Depth learning recognition method of rotor wing type unmanned aerial vehicle based on fractional order domain features

The invention relates to a depth learning recognition method of a rotor wing type unmanned aerial vehicle based on fractional order domain features, belongs to the technical field of radar target classification recognition, and solves the technical problem of poor target classification recognition probability. According to the method, the short-time fractional Fourier transform is used for performing primary feature extraction on target echo signals of the rotor wing type unmanned aerial vehicle; and then on the basis, self-coding depth recognition learning is performed in three dimensions ofa u-v domain, a u-z domain and a v-z domain. The result shows that the target recognition rate of the depth learning recognition method can reach 87 percent, and is much higher than the recognition rate obtained by directly using a depth learning algorithm. Meanwhile, the depth learning recognition method has a small data processing amount, and realizes good fusion application on two conventionalmethods including a featur