Classification Method of Air Flying Target based on Neural Network

The traditional air flying target classification and recognition method is not universally applicable to different radar systems and has a low recognition rate. To solve this problem, this paper proposes a air flying target classification method based on principal component analysis and recurrent neural network (PCA-RNN). Firstly, in order to eliminate the impact of different dimensional data differences on the training model, this scheme preprocesses the track data of air flying targets collected by radar. Then, in order to accelerate the modeling speed and improve accuracy, this scheme uses principal component analysis to reduce the dimensionality of the obtained flight data. Finally, considering the time series of flight data, a recurrent neural network is used to establish a model for the data and classify the test data, thus achieving good classification results. Experimental results show that using this method for air flying target classification does not require additional hardware, and it can achieve classification of radar air flying targets such as light and small unmanned aerial vehicles, flying birds, helicopters, and civil aircraft in a single radar antenna cycle.