作　　者： (刘闯）; (金仁瀚）; (巩二磊）; (刘勇）; (岳孟赫）;

机构地区： 南京航空航天大学能源与动力学院,江苏省南京市210016

出　　处： 《中国电机工程学报》 2017年第16期4731-4738,共8页

摘　　要： 为了预测燃气轮机发电机组的性能,基于大量的机组运行数据,采用多元线性回归分析方法,获得了通过燃机发电机组初始进口条件预测燃机功率和燃机效率的数学模型。以PG9171E型燃机发电机组部分历史运行数据作为训练样本,建立关于燃机性能的多元线性回归数学模型,并对模型进行验证分析。结果表明:基于多元线性回归分析方法,可以通过初始进口条件有效地预测燃机发电机组的性能;从预测结果分析来看,对于燃机功率的预测优于燃机效率。该方法可以为电厂大数据分析和经济性运行提供一定的参考价值。 In order to predict the performance of gas turbine generator units, based on a large number of gas turbine generator unit＇s operation data and the method of multiple linear regression analysis, the predicted mathematical models of gas turbine power and gas turbine efficiency through the initial import conditions were obtained. The part of PG9171E type gas turbine generator unit＇s historical operating data was used as training samples, then the multiple linear regression mathematical model of gas turbine performance was established, validated and analyzed. The results showed： based on multiple linear regression analysis, the performance of gas turbine generator unit can be predicted effectively through the initial import conditions; judging from the analysis of predicted results, multiple linear regression analysis for gas turbine power prediction is superior to the prediction of gas turbine efficiency. This method can provide some reference value for the big data analysis and economical operation of the power plant.

关 键 词： 燃气轮机 发电机组性能 多元线性回归 数学模型