A robust control of nZEBs for performance optimization at cluster level under demand prediction uncertainty

Collaborations among nZEBs (e.g. renewable energy sharing and battery sharing) can improve the nZEBs' performance at the cluster level. To enable such collaborations, existing studies have developed many demand response control methods to control the operation of nZEB systems. Unfortunately, due to lack of consideration of demand prediction uncertainty, most of the demand response control methods fail to achieve the desired performance. A few methods have considered the impacts of uncertainty, but they merely perform simple and limited collaborations among nZEBs, and thus they cannot achieve the optimal performance at the cluster level. This paper, therefore, proposes a nZEB control method that enables full collaborations among nZEBs and takes account of the demand prediction uncertainty. The proposed robust control method first analyzes the demand prediction uncertainty, next optimizes the nZEB cluster operation under uncertainty, and then coordinates single nZEB's operation using the cluster operational parameters. The performance of the robust control has been studied and compared with a deterministic control. Case studies show that the robust control can effectively increase the cluster load matching and reduce the grid interaction with the demand prediction uncertainty existed. The proposed method can achieve robust performance improvements for the nZEB cluster in practice particularly as uncertainty exists.