Improved multi-time scale lumped thermoelectric coupling modeling and parameter dispersion evaluation of lithium-ion batteries. (15th October 2022)

• Record Type: Journal Article
• Authors: Shi, Haotian; Wang, Shunli; Fernandez, Carlos; Yu, Chunmei; Xu, Wenhua; Dablu, Bobobee Etse; Wang, Liping
• Is Part Of: Applied energy. Volume 324(2022)
• Publication Date: 2022-10-15
• Subjects: Modeling battery dynamics -- Lumped thermoelectric coupling model -- Multi-time scale effects -- Multi-feature separation modeling -- Extended Kalman filtering; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2022.119789
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: A lumped thermoelectric coupling model is established at multi-time scale. A multi-feature separation modeling idea suitable for batteries is proposed. A strong coupling adaptive asynchronous identification strategy is designed. Three coupled filters on different time scales are used for model identification. Model performance evaluation is implemented using the concept of dispersion. Abstract: The rapid development of new energy fields such as electric vehicles and smart grids has put forward higher requirements for the power management of battery integrated systems. Considering that internal temperature and parameter consistency are important factors affecting battery safety and state estimation accuracy, a lumped thermoelectric coupling model based on the multi-time scale effects of battery dynamics parameter is established in this paper. On this basis, a new multi-feature separation modeling idea is proposed and adopted to complete the development of the strong coupling adaptive asynchronous identification strategy to realize the solution of the model. Specifically, the high-frequency and low-frequency characteristics of the resistor–capacitor link under different time constants are distinguished on different time scales. Three sub-filters based on forgetting factor recursive least squares, extended Kalman filtering and joint Kalman filtering are used to realize the adaptive asynchronous synergistic estimation of battery high-frequency dynamics parameter, low-frequency dynamics parameter and internal temperature. In addition, the filters at different time scales are strongly coupled through the voltage response on the diffusion impedance, and the time scale drive under slow dynamics depends on the current distribution of the test conditions. The experimental results of two long-term cycles show that the proposed strategy exhibits excellent terminal voltage tracking effect and internal temperature estimation accuracy. Finally, the concept of parameter dispersion is proposed and discussed. Compared with the results under the traditional identification method, the proposed strategy reduces the maximum parameter dispersion by 51.9%.