A model-based state-of-charge estimation method for series-connected lithium-ion battery pack considering fast-varying cell temperature

• Published in: Energy (Oxford) 2019-10, Vol.185, p.847-861
• Main Authors: Huang, Deyang, Chen, Ziqiang, Zheng, Changwen, Li, Haibin
• Format: Article
• Language: English
• Subjects: Adaptive filters; Algorithms; Batteries; Cell culture; Computer applications; Extended Kalman filter; Fast-varying temperature; Inconsistency; Lithium; Lithium-ion batteries; Lithium-ion battery; Management systems; Mathematical models; Model-based; Open circuit voltage; Parameter identification; Power management; Rechargeable batteries; State of charge; Temperature dependence; Temperature effects
• ISSN: 0360-5442; 1873-6785
• DOI: 10.1016/j.energy.2019.07.063

Accurately estimating the state-of-charge (SOC) of lithium-ion batteries under complicated temperature conditions is crucial in all-climate battery management systems. This paper proposes a model-based SOC estimation method for series-connected battery pack with time-varying cell temperature. Systematic battery experiments are conducted to investigate the influences of changing temperature on both cell characteristics and cell-to-cell inconsistencies. A normalized open-circuit voltage (OCV) model is developed and applied in cell Thevenin model to describe the temperature-dependent OCV-SOC characteristic. The battery pack SOC is analyzed considering the effect of passive balance control. Then, a lumped parameter battery pack model is established by connecting cell models in series. To reduce computational complexity, a dual time-scale parameter identification framework is proposed which is supported by an online filtering process of selecting variable reference cell (VRC). An adaptive co-estimator is presented to update pack parameters in dual time-scale using an optimized recursive least squares algorithm, and to estimate the battery pack SOC using an extended Kalman filter. Experimental verifications are conducted under time-varying environmental temperature ranging from −40 °C to 40 °C. Results indicate the established model can well describe the dynamic behavior of battery pack, and the proposed method can estimate the battery pack SOC with considerably high precision. •Investigate cell-to-cell inconsistencies experimentally at different temperatures.•Establish a temperature-dependent normalized OCV model.•Select VRC online and develop a dual time-scale identification framework.•Propose a VFFRLS-EKF based co-estimator considering fast-varying cell temperature.•Prove the accuracy and robustness of proposed method experimentally.