A comparative study of different deep learning algorithms for lithium-ion batteries on state-of-charge estimation. (15th January 2023)

• Record Type: Journal Article
• Authors: Guo, Shanshan; Ma, Liang
• Is Part Of: Energy. Volume 263:Part C(2023)
• Publication Date: 2023-01-15
• Subjects: Lithium-ion battery -- State of charge estimation -- Battery management -- Deep learning; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2022.125872
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: -State-of-charge (SOC) plays a fundamental role in guiding battery management strategies. Recently, a variety of deep learning methods have been successfully applied in SOC estimation with impressive estimation accuracy. Nevertheless, the pros and cons of deep-learning estimators remain unexplored. This work investigates the performance of four state-of-the-art deep learning algorithms in the context of SOC estimation, including the fully connected neural network (FCNN), long short-term memory (LSTM), gate recurrent unit (GRU) and temporal convolutional network (TCN). Two kinds of lithium-ion batteries are tested by using specific devices programmed with dynamic drive cycles. The four methods are then evaluated regarding the accuracy by using experimental data collected at 25 °C. Afterwards, their robustness is evaluated at various temperatures with noise-polluted input data. The battery chemistries are also taken into consideration to assess their generalization performance. Finally, the computational costs are quantified to evaluate the efficiency of the four algorithms. Our results indicate that the LSTM, GRU, and TCN are superior to the FCNN in terms of accuracy. The TCN is the most robust one while the GRU has the shortest time at each time step among the three methods. Highlights: Four state-of-the-art deep learning algorithms on SOC estimation are compared and evaluated. Experiments of two kinds of batteries are carried out comprehensively considering various operation situations. The four methods for battery SOC estimation are explained on mathematical perspective. The results can guide the BMS choosing appropriate algorithms for a certain operation situation.