Variable Inertia Servo System Intelligent Control Technology

This paper takes the project "*** servo system's research and design" as the background. The system is designed for special mission, which need to drive different types of loads and cause the system rotor inertia change. If the controller is designed without treatment, it will lead poor performance of the system's control. In order to solve the variable inertia servo system's control problems, the classic PID control and neural network control are combined for a more in-depth research.This paper introduces the overall design of the servo system and hardware system's engineering. For variable inertia case, the servo control system mathematical model is established. The classical control theory is used to analyse inertia changes on servo system stability and control performance, and MATLAB is used on simulation study of the variable Inertia servo system classic PID control.For the shortcomings of classic PID control, a PID neural network control is proposed, and a PID neural network controller is designed. The PID neural network controller learning algorithm and PID neural network connection weights of initial selection is studied. The PID neural network control system stability is analysed, and theoretical value is given the weight of each layer to learn and modify the step length range. Finally, the variable inertia servo system PID neural network control of a simulation is studied.Simulation results show that the study of PID neural network controller for such variable inertia servo system has good control effect.