Multi-modal framework for battery state of health evaluation using open-source electric vehicle data

Area of Science:

• Electric Vehicle Technology
• Battery Management Systems
• Data Science

Background:

• Reliable battery state of health (SOH) evaluation is critical for electric vehicle (EV) efficiency and safety.
• Limited large-scale, high-quality field data impedes the development of effective battery management systems (BMS) for SOH estimation, lifetime prediction, and fault detection.

Purpose of the Study:

• To analyze real-world EV operational data to understand limitations in current BMS performance.
• To investigate the discrepancies between field and laboratory battery test data impacting SOH estimation.
• To develop an advanced deep learning framework for accurate and cost-effective SOH estimation using historical vehicle data.

Main Methods:

• Analysis of operational data from 300 diverse electric vehicles over three years.
• Comparison of field data with laboratory battery test data.
• Development and application of a deep learning-based multi-modal framework for SOH estimation.

Main Results:

• Identified key factors limiting BMS performance in real-world EV applications.
• Demonstrated the effectiveness of the proposed deep learning framework in leveraging historical vehicle data for SOH estimation.
• The developed framework shows significant potential for SOH estimation and diagnostics in multi-sensor systems.

Conclusions:

• Bridging the gap between field data and laboratory testing is essential for robust BMS development.
• The proposed deep learning approach offers an efficient, accurate, and cost-effective solution for EV battery SOH estimation.
• Publicly releasing the collected field data aims to foster further research and innovation in EV battery management.