Features Extraction and Recognition of Weak Ship-radiated Noise under Impulse Noise Environment

The noise reduction technology of underwater targets is continuously improved, and the target radiation noise is getting more and more weak. It is very important to find out the effective feature extraction and recognition method of target weak radiated noise in the reconnaissance and surveillance of the sea target and the implementation of precision strikes. In this paper, a recognition method based on fractional stochastic resonance is proposed to deal with the problem of weak ship-radiated noise under alpha stable noise distribution. Based on the fractional stochastic resonance algorithm, this method introduces the optimization parameters of artificial-fish swarm algorithm, effectively solves the optimization problem of fractional order, and then uses the power spectrum characteristics as the recognition feature, and uses the support vector machine classifier to classify and identify. Theoretical analysis and simulation experiments verify the effectiveness of the proposed algorithm.