DEVICE of neural network target recognition a set of attributes

摘要

The invention relates to radar engineering and may be used during recognition of aerial targets using radar having a wide variety of probing signals. The purpose of the utility model is to increase the reliability of detection of air targets through the use of neural network classifier in the processing of expanding the number of features. The goal is achieved through the introduction of the known device N analog-to-digital converters and neural network classifier with the corresponding change interblock bonds. This allows you to use in recognizing the advantages of neural network classifier in the processing of the values ​​of different signs and thereby increase the probability of making the right decision in recognizing the class (type) of the target. The advantage of the device is its ability to adapt to work in different radar stations with a wide variety of probing signals.