摘要
针对重型燃气轮机缺乏有效综合性能评价方法的问题,提出了一种基于工况辨识的重型燃气轮机性能评价方法。根据重型燃气轮机效率的影响因素确定了机组性能综合评价特征参数;选取机组多工况运行历史数据进行稳态筛选,并对筛选出的稳态运行数据进行基于多步K-均值聚类的工况辨识,实现机组稳态运行空间的工况划分;采用多元高斯混合模型确定各个工况下性能评价特征参数的基准值和阈值;最后建立基于熵权法的机组性能评价模型,实现了机组的综合性能评价。实例证明了所提方法的可行有效性。
In view of the lack of effective performance evaluation methods for heavy-duty gas turbines,a method of performance evaluation is proposed for heavy-duty gas turbines based on condition identification. Firstly,the characteristic parameters of synthetical performance evaluation are determined according to the influencing factors of gas turbine efficiency. Secondly,full region operation history data are screened,and the selected steady state operation data are identified based on multi-step K-means clustering to realize the division of unit operation space. Then the benchmark value and threshold value of characteristic parameters of performance evaluation are determined by using multivariate Gaussian mixture model. Finally,Entropy weight based performance evaluation model is built to realize gas turbine performance evaluation. Case study has shown that the method is effective for gas turbine performance monitoring and evaluation,provides guarantee for safe operation of new energy power system.
作者
朱俊杰
王晓维
董玉亮
顾煜炯
ZHU Junjie;WANG Xiaowei;DONG Yuliang;GU Yujiong(Key Laboratory of State Monitoring and Control of Power Plant Equipment,Ministry of Education,North China Electric Power University,Beijing 102206,China)
出处
《智慧电力》
北大核心
2020年第7期24-29,共6页
Smart Power
基金
国家自然科学基金资助项目(51676068)
北京市重点研发计划资助项目(Z181100005118005)。
关键词
重型燃气轮机
工况辨识
K-均值聚类
熵权法
性能评价
heavy gas turbine
condition identification
K-means clustering
entropy weight method
performance evaluation
作者简介
朱俊杰(1996),男,江苏南通人,硕士研究生,主要研究方向为重型燃气轮机健康监测与故障诊断。
