Research On Friction Compensation And High-precision Control Strategy Of The Launcher Servo System

The friction compensation and high-precision control methods of launching equipment servo system are carried out in this paper,which is based on the National Natural Science Foundation of China,The main work is as follows:Firstly,establish the mathematic model of launching device servo system.The composition and working principle of servo system are introduced,and the mathematical model of permanent magnet synchronous motor and the launching device servo system are established.Then the current loop,speed loop and position loop of servo system are designed and analyzed.Finally,combined with the controller of the closed loop system,the influence of friction on position servo system is in-depth analysis analyzed,which lays a good foundation for the design of subsequent high performance control strategy.Secondly,robust adaptive control strategy with friction state observer.For the unmeasurable friction state in the LuGre friction model,two types of state observers are designed.One is an improved state observer based on discontinuous mapping and smooth transfer function,which solves the problem of system instability caused by traditional observer at high speed.The other is to utilize the universal approximation property of neural network to design neural network mane observer,which has higher observational accuracy than the first observer,and has strong self-learning and self-adapting advantages.In addition,based on the state observer,the idea of robust adaptive control is fused to solve the problems of unknown friction coefficient,load torque and external time-varying disturbance.Thirdly,research on intelligent compound control strategy based on global friction approximation.In order to get rid of the shackles of complex mathematical models of friction and design a controller which can meet the high performance requirements,an intelligent compound control strategy based on adaptive robust control and neural network control is be brought up.Under the assumption of completely unknown friction characteristics,the whole firiction torque is approximated by radial basis function neural network,and adaptive robust control is adopted to deal with parameter uncertainty,external disturbance and friction approximation error.In addition,in order to enhance the ability of approximating nonlinear friction,a multi-layer neural network based on Sigmoid function is designed to approximate the entire friction torque.Lastly,experimental research.Analyzing the basic components of the experimental system,and design the system's PC softwares.Based on the launching device of servo system experimental platform,the validity and feasibility of the proposed control strategy are verified by experiments.