Project R-16028

Research on Battery Fault Mechanism and Fault-Tolerant Control Based on Electrochemical Models in Cloud BMS (Research)

Lithium-ion battery failures, particularly thermal runaway, pose significant safety risks. The main challenges for current Battery Management Systems (BMS) include a limited understanding of fault mechanisms, challenges in fault diagnosis and prognosis, and inadequate fault-tolerant strategies.This work proposes an integrated framework that studies battery failure mechanisms and combines electrochemical models with cloud-based BMS, utilizing internal physical state information to enhance fault diagnosis, prognosis, and fault-tolerant control. Firstly, an enhanced Single Particle Model with electrolyte dynamics is developed, incorporating aging mechanisms and optimized through advanced numerical methods for real-time cloud applications. Secondly, electrochemical experiments investigate fault mechanisms. Multi-physics models establish relationships between fault characteristics and electrochemical model, enabling understanding of fault evolution. Thirdly, a fault diagnosis and prognosis framework is developed based on electrochemical models and fault mechanisms. The cloud platform enables accurate fault detection and prediction through electrochemical insights. Finally, active fault-tolerant control strategies using Model Predictive Control are implemented via cloud-based systems, enabling dynamic adaptation while maintaining optimal performance and safety. This work advances battery safety by addressing core challenges in fault diagnosis, prognosis, and control.

01 October 2025 - 30 September 2028