Muzzle Vibration Compensation Based on RBF Neural Network Active Disturbance Rejection Control

An active disturbance rejection compensation method for gun control systems of tank based on RBF neural networks is proposed to reduce the horizontal muzzle vibrations during tank high-speed movement caused by external disturbances. A comprehensive analysis and investigation are conducted on a multi-body dynamics model aimed at mitigating the effects of muzzle vibrations under the control of a tank bidirectional stabilizer. The RBF neural network is embedded into ADRC to perform online learning and prediction of the horizontal deviation angle of the muzzle vibrations as well as the horizontal tracking error of the turret. It dynamically generates corresponding compensation signals in real-time to compensate for the output signals of ADRC of turret’s horizontal servo system. The co-simulation results using Recurdyn/Simulink demonstrate the efficacy of the proposed control method in reducing the deviation angle and acceleration of muzzle horizontal vibration by 13.51% and 9.94% respectively. Additionally, it significantly improves the horizontal stability precision of the turret by 14.99%.