Optimization of battery charging strategy based on nonlinear model predictive control. (15th February 2022)

• Record Type: Journal Article
• Authors: Wang, Yujie; Zhou, Caijie; Chen, Zonghai
• Is Part Of: Energy. Volume 241(2022)
• Publication Date: 2022-02-15
• Subjects: Charging optimization -- Thermoelectric coupling model -- Nonlinear model predictive control -- Parameter identification -- Lithium-ion batteries; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2021.122877
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: With the increased applications of lithium-ion batteries in energy storage systems and electric vehicles, there is a growing demand for battery energy storage systems and management systems. Considering that the temperature especially internal temperature significantly can affect the performance and safety of the battery, a triple-objective optimization charging method which can reduce the cell charging time, energy loss, and internal temperature rise is proposed based on a thermoelectric coupling model in this paper. Specifically, a thermoelectric coupling model suitable for a wide temperature range from −5 °C to 45 °C is formulated. On this basis, a nonlinear model predictive control framework is proposed to obtain the real-time charging current by solving the nonlinear optimization problems. The impacts of the objective function weights and internal temperature thresholds on the charging result are discussed through experiments, and another multi-stage constant current charging method is conducted as a comparison. Results show that the nonlinear model predictive control can achieve a good balance between three objectives while satisfying constraints. Compared with the traditional multistage constant current charging method, the proposed strategy can reduce energy loss by 150 J and temperature rise by 1–2 °C in similar charging time. Highlights: A specific thermoelectric coupling model for lithium-ion battery is formulated. The relationship between electrical and thermal sub-models is established and verified. A triple objective function considering different constraints is proposed. A nonlinear model predictive control framework is established for optimal charging.