摘要
以某项目的数据对6种冷水机组模型(Simple Linear (SL)Regression Model、Bi-quadratic (BQ)Regression Model、Multivariate Polynomial (MP) Regression model、Gordon-NG Universal (GNU) Model、Gordon-Ng Simplified (GNS) Model、Lee’s Simplified (LS) Model)的预测精度进行分析与评价,并讨论了当数据量减少时模型预测精度的变化,结果发现GNU模型的性能更优秀,可以作为冷水机组故障诊断的模型。
The author analyzes and values prediction accuracy of 6 chiller plant models(Simple Linear(SL) Regression Model、Bi-quadratic(BQ) Regression Model、Multivariate Polynomial(MP) Regression model、Gordon-NG Universal(GNU) Model、Gordon-Ng Simplified(GNS) Model、Lee’s Simplified(LS)Model) in some project and discusses prediction accuracy changes while data reduces.The results show that GNU model has better performance as chiller plant failure diagnosis model.
作者
王群
WANG Qun(Shanghai Tansuo Energy Service Limited Company)
出处
《上海节能》
2020年第7期734-746,共13页
Shanghai Energy Saving
关键词
冷水机组
回归模型
预测精度
GNU模型
Chiller Plant
Regression Model
Prediction Accuracy
GNU Model
