Figure - available from: Neural Computing and Applications
This content is subject to copyright. Terms and conditions apply.
Source publication
In an intelligent transportation system, accurate bus information is vital for passengers to schedule their departure time and make reasonable route choice. In this paper, an improved deep belief network (DBN) is proposed to predict the bus travel time. By using Gaussian–Bernoulli restricted Boltzmann machines to construct a DBN, we update the clas...
Similar publications
Inversion of hypocenters is the first and most fundamental step in the study of seismic activities. It requires solving the nonlinear relation between the travel time and hypocenter locations, which is heavily dependent on the knowledge of the medium properties, most importantly the velocity structure. In this study, we prove that machine learning...
This study applies response surface methodology (RSM) to the hyperparameter fine-tuning of three machine learning (ML) algorithms: artificial neural network (ANN), support vector machine (SVM), and deep belief network (DBN). The purpose is to demonstrate RSM effectiveness in maintaining ML algorithm performance while reducing the number of runs req...
Citations
... The majority of research in this area has centered on specific bus lines and one-step regression models (i.e., next-stop delay prediction) utilizing realtime bus data (Tiong et al. 2023). Commonly, the models are utilized for proactive realtime management of bus routes based on external information (e.g., temporal, weather) (Achar et al. 2019;Zhang et al. 2021;Zhou et al. 2022;Ma et al. 2019;Rodriguez-Deniz and Villani 2022;Huang et al. 2021;Chen et al. 2020). Limited research has been conducted on the reliability of bus services from a network-level perspective. ...
To effectively manage and control public transport operations, understanding the various factors that impact bus arrival delays is crucial. However, limited research has focused on a comprehensive analysis of bus delay factors, often relying on single-step delay prediction models that are unable to account for the heterogeneous impacts of spatiotemporal factors along the bus route. To analyze the heterogeneous impact of bus arrival delay factors, the paper proposes a set of regression equations conditional on the bus location. A seemingly unrelated regression equation (SURE) model is developed to estimate the regression coefficients, accounting for potential correlations between regression residuals caused by shared unobserved factors among equations. The model is validated using bus operations data from Stockholm, Sweden. The results highlight the importance of developing stop-specific bus arrival delay models to understand the heterogeneous impact of explanatory variables. The significant factors impacting bus arrival delays are primarily associated with bus operations, such as delays at consecutive upstream stops, dwell time, scheduled travel time, recurrent congestion, and current traffic conditions. Factors like the calendar and weather have significant but marginal impacts on arrival delays. The study suggests that different bus operating management strategies, such as schedule adjustments, route optimization, and real-time monitoring and control, should be tailored to the characteristics of stop sections since the impacts of these factors vary depending on the stop location.
... This enables the technique to maintain iterative representation of attributes, which makes DBN a suitable tool for providing network security, reliable and high accuracy prediction of intelligent vehicles drivers' emotions and travel time in 5G/6G-enabled IoV environments. Additionally, study has shown that up to 51% of blockchain attacks in emerging vehicular networks can be prevented with the aid of DBN techniques with regards to the reduction of the delivery time of intelligent vehicular packet transmissions and blockchain puzzle computation time, as well as in the identification of trusted and malicious intelligent mobile stations [48]. ...
Improving transportation efficiency and on-road safety using Intelligent Transportation Systems (ITSs) has become crucial as road congestion and vehicle complexity increase coupled with ongoing rapid development and deployment of electric vehicles across the globe. Recent advances in computer systems and wireless communications have ushered in more possibilities for smart solutions to road traffic safety, congestion reduction, convenience, and overall efficiency. The evolution and deployment of 5G have opened up new technologies and features that can provide the much needed high-mobility wireless networks for the emerging Internet of Vehicles (IoV). The application of AI consisting of Deep Learning (DL), Machine Learning (ML) and Swarm Intelligence (SI) techniques have emerged in both conventional and vehicular wireless networks with strong promises towards enhancing traditional data-centric methods. Particularly, in the application domains of IoV, big data is frequently generated from various sources within the vehicular communication environment. The collected big data is usually processed and used for both safety and infotainment services including routing, broadening drivers' awareness, traffic mobility prediction for hazardous situation avoidance to improve overall safety and passenger comfort, and general quality of road experience. Applying data-driven methods enables AI to address high mobility and dynamic vehicular communications and network issues facing traditional solutions and approaches like network optimization techniques and conventional control loop design. This study provides a concise review of DL, ML and SI techniques and applications that are currently being explored by different research efforts within the application area of vehicular networks. The paper further discusses the strengths and weaknesses of the proposed AI-based solutions for the IoV networks.
