Research And Implementation Of Microseismic Monitoring And Positioning Correction System For Coal Mine Rockburst

As the mining depth of China’s coal resources increases year by year,the frequency of the resulting accidents,such as coal mine impact ground pressure,increases year by year,and these accidental disasters threaten the safe mining of coal resources.Microseismic monitoring in coal mines is a common technique and means.Compared with earthquakes,coal mine microseismic monitoring has the characteristics of fast response,external source,easy cause damage to underground roadways and ground vibration.Efficient microseismic monitoring systems and accurate positioning algorithms have become critical issues that must be addressed urgently.The traditional microseismic monitoring system usually relies on manual operation.It mainly analyzes the low-dimensional characteristics of microseismic waveform signals,which have low positioning accuracy and require much time and effort.Using computer technology instead of traditional methods to monitor impact ground pressure dynamic hazards has become a significant development trend in the coal mining industry.Therefore,this paper proposes the Transformer coal mine microseismic localization algorithm based on conditional adversarial augmentation and designs and implements a coal mine impact ground pressure microseismic monitoring and localization correction system.The system is divided into a microseismic monitoring module,microseismic event report generation module,microseismic event query module,microseismic event location correction module and parameter configuration module.In the microseismic event localization correction module,the Transformer coal mine microseismic localization algorithm based on conditional adversarial enhancement is proposed for the microseismic event source localization problem in the coal mine microseismic monitoring system.The method firstly learns the existing microseismic signals through a conditional adversarial enhancement layer;secondly,it converts the microseismic signals into feature data using the Transformer encoder layer and then further learns the deep-level features and complex inter-station dependencies of the microseismic signals using its attention mechanism,and also uses Gaussian distributed random variables to offset the influence of different geological conditions on the localization accuracy;finally,it obtains the high-density data by introducing a mixed density output layer to get Gaussian distribution parameters and calculate the optimal earthquake source location.The experimental results on a mining dataset in Chile and Liaoning verify the method’s effectiveness.The results show that both the epicentre error and the source error obtained by the method are better than other methods,and the localization error is reduced by 38% and 12% in the two datasets,respectively,which achieves the purpose of improving the source localization accuracy and the robustness of the localization model.The algorithm proposed in this paper has improved accuracy and noise immunity compared with the traditional algorithm,which is of great practical significance for studying microseismic in coal mines.The research in this paper is about the design and implementation of a microseismic monitoring and localization correction system for coal mine impact ground pressure.The system is developed using C/S architecture and relevant development languages and frameworks like Java,Java FX and My SQL.In order to ensure the stability of the system operation,this study uses multi-threading,high concurrency and other related techniques to improve the system running speed.