Research On Sliding Control Algorithms For Uncertain Systems And Its Application On Manipulators

As the applications of robots and manipulators are more and more popular,the control accuracy and anti-jamming capability of the manipulators are becoming much stricter,so it is necessary to design robust controller to completely reject the effect of the systems when there are load's changes and external disturbances,and to enhance the robustness of the robots or manipulators.In the first two chapters,the background,purpose and theoretical basis of the algorithm are introduced,and these content will provide the theoretical basis for the design of the controller below.The third chapter gives a method of designing neural network sliding mode controller for a class of systems with matched uncertainty.Firstly,system parameter variations,coupled uncertainty and external disturbance of the nonlinear interconnected systems are analyzed.Secondly,a recurrent neural network which is to estimate the system uncertainty is given to compensate the impact of the uncertainty on the control performance,and then we design the adaptive laws of the parameters of RBF function.Then,it is assumed that the whole of neural network estimation error,coupled uncertainties and external disturbance are represented by a lumped uncertainty term When the bound of lumped uncertainty can be known from priori-knowledge,auxiliary controller which is similar to reaching law is directly designed to eliminate the lumped uncertainty;on the other hand,a H? performance is proposed and a H? controller is designed to weaken the impact of external disturbance and couping uncertain on the system,at the same time,to eliminate the neural network estimation error,a auxiliary controller is adopted.Finally,the example of Bpolar coordinate parallel manipulator is given,and the simulation by Matlab/Simulink simulation verify the feasibility and effectiveness of the control method.The chapter four proposes a method of disturbance observer-based sliding mode control for a class of systems with unmatched uncertainty.Firstly,the method of constructing general disturbance observer is exhibited to realize the accurate estimate of the unmatched disturbance and unmatched uncertainty,a novel global sliding surface and controller is designed.The simulation example demonstrate that the general method has better performance than traditional sliding mode control,integral sliding mode control.Secondly,a control is proposed to handle a larger class of mismatched uncertainties by extending the disturbance observer and modifying and generalizing the sliding surface.Finally,the stability of the overall system is proved and the results are verified by comparison simulation of the general disturbance observer-based sliding mode control method and extended disturbance observer-based sliding mode control method.At last,the above control methods are summarized,and further expectation for its future development is given.