Development of a battery real-time state of health diagnosis based on fast impedance measurements. (June 2021)

• Record Type: Journal Article
• Authors: Locorotondo, Edoardo; Cultrera, Vincenzo; Pugi, Luca; Berzi, Lorenzo; Pierini, Marco; Lutzemberger, Giovanni
• Is Part Of: Journal of energy storage. Volume 38(2021)
• Publication Date: 2021-06
• Subjects: Lithium-ion battery -- Second life -- State of health -- Electrochemical impedance spectroscopy -- PRBS -- Electric vehicle; Energy storage -- Periodicals; Energy storage -- Research -- Periodicals; 621.3126
• DOI: 10.1016/j.est.2021.102566
• Languages: English
• ISSNs: 2352-152X

Abstract

Highlights: An algorithm suitable for real-time implementation was developed for battery health diagnosis. The estimation of battery SOH is based on fast impedance measurements. EIS test is performed by applying a low-cost prototype device and broadband excitation signals. EIS test is performed on a group of cells with different SOHs. Battery SOH identification is performed through data clustering. Abstract: The capability to assess and monitor the state of health (SOH) of lithium-based cells is a highly demanded feature for advanced battery management systems. Due to the existing relation between SOH and internal impedance, electrochemical impedance spectroscopy (EIS) methods are adopted for SOH diagnosis. Nevertheless, accurate EIS tests demand expensive facilities, long time test procedures, and algorithms with high-computational efforts, which makes them almost unsuitable for on-board systems. This paper presents a new diagnostic method aimed at detecting battery SOH using fast impedance measurements. Key factor is the application of a broadband current signal excitation on the battery; for the application here presented, a pseudo-random binary sequence (PRBS) excitation is adopted. To demonstrate the functionalities of a prototype testbed, several cells of the same manufacturer but presenting different SOHs, due to their past load history, have been subjected to the EIS test, acquiring voltage response under imposed excitation. Finally, test results have been processed: the key step being the clustering of impedance measurements (represented in Nyquist diagram) in different rectangular areas, which are related to actual SOH. The performed experimental test results showed the possibility to determine frequency points in which the impedance measurements dramatically change due to different cell SOH; as a consequence, these peculiar frequencies can be adopted as reference for cluster separation. According to the results here presented, the proposed method is sufficiently accurate and is a promising solution for real-time diagnostic of battery SOH.