A Novel Charging Algorithm for Lithium-Ion Batteries Based on Enumeration-Based Model Predictive Control

Abstract

Lithium-ion (Li-ion) batteries play a substantial role in energy storage solutions for modern-day technologies such as hand-held consumer electronics, aerospace, electric vehicles, and renewable energy systems. For Li-ion batteries, designing a high-quality battery charging algorithm is essential since it has significant influences on the performance and lifetime of Li-ion batteries. The objectives of a high-performance charger include high charging efficiency, short charging time, and long cycle life. In this paper, a model predictive control based charging algorithm is proposed, the presented technique aims to simultaneously reduce the charging time, and the temperature rise during charging. In this study, the coulomb counting method is utilized to calculate the future state-of-charge and an artificial neural network trained by experimental data is also applied to predict the future temperature rise. Comparing with the widely employed constant current-constant voltage charging method, the proposed charging technique can improve the charging time and the average temperature rise by 1.2 % and 4.13 %, respectively.

Keywords

• Model predictive control
• artificial neural network
• lithium-ion battery