Recognition Algorithm, Based On The Type Of Radar Target Of The Ar Model Poles

The radar target recognition is an important field of modern radar signal processing research. The process of the recognition approximately can be divided into two processes: The first is the abstracting of characters; the second is the pattern recognition. Since 1950's, some scholars have presented the radar target recognition techniques and put forward some effective technique methods. First, several main radar target recognition techniques are introduced in this paper,A method is presented in which the feature and the Autoregressive model pole are abstracted to identify the radar target. With dealing with the wave-shape preprecessing, a series of methods are used, such as moving trend. zero-average value shifting, deciding start-point and terminal-point, smoothing etc. this built solid foendation for abstracting better character next. In the target character abstracting algorithm, the modern spectral estimation algrithm, the modern spectral estimation is used to abstract model parameters, the method that model's class is decided is studyed deeply, the observation-method is used to salute in detail, the observation-method is used to salute it,and this creat condition for builting low-dimension character and real-time processing with. In the meantime,a method that use estimation PSD curve to recognite the aircraft'numbers is mentioned.In the light of the character, Using simple statictical pattern recognition method to identify the target.the results show that this method is very effective. In the meantime, the results that identifying the real target noise signal show that the method is simple, feasible, and have some practical values.