Research on Digital Twin Model and Fault Diagnosis of Chiller
In this paper,a fault diagnosis method of chiller based on digital twin is adopted to search the fault diagnosis strategy of refrigerant leakage,condenser water flow reduction,and evaporator water flow reduction faults in the chiller.The reliability of data generated by the joint modeling methods of physical modeling,empirical modeling and data-driven modeling and the improvement effect on fault diagnosis are analyzed,and the reliability of the proposed method is verified by experiments.The results show that:digital twin method can effectively solve the problem that the diagnostic ability of the FDD model decreases under the new operating point.After using the digital twin method,the maximum relative error of all thermal parameters of the system simulation model does not exceed 4%,and the accuracy,recall,and F1-score of the fault diagnosis model were increased by 9.7%,29.4%,and 16.6%,respectively.
国家科技期刊平台
NSTL
万方数据
