基于定性仿真和模糊知识的离心式压缩机排气量不足原因诊断

Qualitative Simulation and Fuzzy Knowledge Based Fault Diagnosis of Centrifugal Compressor Insufficient Discharge

针对引起离心式压缩机排气量不足的原因,本文提出一种基于定性仿真和模糊知识的诊断方法.利用压缩机结构原理和故障机理等定性知识建立故障定性模型库,并推理得到压缩机排气量受不同因素影响时的定性规则库.对系统变量的观测值利用定性趋势提取和模糊化进行定性化处理.提出了基于变量定性趋势和模糊定性值约束的滑动窗口加权匹配策略,并根据匹配结果诊断出导致排气量不足的原因.最后通过两个实例仿真验证了所提方法的有效性.

In order to diagnose the causes of insufficient discharge of centrifugal compressor, a qualitative simulation and fuzzy knowledge based diagnosis method is proposed in this paper. Qualitative models are built according to the qualitative knowledge derived from structure principle and faults mechanism of the centrifugal compressor, and a qualitative rules library is obtained by inference to describe how the discharge flow is influenced by various factors. The online observations of variables are qualitatively processed by qualitative trends extraction and fuzzy approach. A weighted sliding window match strategy based on variables qualitative trends and fuzzy qualitative value constraint is proposed for qualitative state matching, and the matching results is used to diagnose the reason of insufficient discharge of the compressor. Finally,the effectiveness of the proposed method is verified through simulation of two fault conditions.