Researchers have discovered the fundamental mechanism behind battery degradation, which could revolutionize the design of lithium-ion batteries, enhancing the driving range and lifespan of electric vehicles (EVs) and advancing clean energy storage solutions.
For example, only batteries that are not in a rapid degradation stage can be reused; otherwise, they will be recycled directly. More importantly, this approach fills the gap in battery detection regarding degradation stage diagnosis. 4.2. Performance of physics similarity-based in-operando battery life prediction 4.2.1.
The reason for the low accuracy is that there are only two batteries under the CY35–0.5/1 condition in the NCA battery, and the inconsistency between them is large. In addition, the results of the proposed degradation stage detection method on the (LFP)/graphite battery dataset are shown in Fig. A4.
Based on feature engineering, battery degradation stage detection and physical similarity analysis are used as the first two steps of life prediction. Battery data from the same aging stage and similar physics to the test data are clustered into subgroups, and the prediction model is subsequently established. Fig. 1.
After batteries are grouped, the differences among cells cause different attenuation rates of each cell, thus affecting the service life of the battery pack. The life of the battery pack depends on the cell with the shortest life. The health of lithium-ion batteries will continue to deteriorate after long-term recycling.
First, for the first time, a degradation stage detection method that does not involve accessing historical data is proposed; this method can quickly classify retired batteries, particularly by detecting whether the current cycle is in a rapid degradation stage.