... Researchers have used deep learning methods to build subway station passenger flow prediction models, including back propagation (BP) networks [37], convolutional neural networks (CNNs) [38], and LSTM networks [39]. Chen et al. used the travel time of passenger flow as the research object and employed a BP neural network to forecast the travel time of passengers [40]. Zhang et al. [41] used an LSTM network for prediction and then employed a CNN to build a prediction model for the subway [42]. ...
Rational use of urban underground space (UUS) and public transportation transfer underground can solve urban traffic problems. Accurate short-term prediction of passenger flow can ensure the efficient, safe, and comfortable operation of subway stations. However, complex and nonlinear interdependencies between time steps and time series complicate such predictions. This study considered temporal patterns across multiple time steps and selected relevant information on short-term passenger flow for prediction. A hybrid model based on the temporal pattern attention (TPA) mechanism and the long short-term memory (LSTM) network was developed (i.e., TPA-LSTM) for predicting the future number of passengers in subway stations. The TPA mechanism focuses on the hidden layer output values of different time steps in history and of the current time as well as correlates these output values to improve the accuracy of the model. The card swiping data from the Hangzhou Metro automatic fare collection system in China were used for verification and analysis. This model was compared with a convolutional neural network (CNN), LSTM, and CNN-LSTM. The results showed that the TPA-LSTM outperformed the other models with good applicability and accuracy. This study provides a theoretical basis for the pre-allocation of subway resources to avoid subway station crowding and stampede accidents.
... The InterQuartile Range (IQR) was introduced in [63] to estimate the travel time variation. The performance achieved by this approach effectively maximizes the efficiency of the public transport system. ...
Cities play a vital role in promoting commercial growth and wealth. The development of cities is mainly based on
their social, physical, and institutional infrastructure. In this situation, the significance of urban transportation is dominant.
Urban area transportation is a common public bus service and is mostly used to transport many people in urban areas. In
this paper, a survey on important parameters for improving the UBTS is reviewed. At first, the article reviewed the flow of
passenger prediction based on the UBTS. Here, the UBTS issues like the prediction of passenger flow, fleet size, passenger
comfort perception, delay, driver's behaviour, sound level, vehicle breakdowns, and so on are reviewed. To overcome these
issues, the authors have presented some techniques and solutions for urban transportation, which are reviewed. A systematic
literature review is conducted for the urban transportation system from 2011 until 2021. This survey provides a technical
direction for researchers' work, and the potential future aspects have been discussed.
... Using historical or real-time data, multiple prediction methods and mechanisms have been developed to predict bus travel time. These methods can be categorized into statistical methods [8][9][10], machine learning methods [11][12][13][14], and neural network methods [1,5,7,15,16]. Statistical methods can be separated into historical average methods, regression methods and time-series methods [7]. ...
... We also pass the input vector v to another GRN block followed by ELU activation function to get the output # using Eq. (15). After computing K, we use the scaled-dot-product function (MatMul), which is the most used function in attention mechanism, with an ELU to calculate the weight x in Eq. (16). ...
Accurate travel time prediction allows passengers to schedule their journeys efficiently. However, cyclical factors (time intervals of the day, weather conditions, and holidays), unpredictable factors (incidents, abnormal weather), and other complicated factors (dynamic traffic conditions, dwell times, and variation in travel demand) make accurate bus travel time prediction complicated. This paper aims to achieve accurate travel time prediction. To do so, we propose a clustering method that identifies travel time paradigms of different route links and clusters them based on their similarity using the nonnegative matrix factorization algorithm. Additionally, we propose a deep learning model based on CNN with spatial–temporal attention and gating mechanisms to select the most relevant features and capture their dependencies and correlations. For each defined cluster, we train a separate model to predict the travel time at various time intervals over the day. As a result, the travel times of all journey links from related prediction models are aggregated to predict the total journey time. Extensive experiments using data collected from four different bus lines in Beijing show that our method outperforms the compared baselines.
... The ITS passengers require accurate information regarding bus departure/arrival time to reasonably schedule the passenger's departure time and route. By considering this, in Chen et al., 32 an improved deep belief network (DBN) was introduced for predicting the bus travel time. In this method, the Gaussian-Bernoulli-based RBM was introduced in DBN for modeling the continuous data. ...
