Multi-time scale scheduling framework for multi-energy system considering demand response: A self-approaching optimal approach

With energy communities transforming from the traditional vertical provision structure to the interactive competition pattern, how to balance the interests of agents and flexibly manage their internal resources becomes a key challenge, especially in multi-time scale multi-energy scheduling. This study proposes a self-approaching optimal scheduling framework that solves this problem considering two stages, i.e., day-ahead and intra-day, and trade-offs in the benefit allocation among multi-agents. The framework reformulates the Nash game problem to generate a day-ahead scheduling plan. The Lyapunov optimization approach is applied to real-time scheduling of the multi-energy system and flexible load, which effectively copes with the unknown dynamics of the intra-day input data. Different response characteristics of community resources are considered. It is emphasized that scheduling strategies are asymptotically optimal under balance-of-interest deviations. The case study demonstrates that the proposed framework can balance benefits to obtain the day-ahead scheduling plan. In terms of winter, the 90.05% profit loss is avoided to achieve the 13.15% cost saving. The intra-day benefit allocation based on the proposed approach is more equilibrium, e.g., in event 2, it avoids sacrificing 79.98% of the energy profit to reduce 250.15% of the energy cost.