Flight Target Recognition Method Based on RCS Sequenceed

Aiming at the problem of low target recognition rate in traditional methods, this paper presents a flight target recognition method based on radar cross section area (RCS) sequence. Firstly, the dynamic RCS sequence data of flight target is generated by 3D modeling, electromagnetic calculation, flight path, conversion of target in radar line of sight angle and RCS sequence noise simulation. Then, the feature extraction and target recognition of RCS sequences are carried out by using the convolution layer, activation layer, lower adoption layer, random discard layer, full connection layer and classification layer of one-dimensional convolutional neural network (CNN). Finally, the simulation results show that the recognition performance of the proposed method is better than the traditional recognition methods such as SVM and KNN.