Defining 'Better' in the World of BMS Algorithms: Scenarios and Requirements for Automotive Applications

Abstract

This paper investigates the complexities in the development and implementation of Battery Management System (BMS) algorithms. Understanding the performance of BMS algorithms is crucial, as poor algorithms can lead to customer frustration, safety issues, and acceleration of battery degradation, becoming a significant liability for automotive companies. However, a clear definition of "better" for BMS algorithms remains elusive, complicating validation efforts. There are also challenges in hardware selection, data storage, calibration during development and use, and cost constraints. Furthermore, different state estimation principles have varying data requirements, impacting algorithm performance. To address these issues, we derived a comprehensive methodology of automated test scenario creation obtained from expert interviews, considering aspects like computational capacity and specific usage scenarios for algorithm comparison and validation. Our research underscores the importance of robust validation and the nuanced dependencies of algorithm performance on factors such as hardware, stored data, calibration, and specific usage contexts.