Optimal Real Time Evasion against High Speed Pursuer Using Evolutionary Programming

Abstract

This paper discusses the use of evolutionary programming to help a fighter pilot or to guide an UAV out manoeuvre a very agile and fast pursuer such as a missile. There are many techniques used to find the optimal evader’s trajectory but these techniques have their specific drawbacks such as being time consuming and computing intensive. This is not acceptable when the solution requires high reliability and must be generated quickly. An evolutionary technique has been developed to search for an optimal trajectory for an evader against a very agile and fast pursuer. In an air combat scenario, the technique searches for trajectories that conform to aircraft’s performance and aerodynamic constraints. Two scenarios were considered. The pursuer simply home-in to the evader and the evader follows the trajectory suggested from the algorithm. In both scenarios, the evader is able to steer itself from interception.