Predicting the passenger count and rerouting the busses for the demanding routes to reduce the waiting time of passengers is considered the major challenge in the modern public transportation system. Meanwhile, the transportation sector has been revolutionized in recent years due to big data. In this proposed method, the deep learning and fuzzy logic concept is introduced to face such challenges. In this proposed method, the deep attributes extraction and prediction of passenger demand are accomplished using a deep sailfish network (DSFN). Then, the rerouting concept for demanding routes is carried out using the fuzzy logic‐based rerouting process, which reduces the waiting time of passengers. This prediction‐based rerouting framework enhances the efficiency of bus scheduling by satisfying the passenger demand is performed. It increases the quality of transportation with greater comfort for passengers. Meanwhile, the usage of public transportation is also getting increases. Four different routes from Gujarat (India) are considered in this work analysis, they are Railway station terminal (RAST), Sahara Darwaja (SAHD), Katargam Darwaja BRT (KADB), Udhana Darwaja BRTS, and Adajan GSRTC (ADGS). The proposed approach is implemented with Matlab, and the performance is evaluated using standard metrics such as mean absolute percentage error (MAPE) and root mean square error (RMSE). The experimental results prove that the suggested method provides prediction results with high accurate values than other existing methods in terms of MAPE and RMSE. It intends to reduce the waiting time and to improve the comfort perception of each passengers. Initially, cluster the input data that may increase the efficiency of passenger flow prediction. Based on the predicted passenger count, the busses are reroute to the demanding routes using fuzzy logic.
... Since the model parameters are determined by the collected data and not a predetermined distribution, they require a larger amount of data than many other approaches [34]. Non-parametric methods include Support Vector Regression (SVR) [35][36][37][38], the Nearest Neighborhood model [39,40], and deep-learning based models [18,27,34,[41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56][57][58][59]. In the following subsection, different deep learning-based models are reviewed. ...
... Table 1 provides a comprehensive review of the studies predicting the travel time using deep learning models. Deep learning models approach problems through the learning of a very large amount of data, so the GPS trajectories of vehicles with a relatively huge number of data were used to predict travel time [18,27,34,42,43,[45][46][47][48][49][50][51][52][53][54][55][56][57][58][59]. Additionally, since trajectory data contains vehicle location information in chronological order, specialized long short-term memory (LSTM) models for sequence-based classification or for solving regression problems have been widely used [18,27,34,[41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56][57]. ...
... Additionally, since trajectory data contains vehicle location information in chronological order, specialized long short-term memory (LSTM) models for sequence-based classification or for solving regression problems have been widely used [18,27,34,[41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56][57]. Link TT Taxi trajectories LSTM - [43] Route TT Taxi trajectories LSTM - [27] TT between bus stops AVL LSTM - [34] TT between bus stops AVL LSTM Weather, intersections [44] Link TT Vehicle passage records CNN + LSTM - [45] Route TT Taxi trajectories MLP + LSTM Weather, driver rider vehicle Profile, traffic restriction [46] Route TT Taxi trajectories CNN + LSTM Weather, driver profile [47,48] Route TT Taxi trajectories CNN + LSTM - [51] Route TT Vehicle passage records Graph attention + CNN - [52] Route TT Taxi [56] Route TT Taxi trajectories GCN + LSTM - [18] TT between bus stops AVL CNN + LSTM - [57] TT between bus stops AVL LSTM + ANN - [58] Link TT Taxi trajectories Denoising auto-encoder Road segment, contextual features [59] TT between bus stops AVL DBN -1 TT means travel time. ...
With the abundance of public transportation in highly urbanized areas, it is common for passengers to make inefficient or flawed transport decisions due to a lack of information. The exact arrival time of a bus is an example of such information that can aid passengers in making better decisions. The purpose of this study is to provide a method for predicting path-based bus travel time, thereby assisting accurate bus arrival and departure time predictions at each bus stop. Specifically, we develop a Geo-conv Long Short-term Memory (LSTM) model that (1) extracts subsequent spatial features through a 1D Convolution Neural Network (CNN) for the entire bus travel sequence and (2) captures the temporal dependencies between subsequences through the LSTM network. Additionally, this study utilizes additional variables that affect two components of bus travel time (dwelling time and transit time) to precisely predict travel time. The constructed model is then evaluated by the practical application to two bus lines operating in Seoul, Korea. The results show that our model outperforms three other baseline models. Two bus lines with different types of operation show different model performance patterns that are dependent on travel distance. Interestingly, we find that the variable related to the link of the stop location appears to play an important role in predicting bus travel time. We believe that these novel findings will contribute to the literature on transportation and, in particular, on deep learning-based travel time prediction.
