复合控制拦截弹的神经网络姿态控制

Attitude control for interceptor missile by combine control based on neural networks

为了有效地调整复合控制拦截弹的姿态运动,提出了一个姿态控制方法。首先基于动量矩定理建立了拦截弹姿态运动的数学模型,然后针对数学模型和推力器点火逻辑的特点,设计了拦截弹姿态运动系统的神经网络控制器。使用常规方法获取基本的控制器,通过神经网络对常规控制器的控制参数进行调节,最后利用最优原理求出了拦截弹期望的俯仰控制力矩和偏航控制力矩。仿真结果表明,该姿态控制方法能够有效地实现拦截弹的精确姿态控制,并且具有较好的动态性能。

In order to effectively adjust attitude motion of interceptor missile by combine control, an attitude control method was presented. A mathematical model of interceptor missile was constituted based on moment of momentum theorem. And with a view to mathematical model and fire logic of thrusters, a neural network controller of attitude motion system was designed for interceptor missile： acquiring basal controller based on simple method, and adjusting control parameters of simple controller using neural network. A desired pitch control moment and yaw control moment were obtained using optimal principle for interceptor missile. Simulation results show that this attitude control method can implement efficiently precise attitude control of interceptor missile as well as preferable dynamic performance.