Capacity estimation algorithm with a second-order differential voltage curve for Li-ion batteries with NMC cathodes. (15th September 2017)

• Record Type: Journal Article
• Authors: Goh, Taedong; Park, Minjun; Seo, Minhwan; Kim, Jun Gu; Kim, Sang Woo
• Is Part Of: Energy. Volume 135(2017)
• Page Start: 257
• Page End: 268
• Publication Date: 2017-09-15
• Subjects: Differential voltage curve -- Prominence peak -- Curve alignment -- Cycle aging; 00-01 -- 99-00; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2017.06.141
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Accurate diagnosis of battery degradation is important for safe and efficient battery management. Capacity is a reliable index to describe the state of health (SOH) in batteries. In this paper, a capacity estimation algorithm for Li-ion batteries with nickel, manganese, and cobalt (NMC) cathodes based on a second-order differential voltage is proposed. A reference voltage curve was obtained during the CC charging phase from a fresh battery beforehand, and the input voltage curve was measured and compared, under the same operating conditions, from an aged battery. The input voltage curve is aligned to the reference curve to minimize the error of the second-order differential voltage. The compensated charging time of the aligned curve has a linear relation with the battery capacity until capacity reduction reaches 23.5%. From the linear model, the capacity can be estimated easily. This method is verified for five packs aged with different discharge currents. In the aging cycle and the initial SOC variation test, the capacity estimation error is less than 2% until it reaches 76.5% capacity. The proposed method does not require a complete aging test (for the table) to relate the charging time and the capacity. Highlights: Superior battery management calls for accurate diagnosis of battery degradation. A new algorithm was created to provide improved state of health via capacity. A second-order differential voltage curve was used instead of a first order curve. A reference update method reduces error form nonlinear changes in voltage curves. Getting results using the new method does not require complete aging tests.