Distributed Economic Model Predictive Control for the Joint Energy Dispatch of Wind Farms and Run-of-the-River Hydropower Plants

Luca Santosuosso (1) , Simon Camal (1) , Arthur Lett (2) , Guillaume Bontron (2) , Georges Kariniotakis (1)

1 PERSEE - Centre Procédés, Énergies Renouvelables, Systèmes Énergétiques
2 CNR - Compagnie Nationale du Rhône

This study addresses the energy dispatch problem of a virtual power plant (VPP) acting as a price-tacker in the day-ahead electricity market. The VPP comprises wind farms and a cascade of run-of-the-river hydropower plants. Even if the storage capacity of the cascade is limited, it can still be exploited to compensate the variability of wind. This implies dispatching the water reservoirs near to real-time, while accounting for complex constraints and various sources of uncertainty. To this aim, we present a control strategy based on economic model predictive control (MPC), which is then decomposed using the auxiliary problem principle. As a distinctive feature, the proposed algorithm is fully-distributed, i.e. no central coordinator is required. Compared to centralized MPC, the distributed algorithm brings a ∼10% reduction in the average execution time of the controller. Moreover, the joint operation of hydropower and wind is shown to enhance the economic value of both assets.

Keywords: Distributed model predictive control; run-of-the river hydropower; wind power; storage; complex systems; Control

Domains: Mathematics [math]; Optimization and Control [math.OC]; Electric power

Luca Santosuosso, Simon Camal, Arthur Lett, Guillaume Bontron, Georges Kariniotakis. Distributed Economic Model Predictive Control for the Joint Energy Dispatch of Wind Farms and Run-of-the-River Hydropower Plants. PSCC'2024 - XXIII Power Systems Computation Conference, Jun 2024, Paris, France. ⟨hal-04533246⟩