... AutoRegressive Integrated Moving Average (ARIMA) [14] [15] [11] [16], authors have attempted to predict the bus arrival time at Houston, Texas using Historic averages like a model and Artificial Neural Network (ANN) model, and compared the accuracy of models to conclude that the neural network model performed better than the other models. Authors in [17] have proposed a Deep Belief Network (DBN) model to forecasts the arrival time of buses and compared its performance against the basic models such as k-NN, ANN, SVM, and random forests, and the DBN performance showed superior results. ...
Congested roads are a global problem, and increased usage of private vehicles is one of the main reasons for congestion. Public transit modes of travel are a sustainable and eco-friendly alternative for private vehicle usage, but attracting commuters towards public transit mode is a mammoth task. Commuters expect the public transit service to be reliable, and to provide a reliable service it is necessary to fine-tune the transit operations and provide well-timed necessary information to commuters. In this context, the public transit travel time is predicted in Tumakuru, a tier-2 city of Karnataka, India. As this is one of the initial studies in the city, the performance comparison of eight Machines Learning models including four linear namely, Linear Regression, Ridge Regression, Least Absolute Shrinkage and Selection Operator Regression, and Support Vector Regression; and four non-linear models namely, k-Nearest Neighbors, Regression Trees, Random Forest Regression, and Gradient Boosting Regression Trees is conducted to identify a suitable model for travel time predictions. The data logs of one month (November 2020) of the Tumakuru city service, provided by Tumakuru Smart City Limited are used for the study. The time-of-the-day (trip start time), day-of-the-week, and direction of travel are used for the prediction. Travel time for both upstream and downstream are predicted, and the results are evaluated based on the performance metrics. The results suggest that the performance of non-linear models is superior to linear models for predicting travel times, and Random Forest Regression was found to be a better model as compared to other models.
... AutoRegressive Integrated Moving Average (ARIMA) [14] [15] [11] [16], authors have attempted to predict the bus arrival time at Houston, Texas using Historic averages like a model and Artificial Neural Network (ANN) model, and compared the accuracy of models to conclude that the neural network model performed better than the other models. Authors in [17] have proposed a Deep Belief Network (DBN) model to forecasts the arrival time of buses and compared its performance against the basic models such as k-NN, ANN, SVM, and random forests, and the DBN performance showed superior results. ...
... Tan et al. [67] introduced two DBNs, one having Gaussian visible units and hidden binary units and the remaining units being binary, with results showing an improvement in the accuracy, but less robust nonetheless. Chen et al. [68] combined a DBN with Gaussian-Bernoulli restricted Boltzmann machines and a BPNN to improve the accuracy further, but robustness was still a concern. To enhance the prediction accuracy and robustness, Koesdwiady et al. [69] correlated weather parameters and traffic flow by employing a decision-level data fusion scheme. ...
With the rapid growth and development of cities, Intelligent Traffic Management and Control (ITMC) is becoming a fundamental component to address the challenges of modern urban traffic management, where a wide range of daily problems need to be addressed in a prompt and expedited manner. Issues such as unpredictable traffic dynamics, resource constraints, and abnormal events pose difficulties to city managers. ITMC aims to increase the efficiency of traffic management by minimizing the odds of traffic problems, by providing real-time traffic state forecasts to better schedule the intersection signal controls. Reliable implementations of ITMC improve the safety of inhabitants and the quality of life, leading to economic growth. In recent years, researchers have proposed different solutions to address specific problems concerning traffic management, ranging from image-processing and deep-learning techniques to forecasting the traffic state and deriving policies to control intersection signals. This review article studies the primary public datasets helpful in developing models to address the identified problems, complemented with a deep analysis of the works related to traffic state forecast and intersection-signal-control models. Our analysis found that deep-learning-based approaches for short-term traffic state forecast and multi-intersection signal control showed reasonable results, but lacked robustness for unusual scenarios, particularly during oversaturated situations, which can be resolved by explicitly addressing these cases, potentially leading to significant improvements of the systems overall. However, there is arguably a long path until these models can be used safely and effectively in real-world scenarios.
