In the context of the global transition to sustainable transportation, electric vehicles have become an important development direction. However, the lithium coating of lithium-ion batteries (LIBs) poses a threat to the safety and reliability of electric vehicles, affecting their large-scale application. A research team from the University of Shanghai for Science and Technology has recently developed an intelligent lithium coating detection and early warning system based on random forest machine learning algorithms, which provides a new solution for electric vehicle battery safety monitoring.
Ⅰ. The impact of lithium coating on battery safety
Lithium plating refers to the phenomenon that lithium ions are not properly embedded in the graphite structure, but accumulate on the surface of the negative electrode of the battery. This problem mainly occurs in fast charging, low temperature environment, or high state of charge, which can lead to safety issues such as battery capacity attenuation or short circuit. Traditional detection methods have the disadvantages of strong equipment dependence or insufficient accuracy, so more efficient real-time monitoring technology needs to be developed.
Ⅱ. Application of random forest algorithm in lithium coating detection
The system developed by the team of University of Shanghai for Science and Technology uses the random forest algorithm to analyze the pulse charging data and identify the lithium coating phenomenon through the characteristics of the electrical signal. The system can achieve 97.2% detection accuracy with only external electrical measurement, and does not require hardware modification, which has high compatibility and practicability.
1. Multi-dimensional feature extraction improves detection accuracy
The accuracy of the traditional single feature detection method is 68.5%, and the system uses multi-dimensional feature extraction technology to improve the sensitivity and accuracy of the detection by analyzing the normalized internal resistance mode and relaxation voltage characteristics in the pulse charging process, which provides a basis for timely intervention.
Figure: Intelligent monitoring technology builds a strong line of defense for electric vehicle battery safety
Ⅲ The characteristics of technical application
1. Compatible with existing systems
The system requires only standard electrical measurement data and can be integrated into existing battery management systems (BMS) or cloud monitoring platforms through software updates, making it suitable for battery safety upgrades in EV manufacturing and existing vehicles.
Ⅳ The future development direction
1. Expand the scope of application
The research team plans to incorporate more types of lithium-ion batteries into the dataset to improve the versatility of the model under different battery chemistries and morphologies, and expand its application in energy storage systems, consumer electronics, and other fields.
2. Optimize fast charging technology
The team is exploring combining this technology with a fast-charging protocol to dynamically adjust charging parameters through real-time lithium coating risk assessment, so as to improve charging efficiency while ensuring safety.
3. Technical significance
With the popularization of electric vehicles, the development of intelligent lithium coating detection systems will help improve battery safety and promote the sustainable development of the new energy industry. This technology provides an effective guarantee for battery safety through data-driven monitoring and early warning capabilities, and is expected to play a greater role in the energy transition in the future.
