Research And Implementation Of Mine Microseismic Event Magnitude Calculation And Identification System

Coal mine microseism is a dynamic phenomenon of rock mass fracture in the process of mining,which affects the production of coal mine.With the increase of mining area,the number of microseismic disasters in coal mine is increasing.Although there are many analyses on coal mine at home and abroad,the magnitude calculation and identification analysis of coal mine microseismic events are less under the coal mine microseismic monitoring system.Coal mine rock burst,gas outburst and other disasters and accidents are more frequent.Microseismic monitoring is a more commonly used monitoring method.The magnitude and energy of microseismic monitoring events is an important measure of coal mine disasters and accidents.It is particularly important for coal mine production and the formulation of the recovery plan after the accident.The accuracy of the existing magnitude and energy calculation methods needs to be improved The accuracy of microseismic event identification based on the characteristics of P wave and S wave needs to be improved.This paper studies and realizes the calculation and identification system of microseismic event magnitude in coal mine.The B / S architecture is adopted to realize the four functions of the system,including the data preprocessing module of microseismic data in coal mine,the calculation module of microseismic event magnitude in coal mine,the identification module of microseismic event in coal mine and the result display module.The system has been applied in a mining area in Liaoning Province.The pre-processing module of microseismic data mainly includes data reading module,time picking module of large and small energy ratio method and storage module of microseismic events in coal mine.The calculation module of microseismic event magnitude in coal mine mainly includes the calculation of near earthquake magnitude,duration magnitude,unified magnitude and optimized energy calculation module.The magnitude of near earthquake is calculated according to the maximum amplitude of microseismic signal data.Duration magnitude is used to calculate the microseismic signal whose vibration attenuation is equal to twice the variance of background noise.The unified magnitude is calculated according to the characteristics of microseismic surface wave.In order to further improve the accuracy of microseismic event magnitude calculation,this module proposes and implements the EMD energy optimization algorithm based on wavelet de-noising(Wavelet Empirical Mode Decomposition).The algorithm uses wavelet de-noising algorithm to de-noising the interference noise in the microseismic signal,and uses the de-noising microseismic signal for EMD energy calculation,and finally realizes the energy optimization.In the microseismic event recognition module,an on-line recognition algorithm of coal mine microseismic events based on improved OS-ELM is proposed and implemented(Microseism Online Series-Extreme Learning Machine).Firstly,the original microseismic data is preprocessed by using extreme value connection dimension reduction.The preprocessed microseismic data is used as the input of OSELM model,and the output weight obtained from OS-ELM training is used as the next update information to establish the OS-ELM classifier model,Then,the sliding window mechanism is added to realize the online batch training and prediction of massive microseismic signals,so as to improve the speed and accuracy of training and prediction.The experimental results show that the improved OS-ELM algorithm can not only maintain good training and recognition accuracy,but also greatly improve the operation speed.The results display module realizes the results display of magnitude calculation module and microseismic event recognition module,which makes the results more intuitive.The experiment shows that the Improved EMD energy optimization algorithm can achieve the purpose of optimizing energy,improve the energy calculation accuracy of the system,improve the OS elm on-line identification algorithm of coal mine micro earthquake events,not only can maintain a good training and recognition accuracy,but also greatly improve the operation speed.