Optimal Evasive Guidance Against Linear Optimal Guidance Laws

An analytical solution is proposed for optimal bang bang evasive guidance against pursuers governed by linear optimal guidance laws. The command derived involves an exact infinite series representation of the switch function and hence a finite-order approximation scheme that achieves arbitrary accuracy while maintaining computational efficiency is developed. For the special case of ideal evader dynamic, we demonstrate that the switch function of the guidance command can be determined by a finite-order polynomial without any approximation, which requires much less computational burden than the approximate solution. Through comprehensive numerical simulations, we validate the proposed methods and establish their superior computational efficiency and accuracy compared to existing approaches. Its effectiveness in a realistic three-dimensional, nonlinear, speed-varying aerodynamic scenario has been validated by simulations. The results show significant improvements in both real-time performance and solution precision for evasive guidance applications.