Active Disturbance Rejection Sliding Mode Control For Nonlinear Servo System With Dead Zone

In the design of nonlinear system control scheme,dead zone,nonlinear link such as backlash,friction and unknown disturbances are negative influence on the closed-loop output signal tracking effect,main show is backlash moment in a certain period of time delay,the tracking error of the system of the reference signal and gear when making the reversing motion due to the impact of disturbance signal.Therefore,how to use active disturbance rejection control,sliding mode control and other advanced control methods to compensate and suppress nonlinear links to improve the transient and steady-state performance of the system has been one of the focus of control research.In this paper,the relationship between dead zone and backlash nonlinearity is analyzed,and three common backlash models,their characteristics and application scenarios are introduced,and how to realize the accurate and fast tracking of reference signal in backlash servo system is studied.Since the immunity control of the extended state observer is able to "total disturbance" real-time feedback,based on the structural feature of servo system corresponding to improve the extended state observer,or by using neural network to estimate the system states and unknown disturbance signal in real time,and designed the structure of immunity of variable gain controller,the adaptive sliding mode controller,The stability and effectiveness are proved by theoretical analysis and simulation experiment.1.An improved gain-changing switched precise differential observer(SEDO)based on extended state observer is proposed to solve the state estimation problem of mechanical servo system with unknown backlash.In order to compensate the unknown nonlinear backlash accurately,a mathematical model of switched backlash moment was established to describe the input-output relationship of backlash.Then,based on the mathematical model of switching backlash moment,a switching backlash compensation control based on variable gain strategy is proposed,and the system states are estimated accurately according to the actual working conditions.In addition,the designed controller can further reduce the computational burden by feedforward compensation strategy,and part of the feedforward compensation task is shared by the observer.It is proved that all signals in the closed-loop system are bounded and the observation errors and tracking errors converge to a compact set small enough around the origin.2.A tracking filter adaptive sliding mode control scheme based on ADRC was proposed for servo systems with nonlinear backlash.The nonlinear link of backlash and friction is analyzed,which is transformed into system disturbance and the series integral system structure is established.A new type of non-singular sliding mode surface is constructed to make the system state converge faster and chattering less.In order to realize feedforward control,an adaptive weight regulation law is designed for online estimation of unknown nonlinearity.The simulation results show that the control scheme has better control effect on the nonlinear backlash of servo system.