Radar Target Identification System Research And Design

This paper expatiate the importance of the radar targets recognition firstly. Because of more and more missiles and radars equipped the army in modern warfare, there have formed a complex and levity electron confronted circumstance. Besides to measure the targets' distance, orientation and elevation, the tasks of the radar is included to measure the targets' velocity. We can achieve more information from the radar's echo, and identify the targets. Then we can take any actions to protecting ourselves. So we will have the superiority status in modern warfare. Then I analyzed the situations of that in our country. The situation is that we just deal with the signals which sent from the radar systems with the recognition databases directly. This method requires the recognition database's workload very high, but on the other hand, its efficiency is low.Aim at this situations, I put forward a method on the base of the old recognition database is that a system have three identifier parts which using the theories of BPNN(Back Propagation Neural Network),Fuzz Clustering theory and Data Fusion theory. Using BPNN to classify the targets on the base of menace level, we can deal with the signals of the most dangerous targets which will have the superiority status in the war. Then the fuzz clustering system will classify the special targets which the BPNN can't identify. The fuzz clustering system will identify the character parameters of the same class targets, which reduces the working of the recognition databases. The data Fusion system will deal with all the data which sent from each radar terminal. And the last result will be more reliable.With the experiment under the MATLAB, the method was proved a good way to identify the targets of many radar signals. It has been also proved efficiently.