No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Abstract
Networked non-linear control systems (NNCS) are studied in this paper. A control scheme based on predictive control is proposed to address the problems introduced by the network. Based on the theory of the linearization of the non-linear process, a predictive control algorithm is constructed to generate a sequence of control predictions, some of which can suitably be selected to control the non-linear process and further compensate for the time delay. Methods are proposed to deal with the selection of control predictions, the control prediction sequence updating, and data disorders processing. The proposed control scheme is able to avoid the synchronization issue by modelling the predictive controller as an event-driven element. The stability of this NNCS has been studied by discussing the stability of an augmented system, the stability conditions of which can be transformed into a series of inequalities. In addition, the control method is validated by both simulations and experiments.
... Various approaches have been proposed to deal with these problems. In Ouyang et al. (2007) a predictive control algorithm is constructed to generate a sequence of control predictions, some of which can suitably be selected to compensate the induced time-delay. In Ding, Huang (2007), an infinite horizon min-max model predictive control with a polytopic uncertainty and time delay is considered. ...
Paper studies the problem on robust networked static output feedback model predictive controller design that stabilizes uncertain system with guaranteed cost. The upper bound of the time-delay is assumed to be bigger than sampling time. Control design is based on sufficient robust stability condition formulated as a solution of bilinear matrix inequality BMI. The example illustrates the viability of the proposed output feedback design method.
... However, only few results on nonlinear NCSs have been reported to date under such a predictive control based framework [11]. ...
A predictive control based approach is proposed to deal with a Wiener type system which is closed through a network. In this approach, an output feedback predictive controller is designed using delayed sensing data with a specially designed state observer. The network constraints, i.e., the network-induced delay and data packet dropout, are compensated in both the forward and backward channels by taking advantage of the char-acteristics of both the predictive controller and the network transmission. Stability of the closed-loop system is derived by using the separation principle and switched system theory. Simulations illustrate the validity of the proposed approach.
... Remark 3 The networked predictive scheme was also proposed by Ouyang. [30] However, the state observer used a nonlinear change of coordinates in Ouyang's method. In this paper, a nonlinear change of coordinates is not required in state estimation. ...
This paper discusses the model-based predictive controller design of networked nonlinear systems with communication delay and data loss. Based on the analysis of the closed-loop networked predictive control systems, the model-based networked predictive control strategy can compensate for communication delay and data loss in an active way. The designed model-based predictive controller can also guarantee the stability of the closed-loop networked system. The simulation results demonstrate the feasibility and efficacy of the proposed model-based predictive controller design scheme.
This paper studies the problem of robust networked static output feedback model predictive controller design that stabilizes uncertain system with guaranteed cost and Parameter-Dependent Quadratic Stability (PDQS). The upper bound on the time-delay is assumed to be bigger than sampling time. Control design is based on sufficient robust stability condition formulated as a solution of bilinear matrix inequality BMI. The example illustrates the viability of the proposed output feedback design method.
This study is concerned with the control problem of networked non-linear systems with communication delay and data dropout. A networked non-linear predictive control scheme is proposed to compensate for communication delay and data dropout actively rather than passively. The stability analysis shows that the stability problem of the closed-loop networked non-linear predictive control system is converted to the one of a conventional non-linear control system without network. The simulations illustrate that the proposed scheme can compensate for communication delay and data dropout, and achieve the same control performance of the local closed-loop control system without network.
This paper is concerned with the design of networked control systems using the modified generalized predictive control (M-GPC) method. Both sensor-to-controller (S-C) and controller-to-actuator (C-A) network-induced delays are modeled by two Markov chains. M-GPC uses the available output and prediction control information at the controller node to obtain the future control sequences. Different from the conventional generalized predictive control in which only the first element in control sequences is used, M-GPC employs the whole control sequences to compensate for the time delays in S-C and C-A links. The closed-loop system is further formulated as a special jump linear system. The sufficient and necessary condition to guarantee the stochastic stability is derived. Simulation studies and experimental tests for an experimental hydraulic position control system are presented to verify the effectiveness of the proposed method. [DOI: 10.1115/1.4003385]
This paper studies the robust control problem of a class of Networked Predictive Nonlinear Control Systems (NPNCS), where the nonlinear process is linearisable and contains non-vanishing perturbations. The robust control of the NPNCS is considered. An event-driven networked predictive controller is applied to stabilise the system. A method is developed to study the stability of the closedloop unperturbed system by formulating the problem as a sub-optimal problem constrained by a series of inequalities. This method is also generalised to the robust stability analysis of the perturbed NPNCSs. The efficiency of the control scheme is illustrated using a numerical example.
This paper is concerned with the controller design of networked control systems (NCSs). A new model of the NCSs is provided under consideration of both the network-induced delay and the data packet dropout in the transmission. In terms of the given model, a controller design method is proposed based on a delay-dependent approach. The feedback gain of a memoryless controller and the maximum allowable value of the network- induced delay can be derived by solving a set of linear matrix inequalities. Two examples are given to show the effectiveness of our method.
