Receding horizon optimization framework for PV-BESS scheduling

Efficient scheduling of battery energy storage systems (BESS) is critical for maximizing the economic value of photovoltaic (PV) installations. This paper presents a computational framework for PV-BESS scheduling optimization, formulated as a constrained energy sales profit maximization problem. The framework employs a 72-hour Receding Horizon strategy, optimizing BESS scheduling over a rolling three-day window while adopting only the first 24 h as the final dispatch schedule and thereby allowing the optimizer to account for near-future market prices and solar production forecasts. Six state-of-the-art global optimization methods are benchmarked within this framework: Particle Swarm Optimization (PSO), Fireworks Algorithm (FWA), Multi-Scale Grid Descent (MSGD), Artificial Bee Colony (ABC), Grey Wolf Optimizer (GWO), and Squirrel Search Algorithm (SSA). A preliminary study on representative winter and summer periods identifies PSO as the best overall method in terms of solution feasibility, profit, and computational efficiency. Applied to a full year of synthetic solar production and electricity market data for Istria, Croatia, the PSO-based framework achieves a maximum annual profit of 189.78 M€, yielding a 7% improvement over a baseline daily charge/discharge scheduling program. The results demonstrate the robustness and practical applicability of the proposed framework for PV-BESS feasibility studies, system sizing, and automated operation scheduling.