Review on robust optimization theories and applications for energy sustainability

Abstract

The transition to low-carbon, decentralised, and interconnected energy systems increases uncertainties in renewable generation, including climate conditions, technology, prices, and supply chains. Robust optimization (RO) supports decision-making when parameters are uncertain but can be bounded, while their probability distributions are unreliable. Its guarantees depend on the construction and calibration of uncertainty sets and do not extend to conditions excluded from the planning. This review examines the RO from classical set-based formulations to adjustable, distributionally robust, stochastic–robust, and machine-learning-assisted approaches. The applications are classified into policy planning; infrastructure and system design; operational scheduling and control; and energy networks, logistics, and supply chains. These indicates that RO strengthen feasibility, flexibility, security, and resilience among renewable integrations. This is motivated from the perspective of uncertainty protection alongside technological affordability and distributional equity. Future research should coordinate operations, represent climate and cross-sector risks, incorporate circular material flows and ecological justice norms, and establish transparent benchmarks, including physically credible tests.