Swarm guidance using a multi-objective co-evolutionary on-Line evolutionary algorithm

This paper details the application of a multi-objective cooperative co-evolutionary on-line EA (MOC-COLEA) to the guidance of a swarm of multiple missiles, against multiple targets. The multi-objective algorithm is used to develop a dynamic objective front which is used to trade the spatial distribution of missiles at impact, against the time to impact. Each missile optimises its own performance, based on limited information of the current intended actions of the other missiles. The decision making process is thus distributed between the missiles giving distributed coordination. Results demonstrate the algorithm can form effective leader-less distributed control for multiple missiles against multiple targets in noisy and environments.