Stochastic Adaptive Dynamic Surface Control Of Asynchronous Motor Based On Neural Network

Induction motors(IMs)are playing increasingly important roles in AC speed control systems and transmission systems due to their low cost,simple structure and high reliability.Since the IM drive system is a high-order,multivariable,strongly coupling nonlinear system,and the damping torque,the torsional elastic torque and the magnetic saturation will make parameters in the IM system such as motor torque,self-inductance,mutual inductance and winding resistance variable,which means there exits stochastic disturbance in the IM drive system,the dynamic response speed and control precision of the motor system are affected.Although a variety of control schemes based on IM drive system have been proposed by some scholars,only few existing literature on IM speed regulation system has taken stochastic disturbance into consideration.Therefore,the research of control strategy to improve the dynamic and static performance of IM speed regulation system is a valuable research direction in theoretical and practical applications.In this paper,the neural network speed regulation control of IM stochastic system is studied based on dynamic surface technology and adaptive backstepping.During the design process,the radial basis function(RBF)neural networks are used to approximate the unknown nonlinear functions in the systems,and then combine dynamic surface technology with backstepping to construct nonlinear controller.The proposed method can eliminate the influence of stochastic disturbance and achieve the high quality control of the IM speed regulation system.The main research results of this paper can be summarized as follows:First,the neural network adaptive dynamic surface control is proposed for a class of stochastic nonlinear system.During the design process,RBF neural networks are used to approximate the unknown nonlinear functions in the systems,and the dynamic surface control technique is used to overcome the problem of “explosion of complexity” inherent in the traditional backstepping design procedure by introducing first-order low-pass filters,then backstepping is employed to design adaptive controllers.Finally,Lyapunov method is used to analyze the stability of the system.Second,the problem of speed regulation control for induction motor is proposed based on dynamic surface technique and neural network via adaptive backstepping.RBF neural networks are applied to approximate the unknown nonlinear functions in the systems and the dynamic surface control technique is used to overcome the problem of “explosion of complexity” inherent in the traditional backstepping design procedure.The real control law of the whole system is given in the last step of the backstepping control,then the stability analysis shows that the designed controller can eliminate the influence of stochastic disturbance and all the signals in the system remain bounded.The simulation results are provided to demonstrate the effectiveness and robustness of the proposed method.Third,the speed regulation control problem of IM stochastic nonlinear systems with input saturation is further studied.The nonlinear functions of the system are approximated by RBF neural network systems,the computational burden inherent in the traditional backstepping design procedure is greatly reduced by introducing the dynamic surface control technique.As a result,the proposed controller can achieve a good speed regulation performance against the stochastic disturbances and only contains one adaptive parameter to be adjusted,which means it is more effective in practical engineering.