Big Data-Based Early Fault Warning of Batteries Combining Short-Text Mining and Grey Correlation

Considering the battery-failure-induced catastrophic events reported frequently, early fault warning of batteries is essential to safety electric vehicles (EVs). Motivated by this, a novel data-driven method for early-stage battery-fault proposed in this paper fusion short-text mining and grey correlation. In particular, approach exploited identify information recorded maintenance service documents further analyze categories battery faults EVs statistically. The correlation algorithm employed build relevance between vehicle states typical faults, which contributes extracting key features corresponding failures. A fault-prediction model buses based on big data then established feature variables. Different selections kernel functions hyperparameters are scrutinized optimize performance warning. validated with real-world acquired from operation. Results suggest that constructed prediction can effectively predict carry out desired