Cause Analysis Of Gas Explosion Accident Based On Accimap

The main accidents in China’s coal mining include gas accidents,roof accidents,water damage,fire,transportation,mechanical and electrical blasting,etc.Although the frequency of gas explosion is not as high as other accidents,it is one of the most destructive accidents in coal mining.The accident affected a large range,and the average number of deaths in each accident ranks the top,casualties and economic losses are huge,with unexpected,predictable,extreme impact and other typical characteristics.For gas explosion accidents,people hope to find the cause of the event and analyze the relationship between the causes.From relying on traditional experience and subjective decision-making to making model data based on the new information generated by the interaction of conditions and changes in the coal mine gas accident system.The data set in this thesis is from The National Mine Safety Administration and the coal mine safety net.Typical gas explosion accident reports are selected and the cause of the accident is sorted out by text extraction method.Co-word analysis was introduced to determine the correlation between the causative factors and construct the social network of causative factors by using social network analysis as quantitative analysis method.Then the model is conceptualized into six levels: government policy and law,regulatory agency,company management,technology and business management,physical process and participant activities,equipment and surrounding environment.According to the correlation among the causative factors obtained on the basis of quantitative analysis of social network,the strong correlation causative factors are connected by arrows,and Accimap gas explosion accident map is established.The partial Accimap causality diagram of gas explosion accidents,as the structural learning part of Bayesian network,analyzes the influence of different states of each cause factor,and further constructs the Bayesian network of gas explosion accidents.Then the parameters of the Bayesian network are determined based on data and parameter learning methods,the sensitivity analysis and probabilistic inference are carried out to verify the Bayesian network,to Find out the key cause of gas explosion accident and determine the gas accident control strategy under different conditions.In view of the gas explosion accidents in various regions of China,according to the different environmental conditions,equipment configuration,personnel involved,etc.,the corresponding control methods of gas explosion accidents are put forward,and corresponding solutions are put forward for different levels of gas explosion accidents.Taking the coal mines in Henan as the research area and case analysis background,the coal mine hidden danger data is input into the Bayesian network model to realize the analysis of gas explosion accident to verify the effectiveness of the proposed method.Compared with the actual coal mine safety situation,the rationality of the proposed scheme is verified.