A neural-network-based adaptive variable structure control methodology is proposed for the tracking problem of nonlinear discrete-time input-output systems.The unknown dynamics of the system are approximated via radia...A neural-network-based adaptive variable structure control methodology is proposed for the tracking problem of nonlinear discrete-time input-output systems.The unknown dynamics of the system are approximated via radial basis function neural networks.The control law is based on sliding modes and simple to implement.The discrete-time adaptive law for tuning the weight of neural networks is presented using the adaptive filtering algorithm with residue upper-bound compensation.The application of the proposed controller to engine idle speed control design is discussed.The results indicate the validation and effectiveness of this approach.展开更多
基金This project is supported by National Natural Science Foundation of China (No.59806007)
文摘A neural-network-based adaptive variable structure control methodology is proposed for the tracking problem of nonlinear discrete-time input-output systems.The unknown dynamics of the system are approximated via radial basis function neural networks.The control law is based on sliding modes and simple to implement.The discrete-time adaptive law for tuning the weight of neural networks is presented using the adaptive filtering algorithm with residue upper-bound compensation.The application of the proposed controller to engine idle speed control design is discussed.The results indicate the validation and effectiveness of this approach.
