基于聚类分析与加权模糊逻辑的汽轮机组振动故障诊断方法研究

针对汽轮机组振动故障诊断中故障征兆的使用问题,提出了一种基于聚类分析与加权模糊逻辑相结合的故障诊断方法。利用振动的频谱特征对振动故障的几种常见故障模式进行分类,形成故障模式类,从而可以在故障模式类层次区分开属于不同性质的故障模式,解决类间的识别问题,进而缩小故障模式的识别范围。对于同一故障类中的故障模式,采集不同类型的故障征兆,利用粗糙集理论建立故障诊断决策表,提取对故障识别有贡献的故障征兆构建故障诊断规则,再应用知识依赖度为故障诊断规则的前提条件分配权重,克服了主观分配权重存在的不足,减少了故障诊断推理过程中的不确定性影响。再应用加权模糊逻辑对故障诊断规则进行推理,根据推理结果对故障模式进行识别。该方法既充分利用了振动的频谱特征这一重要故障征兆作为故障诊断的初步判断依据,又综合利用了反映故障不同方面信息的不同类型的故障征兆,从而做到更加准确地进行故障识别。

Study of the Methods for Diagnosing Vibration Faults of Steam Turbine Units Based on the Clustering Analysis and Weighted Fuzzy Logic

In the light of the problems in using fault signs to diagnose faults of a steam turbine unit,presented was a method for diagnosing faults by a combination of the clustering analysis and the weighted fuzzy logic.The frequency spectrum characteristics of various vibrations were employed to classify commonly-seen fault modes and form various categories of the fault modes,thus differentiating the fault modes in various natures according to their categories,solving the problem in identifying their categories and...