As the mining depth of coal resources increases, resulting in frequent mine earthquakes during mining. In this study, the rolling window ratio method is firstly chosen as the seismic phase recognition method to read the mine earthquake data received by the microseismic sensor. Secondly, the improved genetic algorithm is used as the optimization algorithm of the objective function to build the algorithmic framework of accurate inverse localization of mine earthquake. Finally, the accuracy of the algorithm for seismic source localization is verified based on actual engineering cases. Results show that the first arrival time extraction by the rolling window ratio method has the advantages of high accuracy and fast algorithm operation speed. The Fast Fourier Transform-Butterworth joint noise reduction method has a good noise reduction effect, which successfully suppresses the noise outside the mine earthquake signal and effectively improves the problem of much noise in the mine earthquake signal. Compared with the microseismic monitoring date, the mine earthquake localization error is within 5%.
Read full abstract