制冷技术2025，Vol.45Issue(5)：42-49,65,9.DOI:10.3969/j.issn.2095-4468.2025.05.201
多核集成学习算法在多联机制冷剂充注量故障中的诊断研究
Research on Diagnosis of Refrigerant Charging Faults for Variable Refrigerant Flow Systems with Multi-Kernel Ensemble Learning
摘要
Abstract
Aiming at the problem of refrigerant charge faults in the variable refrigerant flow(VRF)system,a refrigerant charge fault diagnosis model based on a multi-kernel ensemble learning algorithm is proposed in this paper.Firstly,the operational data of the VRF system is preprocessed,including data deduplication,label assignment,outlier handling,data standardization,and dataset partitioning.Then,the required refrigerant charge fault diagnosis model for the VRF is constructed based on the principles of the multi-kernel ensemble learning.The model is fine-tuned using the preprocessed data to obtain the optimal fault diagnosis model.Finally,the optimized model is evaluated for performance.The results show that the fault diagnosis model is able to effectively diagnose the refrigerant charge faults of the VRF,and the macro-Precision,macro-Recall,macro-F1,and the overall diagnosis accuracy reach 98.26%,97.27%,97.76%,and 98.10%,respectively,which are higher than the rest of the single-kernel fault diagnosis models.关键词
多联机系统/多核集成学习/故障诊断Key words
Variable refrigerant flow system/Multi-kernel ensemble learning/Fault diagnosis分类
建筑与水利引用本文复制引用
ZHANG Xue,CHEN Huanxin,ZHANG Li,CHEN Hengda..多核集成学习算法在多联机制冷剂充注量故障中的诊断研究[J].制冷技术,2025,45(5):42-49,65,9.基金项目
国家自然科学基金(No.51876070,No.51576074). （No.51876070,No.51576074）
