离心压缩机组的模糊诊断决策树

A Fuzzy Diagnosis Decision Tree For Centrifuge Compressor Units

离心压缩机组是大型化工厂的关键设备,其正常运转与故障的迅速诊断对企业生产具有重大意义。文章主要研究利用模糊诊断决策树对离心压缩机组的故障现象进行迅速诊断。文中建立离心压缩机组模糊诊断决策树,针对决策空间中无序分布决策类提出了新的推理策略,并重新定义了多类别问题的信息熵计算公式。工程实践检验结果证明了该算法的有效性。

The centrifuge compressor units are critical assets for large scale chemical factories. Having it operating properly and performing fault diagnosis swiftly is of crucial importance to the production processes of the industry.？ This paper proposed an innovative methodology which utilizes fuzzy diagnosis decision trees to achieve prompt fault diagnosis for the centrifuge compressor units. The paper elaborates the establishment of a fuzzy diagnosis decision tree for the centrifuge compressor unit; proposes a new reasoning strategy based on the random distribution of decision space, and re-defines the calculation of comentropy for the multi-categories classifications. The proposed methodology has been verified through engineering practices, and the results confirmed the validity of the methodology.