A life-prediction method for lithium-ion batteries based on a fusion model and an attention mechanism
- PDF / 848,010 Bytes
- 8 Pages / 595.22 x 842 pts (A4) Page_size
- 79 Downloads / 161 Views
Vol.16 No.6, 1 November 2020
A life-prediction method for lithium-ion batteries based on a fusion model and an attention mechanism* WANG Xian-bao (⦻ᇚ**)؍, WU Fei-teng (伎㞮), and YAO Ming-hai (ည᰾⎧)ķ College of Information Engineering, Zhejiang University of Technology, Hangzhou 310023, China (Received 17 December 2019; Revised 5 February 2020) ©Tianjin University of Technology and Springer-Verlag GmbH Germany, part of Springer Nature 2020 The current life-prediction models for lithium-ion batteries have several problems, such as the construction of complex feature structures, a high number of feature dimensions, and inaccurate prediction results. To overcome these problems, this paper proposes a deep-learning model combining an autoencoder network and a long short-term memory network. First, this model applies the characteristics of the autoencoder to reduce the dimensionality of the high-dimensional features extracted from the battery data set and realize the fusion of complex time-domain features, which overcomes the problems of redundant model information and low computational efficiency. This model then uses a long short-term memory network that is sensitive to time-series data to solve the long-path dependence problem in the prediction of battery life. Lastly, the attention mechanism is used to give greater weight to features that have a greater impact on the target value, which enhances the learning effect of the model on the long input sequence. To verify the efficacy of the proposed model, this paper uses NASA's lithium-ion battery cycle life data set. Document code: A Article ID: 1673-1905(2020)06-0410-8 DOI https://doi.org/10.1007/s11801-020-9214-y
Lithium-ion batteries possess many advantages over traditional batteries, such as high output voltage, high energy density[1], low self-discharge rate, long cycle life, and high reliability[2]. They have been widely applied in vehicles, household equipment, communications, the aerospace industry, and so on[3]. Lithium-ion batteries that exceed their service life often lead to accidents, such as fires and explosions. The accurate prediction of lithium-ion batteries’ remaining useful life (RUL) plays an important role in state estimation and health management of lithium-ion batteries. The methods for predicting a lithium-ion battery’s RUL can be divided into three categories: experience-based, model-based, and data-driven. The experience-based method uses the battery’s history data to estimate its service life, but it cannot describe the physical and chemical changes inside the battery. This method is applicable only under special conditions and has gradually been abandoned in favor of other methods. The model-based method integrates the material characteristics, the decay mechanism, and the operating environment inside the battery[4]. However, due to the complexity of chemical reactions in lithium-ion batteries and the fact that the state of lithium-ion batteries is affected by temperature[5] and humidity in the working environment[6], predictions from th
Data Loading...
