NextETRUCK presented a paper on the optimised charging management for megawatt stations using model predictive control during the Speedam 2024 Conference, which took place from 19-21 June 2024 on the Italian island of Ischia.

CIDETEC assisted the 3rd. Scientific meeting: Battery based energy storage Systems organized by Mondragon Unibertsitatea. The scientific meeting was 1 whole day event where 6 reference speakers presented the advances of I+D+I of battery systems they were involved. The meeting gathered around 60 people, where representatives from University of Colorado Spring, Universitat Politècnica de Catalunya, and Aalborg University where of huge interest. The meeting started with a deep review of BMS, which allowed the audience to have a nice overview and contextualization of nowadays state of the art and future upcoming developments. Afterwards, the presentations treated different specific I+D+I developments where topics such as physic-based models, physic informed neural network models, control strategies using physic based models, model predictive control or state of health estimation were described and discussed.

Paper summary

The rapid electrification of heavy-duty vehicles (HDEVs) demands innovative solutions to address two critical challenges in megawatt charging stations (MCS): maximizing efficiency and minimizing charging time. This study proposes a Model Predictive Control (MPC)-based high-level charging strategy to optimize the charging process in modular megawatt stations.Heavy-duty vehicles, with their large battery capacities (500–700 kWh), require ultra-fast charging to meet operational demands. Modular charging station designs allow for higher reliability, efficiency, and simplified maintenance. By leveraging MPC, the study achieves dynamic power allocation across the station’s modules, ensuring optimal efficiency while meeting user-specified state of charge (SoC) levels within desired timeframes.

The proposed algorithm was simulated in MATLAB, with results compared against two conventional strategies: Average Power (AP) charging and Minimum Time (MT) charging. MPC outperformed both approaches in efficiency and energy loss. Over 2,000 charging simulations, the MPC method achieved an average efficiency of 97.84%, compared to 95.95% for AP and 97.28% for MT. Charging time was reduced by 50% compared to AP, while energy losses were minimized to 6.80 kWh on average—significantly better than MT’s 8.13 kWh and AP’s 11.86 kWh.This trade-off between charging speed and energy optimization makes MPC ideal for scalable real-world applications in MCSs. While the MT method excels in speed, it sacrifices efficiency, while the AP method is less efficient and slower. The MPC-based solution delivers a balanced approach, combining reduced charging time, lower energy loss, and operational efficiency.

This research highlights MPC’s potential to revolutionize high-power EV charging, paving the way for more efficient, user-friendly, and environmentally sustainable heavy-duty vehicle charging infrastructure. Future work will focus on refining the model with realistic vehicle data to further enhance its practical viability.