船用核动力装置的Bayesian融合诊断方法研究

Study of Bayesian fusion method of fault diagnosis in marine nuclear power plants

船用核动力装置发生故障时很多参数会发生变化,且参数的变化速度不同,因此有必要对这些参数进行融合处理,得到用于故障诊断的可靠征兆.在得到可靠征兆的基础上,利用基于Bayesian理论的融合诊断方法对故障进行诊断,随着故障现象的逐渐出现,诊断系统能够准确地判断故障类型.将该诊断方法应用于船用核动力装置的失水事故的诊断,利用安全分析得到充分性因子(Ls),用Bayesian公式进行融合计算,最终能够得到可靠的诊断结果,满足了船用核动力装置的故障诊断要求,具有一定的实际应用价值.

Many parameters will change when marine nuclear power plants have faults, and they change in different speed. So it is necessary to fuse the parameters in order to gain the credible features of diagnosing the faults. Based on the credible features, Bayesian fusion method is applied to fault diagnosis. With emergence of the fault phenomena,the fault diagnosis system can judge the type of fault exactly. The fault diagnosis method is applied to diagnose the LOCA fault of marine nuclear power plants. After getting the Ls through safety analysis, and fusion calculations using Bayesion formula,the accurate diagnosis results can be obtained. The fault diagnosis method can satisfy the demands for marine nuclear power plants fault diagnosis, and it is practical.