Simulation and adaptive control of back propagation neural network proportional–integral–derivative for special launcher using new version of transfer matrix method for multibody systems

Abstract

Rocket launcher system, as a special launcher placed on tactical vehicles, is a very complex mechanical system with characteristics of strong shock and vibration. In order to improve position accuracy, as well as reduce vibration, this paper creates a nonlinear dynamics model of the launcher system by using a new version of the transfer matrix method for multibody systems. The overall transfer equation of the nonlinear model is deduced. Combining with general kinematics equations of the rocket, the system launch dynamics are simulated and compared with experiments to verify the correctness of the model. On this basis, a backpropagation neural network proportional–integral–derivative adaptive control system is designed to improve servo control of the launcher. Then, the effectiveness of this method is verified by comparing with the traditional proportional–integral–derivative control method. Simulated results show that the backpropagation neural network proportional–integral–derivative control system makes the azimuth and elevation angles reach the target values smoothly and quickly, with higher accuracy. The results prove that the proposed method prominently reduces vibrations of the launcher, by adjusting the control parameters online according to the operation state of the system, presenting a better stability and robustness.