概率因果网络在汽轮机故障诊断中的应用

APPLICATIONS OF PROBABILISTIC CAUSAL NETWORK IN FAULT DIAGNOSIS OF TURBINE MACHINERY

在分析了汽轮机振动故障特点的基础上 ,提出了用遗传算法进行汽轮机等旋转机械故障诊断问题 ,定义了遗传算法求解故障诊断问题的概率因果网络 ,给出了求解故障诊断的数学表达式和适合汽轮机等旋转机械的故障集、征兆集、因果强度和先验概率表。建立了汽轮机故障诊断模型 ,指出表达式的最小值的集合对应于故障集和征兆集。以某汽轮机故障为例验证了该方法的有效性。该模型能有效地识别出汽轮机的多故障 ,弥补了专家系统和神经网络等诊断方法不能正确诊断多故障的不足。

Based on analysis of vibration fault of the turbine machinery, applications of genetic algorithm in fault diagnosis were researched, the probabilistic causal network of fault diagnosis was made. The mathematics equation to solve the problem of fault diagnosis, the fault sets、manifestation sets、relation intensity and experience probability of the rotating machinery, such as turbine machinery were given. The mathe matics model of fault diagnosis of turbine machinery by genetic algorithm was established, and the minimum value sets of formula corresponding to the relating fault and manifestation sets were put forwarded. The availability of this method was proved by a fault diagnosis examples of turbine machinery.The results show that the genetic diagnosis model proposed in this paper can be used for multi fault diagnosis together, it may make up shortage for some of expert systems and neural networks in some aspect. From the practice, this is a good way of fault diagnosis of rotating machinery.