Accurate state-of-health estimation can give valuable guidelines for improving the reliability and safety of energy storage system. In this article, a novel battery degradation tracking method is proposed through the fusion of significant health features with Gaussian process regression (GPR). First, an advanced filter method is used to smooth differential thermal voltammetry (DTV) curves. Thereafter, considering the relationship between battery degradation and DTV curves, some health factors are extracted from DTV curves. In this article, these health factors involve different dimensions of the DTV curve, including peak position, peak, and valley values. Third, a correlation analysis method is employed to select four high-quality features from health factors, which are fed into GPR to learn and establish a battery degradation model. Finally, the estimation accuracy, robustness, and reliability of the proposed model are verified using four batteries with different aging test conditions and health levels. The results demonstrate that the proposed model can provide accurate battery health status forecasting.

A thermal runaway happened in a battery cell of an electric vehicle during driving, and the fire spreads to other batteries in a few minutes. Based on the recorded battery operation data, this paper analyzes the battery thermal runaway process, and identifies the critical time points of fault signals like cell voltages and temperatures, etc., which depict the start and spread of thermal runaway in the battery pack. By tracing back the clues, it is concluded that the cause of the fire is battery internal short circuit (ISC), and battery overcharge is a potential cause of ISC. Then, three different patterns of battery cell voltage abnormalities are compared, including battery external short circuit, internal short circuit, and poor signal wire connection. A cell voltage difference-based method is proposed to diagnose ISC online, which is simple and reliable to implement in engineering.

Battery remaining useful life (RUL) estimation is critical to battery management and performance optimization of electric vehicles (EVs). In this paper, we present an effective way to estimate RUL online by using the support vector machine (SVM) algorithm. By studying the characteristics of the battery degradation process, the rising of the terminal voltage and changing characteristics of the voltage derivative (DV) during the charging process are introduced as the training variables of the SVM algorithm to determine the battery RUL. The SVM is then applied to build the battery degradation model and predict the battery real cycle numbers. Experimental results prove that the built battery degradation model shows higher accuracy and less computation time compared with those of the neural network (NN) method, thereby making it a potential candidate for realizing online RUL estimation in a battery management system (BMS).

Accurate detection and diagnosis battery faults are increasingly important to guarantee safety and reliability of battery systems. Developed methods for battery early fault diagnosis concentrate on short-term data to analyze the deviation of external features without considering the long-term latent period of faults. This work proposes a novel data-driven method to detect long-term latent fault and abnormality for electric vehicles based on real-world operation data. Specifically, the battery fault features are extracted from the incremental capacity curves, which are smoothed by advanced filter algorithms. Secondly, principal component analysis algorithm is utilized to reduce dimensionality and the cumulative percent variance is to determine the number of significant features. Based on the features, a cluster algorithm is employed to capture the battery potential failure information. Moreover, the cumulative root-mean-square-deviation is introduced to quantificationally analyze the degree of the battery failures using large-scale battery data to avoid the missing fault reports using short-term data. In addition, a battery system failure index is proposed to evaluate battery fault conditions. The results indicate that the proposed long-term feature analysis method can effectively detect and diagnose faults.

This paper focuses on state of charge (SOC) estimation for the battery packs of electric vehicles (EVs). By modeling a battery based on the equivalent circuit model (ECM), the adaptive extended Kalman filter (AEKF) method can be applied to estimate the battery cell SOC. By adaptively setting different weighed coefficients, a battery pack SOC estimation algorithm is established based on the single cell estimation. The proposed method can not only precisely estimate the battery pack SOC, but also effectively prevent the battery pack from overcharge and over-discharge, thus providing safe operation. Experiment results verify the feasibility of the proposed algorithm.

This paper develops an efficient fault diagnostic scheme for battery packs using a novel sensor topology and signal processing procedure. Cross-cell voltages are measured to capture electrical abnormalities and the recursive correlation coefficients between adjacent voltages are calculated to embody system state. Then the discrete wavelet packet transform is applied on the correlation sequences to extract diverse characteristic indexes, wherein the most representative components are refined as fault features by the principal component analysis. Afterwards, resorting to the multiclass relevance vector machine, sparse classification models are constructed to cognize fault patterns, and accordingly, fault types and grades are evaluated. Common faults, including external and internal short-circuit, thermal abuse and loose connection, are physically triggered on a series pack to acquire realistic dataset. Experimental verifications under different conditions and algorithmic configurations suggest that the proposed diagnosis scheme can give accurate and reliable assessments on different fault specifics, with a fault isolation success rate of 84% and a fault severity grading success rate of 90%.