Towards Safer Electric Vehicles: Autoencoder-Based Fault Detection Method for High-Voltage Lithium-Ion Battery Packs

Area of Science:

• Electrical Engineering
• Materials Science
• Automotive Engineering

Background:

• The expanding battery electric vehicle (BEV) market relies heavily on lithium-ion battery (LIB) packs, prized for their energy and power density.
• LIBs operate within a narrow safety window and are vulnerable to environmental factors, operational conditions, and manufacturing defects, posing risks like thermal runaway.
• Current fault detection methods require improvement to address the safety challenges in high-voltage LIB systems.

Purpose of the Study:

• To develop and present an advanced autoencoder-based fault detection methodology for lithium-ion battery packs.
• To enhance the safety of battery electric vehicles by proactively identifying potential battery faults.
• To leverage computational intelligence and machine learning for robust battery fault diagnosis.

Main Methods:

• An autoencoder-based fault detection method was developed, integrating computational intelligence and machine learning techniques.
• The method utilizes optical liquid detection systems (OLDSs) for monitoring immersion-cooled battery packs, employing optical signals within high-voltage areas.
• Performance evaluation involved real-life datasets, encompassing both faultless and simulated fault conditions, using specific performance indicators.

Main Results:

• The autoencoder-based method demonstrated effectiveness in detecting faults within LIB packs.
• The use of optical signals via OLDSs proved viable for monitoring in high-voltage environments.
• Performance indicators confirmed the system's capability in identifying both normal and fault states.

Conclusions:

• The developed autoencoder-based fault detection method offers a promising approach to enhance the safety of lithium-ion battery packs in electric vehicles.
• The integration of OLDSs provides a novel, safe alternative for fault detection in high-voltage battery systems.
• This research contributes to improving the reliability and safety standards for the rapidly growing BEV market.