基于多态模糊贝叶斯网络的高校实验室安全致险因子诊断与防范策略

Diagnosis and prevention strategy of safety risk-causing factors in university laboratories based on polymorphic fuzzy Bayesian network

摘要:

致险因子诊断是防控高校实验室危险源的重要关口，为解决已有研究在致险因子识别完备性、诊断精度与效率上的不足，基于扎根理论编码技术对33例案例进行事故成因解构，萃取出23个主要致险因子；针对高校实验室风险的动态性、多态性和模糊性特征，构建多态模糊贝叶斯网络模型，利用其双向推理和重要度分析技术进行致险因子诊断，研究结果表明：安全意识薄弱、职责不清、安全检查与隐患整改不到位等为关键致险因子，应对其进行重点管控与治理；以某高校实验室安全事故为实例进行方法应用，验证了该方法的科学性与有效性；立足模型研究结果与我国高校实验室建设现状提出科学的风险防范策略，为提升高校实验室安全管理水平提供支撑。

Abstract:

Diagnosis of risk-causing factors is an important gate for the prevention and control of hazardous sources in university laboratories, in order to solve the deficiencies of existing research in the completeness of risk-causing factor identification, diagnostic accuracy and efficiency, based on the coding technique of rooted theory, the cause of the accident is deconstructed in 33 cases, and 23 main risk-causing factors are extracted; in view of the dynamics, polymorphism, and ambiguity of the risk of university laboratories, the polymorphic fuzzy Bayes network model is constructed, and its bidirectional inference and importance analysis techniques are used to diagnose risk-causing factors. In order to diagnose the risk-causing factors by using its bidirectional reasoning and importance analysis technology, the research results show that: weak safety awareness, unclear responsibilities, safety inspections and hidden dangers are not in place for the key risk-causing factors, which should be focused on control and management; a university laboratory safety accidents as an example of the application of the method, to validate the scientific and effectiveness of the method; based on the results of the model and the current situation of China's university laboratory construction to put forward scientific risk prevention strategies. Based on the results of the model research and the current situation of China's university laboratory construction, scientific risk prevention strategies are proposed to provide support for improving the safety management level of university laboratories.