基于事故致因和分层感知的安全风险预警系统研究

Research on risk management and control system based on hierarchical perception and accident cause

提出基于分层感知和事故致因要素的分层感知模型、交叉识别模型及分类学习算法。利用基于AI的热网安全风险预警系统验证了分层感知和交叉识别模型的可行性,推导出基于分层感知和事故致因要素的数学模型和风险分级策略,实现了事故致因要素理论的逆向应用,且朴素贝叶斯等机器学习算法适应性良好。

A hierarchical perception model,a cross-identification model and a classification learning algorithm based on hierarchical perception and accident causal factors are proposed.The feasibility of the hierarchical perception and cross-recognition models is verified using an AI-based thermal network safety risk warning system,and the mathema-tical model and risk classification strategy based on hierarchical perception and accident causal factors are derived.The inverse application of accident causal factors theory is realized,and machine learning algorithms such as plain Bayes are well adapted.