Research On The Control Strategy Of The AC Servo System Of The Rocket Obstacle-breaking Weapon

With the improvement of the modern military level,higher requirements have been put forward on the combat performance of rocket destroying obstacles weapon.The rocket destroying obstacles weapon have non-linear and time-varying characteristics due to factors such as the external environment,car body disturbance,and the impact moment generated when the rocket is launched.Therefore,the research on the control strategy of the AC servo system of the rocket destroying obstacles weapon has important practical value.This article takes the rocket destroying obstacles weapon as the engineering background,and the research work mainly includes the following aspects.(1)The structure and working principle of the AC servo system of the rocket obstaclebreaking weapon were analyzed;a mathematical model of the AC servo system of the rocket obstacle-breaking weapon was constructed,and the uncertain factors of the model were analyzed in detail.(2)Based on the equivalence principle of the RBF neural network,aiming at the nonlinear time-varying characteristics of the rocket barrier weapon system,using the TS-type fuzzy logic system,a TS-based PID controller(TS-type fuzzy RBF for short)was designed Neural network PID controller).The three parameters of the PID controller were adjusted mainly by using the T-S RBF neural network.Build MATLAB/Simulink simulation models for traditional PID controllers and T-S PID controllers,and conduct simulation comparison experiments.The simulation results showed that the overall control performance of the T-S fuzzy RBF neural network PID controller was much better than that of the traditional PID controller.(3)Aiming at the nonlinear and time-varying characteristics of the rocket obstaclebreaking weapon system and the principle that the TS fuzzy logic system and the RBF neural network were equivalent,a RBF neural network PID controller based on the TS fuzzy logic system was designed.(TS-FNN-PID for short).The three parameters of the PID controller were continuously adjusted mainly through the T-S fuzzy RBF neural network.A MATLAB/Simulink simulation model was built for the traditional PID controller and TS-FNNPID controller,and then the simulation comparison experiment was carried out.The simulation results showed that the overall control performance of the TS-FNN-PID controller was better than that of the traditional PID controller.(4)Since the training algorithm of the T-S fuzzy RBF neural network was a gradient descent method,this gradient descent method had the problem of slow convergence and easy to fall into the local minimum of the error surface.In order to avoid the above problems and ensure the compactness of the network at all times,an LM online adaptive network structure optimization algorithm based on the sliding window idea was proposed,and the algorithm was applied to the T-S fuzzy RBF neural network PID controller.The effect of TS type fuzzy RBF neural network PID controller(referred to as LM-SW-TS-FNN-PID,based on LM online adaptive network structure optimization algorithm)and TS type fuzzy RBF neural network PID controller(referred to TS-FNN-PID,Based on the gradient descent method)were simulated and compared through MATLAB/Simulink.The simulation results showed that the use of this training algorithm further improved the control performance of the system.