State of charge and state of health online estimation algorithm for lithium-ion batteries applied to hybrid electric vehicles based on adaptative filtering

Nowadays, the need to use alternative and renewable energy sources has become a frequent agenda in the technological development of various segments. This includes the automotive sector, which has presented an exponential increase in the production and demand for electric vehicles in the last few years. Hybrid electric vehicles can be considered as an intermediate step for the transition from internal combustion vehicles to 100% electric vehicles once they use both electric and combustion engines. One of the biggest challenges currently observed in vehicle electrification relies on the energy storage system composed of batteries. The lithium-ion technology is the most used due to its efficiency, safety, and useful life. For lithium-ion batteries to operate safely and efficiently, a battery management system is required, which must be able to accurately estimate the state of charge and state of health, thus preventing the battery from being exposed to dangerous conditions adverse such as underdischarges or overcharges. Considering this focus, in this work the development of an algorithm for estimating the state of charge and state of health of lithium-ion batteries in applications with hybrid vehicles is presented. The methodology used was based on adaptive filtering algorithms through the implementation of a dual extended Kalman filter embedded in a Battery Management System and bench tested. The state of charge estimates resulted in a root mean square error of 0.8657% and a state of health mean absolute error of 1.4116%. Through this validation, the methodology proved efficient and robust even under operating conditions with high current variations in short periods, indicating that it is possible to use it in real applications with hybrid electric vehicles.