An Efficient Compression Method for Lightning Electromagnetic Pulse Signal Based on Convolutional Neural Network and Autoencoder

• DOI: 10.3390/s23083908
• Publication Date: 2023-04-12T06:57:08Z
• Authors: Jinhua Guo; Jiaquan Wang; Fang Xiao; Xiao Zhou; Yongsheng Liu; Qiming Ma
• Keywords: Autoencoder; Lightning; Pulse compression; SIGNAL (programming language)

Advances in technology have facilitated the development of lightning research and data processing. The electromagnetic pulse signals emitted by (LEMP) can be collected very low frequency (VLF)/low (LF) instruments real time. storage transmission obtained is a crucial link, good compression method improve efficiency this process. In paper, convolutional stack autoencoder (LCSAE) model for compressing LEMP was designed, which converts into low-dimensional feature vectors through encoder part reconstructs waveform decoder part. Finally, we investigated performance LCSAE under different ratios. results show that positively correlated with minimum neural network extraction model. When compressed 64, average coefficient determination R2 reconstructed original reach 96.7%. It effectively solve problem regarding sensor remote transmission.