A signal de-noising method for gas switch discharge based on EMD and energy ratio

In order to solve signal de-noising problem in gas switch discharge, an EMD-based de-noising method with energy ratio is proposed. Aiming at the end effect shortcoming of the traditional empirical mode decomposition (EMD), an idea of screening and extension is introduced according to the boundary feature of discharge current signal, and then the intrinsic mode functions (IMF) are acquired by the improved EMD. The energy ratio of each IMF is selected as the signal feature parameter with transient features of gas switch discharge, through analyzing frequency spectrum and energy ratio of the current signal under different conditions, the signal noise is recognized and filtered effectively, and then the de-noising signals were reconstructed with the rest effective IMF components. The result of the statics experiment of gas switch discharge verifies the effective and feasibility of the proposed method, which provides a reliable theory for engineering applications of Marx generator.