Data-driven lifetime extension and performance optimization for vehicle battery systems

The lifetime of traction batteries is a major bottleneck for more penetration of the market share for electric vehicles, while prolonging the battery's lifetime is even more valuable and meaningful.

The existing methods are still focused on the single-cell level and treat the battery pack as a bulky cell, without systematically studying the compound effects in cell connection.

This project aims to bridge these two critical gaps by investigating the complex problem of pack-level battery aging estimation and prediction.

By utilizing vehicle field data and laboratory cycling data, machine learning algorithms will be developed to comprehend and predict battery aging behavior.

Subsequently, innovative methods will be developed to optimize the proper usage window and optimally control the battery operating constraints, leading to significantly extended battery lifetime.