Research Review on the Scheduling Optimization of Production and Energy Systems under Uncertain Environments

Abstract

The uncertainties in actual production and energy systems pose severe challenges to the stable, efficient, and low-carbon operation of the systems. Centering around the core of "uncertainty handling," this paper systematically reviews the mainstream modeling and solution methods for dealing with uncertainties in the current scheduling optimization field. At the measurement level, uncertainties can be characterized from two dimensions: parameter fluctuations and event perturbations, and there are essential differences in the impact mechanisms of the two on the scheduling scheme. From the perspective of uncertainty modeling, the theoretical foundations and applicable scenarios of methods such as fuzzy programming, stochastic programming, robust optimization, and distributionally robust optimization are reviewed; from the perspective of solution strategies, the evolution of methods from exact algorithms to heuristic, meta-heuristic, and then to intelligent decision-making methods based on learning and simulation is analyzed. At the effect level, uncertainties will significantly weaken the economic benefits and environmental performance of the scheduling scheme, but there are significant scenario heterogeneities in the degree of their impact. Research shows that distributionally robust optimization has attracted much attention due to its good balance between conservatism and economy, and the hybrid-driven method that combines simulation and learning is becoming an important trend to improve the adaptability of the scheduling system. Future research needs to further focus on the decision-making requirements in complex scenarios such as the coupling of multiple uncertainties, multi-agent games, and real-time dynamic responses.