Research on Detection and Material Identification of Particles in the Aerospace Power

The aerospace power is widely used in the aerospace system. Its reliability directly affects the safety of the whole system. However, the particles generated in the production process usually cause failures to the aerospace power. In this paper, a novel automatic detection method for particles in the aerospace power is proposed based on Particle Impact Noise Detection (PIND) test. Firstly, stochastic resonance algorithm is presented to detect the existence of tiny particles. Secondly, in order to obtain the sources of particles, wavelet packet transform is used to extract energy distribution vectors of different material particles, and Learning Vector Quantization (LVQ) network is brought in for material identification of particles. Finally, the results indicate that the accuracy meets the requirements of practical application.