Research On Prediction And Early Warning Of Coal Mine Gas Risk Situation

Coal occupies a pivotal position in my country's energy structure,and China's demand for coal will remain huge for a long time to come.The occurrence of coal mine accidents will seriously affect the normal production of coal mines and the safety of people's lives and properties.Among them,gas accidents are one of the most dangerous and harmful coal mine accidents.Therefore,the research on pre-control of gas accidents has important theoretical and application value.Under the general trend of intelligent transformation of coal mines,many coal mining companies have successively upgraded the mine information infrastructure and built a complete safety monitoring system that can monitor and record underground information in real time.Based on the massive data in the underground safety monitoring system,this paper introduces the concept of situation into the field of coal mines,and proposes a coal mine safety situation prediction and early warning research method based on the risk of gas accidents.The main work and contributions of this paper are as follows.Clarified the definition of coal mine safety situation.The coal mine safety situation reflects the current safety state of the entire coal mine production system and how this state will change in the future through the risk of gas explosion.First,starting from the cause of the gas explosion,combined with accident cases and field experts' experience,the situation characteristic indexes are extracted,and the coal mine safety situation risk index system is constructed,and the situation indexes in the system are specifically quantified and the risk level regional division is carried out.Then,a situation assessment model based on FCE-AHP is constructed,K-th parabolic membership function is used to evaluate the situation,and the weight of each situation characteristic index is determined by the analytic hierarchy process.The prediction of the security situation is transformed into the prediction of the characteristics of the situation.A time series forecast model of situation characteristic indicators based on autoregressive difference moving average method and recurrent neural network is constructed,and corresponding forecast models are used to predict situation characteristic indexes according to different sample situations.The predicted values of characteristic indicators are integrated to obtain the final coal mine safety situation forecast value,and a corresponding safety situation early warning system is established from the perspectives of single indicator early warning and overall comprehensive early warning.Take the historical data of a coal mine in Shanxi as a sample for case analysis.The K-nearest neighbor method is used to process the missing values of the data,and the original data is denoised by the wavelet threshold noise reduction algorithm,and the final cleaned data sample is obtained for situation prediction.Analyzing the results of the situation prediction,it can be found that the coal mine safety situation prediction results based on the risk of gas accidents are consistent with the actual conditions of the coal mine.At the same time,it has high accuracy and can provide auxiliary decision support for the safety production management of the coal mine.The thesis has 29 pictures,20 tables,and 77 references.