Development of Advanced Diagnostic and Control Strategies for Li-ion Batteries using Machine Learning - Next-Generation BMS

虽然锂离子电池化学继续快速发展，但在组装成电池组时，控制和监控电池的能力在许多方面都缺乏。电池管理系统(BMS)监视和控制电池组中的电池，确保它们不会过热、平衡电荷、处理安全问题等。电池工程的这一方面往往会让在电池级别提高电池性能的重大努力付诸东流。该研究项目的目标是：-开发基于机器学习的新方法，以准确描述锂离子电池的性能。-开发可部署在电池资产上的诊断工具，以基于实时和历史数据预测未来性能。-生产一种先进的BMS，能够将电池组控制到一个新的精度和功能水平，准备在商业车辆上实施。为了实现这一目标，该项目将获得商业模块，与业界(例如Horiba Mira)密切合作，学习如何将新算法应用到商业BMS系统中，并将建模洞察力与实验实践相结合，将新的BMS技术带入投资/商业化潜力点。

While Li-ion battery chemistry continues to advance apace, the ability to control and monitor batteries when assembled into packs is lacking in many respects. The battery management system (BMS) monitors and controls cells within a pack, ensuring that they do not get too hot, balancing charge, dealing with safety aspects, etc. This aspect of battery engineering often lets down the significant effort that goes into improving battery performance at the cell level. The aims of this research project are:- Develop new methodologies based on machine learning to accurately describe the performance of Li-ion cells.- Develop diagnostic tools that can be deployed on battery assets to predict future performance based on real-time and historical data.- Produce an advanced BMS capable of controlling a battery pack to a new level of accuracy and functionality, ready for implementation in a commercial vehicle.To achieve this, the project will acquire commercial modules, work closely with industry (e.g. Horiba MIRA) to learn how to implement new algorithms into commercial BMS systems and combine modelling insight with experimental practice to bring new BMS technologies to a point of investment/commercialisation potential.