径向基神经网络在优化导引律中的应用

应用神经网络的非线性的特点,在纯比例导引律的导引下,分析战机的运动轨迹,通过采样取得的一系列载机的控制输入数据,根据追踪导引控制的内在要求(目标视线角速度的绝对值减小并趋于零),设计载机导引控制器的相应的离散输出值,离线训练一个径向基神经网络模块,嵌入到载机控制回路中,作为战机导引的控制输入,实现载机在纯比例导引下的性能优化;仿真结果表明,在节能及省时方面,能达优化纯比例导引律的目的。

Study on RBF Optimization for The Guidance Law

Guidance laws of an aircraft may be optimized based on Neural Network.The corresponding one-dimensional discrete input signal to the guidance controller is designed using the sampled two-dimensional controlling parameters of the aircraft under the PPN(Pure Proportional Navigation) guidance law,according to the inherent requirements for tracking controls,and the off-line training of a module of RBF(Radial Basis Function) with two inputs and one output is performed.Then,the module is inserted into the control system of the aircraft and used as its control input to implement the aircraft performance optimization under the PPN guidance law.The simulation shows that the proposed guidance law with RBF is superior to the PPN guidance law in target track and acquisition time and energy management.