Thermal Model Identification of Commercial Building based on Genetic Algorithm

In recent years, there is dramatically increasing in energy consumption. As the main power consuming in demand side, buildings can distribute the thermal storage to help decreasing the peak loads. To predict the demand ability of the power grid, thermal storage model of commercial building is strongly needed. In the literatures, several researchers have presented many different approaches to model the thermal buildings for the estimation of heating/cooling loads. However, many of the models are complex and they are very difficult to be realized. Therefore, it is very essential to find a simple-to-implement building thermal model and it is an important part of building energy management system. A popular approach of building model is gray model, which combines the physical knowledge with the experimental data. One of most popular gray models, Resistance-Capacitance (RC) network can simplify the thermal model of commercial building in some cases. In this paper, we regard the building as a whole, and genetic algorithm is applied to the 2R2C network. The parameters of the 2R2C are optimized by historical data. Lastly, we verify the effectiveness of the proposed approach by the estimated indoor air and ambient temperatures.