A Thunderstorm Cloud Point Charge Localization Method Based on CEEMDAN and SG Filtering

The current thunderstorm monitoring methods ignore the nonlinear and non-stationary characteristics of atmospheric electric field signals, which has a negative effect on results. Based complementary ensemble empirical mode decomposition with adaptive noise Savitzky-Golay filtering (CEEMDAN-SG), point charge localization method for cloud is proposed. After CEEMDAN used to decompose signal into series intrinsic function (IMF) components, reconstructed after SG those noise-dominant components. Then, correction. By changing samples, order SNR CEEMDAN-SG, CEEMDAN, etc. are compared, performance analyzed. Experiments show that compared before reconstruction, reconstruction improved by about 3%. At same time, results can match radar chart well time scale. using CNN-LSTM network model, it found original signal, amplitude absolute error larger, so be more displayed. This once again proves effects.