Aggregate Model of Distributed Multi-Energy Flexible Resources for Power System Operation: A Decomposition Approach

Distributed multi-energy flexible resources (DMEFR) can exploit the complementarity of electricity, gas, and heat to enhance power flexibility, demonstrating significant potential to promote renewable energy consumption and improve operational economics. An accurate model describing the aggregate feasible region (AFR) of DMEFR is essential for power systems to leverage this flexibility effectively. However, the time-coupling characteristic of DMEFR results in a prohibitive computational burden when computing the exact AFR. To address this challenge, this paper proposes a novel decomposition-based aggregate modeling method for DMEFR that features high computational efficiency and clear physical interpretation. First, we prove that the AFR of DMEFR is equivalent to the Minkowski sum of two decoupled components: a time-decoupled component capturing power boundaries via linear programming, and a time-coupled component (TCC) characterizing storage dynamics. Second, for the TCC, we develop an adjustable homothetic polytope (AHP)-based method that jointly optimizes the basic homothetic polytope, scaling factors, and translating factors, reducing the conservatism of fixed homothetic polytope-based methods. Third, we formulate the AHP optimization as a bilinear problem and propose a block coordinate descent algorithm with a convergence guarantee for it. Case studies validate the effectiveness and superiority of the proposed method.