Study on the improvement of the microseismic network configuration for San He-jian coal mine

摘要

Two seismically active regions and also the occurrence probabilities of strong mining-induced tremors of them are determined by the comprehensive index method to improve the monitoring efficiency of seismological observation system in San He-jian coal mine. The set consisted of 91 candidate points is formed based on the general choosing candidate points of geophone stations. To this kind of large-scale network configuration of combinatorial optimization problem in coal mine, natural number coding method is used to construct the individual genes expression of microseismic network configuration plan in genetic algorithm and the individual fitness function is built to evaluate the performance of configuration plan based on the theory of D-optimal design. Crossover and mutation operators are mixed and improved to prevent the premature convergence of algorithm thus increase its performance. The application results from situation show that: (i) the optimum solution can be quickly solved by the genetic algorithm;(ii) the improved network configuration is better than the currently used one, which is proved by using numerical emulation technique, because it significantly decrease the hypocenter location errors in the key monitoring areas, and the maximum error is reduced by 230m.