Cloud based adaptive battery management system for electric vehicles

An Electric Vehicle’s (EV’s) battery pack is environmentally and financially expensive. If not managed well, it can result in reduced battery pack life-time. A battery pack’s State of Health (SOH), a degradation indicator, depends on the temperature, State of Charge (SOC), voltage, current and age. The SOH and SOC are not measured, but estimated. Their estimation, along with other relevant indicators, is called battery state estimation, crucial in a battery management system (BMS).

Adaptive methods such as Kalman filters are used in battery state estimation. Though effective, their accuracy depends on the system model. However, as a battery pack’s model is uncertain and time-variant, it is challenging to maintain BMS performance, especially as EVs age. Therefore, cloud-based adaptive and integrated BMS with the goal of improving durability of battery packs is an important research topic.

In this project, you will:

• design a cloud based adaptive system to identify dynamic battery pack models on regular intervals to maintain model accuracy
• design and test a cloud based adaptive system to re-design and update the BMS on regular intervals to maintain its performance
• analyse the effect of your system on battery pack life-time
• publish your findings in two high quality research articles

Your research outcomes could assist in improving the life-time of EV battery packs, resulting in higher cost effectiveness of EVs, which could help with market adoption of EVs. They could also help with reducing environmental pollution and waste from EV batteries.

In addition to the supervisors from the University of Southampton, this project includes the following external supervisor:

• Dr Othman Maganga (Jaguar Land Rover)