Incremental capacity analysis of battery under dynamic load conditions

Area of Science:

• Electrochemistry
• Battery Technology
• Data Science

Background:

• Electric vehicle (EV) battery State of Health (SOH) assessment is difficult due to inconsistent charging/discharging and varying operational parameters.
• Conventional SOH estimation methods require controlled calibration cycles, which are not feasible in real-world EV applications.
• Accurate SOH monitoring is crucial for EV performance, safety, and lifespan management.

Purpose of the Study:

• To develop an analysis-based method for obtaining labeled capacity and SOH values from EV batteries under real-world operating conditions.
• To extract battery health features using Incremental Capacity Analysis (ICA) that correlate with battery age.
• To enable data-driven prediction of EV battery capacity and SOH using machine learning or deep learning models.

Main Methods:

• Application of the Incremental Capacity Analysis (ICA) method to electric vehicle (EV) battery operational data.
• Extraction of battery health indicators as a function of battery age using the ICA method.
• Calculation of State of Health (SOH) for EV batteries utilizing the proposed analysis-based approach.

Main Results:

• The proposed method successfully obtains labeled capacity and SOH values from EV batteries operating under dynamic conditions.
• Identified and extracted key health features from battery data using ICA, demonstrating a correlation with battery aging.
• Validated the calculation of SOH for a vehicle battery using the developed analysis-based technique.

Conclusions:

• The presented analysis-based method offers a practical solution for EV battery SOH assessment without requiring controlled calibration.
• Extracted health features provide valuable inputs for advanced machine learning models to predict future battery capacity and SOH.
• This research contributes to more reliable and efficient battery management systems for electric vehicles.