Evaluation method for consistency of lithium-ion battery packs in electric vehicles based on the Mahalanobis-Taguchi system

The consistency of battery packs is vital for safety and reliability during electric vehicle (EV) operations. Many consistency evaluation methods based on laboratory conditions are time-consuming and difficult to implement in the real-world. This study proposes an evaluation method for the consistency of lithium-ion battery packs in EVs based on the Mahalanobis-Taguchi system (MTS). First, a Douglas-Peucker (DP) algorithm was developed to compress high-dimensional cell voltage data, which reduced the feature extraction time by 81.64 %. Next, the consistency features were extracted based on sample entropy and fast-dynamic time warping (Fast-DTW) to quantify inconsistencies in battery parameters. Finally, the consistency evaluation method of the two-level warning based on MTS was established, and the hierarchical inconsistency warning threshold was identified based on the Chebyshev theorem. The results showed that the inconsistencies in battery packs were identified and quantitatively evaluated, and the occurrences of inconsistencies are occasionally discontinuous. The inconsistencies in battery packs were detected at high state of charge (SOC) levels at the end of charging. This method can evaluate the consistency of battery packs online based on EV operation data to monitor battery safety and provide detailed information for maintenance.