基于多智能体优化的水电主导型水-风-光供应系统短期发电调度模型

Abstract Driven by the increasing demand for electricity and the rapid development of hydropower, wind, and solar energy, the joint scheduling of hydropower-dominated hydro-wind-solar supply systems (HHWSSSs) has emerged as a key research focus in power systems . However, the unique characteristics of the delicate scheduling requirements, complex constraints, and spatiotemporal coupling connections within HHWSSSs present significant issues for short-term generation scheduling (STGS), including excessive computation times for overall optimization, difficult model formulation, and low solution accuracy. Therefore, this paper develops a multiagent optimization model for STGS to address these issues efficiently. First, the centralized optimization of HHWSSS is transformed into the cooperative operation of multiagent systems based on the large system decomposition principle, which uses the hierarchical decomposition strategy rooted in basin characteristics to enhance the solution efficiency. Second, a bilevel nested multiagent optimization model aimed at minimizing water consumption, subsequent to the integration of the forecasted WSP output , is formulated by combining the unified optimal operation theory of multiagent systems to enable intelligent model construction. Finally, to improve solution quality, an improved alternating direction method of multipliers (IADMM) algorithm is employed to coordinate interactions among agents by implementing a modification strategy based on inflow balance and load matching principles. The feasibility and effectiveness of the proposed model are validated via typical daily datasets from different seasons. The results demonstrate that the proposed model can effectively determine the generation plan for HHWSSSs within a reasonable solution time and offer decision-makers flexible options to weigh the trade-off between operational efficiency and costs.