Autoregressive Wavelet Neural Network Predictive Control Of A Shipborne Artillery

With the development of military science and technology,modern war requires high precision and fast response.The best way to improve the firing accuracy of the rocket is to realize the automatic operation and aiming of the weapon,so the rocket control system plays a key role.This paper takes a shipborne rocket launcher system a s the engineering background.The high-precision position control can be realized under strong interference and parameter perturbation through the designed rocket launcher's position servo system control strategy.The research work of this paper mainly focuses on the following aspects:(1)This paper introduces the composition of the servo system of a shipborne rocket gun.The PMSM model is derived based on the vector control method,and the mathematical model of the AC servo system of the shipborne rocket gun is established.The nonlinear factors affecting the system are analyzed.(2)The off-line identification of the servo system is studied,and the model prediction algorithm is introduced to modify the identification output.Wavelet neural network and autoregressive wavelet neural network(SRWNN)are used to identify the model.The simulation results show that the scheme has good dynamic characteristics and can better adapt to the rapid changes of the system.(3)An on-line identifier is designed to provide gradient information for the parameter adjustment of neural network in the controller.According to the prediction characteristics of SRWNN,an adaptive autoregressive wavelet sliding mode position loop control strategy is proposed.Based on the fast terminal sliding mode function,the control rate of the system is derived,and the SRWNN is used as the approximation algorithm to approach the fast terminal sliding mode control law.The adaptive rate of the on-line adjusting parameters is given,and the stability of the system is proved.(4)The hardware circuit and the software control interface of the host computer are designed based on STM32 and DSP dual processor architecture.A semi physical simulation experiment platform is built to verify the adaptive autoregressive wavelet sliding mode control strategy designed in this paper.The experimental results show that the control strategy can meet the performance index of the system,and the system design scheme is correct.