Volume 22: Sustainable Energy Solutions for a Post-COVID Recovery towards a Better Future: Part V

Variable Refrigeration Flow System Simultaneous Fault Diagnosis Based on Deep Convolutional Network Zhenxin Zhou, Huanxin Chen



The potential for saving on energy related cost with timely and accurate Fault Detection and Diagnosis (FDD) in the air-conditioning system which is as one of the major energy consumer in buildings has been estimated to huge. In recent years, the research on fault diagnosis of air-conditioning systems has mostly focused on single fault or multiple faults. While, there is less research on simultaneous faults because of the complexity of simultaneous faults. This paper proposes a diagnosis method based on deep convolutional neural network, which can effectively diagnose the common two simultaneous faults and three simultaneous faults in variable flow systems. The results show that this method can effectively isolate faults in the case of multiple faults and multiple-simultaneous faults. The diagnostic accuracy is over 98%, and the Hamming loss value is lower 1. 5%.

Keywords deep convolutional neural network, Variable refrigerant flow system, simultaneous faults, Fault diagnosis, Artificial intelligence

Copyright ©
Energy Proceedings