A new method to estimate the state of charge of the green battery. (December 2017)

• Record Type: Journal Article
• Authors: Li, Ling-Ling; Ren, Yu-Han; Wang, Ching-Hsin; Jen, Ching-Tsung
• Is Part Of: Microelectronics and reliability. Volume 79(2017)
• Page Start: 306
• Page End: 313
• Publication Date: 2017-12
• Subjects: Estimation -- State of charge -- Open-circuit voltage -- Extended Kalman filter; Electronic apparatus and appliances -- Reliability -- Periodicals; Miniature electronic equipment -- Periodicals; Appareils électroniques -- Fiabilité -- Périodiques; Équipement électronique miniaturisé -- Périodiques; Electronic apparatus and appliances -- Reliability; Miniature electronic equipment; Periodicals; 621.3815
• DOI: 10.1016/j.microrel.2017.07.031
• Languages: English
• ISSNs: 0026-2714

Abstract

Abstract: Green batteries have attracted great attention due to the characteristics of its high performance and non-pollution. In order to understand the working condition of the batteries and get a better estimation effect on the state of charge (SoC), the following works had been done in NMC18650 lithium ion battery. Firstly, the hybrid pulse power characteristic (HPPC) test was carried out on the battery with different currents. The extended Kalman filter (EKF) was used to estimate the SoC of the battery based on combined model and Thevenin model whose parameters were identified in advance; furthermore, the estimation results of the two models were compared. Secondly, an improved open circuit voltage (OCV) based method was proposed. Its improvements were as follows: the changes of OCV on battery were recorded during the current interruption, and it was assumed that the OCV had been restored to a certain degree if the change of OCV did not exceed 0.001 V in 10 s. Finally, two new improved methods were proposed based on the combined model, and the estimation effects of the above methods were compared under dynamic condition. The results showed that the accuracy of the Thevenin model was slightly higher than that of the combined model, and the accuracies of the two improved methods were both improved. Especially the second improved method had the least error and the best adaptability; the maximum error under dynamic conditions was 3.07%, and the average error was less than 1%, which only accounted for 22.46% of the un- improved. The improved OCV based method proposed in this study is applied to the SoC estimation of batteries, which greatly improves the accuracy of the estimation; moreover, the method is easy to implement and suitable for estimating SoC in real time. Highlights: The hybrid pulse power characteristic test is conducted in the lithium battery to identify the parameters of its model and study the changing trends of the parameters. Two equivalent circuit models are selected to estimate SoC of the battery. An improved open-circuit voltage based method is proposed for the EMF estimation. Two methods based on extended Kalman filter and the improved open-circuit voltage based method are proposed to estimate SoC of the battery.