Capacity Fade Estimation for Li-ion Battery Packs

As the electric vehicle market is growing at a rapid pace, the need for accurate state of charge (SOC), state of health (SOH) estimation of the battery pack is required. Moreover, the battery management system should be able to estimate the capacity precisely with limited data from the vehicle. Currently, the different tests performed on the battery take from several hours to a few days to collect the data necessary for such estimation. In order to achieve precise OCV-SOC curves, characterization tests are executed at lower C rates from 100% to 0% SOC, which requires prolonged testing time. Additionally, more data takes longer time to process and train the system for estimating the battery states. This would prove inefficient for practical implementation. Therefore, carrying out tests at much higher C rates and reducing the SOC range will considerably reduce the time required for characterization. This thesis aims to model, predict the open circuit voltage (OCV) curve for incremental capacity analysis (ICA), and estimate capacity fade in Li-ion battery packs using relatively few data points. The study is carried out for battery packs at various temperatures.