Evolutionary generation of fuzzy guidance laws

A wide range of techniques have been applied to the missile guidance problem. These range from the classical proportional navigation and augmented proportional navigation (APN) to modern techniques using optimal control, feedback linearisation and differential game theory. Soft computing techniques have also been applied including neural networks and fuzzy logic. In this paper an evolutionary computing based approach is used to generate a fuzzy based guidance law. The performance of each fuzzy guidance law is assessed using a competition based approach. The individual fuzzy guidance laws compete with one guidance law, a differential game theory based law as well as with each other. In order to produce robust laws a wide range of engagement geometries, target velocities and target manoeuvre levels were used. The fuzzy guidance laws produced by the evolutionary algorithm produced encouraging results in comparison to an APN and a differential game theory based guidance law. The evolutionary technique used in this paper is seen as an extra tool which will become more relevant as the power of computers increases in aiding the missile guidance law designer develop new laws