In this paper we introduce a novel control network protocol, Try-Once-Discard (TOD), for networked control systems (NCS), and provide, for the first time, an analytic proof of global exponential stability for both the new protocol and the commonly used statically scheduled access methods. Controllers are designed without regarding the presence of the network in the feedback loop, so consequently many controller design techniques may be employed. The performance of the new network protocol and the statically scheduled protocols are compared in simulations.
In this paper, stochastic optimal controllers of networked control systems whose network-induced delay is longer than a sampling period are designed for the two cases of the system with either full state information or partial state information. The controllers are shown to render corresponding networked control systems exponentially mean square stable. The optimal estimator of the system state is presented when the system has partial state information and network-induced delay is longer than a sampling period. The separation theorem is proved to still hold in such networked control systems.
We discuss the effect of parameter uncertainty on the stability of
the closed-loop system regulated by a state feedback that linearizes a
nominal discrete-time nonlinear system which includes an observer. We
show that the stability of the perturbed closed-loop system is assured
if the closed-loop nominal system has an exponentially stable
equilibrium point including an exponential observer, and provided some
other conditions on the perturbation magnitude are also satisfied. An
analysis of the effect of parameter perturbations on the stability of
input-state linearizing controllers that make use of a state estimator,
is presented
We introduce a control network protocol, try-once-discard (TOD),
for networked control systems (NCS), and provide, for the first time, an
analytic proof of global exponential stability for both the new protocol
and the commonly used statically scheduled access methods. Controllers
are designed without regarding the presence of the network in the
feedback loop, so consequently many controller design techniques may be
employed. The performance of the new network protocol and the statically
scheduled protocols are compared in simulations
A survey on networked control systems (NCSs), published in a previous paper, is updated and extended. A simple framework and some general formulations for the study of NCSs are proposed. In addition to the survey on NCSs, the impact of NCSs on traditional large-scale system control methodologies with a related application is also reviewed.
First, we review some previous work on networked control systems
(NCSs) and offer some improvements. Then, we summarize the fundamental
issues in NCSs and examine them with different underlying
network-scheduling protocols. We present NCS models with network-induced
delay and analyze their stability using stability regions and a hybrid
systems technique. Following that, we discuss methods to compensate
network-induced delay and present experimental results over a physical
network. Then, we model NCSs with packet dropout and multiple-packet
transmission as asynchronous dynamical systems and analyze their
stability. Finally, we present our conclusions
Many different network types have been promoted for use in control
systems. In this article, we compare three of them: the Ethernet bus,
with carrier sense multiple access with collision detection,
token-passing bus (e.g., ControlNet), and controller area network bus
(e.g., DeviceNet). We consider how each control network can be used as a
communication backbone for a networked control system connecting
sensors, actuators, and controllers. A detailed discussion of the medium
access control sublayer protocol for each network is provided. For each
protocol, we study the key parameters of the corresponding network when
used in a control situation, including network utilization, magnitude of
the expected time delay, and characteristics of time delays. Simulation
results are presented for several different scenarios, and the
advantages and disadvantages of each network are summarized
The defining characteristic of a networked control system (NCS) is having a feedback loop that passes through a local area computer network. This paper considers nonlinear systems controlled in this manner, and demonstrates that for sufficiently high transmission rates, the network may be considered transparent. Three methods of scheduling data packets are compared: a static scheduler (token ring), the Try-OnceDiscard (maximum-error-first) scheduler with continuous priority levels, and the Try-Once-Discard scheduler with discrete priority levels. The third method is of particular interest when only a small number of bits are available for collision resolution. Asymptotic stability is guaranteed in the first two cases, and ultimate uniform boundedness in the third. In the final section, simulations demonstrate the theoretical results. The contributions of this paper are two-fold: first, it extends the earlier results on NCS to nonlinear systems, and second, it allows for finite word-len...
An analysis of the effect of parameter perturbations on the stability of input–output linearizing controllers for a class of MIMO discrete-time nonlinear systems is presented. A static state feedback is designed to input–output linearize a system without perturbations, and it is applied to the same system with an uncertain parameter vector, thus introducing an uncertain parameter term in the nominal input–output linearized system. It is shown that the state trajectories of the closed-loop perturbed system are ultimately bounded provided the closed-loop nominal system is exponentially stable and some conditions on the bounds of the perturbation terms are satisfied.
The robust control of networked predictive control systems with random network delay in the feedback channel is studied in this paper. The stabilisation of systems with constant time-delay is discussed by converting the corresponding Lyapunov inequality to a non-linear inequality. To obtain the maximum domain of uncertainties, the non-linear inequality is evolved as a non-linear optimisation control problem. After the optimisation problem is solved, it yields a controller that can stabilise the system and the domain of uncertainties. Furthermore, for the case of random network induced time-delay, robust stabilisation problem can be formulated as a set of inequalities, which are related to the corresponding constant time-delay, respectively. This result is verified by a numerical example.
We consider a single-input nonlinear discrete-time system of the form ¿: x(t+l) = f(x(t),u(t)) where xeIRN, ueIR, and f(x,u): IRN+1¿IRN is a C¿ IRN-valued function. Necessary and sufficient conditions for approximate linearizability are given for ¿. A sufficient condition for local linearizability is also given.
