Abstract：Lithium-ion batteryaging tests are primarily used to study battery aging performance. However, aging experiments for a large number of batteries take time and effort. Therefore, combined with Wasserstein GAN and gradient penalty (WGANGP), the Variational deep embedding (VaDE) model is used to form a VaDE-WGANGP architecture. Then, a method of battery aging performance modelling and data generation is designed. This paper aims to generate simulated data onexternal characteristics (such as battery voltage and charge/discharge current during working). The VaDE-WGANGP generation model is effective during different battery SOH intervals, which indicates that simulated battery data can be obtained across the whole battery life span (SOH∈[100%, 80%]) and coversbattery aging performance. Furthermore, battery external characteristics differ for different battery cells even under the same SOH state. Accordingly, the diversities in battery external characteristics among different battery cells under the same SOH states are considered. An open-source battery test data set, in cludingdata from Toyota Research Institute (TRI) in partnership with MIT and Stanford and working data on LFP/graphite cells, is used. Firstly, the three external characteristics (voltage, current, and discharge capacity during the discharge process) are selected as the input of the model. The latent space by the VaDE encoder can be mapped, and its distribution with specific rules can be obtained through optimization. Then, this distribution space is sampled, and the latent variables obtained by sampling are used as input into the VaDE decoder for data generation. SOH estimation tests show that the VaDE-WGANGP has good generation performance for battery external characteristics, and the primary battery working performance is simulated during the aging process. Besides, effective and extended battery data resources can be provided by this measure for data-driven algorithms, especially in case the primitive data amount is insufficient. The following conclusions can be drawn. (1) Regarding battery data clustering and battery extrinsic characteristics simulation generation, the integrated model of VaDE-WGANGP has better performance than VAE and VaDE. The statistical characteristics of battery extrinsic characteristics in different SOH intervals are accurate. (2) VaDE-WGANGP can generate high-quality simulation data of battery external characteristics. The SOH estimation results prove that the data generated by VaDE-WGANGP is of high quality. This paper provides a novel idea for analyzing the external characteristics of batteries during aging. Given the shortage of original data in data-driven technology, this scheme can also provide an effective method for data extension and is helpful for battery screening.
李弈, 张金龙, 漆汉宏, 魏艳君, 张迪. 基于变分深度嵌入-带有梯度惩罚的生成对抗网络的锂离子电池老化特性建模[J]. 电工技术学报, 2024, 39(13): 4226-4239.
Li Yi, Zhang Jinlong, Qi Hanhong, Wei Yanjun, Zhang Di. Ageing Performance Modeling of Li-Ion Batteries Based on Variational Deep Embedding-Wasserstein GAN with Gradient Penalty. Transactions of China Electrotechnical Society, 2024, 39(13): 4226-4239.
[1] 郭向伟, 邢程, 司阳, 等. RLS锂电池全工况自适应等效电路模型[J]. 电工技术学报, 2022, 37(16): 4029-4037. Guo Xiangwei, Xing Cheng, Si Yang, et al.RLS adaptive equivalent circuit model of lithium battery under full working condition[J]. Transactions of China Electrotechnical Society, 2022, 37(16): 4029-4037. [2] 吴立峰, 刘昊, 林仲钦, 等. 低温环境下锂离子电池荷电状态与超声透射飞行时间的关系研究[J]. 电工技术学报, 2022, 37(21): 5617-5626. Wu Lifeng, Liu Hao, Lin Zhongqin, et al.Relationship between state of charge of lithium-ion battery and ultrasonic transmission flight time at low tempera- ture[J]. Transactions of China Electrotechnical Society, 2022, 37(21): 5617-5626. [3] Iurilli P, Brivio C, Wood V.On the use of elec- trochemical impedance spectroscopy to characterize and model the aging phenomena of lithium-ion batteries: a critical review[J]. Journal of Power Sources, 2021(505): 229860. [4] Tian Huixin, Qin Pengliang, Li Kun, et al.A review of the state of health for lithium-ion batteries: Research status and suggestions[J]. Journal of Cleaner Production, 2020(261): 120813. [5] 武龙星, 庞辉, 晋佳敏, 等. 基于电化学模型的锂离子电池荷电状态估计方法综述[J]. 电工技术学报, 2022, 37(7): 1703-1725. Wu Longxing, Pang Hui, Jin Jiamin, et al.A review of SOC estimation methods for lithium-ion batteries based on electrochemical model[J]. Transactions of China Electrotechnical Society, 2022, 37(7): 1703-1725. [6] Liu Lin, Park J, Lin Xianke, et al.A thermal- electrochemical model that gives spatial-dependent growth of solid electrolyte interphase in a Li-ion battery[J]. Journal of Power Sources, 2014(268): 482-490. [7] Dong Guangzhong, Wei Jingwen.A physics-based aging model for lithium-ion battery with coupled chemical/mechanical degradation mechanisms[J]. Electrochimica Acta, 2021(395): 139133. [8] Ma Lili, Xu Yonghong, Zhang Hongguang, et al.Co-estimation of state of charge and state of health for lithium-ion batteries based on fractional-order model with multi-innovations unscented Kalman filter method[J]. Journal of Energy Storage, 2022(52): 104904. [9] 孙金磊, 唐传雨, 李磊, 等. 