| With the development of the society and the progress of science and technology,robot has been used in more and more fields,including medical,Marine,construction,service,agriculture and other fields.At the same time,the people of the robot system control accuracy,response speed and resistance to disturbance is becoming more and more demanding.At the core of the whole robot control system,on the whole performance of the robot has played a vital role.So the robot system of automatic control has very important theory meaning and realistic meaning.Robot system includes the following four parts,respectively control the manipulator mechanism servo drive system and signal system,robot can imitate human all kinds of action,can carry on the automatic control through the PC software,according to the requirements of the different procedures to complete the corresponding work,and belongs to the modern mechanical and electrical integration equipment.But the control process of multi-joint robot is very complex,common multi-joint robot will have more input,and design a multiple output according to the program,and belongs to the nonlinear system For multi-joint robot control mode mainly divided into the following five kinds,respectively is the robust control of variable structure PID adaptive control and intelligent control.However,PID control is difficult to handle nonlinear features in the robot system.Adaptive control,robust control,variable structure control mostly depends on the precise mathematical model or nominal model,makes it hard to applied in actual robot system;Intelligent control often require previous knowledge as the foundation to support or order need known system,the practical application of intelligent control methods to bring very great difficulty.Therefore,in order to further improve the advanced control methods in the robot system is practical,for a class of nonlinear model is unknown,order time unknown robot trajectory tracking system,this paper studies the robot sliding mode variable structure control method based on data driven technique,completed the following work:(1)The dynamic mathematical model and complex characteristics of the robot system are expounded by looking at the relevant literature at home and abroad.Trajectory tracking control for robot system with unknown disturbance,uncertainty,nonlinear,and difficult to control the problem,this paper proposes a data driven quasi sliding mode decoupling control scheme.First,the non-parametric dynamic linearization of the unknown nonlinear robot system is implemented using data-driven technology.Secondly,the control law of quasi-sliding mode of data drive is designed by using the thinking of approach law.Finally,the decoupling control of the data driven quasi-sliding mode control system is realized by further combining the discrete expansion state observer(DESO).(2)In order to alleviate the vibration of sliding mode control system,a new method of decoupling control of the second order sliding mode is proposed.First,using data-driven technology to deal with the model's unknown and nonlinear properties of the robot system;Secondly,using the second order sliding mode control thinking,this paper presents a data driven two-stage sliding mode decoupling control scheme.Finally,the theoretical analysis shows that this design data data driven second-order sliding mode control scheme can guarantee the robot trajectory tracking system small area near the tracking error converges to zero.(3)In order to improve data drive decoupling control method of sliding mode control performance,enhance the adaptability of this method,based on PSO(Particle Swarm Optimization)algorithm to optimize the key parameters of the designed controller,further proposes the data driven sliding mode decoupling controller based on PSO Optimization,for the controller parameter selection provides an effective solution.(4)In order to analyze the effectiveness of the design control scheme,the closed-loop simulation experiment of robot trajectory tracking control system is carried out simultaneously.Firstly,the simulation platform is built in matlab environment,combining with the mechanism model of the existing robot track tracking system.Second,will the design of the data driven quasi sliding mode decoupling control method,the data driven second-order sliding mode decoupling control method and data driven sliding mode decoupling control based on PSOoptimization method is applied to the nonlinear robot trajectory tracking system,at the same time and existing PID control method simulation analysis;Finally,the simulation curve and performance evaluation index validate the effectiveness of the proposed control scheme. |
