BMS constrained PSO optimized multistage fast charging of an electric vehicle battery in a PV grid hybrid DC charging station

Abstract Fast charging of electric vehicles (EVs) is limited by the cell voltage and temperature limits enforced by the battery management system (BMS), while hybrid charging stations must also keep grid power quality within standards. This paper presents a BMS constrained multistage constant current (MSCC) charging algorithm optimized by particle swarm optimization (PSO) for a PV grid hybrid DC fast charging station. The station, modelled in MATLAB/Simulink, comprises a unity power factor front end converter, an isolated DC to DC charger, a 14.4 kWp photovoltaic array with perturb and observe maximum power point tracking (MPPT) on an 800 V DC link, and a 54 Ah pack of temperature dependent cells with cell to cell variation. The BMS combines four current stages, a 4.2 V constant voltage loop on the highest cell and thermal derating between 40 and 45 °C; PSO tunes the stage currents and state of charge breakpoints. The optimized profile charges the pack from 20 % to 80 % in 22.6 min, 38.1 % faster than 1C constant current constant voltage (CCCV) charging and 6.1 % faster than CCCV at equal peak current, with all cells below 4.201 V and 43.5 °C. At 45 °C ambient the reduction reaches 29.8 %. Rerating the front end converter lowers the grid current total demand distortion from 9.71 % to 2.66 % (2.91 % with PV), a power factor of 0.991. The MPPT extracts over 98.2 % of the available PV power, supplying up to 34.9 % of the station input without affecting the charging current.