Experimental analysis of an innovative RC-Mapping model for flexibility quantification of building air conditioning systems

The integration of renewable energy sources underscores the growing need for enhanced grid flexibility. In this context, the grey-box model emerges as a practical solution for flexibility quantification of building air conditioning systems.

Nonetheless, developing a precise grey-box model faces multiple challenges. Firstly, current resistance-capacitance (RC) models primarily consider buildings’ thermal storage while neglecting air conditioning systems’ thermal storage which is unignorable especially during demand response. Secondly, there is a lack of research identifying optimal conditions for collecting high-quality data to train these models effectively. Thirdly, internal heat gains are rarely measured directly, necessitating robust and experimentally validated estimation methods.

To address these limitations, this study proposes an RC-Mapping model. The RC component incorporates a novel structure that accounts for the thermal storage of both buildings and air conditioning systems, providing a more comprehensive representation. Meanwhile, the Mapping component models internal heat gain schedules. Experimental evaluations highlight the advantages of this approach: the RC component reduces indoor temperature prediction errors, decreasing the RMSE by 0.21℃ (from 0.826℃ to 0.616℃) compared to conventional RC models.

Guidelines for obtaining high-quality training data are proposed, emphasizing the importance of significant indoor-outdoor temperature differences and substantial indoor temperature fluctuations. The Mapping component effectively estimates internal heat gains with a CV-RMSE of 26.56%. Additionally, the model estimates the flexibility of building air conditioning systems with an error margin below 11%.

The study further investigates how various prediction errors—such as those in the model, COP predictions, and internal heat gain predictions—impact flexibility quantification of building air conditioning systems.