Data-driven robust optimization for energy and reserve scheduling considering multi-uncertainty

The intermittency of renewable energy sources and unscheduled outages threaten power system security, increasing the need for operational flexibility. This paper addresses the energy and reserve scheduling problem under photovoltaic power output uncertainty and N-k transmission line security criteria by proposing a data-driven two-stage robust optimization model. The first stage involves day-ahead scheduling for energy and reserve, while the second stage minimizes re-scheduling cost under the worst-case scenario. The model is decomposed into a master problem and a subproblem, using the column-and-constraint generation method for alternating solutions. A data-driven method constructs an interval-partitioned uncertainty set, effectively capturing the characteristics of photovoltaic power output uncertainty while balancing robustness and economy. The model also integrates generator, energy storage and both price-based and incentive-based demand response strategies to enhance reserve capacity, improving risk mitigation. Simulation results confirm the model’s feasibility and effectiveness for reliable and economical operation.<br/><br/>© 2025 The Authors.