Energy management in a domestic microgrid by means of model predictive controllers. (1st August 2016)

• Record Type: Journal Article
• Authors: Bruni, G.; Cordiner, S.; Mulone, V.; Sinisi, V.; Spagnolo, F.
• Is Part Of: Energy. Volume 108(2016)
• Page Start: 119
• Page End: 131
• Publication Date: 2016-08-01
• Subjects: Distributed generation -- Microgrids -- Model predictive control -- Fuel cells; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2015.08.004
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: The need to increase renewable energy sources deployment and to reduce consumption of fossil fuels has led to the diffusion of small-scale DG (Distributed Generation) systems, which may be effectively integrated in micro-grids. The role of control logic in defining microgrid performances and reliability is predominant and can be improved by using advanced control logics such as the ones based on MPC (Model Predictive Control). In a previous paper, a MPC logic, based on the use of weather forecasts to improve the performances, has been applied to the analysis of power management in a domestic micro-grid system composed by: PV (Photovoltaic panels), FC (Fuel Cells) and a battery pack; in that case, the system was not affected by real uncertainties. In this paper the same system has been considered for domestic microgrid applications. The system control logic has been implemented by assuming real weather forecast as input data. DMPC and SMPC (Deterministic and Stochastic Model Predictive Control) concepts have been applied to the system and results have been compared to both MPC and to a standard RBC (Rule Based Control) logic. The impact of forecast uncertainties has been evaluated showing the advantages of a stochastic approach. In that case, the SMPC showed encouraging performances compared to standard control logics, primarily in terms of primary energy savings and downsizing potential of the power-delivering sub-systems using programmable energy sources. Highlights: A Model Predictive Controller has been used to optimize the behaviour of a microgrid. A stochastic approach has been used to model weather forecast uncertainties. 10% savings of HVAC required energy have been demonstrated. Savings in primary fossil energy range between 10% and 30%. Power sub-systems using programmable sources can be consistently downsized.