The teams around the NextETRUCK partners CIDETEC Energy Storage and TECNALIA, under the leadership of Battery Diagnosis Team Leader Mikel Arrinda, recently published a paper on battery thermal management systems in cooperation with other Spanish scientists.
The paper titled ‘Model Predictive Control as the Cloud Control Strategy for a Battery Thermal Management System’ presents an innovative approach to managing the thermal behaviour of batteries in electric vehicles. This particularly applies to high-stress scenarios, such as fast charging. The study proposes using a cloud-based Model Predictive Control (MPC) system to optimise the Battery Thermal Management System (BTMS), balancing energy consumption with thermal safety. In general, the work contributes to the evolving field of EV thermal management by showcasing how cloud computing and predictive control can enhance battery safety, performance, and energy efficiency in next-generation vehicles.
The rise in EV adoption has intensified the need to address range anxiety and safety, where effective battery thermal management plays a key role. Traditional control methods often fall short under complex, multi-variable scenarios. Model Predictive Control offers predictive and multi-input, multi-output optimisation capabilities, making it well-suited for BTMS control.
Central to the MPC framework are digital models of the battery system, BMS, ECU, and BTMS. The battery model incorporates both electrical and thermal characteristics, derived from empirical data and validated through extensive testing. The BTMS model simulates various operational modes—recirculation, heating, and cooling—based on component-level thermal responses. A discrete optimisation method is used, and soft constraints are applied via exponential barrier functions, avoiding abrupt constraint violations and ensuring stability.
Overall, the cloud-based MPC demonstrated a marginal 1% energy saving across the full simulation while improving thermal safety. Its structure supports integrating black-box models and third-party modules, promoting collaborative development in co-simulation environments.
Read the full paper here.