Energy Management of the Grid-Connected Residential Photovoltaic-Battery System Using Model Predictive Control Coupled with Dynamic Programming

Abstract

Appropriate energy management strategy is of great importance for the photovoltaic-battery (PVB) system to achieve desirable performance. This study developed a new method coupling model predictive control (MPC) with dynamic programming (DP) for optimal scheduling of a residential PVB system. The actual power data of a household were used, and both the feed-in-tariff (FiT) and time-of-use pricing (TOU) were considered. Three different strategies, including the economic optimization strategy ( OP C ), the grid-power optimization strategy ( OP P ), and the maximizing self-consumption strategy ( MSC ), were proposed, and compared experimentally by reproducing the historical PV generation and electrical load conditions. It was found that all the developed strategies could be implemented well in experiment, with the maximum relative deviation of 8.89% to the simulated results. The OP C strategy reduced the operation costs at the expense of weakening grid stability and lowering SCR and SSR , while the OP P strategy realized the grid-friendliness at the expense of increasing both the operation cost and battery aging. The MSC strategy has the compromise performance in both the operational economy and the grid-power stability. The operation cost was reduced significantly by increasing the PV capacity, while slightly affected by the battery capacity. Limited by the economic constraints, the battery capacity cannot be fully used for the OP C strategy. As for the OP P strategy, there is an optimal PV capacity for any battery capacity to minimize the grid power fluctuation. The optimized strategy that treats equally the economy and the grid-power stability ( λ 1 = λ 2 =0.5) has similar operation costs, SCR and SSR as the MSC strategy, but with much smaller grid power fluctuation.

Keywords: photovoltaic-battery (PVB) system, Energy management, model predictive control (MPC), dynamic programming (DP), Experiment, parametric analysis