基于HAZOP分析和贝叶斯网络的压缩机故障定位方法研究

Research on Fault Location Method of Compressor Based on HAZOP Analysis and Bayesian Network

燃气式离心压缩机结构复杂、子系统较多,引起故障的原因较为复杂、类型多样,查找难度较大。深入研究危险与可操作性(HAZOP)分析和贝叶斯网络原理,建立一种基于HAZOP分析和贝叶斯网络的压缩机故障定位方法。该方法可对某压气站压缩机进行故障定位,计算出引起故障的原因的概率,对概率最大的原因进行检查,准确、快速地找到故障点。该方法可推广运用于离心式泵机组等设备的故障快速定位,可提高故障诊断效率。

Gas-fired centrifugal compressors have complex structures and more subsystems,which cause more complex and diverse types of faults and are more difficult to find.In-depth study of hazard and operability(HAZOP)analysis and Bayesian network principle,to establish a compressor fault location method based on HAZOP analysis and Bayesian network,The method can realize the fault location of the compressor of a pressure station,calculate the probability of the cause of the failure,check the cause of the greatest probability to and find the fault point accurately and quickly.The method can be extended and applied to the rapid fault location of centrifugal pump units and other equipment,which can improve the fault diagnosis efficiency.