Research on air-to-ground missiles autonomous recognizing targets online based on similar competitive neural networks

Autonomous recognizing attacking target online plays an important role in the characters of the air-to-ground missile´s battle performance. Because of the poor calculation capability and the short reaction time of onboard computer, this paper aims to solve the problem that how to recognize targets quickly. In order to reduce the onboard computer´s calculated amount, a lot of autonomous target recognition works need to be done on ground by means of template matching. To improve the precision and the speed of the online autonomous target recognition, the similar competitive neural network is proposed, which improves insufficient such as poor anti-jamming capability of the competitive networks and targets which cannot be eliminated exclude the template type, well solving the problem of the online autonomous target recognition algorithm. At last, the algorithm is verified by simulation.