Based on Wavelet Packet Fractals and PSO-LSSVM Nuclear Explosion and Lightning Recognition
Xin Dong, Xinbo He, Xudong Han, Shu Zhang, Zhaomin Li, Bing Wei · 2024
The electromagnetic pulses (EMP) from nuclear explosions and lightning are non-stationary and nonlinear. To effectively identify nuclear explosion EMP (NEMP) and lightning EMP (LEMP), we propose a least squares support vector machine optimized by particle swarm optimization (PSO-LSSVM), combining wavelet packet fractal analysis and particle swarm optimization. Wavelet packet fractal analysis extracts features and calculates the fractal box dimension of each wavelet packet layer. The PSO-LSSVM classifier uses these dimensions to classify the signals. Experimental results show that the fractal box dimension of the 2-layer wavelet packet decomposition effectively distinguishes NEMP from LEMP, achieving a recognition rate of 95.25%.