Multi-source Information Fusion Forecasting Method For Coal And Gas Outburst Risk

Coal and gas outburst is one of the most serious disasters in coal mine safety production.The occurrence mechanism of coal and gas outburst is complex,and the prediction methods based on various sensitive parameters are often inaccurate and fuzzy.Based on the main problems existing in the prevention of coal and gas outburst and the development direction of monitoring and early warning technology innovation,this paper analyzes the measurement of coal and gas outburst parameters and sensitive indicators,studies the prediction method of gas emission,studies the risk identification and monitoring and early warning method of coal mine gas outburst based on intelligent algorithm,in order to realize on-line monitoring and judgment of coal and gas outburst hazards.Identify and predict the technical path of early warning.This thesis analyzes the sensitive indexes of coal and gas outburst disasters in the process of occurrence and development,including static sensitive parameters and dynamic sensitive parameters.On this basis,considering that there are many influencing factors of coal and gas outburst and the role and influence degree of different factors in the process of occurrence and evolution of outburst disasters have both similarities and obvious differences,the method based on correlation gain is adopted.By improving the gray correlation method,and through the gray correlation analysis based on the entropy correlation degree gain method,the weight expression of each characterization index is obtained,so as to determine the nine main characterization factors that affect the coal and gas outburst,reduce the information redundancy,and achieve the extraction of the most effective characterization and identification of the gas outburst disaster from the multi-source gas outburst disaster multi parameter information characteristics.The main factors index acquisition and detection prediction methods are studied.The measurement method and characterization function of gas basic index parameters,acoustic emission signal and electromagnetic radiation signal are given.For the process that can not be modeled,the prediction method of gas emission based on improved universal gravitation algorithm is studied with gas emission prediction as the research object.The test results show that the absolute gas emission error of the prediction result is within 0.1m3/min,and the prediction accuracy is relatively high.The continuous monitoring signal based on the gas concentration contains traces and characteristic information of the dynamic evolution process of the system.The multi sparse echo state network model is applied to the prediction of the gas concentration time series.The test results show that the RMS error of the proposed method is small and it can accurately express the actual situation of the gas concentration change.In order to effectively integrate multi-source information and improve the comprehensive decision-making ability of multi-source information,this paper studies the decision-making fusion method of improved evidence theory,proposes the evidence body representation method based on the similarity measurement of fuzzy sets,introduces the similarity function of Vague sets to modify the evidence source,and modifies the evidence of multi-source indicators.The test results show that this method can effectively integrate multiple feature information,better deal with uncertainty and external interference,and improve the reliability of coal and gas outburst prediction.The risk early warning platform of coal and gas outburst disaster is developed.Based on the main sensitive indexes,the risk evaluation and early warning of coal and gas outburst disaster are carried out.The system function is perfect,which can improve the ability of preventing and controlling gas outburst disaster in coal mine and ensure the safety of production in coal mine.There are 44 pictures,30 tables and 222 references.