Research on the correction method of fault event occurrence probability in the system fault evolution process
To correct fault event probability in the system fault evolution process,a correction method for fault event occurrence probability is proposed.Based on a combination of previous experience gathered in this field,a set of evaluation values for fault event occurrence probability under different factors,as previously analyzed by experts,are modified.These methods involve DS evidence theory,Pearson correlation,information entropy,and others.This paper discusses the relationship between the system fault evolution process and the fault event occurrence probability;the causes of fault event occurrence probability are then analyzed,after which a mathematical model for the fault event occurrence probability correction method is constructed.This method combines the evaluation values for fault event occurrence probability based on evaluations carried out by numerous experts in the field,using evidence theory to form a probability evaluation matrix for fault events.Pearson correlation is used to establish a correlation matrix for the occurrence of fault events;the information vector for fault event occurrence probability is then obtained using this correlation matrix.Using the evaluation matrix to obtain the information entropy vector of fault event occurrence probability,the information vector and information entropy vector can be synthesized to obtain a correction vector for fault event occurrence probability.Finally,the evaluation value for fault event occurrence probability is corrected to obtain a correction vector.The analytical process is then demonstrated using an example.Existing research shows that the fault event occurrence probability is the basis for obtaining a quantitative description of the system fault evolution process.The reason for the inaccuracy of this probability lies in methodology,system,and factors,as well as technical limitations.During the correction process,relevant coefficients such as the importance of evaluation data,the importance of probability fluctuation,the importance of expert opinion,and the importance of other factors can also be obtained.The research can thus provide an analytical method for the relevant parameters.
basic discipline of safety science and technologysystem fault evolution processprobability of failure eventintegration of expert experiencecorrection method
万方数据
维普