基于状态与模型参数联合估计的老化电池可充入电量估计方法[J]. 电工技术学报, 2022, 37(22): 5886-5898. Sun Jinlei, Tang Chuanyu, Li Lei, et al.An estimation method of rechargeable electric quantity for aging battery based on joint estimation of state and model parameters[J]. Transactions of China Electrotechnical Society, 2022, 37(22): 5886-5898. [10] Li Shi, Pischinger S, He Chaoyi, et al.A comparative study of model-based capacity estimation algorithms in dual estimation frameworks for lithium-ion batteries under an accelerated aging test[J]. Applied Energy, 2018(212): 1522-1536. [11] 王义军, 左雪. 锂离子电池荷电状态估算方法及其应用场景综述[J]. 电力系统自动化, 2022, 46(14): 193-207. Wang Yijun, Zuo Xue.Review on estimation methods for state of charge of lithium-ion battery and their application scenarios[J]. Automation of electric power systems, 2022, 46(14): 193-207. [12] 黄凯, 丁恒,郭永芳, 等. 基于数据预处理和长短期记忆神经网络的锂离子电池寿命预测[J]. 电工技术学报, 2022, 37(15): 3753-3766. Huang Kai, Ding Heng, Guo Yongfang, et al.Prediction of remaining useful life of lithium-ion battery based on adaptive data preprocessing and long short-term memory network[J]. Transactions of China Electrotechnical Society, 2022, 37(15): 3753-3766. [13] Wang Jiwei, Deng Zhongwei, Yu Tao, et al.State of health estimation based on modified Gaussian process regression for lithium-ion batteries[J]. Journal of Energy Storage, 2022(51): 104512. [14] Yang Fangfang, Wang Dong, Xu Fan, et al.Lifespan prediction of lithium-ion batteries based on various extracted features and gradient boosting regression tree model[J]. Journal of Power Sources, 2020(476): 228654. [15] Severson K A, Attia P M, Jin N, et al.Data-driven prediction of battery cycle life before capacity degradation[J]. Nature Energy, 2019, 4(5): 1-9. [16] Kingma D P, Welling M.Auto-encodingvariational Bayes[C]//Proceedings of the International Con- ference on Learning Representations (ICLR), Banff, 2014: 1-14. [17] Goodfellow I J, Pouget-Abadie J, Mirza M, et al.Generative adversarial networks[J]. Advances in Neural Information Processing Systems, 2014, 3: 2672-2680. [18] Kim S, Choi Y Y, Choi J.Impedance-based capacity estimation for lithium-ion batteries using generative adversarial network[J]. Applied Energy, 2022(308): 118317. [19] Jiao Ruihua, Peng Kaixiang, Dong Jie.Remaining useful life prediction of lithium-ion batteries based on conditional variational autoencoders-particle filter[J]. IEEE Transactions on Instrumentation and Mea- surement, 2020, 69(11): 8831-8843. [20] Huang Y, Tang Y, Vanzwieten J.Prognostics with variational autoencoder by generative adversarial learning[J]. IEEE Transactions on Industrial Elec- tronics, 2022, 69(1): 856-867. [21] 梁凤勤. 基于迁移学习的锂离子电池健康状态估计方法研究[D]. 成都: 电子科技大学, 2021. Liang Fengqin.Research on the method of estimating the health state of lithium-ion battery based on transfer learning[D]. Chengdu: University of Elec- tronic Science and Technology, 2021. [22] 李龙, 燕旭朦, 张钰生, 等. 小样本锂电池数据SOC估算方法[J/OL]. 西安交通大学学报, 2023. 网络首发:https://kns.cnki.net/kcms2/detail/61.1069.T. 20230525.1747.007.html. Li Long, Yan Xumeng, Zhang Yusheng, et al. Lithium battery SOC estimation method based on transfer learning and deep learning[J/OL]. Journal of Xi’an Jiaotong University, 2023. https://kns.cnki.net/kcms2/ detail/61.1069.T. 20230525.1747.007.html. [23] Ni Zichuan, Li Biao, Yang Ying.Deep domain adaptation network for transfer learning of state of charge estimation among batteries[J]. Journal of Energy Storage, 2023(61): 106812. [24] Ma Jian, Shang Pengchao, Zou Xinyu, et al.A hybrid transfer learning scheme for remaining useful life prediction and cycle life test optimization of different formulation Li-ion power batteries[J]. Applied Energy, 2021(282): 116167. [25] Jiang Zhuxi, Zheng Yin, Tan Huachun, et al.Variational deep embedding: an unsupervised and generative approach to clustering[C]//26th Inter- national Joint Conference on Artificial Intelligence (IJCAI), Melbourne, 2017: 1965-1972. [26] Gulrajani I, Ahmed F, Arjovsky M, et al.Improved training of wasserstein GANs[C]//Advances in Neural Information Processing Systems, Long Beach: CA, USA, Curran Associates, Inc, 2017: 5767-5777. [27] Larsen A, Sønderby S K, Larochelle H, et al.Auto- encoding beyond pixels using a learned similarity metric[C]//Proceedings of Machine Learning Research, New York, 2016: 1558-1566. [28] Arjovsky M, Chintala S, Bottou L.Wasserstein GAN[C]//Proceedings of the 34th International Confer- ence on Machine Learning, PMLR, Sydney, Australia, 2017, 70: 214-223. [29] Van Der Maaten L, Hinton G. Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008(9): 2579-2605.
2024, Vol. 39 
