A model-based predictive control for power distribution grids with prolific distributed generation: a case study

Over the past few years, the deployment of distributed generation, namely photovoltaic (PV) power generation, into power distribution grids has generated multiple challenges for grid operators in terms of planning, monitoring, and control of these grids. In the present paper, a model-based predictive control (MPC) strategy is proposed to minimise the discrepancy between supply and demand in a low-voltage power distribution grid while upholding voltage constraints stipulated by regulations. This is done by controlling third-party-owned distributed generators and storage systems (biogas plant and water tower). The main contribution of this paper is the introduction of a modelling approach that handles the ON/OFF water tower control without using mixed-integer non-linear programming and without having recourse to relaxation, which yields efficient results while maintaining reasonable computational complexity.