Sliding Mode Variable Structure Control Method And Its Application In Robot Trajectory Tracking

At present, the reserch of robot trajectory tracking control has become the forefront and hot problem domestically and abroad. Sliding Mode Variable Structure Control(SMVSC) is applicable to solve the problem of robot trajectory tracking, because the sliding mode motion of SMVSC has strong robustness against system parameter perturbation, external disturbances and system uncertainties. However, when the system states reach the sliding surface, it will cause high-frequency oscillation near the sliding surface,and its application and development are limited by the chattering of the SMVSC systems. The main works of this paper are as follows:a) The traditional SMVSC is applied to n-joint robot trajectory tracking control. According to the chattering of sliding mode control, this paper designed a model-based disturbance observer to compensate the uncertainties. However, the observer for slow time-varying interference signal, and depends on the accurate system model , so its application and development are limited to a certain extent.b) The backstepping control method consists of applying a single-variable control scheme to a multivariable control system, so the complex multi-variable system can be simplified. The backstepping control can be used to robot trajectory tracking, because the n-link manipulator is easy to split into multiple sub-systems.In this paper, a kind of backstepping sliding mode controller is designed using the advantages of combining backstepping control simplifying complex multi-joint movement and sliding mode control to solve the nonlinear characteristics.c) Most control methodologies of robot system assume that the states are available for feedback, However, it is difficult to obtain the movement modal of the robot system. According to the problems, a Luenberger-type observer is designed for the robot system with simple structure and better observation, but it is greatly dependent on the accurate model. Robot system is nonlinear and uncertain system, whose accurate models are difficult to obtain, so a Neural Network Observer for nonlinear systems was designed. The observer is less dependent on the accurate model, suitable for nonlinear system state observation. Simulation experiment tested its effectiveness and realized the sliding mode variable structure control based on state observer of robotic tracking, combining the two observers with sliding mode control respectively. d) According to the chattering problem, the paper proposed a method of inhibiting chattering of SMVSC. The main reasons of sliding mode chattering are inappropriate choice of switch gain and the switching characteristic of the controller. The method of restraining chattering designed in this paper is using neural network technology to adjust the switch gain of online adjustment to avoid great chattering due to switch gain choice caused by compensation of uncertainty. Besides, it uses saturation function instead of the sign function of the traditional sliding mode control, softening the switching characteristics of the system, thereby suppress chattering effectively.e) The robot system working environment is changeful and the robot system operates in multiple environments which may change abruptly from one to another by multiple models controller.Therefore,a good way to improve the performance of the controller is to use multiple models control if models are approximately available for different environments.This paper designed a multi-model backstepping sliding mode controller and proposed a PID-type switching index function, which increased forgetting factor making the witching more accurate and reasonable. The simulation results demonstrate the effctiveness and fasibility of he proposed control strategy.