Research On Control Method Of Manipulator Based On Sliding Mode Variable Structure

Manipulator plays an increasingly important role in the field of modern industry.It can replace human beings to engage in repetitive production activities in high-risk and harsh environments.Therefore,it is of great significance to study the control of the manipulator to promote the development of advanced manufacturing industry.Robustness is one of the key indicators to measure the performance of a manipulator joint trajectory tracking system.In order to obtain strong robustness of the control system,this paper combines sliding mode control with other intelligent control methods,and studies the trajectory tracking control method under the influence of two factors: inaccurate system model and non-linear dead zone.For the problem that the precise mathematical model of the manipulator system is difficult to obtain and the control system performance is not good due to external interference,an improved sliding mode control algorithm based on neural network is proposed in this paper.The system is divided into nominal part and non-nominal part.For the former,the direct torque method is used to design the control law.while for the non-nominal part of the system,the sliding mode compensation control law is designed.At the same time,the learning ability of RBF neural network is used to approximate the uncertainties adaptively.And this paper chooses genetic algorithm to optimize the base width and base function center.For the output error and limit cycle oscillation of joint tracking control system of manipulator caused by dead zone,a neural network adaptive sliding mode controller with dead zone compensation is designed in this paper.Firstly,the mathematical relationship among the controller,dead-zone and dead-zone compensator is deduced according to the compensation principle,and the inverse model of dead-zone is obtained.Then,a reasonable membership function is constructed from the inverse mathematical model of dead-zone and corresponding fuzzy rules are formulated.The output of the fuzzy system is obtained by the method of defuzzification.Finally,the output of the fuzzy system and the output of the adaptive sliding mode controller can be transmitted.The output of the dead-time compensator is obtained by summing up.The system can track the given joint trajectory unbiased and fast in the presence of dead-time.To prove the effectiveness of the designed algorithm,a simulation model of two-joint manipulator with actual component parameters is built on the platform of Mat lab/Simulink,and simulation experiments are carried out with the above control strategies.The simulation results show that the optimization of network parameters by genetic algorithm can quicken the convergence of error,improve the accuracy of neural network mapping,and improve the sliding.The model control algorithm can reduce the chattering of the system and improve the stability of the control system.The sliding mode controller with fuzzy logic compensation reduces the influence of dead time and improves the tracking accuracy of the manipulator joint.