基于主因子模型的船用燃气轮机监测参数优化

Measure Parameter Selection for Marine Gas Turbine Based on Main Factors Model

针对燃气轮机气路故障模型存在多重共线性,使用主成分方法分析了主因子模型中故障系数矩阵的特性,并研究了多重共线性对诊断准确性的影响,结合敏感性分析,优化选择了测量参数。结果表明:虽然选择的测量参数数目少于故障因子数目,但故障诊断效果较好,能有效辨别相似故障。

Considering the problem of multicollinearity in gas path fault model of Gas turbine , the method of principal component anal-ysis （ PCA） is used to study the characteristic of fault coefficient matrix of main factors model and research the impact of multicollinear -ity on diagnostic accuracy , combined with sensitivity analysis , optimize the selection of measurement parameters .Results show that the number of selected measurement parameters sequence is less than the number of failure factors , but these parameters supply a better fault diagnosis , which can effectively identify similar faults .