A market-based load regulation method for heterogeneous residential air-conditioning loads under cloud-edge collaboration

Residential air-conditioning load on the power demand side has gradually increased. The air-conditioning loads have thermal storage characteristics, and adjusting their temperature settings will not cause significant negative impacts to users, making them well-suited for demand response. In existing studies, load aggregators usually serve as demand response implementors, and reducing the energy consumption levels of air conditioning loads directly. However, this method often lacks a clear definition of the flexibility of individual air-conditioning units and fails to account for their variations. The absence of well-defined commodity characteristics for the flexible regulation potential complicates pricing, making it difficult to safeguard the interests of the load aggregator and each air-conditioning user. Furthermore, the diversity among residential air-conditioning loads introduces high-dimensional variables into demand response optimization, making it challenging to balance data security with computational efficiency. To address these challenges, this paper proposes a market-based load reduction method for heterogeneous residential air-conditioning loads using cloud-edge collaboration. First, the flexible adjustment capability of air-conditioning loads is defined as a tradable commodity. Then, a cloud-edge collaborative trading scheme is proposed to guarantee the benefits of the load aggregator and each air-conditioning user in a data-neutrality manner. Finally, a hybrid method based on encrypted clustering is developed to compute the optimal trading strategy, ensuring data security and computational efficiency. Simulation results based on data from the Austin area show that Pareto improvement is achieved compared to other load reduction schemes, and the proposed encrypted clustering method reduces computational complexity and data security risks compared to traditional analytical and iterative approaches.