Discharge capacity estimation for Li-ion batteries based on particle filter under multi-operating conditions. (15th June 2015)

• Record Type: Journal Article
• Authors: Li, Junfu; Wang, Lixin; Lyu, Chao; Zhang, Liqiang; Wang, Han
• Is Part Of: Energy. Volume 86(2015)
• Page Start: 638
• Page End: 648
• Publication Date: 2015-06-15
• Subjects: Li-ion rechargeable batteries -- Discharge capacity estimation -- Particle filter -- Feature parameters -- Artificial neural network -- Multi-operating conditions; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2015.04.021
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: In recent years, Li-ion rechargeable batteries are well liked to be used in BMS (battery management system) of EV (electrical vehicle) and satellite due to various advantages. As battery is aging during the whole life cycles, it is essential to estimate discharge capacity to ensure high performance. This paper presents a discharge capacity estimation model for Li-ion battery based on PF (particle filter). To discover effects of different operating conditions on capacity, LiCoO2 cells are designed to experience aging and characteristic tests alternatively. The contributions of this paper are listed below: (i) four feature parameters extracted from charging voltage curves are selectively used for modeling; (ii) under certain aging condition, the model verifies the applicability for LiCoO2 battery with high estimation accuracy; (iii) the adoption of ANN (artificial neural network) helps to mine the nonlinear relationship between discharge capacities and multi-operating conditions. Validation result indicates that the proposed method is able to accurately estimate discharge capacity under multi-operating conditions. Highlights: Four parameters extracted from voltage curves are selectively used for modeling. PF model provides high accuracy of estimations under certain aging conditions. ANN is used to mine relationships between capacity biases and operating conditions. Validation results for multi-operating conditions is presented.