Research On Control Strategy Of Turntable Servo System Based On RBF Neural Network

This paper will study the problems existing in the three-axis turntable.The turntable is an indispensable equipment in the development of aerospace field.It can simulate the flight attitude of missiles and other equipment.However,there is a serious coupling phenomenon between the three-axis turntable frames,which makes the three frames unable to move according to the specified speed and position during the simultaneous movement,which seriously affects the accuracy of the turntable,The control method has an important influence on the tracking error and response time of the turntable;Friction exists in all movements.In view of the friction phenomenon existing in the operation of the turntable,it is necessary to design corresponding methods to suppress it,and finally make the turntable achieve the ideal movement effect.(1)Firstly,the three-axis turntable is modeled,the transformation relationship of coordinate system is established according to the coupling phenomenon of the three frames,and the dynamic model is established;Then the output torque of the motor is established according to the mathematical model of the motor;Then the decoupling control law of the inverse system and the construction of the model are designed.The decoupling effect can be seen through simulation.(2)The control strategy of the turntable is studied.In view of the deterioration of the control effect of classical control methods due to the change of model parameters during the operation of turntable,this paper uses RBF neural network to adjust the PID parameters,so that it can solve the influence of interference factors such as external environment change on the system;Firstly,the turntable system under PID control is simulated,then the RBF neural network is studied,the RBF neural network PID controller is designed,the neural network iterative learning program is written in Matlab using s function,the joint simulation is carried out with Simulink,and the control effect is compared with PID control effect,so that the control effect of RBF-PID controller can be seen.(3)According to the friction phenomenon of the turntable,the corresponding friction suppression method is designed to suppress the friction.In this paper,acceleration feedback is used to improve the servo stiffness of the system to compensate the friction.Stribeck friction model is selected.After determining the friction parameters,Matlab/Simulink is used for system modeling and simulation.The simulation results show that the friction suppression method used in this paper has achieved good results.(4)Using Simscape to build and simulate the physical model of motor is different from the simulation of mathematical model and transfer function in Simulink.Through the modeling and Simulation of the physical model of the motor,it is verified that the RBF neural network PID control method used in this paper has higher reliability.