligent Vehicles, 2017, 2(2): 133–146.
[141] WANG Fei-yue. Parallel System Method for Management and Control
of Complex Systems [J]. Control and Decision, 2004, 19(5): 485–489 (in
Chinese).
[142] WANG Fei-yue. Computational Experiments for Behavior Analysis
and Decision Evaluation of Complex Systems [J]. Journal of System
Simulation, 2004, 16(5): 893–897 (in Chinese).
