A quantitative Fuzzy Causal Model for hazard analysis of man–machine-environment system

• A Fuzzy Causal Model is developed and the analytical solution for the model is put forward.
• FCA can convey uncertainty relations between events for man–machine-environment system.
• The computation result of FCA presents the probability of consequent event with confidence level.
• FCA can get exact membership function curves of computation results compared with FFT.

There are few quantitative analytical models for system safety which considering not only the probability interval of events occurrence but also the influencing degree of incidence relations between different events at present. In this paper, a Fuzzy Causal Model (FCA) for man–machine-environment system hazard analysis which embodying the both as above is brought forward and discussed and applied into hoisting operation superfluities risk computation. Primary achievements are obtained as followed:A FCA based on fuzzy numbers is constructed, and computation analytic solution for the model on account of fuzzy number is put forward, in addition, the computation results are compared with those of α-cut set simplified computation method and Fuzzy Fault Tree (FFT), which shows computation scientificalness and accuracy of the FCA. Compared to FFT, FCA can better to describe fuzzy logic relations of events which lead to accidents; while compared to Bayesian Network, the inputs and outputs of FCA are defined as fuzzy numbers which can reflect data uncertainty and keep fuzzy information of data. Not only Accident occurrence possibility fuzzy values but also their membership degrees are computed based on the models which may give richer, more useable information for risk management of accidents. FCA could be applicable to safety assessment of system which implicates man–machine-environment interfering factors especially.