Micro-seismic data noise reduction method based on VMD and time sequence prediction

• Main Authors: MENG FANBIN, LUO SHILING, REN CHUAN, JIN XUELIANG, ZHOU GUANQUN, HU ZEAN, WANG BO, LI NING, ZENG LINFENG, LU XINGZHOU, WANG YAFEI
• Format: Patent
• Language: Chinese; English
• Published: 04.06.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a micro-seismic data noise reduction method based on VMD and time sequence prediction. The method comprises the following steps: carrying out full-time acquisition in a region needing to be monitored by utilizing each geophone; then microseismic data collected by all geophones in the same day are obtained, whether a microseismic event exists or not is judged in an amplitude sum mode, and the microseismic data are divided into a signal and noise mixed time period or a pure noise time period through a specific method; performing VMD decomposition on the pure noise time period, and training an LSTM neural network by using decomposition data to obtain a model for predicting environmental noise; subsequently selecting micro-seismic data containing a micro-seismic event, inputting a time period before the micro-seismic event occurs into the model, so that the model can predict noise data in the time period of the micro-seismic event, and finally subtracting the noise data obtained by subtrac