In this paper we present a novel methodology for leveraging Receding HorizonControl (RHC), also known as Model Predictive Control (MPC) strategies fordistributed battery storage in a planning problem using a Benders decompositiontechnique. Longer prediction horizons lead to better storage placementstrategies but also higher computational complexity that can quickly becomecomputationally prohibitive. The here proposed MPC strategy in conjunction witha Benders decomposition technique effectively reduces the computationalcomplexity to a manageable level. We use the CIGRE low voltage (LV) benchmarkgrid as a case study for solving an optimal placement and sizing problem fordifferent control strategies with different MPC prediction horizons. Theobjective of the MPC strategy is to maximize the photovoltaic (PV) utilizationand minimize battery degradation in a local residential area, while satisfyingall grid constraints. For this case study we show that the economic value ofbattery storage is higher when using MPC based storage control strategies thanwhen using heuristic storage control strategies, because MPC strategiesexplicitly exploit the value of forecast information. The economic merit ofthis approach can be further increased by explicitly incorporating a batterydegradation model in the MPC strategy.
展开▼
