An Empirical Capacity Degradation Modeling and Prognostics of Remaining Useful Life of Li-ion Battery using Unscented Kalman Filter

In battery power, electric vehicle lifetime and the cost of the battery are the primary concern and challenge in the fast evolving and promising uptake of an electric vehicle. Battery aging leads to gradually deteriorates in battery performance hence identification of battery states and health is essential. Battery health identification helps the customer to maintain and replace batteries in advance to prevent the loss caused by the unexpected failure of these batteries and help in reducing maintenance cost. This paper presents and compares two empirical models to describe battery degradation behaviors over its lifespan. Validation of the accuracy of the empirical model is performed by using experimental life-cycle test data. Subsequently, the most accurate model has been used in an Unscented Kalman Filter to predict battery remaining useful life (RUL). This work provides the initial step towards the development of battery health management system.