Optimisating evader's trajectory in 6DOF pursuit-evasion problem using parallel evolutionary programming

The optimisation of air combat manoeuvre using standard evolutionary programming (EP) algorithm is discussed. The objective is to increase the level of realism in the simulation. This is achieved via employing the nonlinear six degree-of-freedom equations of motion to represent the vehicle. The evader is modelled as a six degree of freedom generic jet fighter aircraft. The aileron, elevator, rudder and throttle setting of the aircraft are set as control variables. The pursuer is a medium range generic air-to-air missile and modelled as a point-mass. The air combat is played in three dimensions. The pursuer seeks to intercept the evader and the evader seeks to avoid interception. The search for optimal evasion solution is conducted utilising evolutionary programming (EP). The optimisation algorithm developed aims to maximise the objective function. The objective function is a self-play simulation between the players. The value of the game is demonstrated as the outcome of the game. The solution that produces the maximum value of fitness is considered to be the best. The optimal solution found is found to be highly dependent on the initial condition. A slight change of the initial condition will result a completely different set of optimal solutions. The computing time is further improved through parallel computing. This is achieved via dividing the solutions into smaller groups and sending each group to different processors for evaluation. As the numbers of processors increase, the time taken to search for optimal solution was found to decrease.