Conference Year
2018
Keywords
electronic expansion valve, correlation, variable refrigerant flow system, ANN
Abstract
Split-type air-conditioning systems with multiple indoor units and variable capacity feature, also known as variable refrigerant flow systems (VRF), are growing in both residential and commercial buildings for its flexibility in configurations and high energy efficiency through zoning control and variable capacity control. The VRF system controls its capacity by the combination of drive frequency of variable speed compressor and opening of electronic expansion valves (EEV). According to fluid dynamics analysis, the mass flow rate is a function of the EEV opening. Therefore, the accurate correlation of mass flow rate through an EEV is important in accurate prediction of the mass flow rate in the VRF systems and design the control system. The simulation and experiment studies are reviewed in this paper. Literature study showed that the most commonly used approach is power-law correlation which is recommended by ASHRAE. The relative deviation of the predicted mass flow of this method is around 5% to 40% depending on the refrigerants used. Different refrigerants include R-22, R-245a, R-134a, R-410A, R-407C and R-744 have been studied since 2005. Second type is polynomial fit correlation is a simplified model, which has a relative deviation of 5% for 98% of the data and developed in 2013. Third type is using Artificial Neural Network (ANN) algorithm, which has a relative deviation from 4% to 7% and has been used since 2015. In addition, a dynamic model of EEV has been developed since 2016. In conclusion, the most accurate method is ANN curve fitting, which is a novel approach to be utilized in EEV correlation. However, only a few researches have been conducted on this method. Therefore, more research is needed.