燃气轮机气路测量参数的模糊智能选择

针对燃气轮机气路性能的监测诊断需求,基于气路小偏差分析得到的影响系数矩阵,首先从测量参数选择的相关性要求出发,应用模糊聚类分析方法得到了要求数目的目标聚类;进一步从参数选择的敏感性等要求出发,通过模糊综合评价方法从多元素聚类中筛选出了各自的代表测量参数,从而实现了燃气轮机气路测量参数的模糊智能选择.实例表明:利用模糊聚类分析方法可直接实现测量参数的相关性选择,并可以方便可靠地根据目标要求选择合理的测量参数组合,利用模糊综合评价方法的测量参数的敏感性选择更为有效可靠.

Fuzzy intelligent selection of gas turbine gas path measurement parameters

According to the needs of gas turbine gas path performance monitoring and diagnosis, based on the influence coefficient matrix by gas path small deviation analysis, the measurement parameters were classified into target clustering with required number by fuzzy clustering analysis method starting from the correlation requirements of parameters selection. Furthermore, from the sensitivity requirements of parameters selection, the representative measurement parameters of each multi-element clustering were picked out by fuzzy comprehensive evaluation method. Thereby the fuzzy intelligent selection of gas turbine gas path measurement parameters was realized. The case study shows that fuzzy clustering analysis method can carry out the parameter correlation selection directly, so it is reliable and convenient to select the reasonable measurement parameter group for a target request; it's also more efficient and reliable for the measurement parameter sensitivity selection with fuzzy comprehensive evaluation method.