Research Of Radar Radiant Point Recognition Arithmetic

The paper firstly analyzes the threat facing radar antagonism and the methods of radar recognition at present, and discusses the important function of the artificial neural network, the most blasting subject in future military system. With the development of science and technology, modern radars have possessed the characteristics of advanced counter-reconnaissance and anti-jamming. Modern warfare still uses database or knowledgebase query method, which is advantageous in the rapid recognition speed and the simple implementation, but disadvantageous in reliability and fault tolerance so do not to meet the need of it. Furthermore a three-layer model for radar type recognition in data fusion is put forward in the paper, in which artificial neural network theory(ANN)* distance/fuzzy theory , uncertainty reasoning and decision-making theory are combined effectively. Artificial neural network is used to recognize radar system, which is basic for the further recognition of radar signal threat; Distance/fuzzy matching are used to recognize radar type are of the radar system ANN has recognized; Uncertain reasoning and decision-making is used to fuse the data in the different space and time fields. The result of recognition demonstrates that this model efficiently improves the recognition rate, the reliability and the ability of fault tolerance and shows the better application foreground in the environment with strong noise and increasingly high-density electromagnetism signal.With radar numbers increase, the method of radar system ANN classification in this paper has the problem of longer learning time and worse capability of artificial neural network. This problem can be solved through using several small and simple neural networks instead of a big and complex neural network, which will be a new research direction of this thesis in future.