SFEP中各因素重要度及未知因素数值的确定方法研究  

Research on determination method of importance of factors and value of unknown factor in SFEP

作　　者：崔铁军 李莎莎 CUI Tiejun;LI Shasha(School of Environmental and Chemical Engineering,Shenyang Ligong University,Shenyang Liaoning 110159,China)

机构地区：[1]沈阳理工大学环境与化学工程学院,辽宁沈阳110159

出　　处：《工业安全与环保》2024年第12期20-25,共6页Industrial Safety and Environmental Protection

基　　金：国家自然科学基金(52004120)。

摘　　要：为研究系统故障演化过程(System fault evolution process,SFEP)中各因素重要度,特别是未知因素的数值和重要度,提出一种基于神经网络的确定方法。阐述了SFEP与因素的关系,解释了未知因素的含义和作用,并构建了因素数值和重要度确定方法。研究认为已知因素和未知因素共同影响了SFEP,除已知因素外的所有因素都可合成一个未知因素。以未知因素数值枚举方式得到多组训练数据,根据最优网络判断标准确定最优训练神经网络,得到未知因素最优数值。根据最优未知因素数值再进行训练得到所有因素的重要度。将所有因素划分为影响SFEP的正负作用因素,这些因素的作用是相反的。实例分析展示了方法的具体过程,得到了定性和定量结果,最终确定了因素重要度及未知因素数值。研究结果可为研究系统故障和安全问题时的因素未知情况提供分析方法。To study the importance of factors in the system fault evolution process(SFEP),especially the value and importance of unknown factor,a determination method based on neural network is proposed.The paper expounds the relationship between SFEP and factors;Explain the meaning and function of unknown factor;A method for determining the value and importance of factor is constructed.The research shows that known factors and unknown factors jointly affect SFEP,and all factors except known factors can be synthesized into an unknown factor.Multiple groups of training data are obtained by enumerating the values of unknown factor,and the optimal training neural network is determined according to the optimal network judgment standard to obtain the optimal value of the unknown factor.The importance of all factors is obtained by training according to the optimal value of unknown factor.All factors are divided into positive and negative factors that affect SFEP,and the effects of these factors are opposite.Example analysis shows the specific process of the method,obtains qualitative and quantitative results,and finally determines the importance of factors and the value of unknown factor.This paper can provide an analysis method for the unknown factors when studying system faults and safety problems.

关 键 词：安全系统学 神经网络 未知因素 因素数值 重要度 确定方法 

分 类 号：X913[环境科学与工程—安全科学]