Research On Human Factors Inference Of Safety Accidents In Coal Mine

Although great attention has been paid to safety issues in China, the security situation coalmine production is still grim at present. A plenty of research results show that human errors are the main contribute factor for coalmine accidents. However, the human factors, such as operators' mental and physiological statuses, management culture et al, are rarely depicted in the investigation reports of coalmine accidents. Consequently, the human factors analysis results are not accurate. Hence, it is very significant to analyze the deep human errors causing coal mine accidents.In order to analyze the main human factors causing coal mine accidents, the theoretical framework of the HFACS combined with the coal mine accidents reports of Fenxi Coalmine Group Co. are used to establish the graphical structure of Bayesian network of human errors. Furthermore, parameters of the graphical network are elicited by maximum likelihood estimation and fuzzy verbal-numerical probability estimation. And then, entropy method, sensitivity analysis and consistency check are used to test the reliability of established Bayesian network. Furthermore, the main factors that cause operators' unsafe acts are analyzed by Bayesian network inference. The results show that the poor state of personal readiness is the most fundamental factor causing operator's unsafe acts of coal mine, that the deep factors resulting in the operator' unsafe acts are are planned inappropriate operations, inadequate supervision and resource management, and that operator's status has a greatly impact on the unsafe acts. Finally, a coalmine accident of Fenxi Coalmine Group Co. is used test the reliability of Bayesian network. Overall, to improve the human factors analysis of coal mine accidents reports and to provide recommendations for prevention of coal mine accidents.