抄録
The daily performance of a water heating system composed of a CO2 heat pump and a hot water storage tank is affected by the history of ambient conditions, hot water demand, and operating conditions as well as the resultant history of temperature distribution of the storage tank. To operate the system optimally under daily changes in these items, it is important to accurately estimate daily changes in the values of performance criteria such as COP,storage and system efficiencies, and volumes of stored and unused hot water. In this paper, the daily changes in the values of performance criteria corresponding to those in hot water demand and one of two operating conditions are estimated under daily constant ambient and the other operating conditions by multilayered neural network models. In addition, the former operating condition is determined based on the values of performance criteria estimated by the models. A numerical study is conducted using the values of performance criteria obtained by a numerical simulation for a simulated monthly hot water demand, and the validity and effectiveness of this
