Application of Bayesian Network to the Assessment of Open Press Electrical Safety Accident Risk
Since multiple state events could not be described and accident probability could not be calculated by fault tree quantitatively for safety assessment, applications of Bayesian network (BN) in safety and reliability fields for the past few years were introduced and basic principle and inference algorithm of Bayesian network were presented. A Bayesian network was developed to model open press electric shock accident as a result of charged press enclosure. The multi-state nodes of the network were illustrated and accident probability was computed finally. The results show that Residual Current Operated Circuit-Breaker with Integral Over-current Protection (RCBO) installation can reduce accidents probability of phase-voltage and under-phase-voltage electric shock sharply for open press working in T -T system of low voltage power supply system. In a word, Bayesian network can model system which contains multiple state events; furthermore, accidents probability of that system can be computed correctly, in addition, it offers a comparative tool for various safety design of machine system to guarantee machine inherent safety.
Chen Wenying Niu Yingjian
State Key Laboratory of Explosion Science & Technology Beijing Institute of Technology Beijing, Chin College of Safety & Environmental Engineering Capital University of Economics & Business Beijing, Ch
国际会议
北京
英文
616-619
2011-09-23（万方平台首次上网日期，不代表论文的发表时间）