Energy Management Algorithm for Battery Sharing based on Model Predictive Control integrating Electro-Thermal Simulation and Demand Forecast

This paper presents the energy management algorithm implemented in an e-bike battery sharing station to control the parallel charging of many tens of batteries. The charging control targets energy-self-sufficient operation from solar energy only. To achieve that, model-based forecasts of the user demands, solar energy production as well as energy demand for thermal conditioning are employed. The latter is generated from electro-thermal simulation considering the self-heating of battery cells as function of charging current. Alternatively, exploiting bi-directional charging capabilities of the used DC/DC converters, the control algorithm predicts the best scheme of grid-oriented operation to maximize profit from energy micro trading, facilitating the batteries in the station as buffer storage and respecting time-variant stock market prices.