Despacho económico robusto con MPC ante incertidumbre de precios y generación renovable

Abstract

A real-time optimal economic dispatch scheme was developed based on receding-horizon model predictive control and a scenario-based robust formulation to represent uncertainty in electricity prices and renewable generation. The model incorporated standard operational constraints of the dispatch problem, including power balance, generation limits, ramprate constraints, reserve requirements, the dynamics of a reservoir-type hydropower unit, and the operation of an electrochemical energy storage system. At each decision instant, a prediction horizon was built, a set of scenarios was generated around the available forecasts, and a constrained optimization problem was solved; only the first control action was implemented and the system states were updated using the real dynamics. Performance was assessed through simulation and benchmarked against an equivalent deterministic MPC approach. Results showed operational feasibility and compliance with terminal constraints for the storage states, with no unserved energy over the analyzed horizon. The comparison yielded marginal cost differences in the base case, consistent with a small system size and moderate uncertainty levels; however, the robust formulation provided a systematic framework to manage operational risk and preserve stable decisions under more severe price and renewable variability in larger-scale systems. Keywords: model predictive control, economic dispatch, robust optimization, renewable generation, energy storage