This paper presents an application of adaptive neural network model-based predictive control (MPC) to the air-fuel ratio of an engine simulation. A multi-layer perceptron (MLP) neural network is trained using two on-l...This paper presents an application of adaptive neural network model-based predictive control (MPC) to the air-fuel ratio of an engine simulation. A multi-layer perceptron (MLP) neural network is trained using two on-line training algorithms: a back propagation algorithm and a recursive least squares (RLS) algorithm. It is used to model parameter uncertainties in the nonlinear dynamics of internal combustion (IC) engines. Based on the adaptive model, an MPC strategy for controlling air-fuel ratio is realized, and its control performance compared with that of a traditional PI controller. A reduced Hessian method, a newly developed sequential quadratic programming (SQP) method for solving nonlinear programming (NLP) problems, is implemented to speed up nonlinear optimization in the MPC. Keywords Air-fuel ratio control - IC engine - adaptive neural networks - nonlinear programming - model predictive control Shi-Wei Wang PhD student, Liverpool John Moores University; MSc in Control Systems, University of Sheffield, 2003; BEng in Automatic Technology, Jilin University, 2000; Current research interests automotive engine control, model predictive control, sliding mode control, neural networks.Ding-Li Yu obtained B.Eng from Harbin Civil Engineering College, Harbin, China in 1981, M.Sc from Jilin University of Technology, Changchun, China in 1986 and PhD from Coventry University, U.K. in 1995, all in control engineering. He is currently a Reader in Process Control at Liverpool John Moores University, U.K. His current research interests are in process control, engine control, fault detection and adaptive neural nets. He is a member of SAFEPROCESS TC in IFAC and an associate editor of the IJMIC and the IJISS.展开更多
This paper introduces a self-learning control approach based on approximate dynamic programming. Dynamic programming was introduced by Bellman in the 1950's for solving optimal control problems of nonlinear dynami...This paper introduces a self-learning control approach based on approximate dynamic programming. Dynamic programming was introduced by Bellman in the 1950's for solving optimal control problems of nonlinear dynamical systems. Due to its high computational complexity, the applications of dynamic programming have been limited to simple and small problems. The key step in finding approximate solutions to dynamic programming is to estimate the performance index in dynamic programming. The optimal control signal can then be determined by minimizing (or maximizing) the performance index. Artificial neural networks are very efficient tools in representing the performance index in dynamic programming. This paper assumes the use of neural networks for estimating the performance index in dynamic programming and for generating optimal control signals, thus to achieve optimal control through self-learning.展开更多
The design of a wireless water level control system is introduced and discussed in detail. In this system, the wireless Proportional Integral (PI) controller is developed using the LabVIEW graphical user programming l...The design of a wireless water level control system is introduced and discussed in detail. In this system, the wireless Proportional Integral (PI) controller is developed using the LabVIEW graphical user programming language. Zigbee wireless technology is chosen for the wireless data transfer system. The experimental testbed was built and the system software and hardware were implemented. In order to compare the performance of the wired and wireless system, a corresponding wired water level control system was built. Experimental results show that under the same PI parameters, the settling time of the wired system is 3.3 times faster than the wireless system. However, the percent overshoot using the wireless controller is 4% smaller.展开更多
Design of control strategies for gene regulatory networks is a challenging and important topic in systems biology. In this paper, the problem of finding both a minimum set of control nodes (control inputs) and a contr...Design of control strategies for gene regulatory networks is a challenging and important topic in systems biology. In this paper, the problem of finding both a minimum set of control nodes (control inputs) and a controller is studied. A control node corresponds to a gene that expression can be controlled. Here, a Boolean network is used as a model of gene regulatory networks, and control specifications on attractors, which represent cell types or states of cells, are imposed. It is important to design a gene regulatory network that has desired attractors and has no undesired attractors. Using a matrix-based representation of BNs, this problem can be rewritten as an integer linear programming problem. Finally, the proposed method is demonstrated by a numerical example on a WNT5A network, which is related to melanoma.展开更多
