Method and device applied to laboratory seismic wave prediction

The invention discloses a method and a device applied to laboratory seismic wave prediction. The method comprises the following steps of extracting continuous seismic wave time sequence signals generated in the laboratory, and performing oversampling on the continuous seismic wave time sequence signals to obtain a continuous seismic wave time sequence signal sample, preprocessing the continuous seismic wave time sequence signal sample to obtain a first processing signal, extracting effective features of seismic waves, statistical features of the seismic waves and basic features of the seismicwaves in the first processing signal, carrying out lifting calculation on the effective features of the seismic waves, the statistical features of the seismic waves and the basic features of the seismic waves, and then fusing a built CNN convolutional neural network model to carry out prediction, outputting the predicted remaining time of the laboratory before the next earthquake. Laboratory datais helped to move to an ac