This paper describes a novel control technique to deal with networked control systems with random communication time delay, which is known to highly degrade the control performance of the controlled system. This problem can be solved using a modified model predictive control, which uses the future control sequence to compensate for the forward communication time delay. Also, using a model predictor, the time delay in the backward channel can be compensated as well. Another key part of this paper is to analyse the stability of the networked control systems. The analytical criteria are obtained for both fixed and random communication time delays. The simulation results and practical experiments illustrate that the proposed controller design is realistic.
Necessary and sufficient conditions are given for a nonlinear discrete-time system to be feedback equivalent to a controllable linear system. Some preliminary work on the effects of sampling on feedback linearizability is reported.
A new class of Lyapunov uniformly globally asymptotically stable (UGAS) protocols in networked control systems (NCS) is considered. It is shown that if the controller is designed without taking into account the network so that it yields input-to-state stability (ISS) with respect to external disturbances (not necessarily with respect to the error that will come from the network implementation), then the same controller will achieve semi-global practical ISS for the NCS when implemented via the network with a Lyapunov UGAS protocol. Moreover, the ISS gain is preserved. The adjustable parameter with respect to which semi-global practical ISS is achieved is the maximal allowable transfer interval (MATI) between transmission times.
The use of a data network in a control loop has gained increasing attentions in recent years due to its cost effective and flexible applications. One of the major challenges in this so-called networked control system (NCS) is the network-induced delay effect in the control loop. Network delays degrade the NCS control performance and destabilize the system. A significant emphasis has been on developing control methodologies to handle the network delay effect in NCS. This survey paper presents recent NCS control methodologies. The overview on NCS structures and description of network delays including characteristics and effects are also covered.
In this paper we present an approach to decoupled force/position control of the end-effector along the same direction for redundant robots, and an approach to nonholonomic cart pushing with mobile manipulators. The mobile manipulator is considered as a redundant robot, and a unified dynamic model for an integrated mobile platform and on-board manipulator is developed. The dynamic model is decoupled and linearized using the nonlinear feedback technique in a unified frame. Combining the event-based planning and control method with singularity analysis of the robot arm, a task level action controller is designed and an online kinematic redundancy resolution scheme is developed. The system is stable during normal operation as well as at the occurrence of unexpected obstacles. In addition, explicit force/position control along the same task direction for redundant robots is proposed. The kinematic redundancy of mobile manipulators enables independent control of force and position along the same task directions. To verify the decoupled force/postion scheme, an integrated task planning and control approach is further proposed for the mobile manipulator to complete complicated tasks by regulating its output force. A cart pushing task, which requires both force and position control along the same task direction, is discussed. The cart manipulation task fully integrates trajectory and force planning of the cart, and planning and control of the mobile manipulators. The approaches have been tested on a mobile manipulator consisting of a Nomadic XR4000 and a Puma 560 robot arm. The experimental results demonstrate the efficacy of the approach for the mobile manipulation of a nonholonomic cart.
In this paper we study the input-output decoupling problem for discrete time nonlinear systems. Both the static state feedback decoupling problem and the dynamic decoupling problem are stated and locally solved under conditions which are directly verifiable in terms of the system's dynamics and output. This paper is the discrete time counterpart of the analogous continuous time theory.
We propose and validate algorithms for choosing finite word length
priorities for dynamically scheduled networked control systems. Two
schemes for selecting priority levels are studied, the first with a
fixed arbitrary grid and the second with an auto-scaling grid. We prove
that the system is uniformly and ultimately bounded in the case of the
static encoding, and asymptotically stable with the auto-scaling
methodology. Simulations of an inverted pendulum controller using these
schemes are compared against a controller with point-to-point wiring
The problem of robust stabilisation is considered for uncertain systems using linear static output feedback (SOF) controllers and linear dynamical output feedback (DOF) controllers with fixed order. The existence conditions of output feedback controllers are given by matrix inequalities and result in some nonlinear optimisation problems. A class of uncertain plants whose nominal systems are minimal phase with a full row rank can be robustly stabilised by an SOF controller and DOF controller with any order, and controllers can be obtained by solving some linear matrix inequalities (LMIs). The method is demonstrated by two numerical examples
Results on input-output L<sub>p</sub> stability of networked control systems (NCS) are presented for a large class of network scheduling protocols. It is shown that static protocols and a recently considered dynamical protocol called try-once-discard belong to this class. Our results provide a unifying framework for generating new scheduling protocols that preserve L<sub>p</sub> stability properties of the system if a design parameter is chosen sufficiently small. The most general version of our results can be used to treat NCS with data packet dropouts. The model of NCS and, in particular, of the scheduling protocol that we use appears to be novel and we believe that it will be useful in further study of these systems. The proof technique we use is based on the small gain theorem and it lends itself to an easy interpretation. We prove that our results are guaranteed to be better than existing results in the literature and we illustrate this via an example of a batch reactor.
Linearizable nonlinear discrete-time systems are described as a
composition of diffeomorphisms. Linearizability of a general nonlinear
system is obtained as an extension of linearizability of the system
expressed in terms of the diffeomorphisms