In this paper, an online optimal distributed learning algorithm is proposed to solve leader-synchronization problem of nonlinear multi-agent differential graphical games. Each player approximates its optimal control p...In this paper, an online optimal distributed learning algorithm is proposed to solve leader-synchronization problem of nonlinear multi-agent differential graphical games. Each player approximates its optimal control policy using a single-network approximate dynamic programming(ADP) where only one critic neural network(NN) is employed instead of typical actorcritic structure composed of two NNs. The proposed distributed weight tuning laws for critic NNs guarantee stability in the sense of uniform ultimate boundedness(UUB) and convergence of control policies to the Nash equilibrium. In this paper, by introducing novel distributed local operators in weight tuning laws, there is no more requirement for initial stabilizing control policies. Furthermore, the overall closed-loop system stability is guaranteed by Lyapunov stability analysis. Finally, Simulation results show the effectiveness of the proposed algorithm.展开更多
A support vector rule based method is investigated for the construction of motion controllers via natural language training. It is a two-phase process including motion control information collection from natural langu...A support vector rule based method is investigated for the construction of motion controllers via natural language training. It is a two-phase process including motion control information collection from natural language instructions, and motion information condensation with the aid of support vector machine （SVM） theory. Self-organizing fuzzy neural networks are utilized for the collection of control rules, from which support vector rules are extracted to form a final controller to achieve any given control accuracy. In this way, the number of control rules is reduced, and the structure of the controller tidied, making a controller constructed using natural language training more appropriate in practice, and providing a fundamental rule base for high-level robot behavior control. Simulations and experiments on a wheeled robot are carried out to illustrate the effectiveness of the method.展开更多
With the increasing popularity of the computer network, more and more attention is paid to the reliability of the network. To improve the reliability of the communication network in the stored program control exchange...With the increasing popularity of the computer network, more and more attention is paid to the reliability of the network. To improve the reliability of the communication network in the stored program control exchange, the author propounds a redundant method supported by a kind of the double local area network. Fault detection, hardware switching and system rebuilding are processed automatically by the administration of the software. The method can improve not only the system′s reliability but also its communication efficiency. It has been implemented in the pSOSystem environment in a super stored program control exchange.展开更多
文摘This paper presents an application of adaptive neural network model-based predictive control (MPC) to the air-fuel ratio of an engine simulation. A multi-layer perceptron (MLP) neural network is trained using two on-line training algorithms: a back propagation algorithm and a recursive least squares (RLS) algorithm. It is used to model parameter uncertainties in the nonlinear dynamics of internal combustion (IC) engines. Based on the adaptive model, an MPC strategy for controlling air-fuel ratio is realized, and its control performance compared with that of a traditional PI controller. A reduced Hessian method, a newly developed sequential quadratic programming (SQP) method for solving nonlinear programming (NLP) problems, is implemented to speed up nonlinear optimization in the MPC. Keywords Air-fuel ratio control - IC engine - adaptive neural networks - nonlinear programming - model predictive control Shi-Wei Wang PhD student, Liverpool John Moores University; MSc in Control Systems, University of Sheffield, 2003; BEng in Automatic Technology, Jilin University, 2000; Current research interests automotive engine control, model predictive control, sliding mode control, neural networks.Ding-Li Yu obtained B.Eng from Harbin Civil Engineering College, Harbin, China in 1981, M.Sc from Jilin University of Technology, Changchun, China in 1986 and PhD from Coventry University, U.K. in 1995, all in control engineering. He is currently a Reader in Process Control at Liverpool John Moores University, U.K. His current research interests are in process control, engine control, fault detection and adaptive neural nets. He is a member of SAFEPROCESS TC in IFAC and an associate editor of the IJMIC and the IJISS.
基金Supported by the National Science Foundation (U.S.A.) under Grant ECS-0355364
文摘This paper introduces a self-learning control approach based on approximate dynamic programming. Dynamic programming was introduced by Bellman in the 1950's for solving optimal control problems of nonlinear dynamical systems. Due to its high computational complexity, the applications of dynamic programming have been limited to simple and small problems. The key step in finding approximate solutions to dynamic programming is to estimate the performance index in dynamic programming. The optimal control signal can then be determined by minimizing (or maximizing) the performance index. Artificial neural networks are very efficient tools in representing the performance index in dynamic programming. This paper assumes the use of neural networks for estimating the performance index in dynamic programming and for generating optimal control signals, thus to achieve optimal control through self-learning.
