汽轮发电机组转子振动故障的模糊聚类分析

Fuzzy Set Analysis of Turbine Rotator Vibration Malfunction

针对汽轮发电机组多个振动故障与多个振动信号之间综合关系研究的不足，阐述了模糊聚类分析的理论、方法及聚类效果的判别原则，并将其应用于汽轮发电机组转子振动故障的诊断研究中．提出了合适的聚类参数，从而对多故障发生与多故障特征情况下的故障进行分层，指出应优先诊断的范围，从而使诊断工作量和诊断时间大为减少．理论计算与现场检查结果相符，并且与神经网络方法相比。

For the deficiency of the research on the relationship between multi malfunctions and multi portents, fuzzy set analysis theory and method were described and applied to the diagnosis of turbine rotator vibration malfunction. According to the fuzzy set judgement principle and the fuzzy set parameters proposed in this paper, the malfunction classification and prior diagnosis range among multi malfunctions and multi portents were presented. The fuzzy set calculation result is in good agreement with the actual inspection one. Compared to the neural network method, the diagnosis work and time are decreased greatly.