A Hybrid Approach of Using Both Simulation plus Neural Networks for Window Design Optimization and HVAC Energy Consumption Prediction Modeling

Abstract

Energy use in the building sector accounts for a large percentage of the world's total energy consumption. Specifically, the energy consumption from the whole life cycle perspective of building is 0.4% at design stage, 16% at construction stage, 83.2% at operation stage, and 0.4% at disposal stage. There have been many studies focusing on the design stage to find alternatives to enhance the energy efficiency of buildings. However, there have been few studies considering both of the efficient energy management of the building operation stage for the optimum design model at the same time. As a result of the design phase study, we proposed an improved window design alternative that could save 2736.06 kW of heating and cooling energy per year compared to the base case building. As for optimum window design, we proposed an ANN (Artificial Neural Network) model which predicts the heating and cooling loads. It satisfied the content of ASHRAE (American Society of Heating, Refrigerating, and AirConditioning Engineers) Guideline 14-2002 and IPMVP (International Performance Measurement & Verification Protocol). Based on this study, it would be possible to save energy from the perspective of a building’s entire life cycle if window selection options standard that can be referenced at building design stage and heating and cooling system control algorithm applicable to the operation stage are developed together. 