Neuro-fuzzy techniques for airborne target tracking

In air defence system, surveillance radar with conventional tracking algorithm may not be able to track multiple targets accurately especially in a dense and cluttered environment, jamming and electronic counter measures. In this paper, an investigation has been carried out to implement the conventional tracking algorithms in conjunction with the neuro-fuzzy technique in such a way that the tracking error can be minimised and the manoeuvring target trajectories can be predicted. The tracking performance of the proposed approaches, called neuro-fuzzy aided joint probabilistic data association, and neuro-fuzzy aided nearest neighbour probabilistic data association, have been simulated. Results show that the proposed approaches yield improvement in tracking accuracy as well as resolving closely spaced tracks