Track: High School and Middle School STEM Competition

Abstract

This paper is to develop a performance prediction model of an integrated system that combines a solar heat pump and an air heat pump based on the DNN (Deep Neural Network) model. In this paper, we describe the overall procedure for building a DNN model to predict the performance of an integrated system, including the data collection method, data set construction, and DNN model structure. To validate the reliability of the performance prediction model based on the DNN model, the mean square error of the coefficient of variation CV(RMSE) proposed by the American Society of Heating, Refrigeration and Air Conditioning Engineers Guideline 14 was used. The RMSE between the prediction result of the CV DNN model and the output variable was calculated as 5%. Therefore, the reliability of the performance prediction model based on the DNN model was verified, and the performance prediction accuracy was similar to the energy simulation model.

Keywords

deep neural network, Integrated system, solar source heat pump, Air source heat pump, DNN