Abstract: A novel algorithm of neural network based information fusion and parallel adaptive tracking of maneuvering targets (NIFPAT) is proposed. The back propagation neural network applied to the information fusion is trained off line based on the feedback information of all state,and the output of the network is adjusted on line automatically based on the feature vector to fit the different movements of targets. The results of simulation show that the NIFPAT algorithm proposed has an excellent capacity for adaptively tracking maneuvering targets.

Key words: info rmation fusion, neural net, target tracking