인공신경망 기반 수열 히트펌프 시스템의 성능예측모델 개발

Abstract

Purpose: Recently, core technologies of the 4th industrial revolution such as machine learning and artificial intelligence are being combined with building technologies to predict the energy consumption and performance of the air-conditioning system of buildings. On the other hand, the energy independence of buildings is reinforced according to government roadmap of zero energy building. Accordingly, a water source heat pump system(WSHP) as one of the renewable energy systems is actively introduced. Method: In this study, a performance prediction model for the WSHP system is developed based on artificial neural network(ANN). This paper is described the process of constructing the ANN model such as analysis of Pearson correlation coefficient and construction of each layers, and the verification method and error analysis of the development model through coefficient of variation root mean square error(Cv(RMSE)). Result: The Cv(RMSE) was calculated as 0.09% between the prediction result of the ANN model and the calculation result of the integrated simulation model. The results predicted by the artificial neural network model met the Cv(RMSE) standard that proposed by American Society of Heating, Refrigerating and Air-conditioning Engineers Guideline 14, and the prediction result was very low error rate.