Study and design of a management system based on an electrochemical model (EChM) for fault tolerant pack of paintable batteries

Contact: Michel KINNAERT

This project aims at developing a fault tolerant battery pack made of a new kind of battery cells, namely lithium-ion paintable batteries. Such batteries can be adjusted to almost any support. Series/parallel arrangements of hundreds of cells can be necessary to reach the voltage/power requirements of certain applications, hence the need to ensure that the battery pack can keep working even upon degradation or failure of a few cells. To achieve this goal, a battery management system (BMS) that performs a suitable reconfiguration upon fault occurrence is needed. Such a BMS should have a modular structure to ensure easy scalability. The research work will focus on the development of the three layers of such a BMS: the supervision at the cell level, at the pack level, and the control of the charge/discharge operation at the pack level. The supervision at the cell level aims at detecting and localizing significant capacity and/or power fade of individual cells and monitoring their state of charge. It will be based on an electrochemical model to ensure proper operation in a wide operating range. The challenges for its design include taking into account thermal effects and being able to detect in a reliable way small changes in state of health despite measurement noise. The supervision at the pack level aims at detecting sensor faults, like drifts notably, by analyzing the correlations between different measurements, and possibly accommodating sensor faults through software sensors. To achieve this goal the right trade-off must be found between algorithm complexity, sensitivity to faults and ease of tuning of the design parameters. Finally health conscious control of the charge/discharge must be achieved while allowing for battery pack reconfiguration. This requires a proper management of the transient to meet operating constraints even during the reconfiguration phase.