Security-constrained unit commitment in the presence of smart devices using a two-stage robust optimization model

This paper introduces a novel resilience-enhanced robust optimization framework for security-constrained unit commitment, integrating adaptive energy storage and demand-side flexibility to improve power system stability under extreme and uncertain conditions. Unlike conventional two-stage robust optimization methods, our approach incorporates a resilience-driven formulation that explicitly accounts for severe renewable generation fluctuations, sudden demand variations, and unforeseen grid disturbances. The proposed model leverages a mixed-integer programming formulation, solved using a decomposition algorithm, to ensure computational efficiency in large-scale networks. The method optimally coordinates charging and discharging strategies for storage systems and electric vehicles, alongside flexible demand management, to minimize operational costs, startup/shutdown and no-load expenses, unserved load costs, and emissions. Simulations conducted on multiple test networks (6-bus, 24-bus, 118-node, and 300-node systems) validate the effectiveness of the approach. The results demonstrate optimal objective function values of $608,976 for the 6-bus network, $3399,185 for the 24-bus network, $21.58 million for the 118-node network, and $62.41 million for the 300-node network. Comparisons with established methods such as Benders decomposition and evolutionary algorithms reveal cost reductions of approximately 4.85 % and 9.47 % for the 6-bus network and 5.66 % and 9.15 % for the 24-bus network, respectively. These findings highlight the model’s ability to enhance economic efficiency, system adaptability, and resilience in modern power grids. • Introduces a framework for security-constrained unit commitment. • Incorporates electric vehicles, energy storage, and flexible loads. • Employs MILP with a CC&G algorithm. • Minimizes operating costs, emissions through a multi-objective function. • Demonstrates of 4.85–9.47 % in operating costs over traditional methods.