柔性身管的RBF神经网络PID控制

摘要

由于舰炮身管的弯曲振动,传统的利用炮管轴角位移进行反馈控制的方法无法精确控制炮口指向.针对这一缺点,提出了一种利用炮口角位移作为随动系统位置环反馈信息的控制结构.为模拟身管的振动,应用多刚体动力学对舰炮身管进行了柔性体建模.为改进后的控制结构设计了基于RBF神经网络整定PID参数的控制器,并与传统PID控制器的控制效果进行了仿真对比,验证其对于提高随动系统控制精度的有效性.%The traditional servo system can not provide the accurate control of the muzzle directionality because of the tube vibration, so aimed at the defect, a control structure which collected the feedback digital information of position loop from muzzle was put forward. A PID controller based on RBF neural network was designed for realizing the purpose of controlling the directionality of muzzle. Whose control effectivities were compared with effectivities of the traditional PID controller in simulation tests.