Key points are not available for this paper at this time.
The development of smart grids allows residential customers to participate in demand response (DR) programs to aid power grid management through HEMS (Home Energy Management Systems), but similar electricity consumption behavior among customers based on time-of-use electricity prices may lead to the problem of peak load shift, also known as peak rebounds. This article proposes a multi-level interactive optimization model considering individual sensitivity for DR. The model consists of community energy aggregators (CEAs), which perform as an intermediate processing layer between customers and power grid. Customer terminals perform energy management for home appliances, electric vehicles, energy storage systems, and renewable energy generation. The scheduling problem is decomposed into smaller parallel decision problems that are easier to solve. Renewable generation especially photovoltaic power generation is predicted and used to mitigate the influence of energy generation uncertainty. By introducing numerical responsiveness of customers, the model deals with uncertainty on the subjective level of customers. As indicated in numerical analyses, the model is a good compromise between stochastic optimization depending on idealized probability models and robust optimization sacrificing cost to meet worst case scenarios. The proposed method was compared with existing optimization-based methods for peak shaving. Compared with coordinated load management, our method reduced the peak load and average cost by 13.36% and 18.96%, respectively. Compared with robust optimization, our method achieved similar effect while handling the uncertainty in customers and PV.
Chen et al. (2023) studied this question.