文摘The design of a wireless water level control system is introduced and discussed in detail. In this system, the wireless Proportional Integral (PI) controller is developed using the LabVIEW graphical user programming language. Zigbee wireless technology is chosen for the wireless data transfer system. The experimental testbed was built and the system software and hardware were implemented. In order to compare the performance of the wired and wireless system, a corresponding wired water level control system was built. Experimental results show that under the same PI parameters, the settling time of the wired system is 3.3 times faster than the wireless system. However, the percent overshoot using the wireless controller is 4% smaller.
基金Supported by National High Technology Research and Development Program of China （863 Program） （2006AA04Z183）, National Nat- ural Science Foundation of China （60621001, 60534010, 60572070, 60774048, 60728307）, and the Program for Changjiang Scholars and Innovative Research Groups of China （60728307, 4031002）
文摘Design of control strategies for gene regulatory networks is a challenging and important topic in systems biology. In this paper, the problem of finding both a minimum set of control nodes (control inputs) and a controller is studied. A control node corresponds to a gene that expression can be controlled. Here, a Boolean network is used as a model of gene regulatory networks, and control specifications on attractors, which represent cell types or states of cells, are imposed. It is important to design a gene regulatory network that has desired attractors and has no undesired attractors. Using a matrix-based representation of BNs, this problem can be rewritten as an integer linear programming problem. Finally, the proposed method is demonstrated by a numerical example on a WNT5A network, which is related to melanoma.
文摘In this paper, an online optimal distributed learning algorithm is proposed to solve leader-synchronization problem of nonlinear multi-agent differential graphical games. Each player approximates its optimal control policy using a single-network approximate dynamic programming(ADP) where only one critic neural network(NN) is employed instead of typical actorcritic structure composed of two NNs. The proposed distributed weight tuning laws for critic NNs guarantee stability in the sense of uniform ultimate boundedness(UUB) and convergence of control policies to the Nash equilibrium. In this paper, by introducing novel distributed local operators in weight tuning laws, there is no more requirement for initial stabilizing control policies. Furthermore, the overall closed-loop system stability is guaranteed by Lyapunov stability analysis. Finally, Simulation results show the effectiveness of the proposed algorithm.
基金This work was partially supported by the Royal Society of UK and the National Natural Science Foundation of PRC (No. 60175028).
文摘A support vector rule based method is investigated for the construction of motion controllers via natural language training. It is a two-phase process including motion control information collection from natural language instructions, and motion information condensation with the aid of support vector machine （SVM） theory. Self-organizing fuzzy neural networks are utilized for the collection of control rules, from which support vector rules are extracted to form a final controller to achieve any given control accuracy. In this way, the number of control rules is reduced, and the structure of the controller tidied, making a controller constructed using natural language training more appropriate in practice, and providing a fundamental rule base for high-level robot behavior control. Simulations and experiments on a wheeled robot are carried out to illustrate the effectiveness of the method.
文摘With the increasing popularity of the computer network, more and more attention is paid to the reliability of the network. To improve the reliability of the communication network in the stored program control exchange, the author propounds a redundant method supported by a kind of the double local area network. Fault detection, hardware switching and system rebuilding are processed automatically by the administration of the software. The method can improve not only the system′s reliability but also its communication efficiency. It has been implemented in the pSOSystem environment in a super stored program control exchange.
文摘一般非线性离散系统存在高度非线性、模型不确定性和动态未知性特性,传统方法应对其控制一般用一个简化或线性化的模型来代表真实系统,会导致固有误差。为此,文章针对一般非线性离散系统,提出了一种基于多维泰勒网(multi-dimensional Taylor networks ,MTN)的近似最优迭代动态规划方法,其所有控制过程均在线进行,无须离线训练步骤。该方法采用actor-Critic框架,并引入3个MTN网络:效用MTN用于在不依赖系统内部动态信息的条件下确定性能指标;Critic MTN用于逼近性能函数;执行MTN则在动态规划框架下在线调整控制策略。整套控制系统采用双闭环控制结构,外环以主要反馈信号实现跟踪控制,内环通过辅助反馈信号进一步提升动态性能。本文充分利用MTN的结构特性,显著降低了控制器的计算复杂度,大幅提升了迭代自适应规划算法的动态响应速度;以液压伺服系统为对象,开展了阶跃信号与正弦信号的跟踪仿真实验。仿真结果表明,基于迭代学习的多维泰勒网离散自适应最优控制器具有良好的跟踪性能和动态响应特性,验证了本文所提方法的有效性与实用性。
