Abstract
Optimal scheduling of integrated energy systems (IESs) remains challenged by the fragmented treatment of heterogeneous energy modeling, uncertainty-aware decision-making, and coordinated control, which limits a unified understanding of how system representation, optimization, and autonomous operation are intrinsically connected. This review develops a three-dimensional analytical framework encompassing system modeling, uncertainty-aware optimization, and coordinated control to systematically synthesize the evolution, methodological foundations, and emerging paradigms of IES scheduling. It examines multi-energy flow modeling, energy hub representations, source–grid–load–storage coordination, and physics-informed component models, highlighting the fundamental trade-off between physical fidelity and computational tractability. Optimization strategies are then reviewed from stochastic and robust formulations to distributionally robust and data-driven approaches, with emphasis on jointly addressing economic efficiency, decarbonization, reliability, and operational flexibility under uncertainty. The transition from centralized scheduling toward hierarchical and distributed coordination, multi-agent control, virtual power plant aggregation, and autonomous operation is further analyzed from the perspectives of scalability, real-time responsiveness, and adaptability. The review identifies key scientific gaps in cross-scale coupling, interpretable and transferable data-driven modeling, market–physics integration, and the unified treatment of low-carbon objectives, resilience, and cybersecurity. Emerging directions are highlighted in physics–data collaborative modeling, digital twin-enabled optimization, intelligent multi-agent coordination, and cross-regional multi-energy markets. By unifying modeling, uncertainty management, optimization, and control within a coherent analytical framework, this review recasts IES scheduling as a cross-scale, adaptive, and trustworthy system intelligence problem, establishing a conceptual foundation for the transition from isolated optimization algorithms to autonomous and coordinated energy system operation.
