Battery Fault Diagnosis for Electric Vehicles Based on Voltage Abnormality by Combining the Long Short-Term Memory Neural Network and the Equivalent Circuit Model

Da Li,Zhaosheng Zhang,Peng Liu,Zhenpo Wang,Lei Zhang

DOI: https://doi.org/10.1109/tpel.2020.3008194

IF: 5.967

2021-02-01

IEEE Transactions on Power Electronics

Abstract:Battery fault diagnosis is essential for ensuring safe and reliable operation of electric vehicles. In this article, a novel battery fault diagnosis method is presented by combining the long short-term memory recurrent neural network and the equivalent circuit model. The modified adaptive boosting method is utilized to improve diagnosis accuracy, and a prejudging model is employed to reduce computational time and improve diagnosis reliability. Considering the influence of the driver behavior on battery systems, the proposed scheme is able to achieve potential failure risk assessment and accordingly to issue early thermal runaway warning. A large volume of real-world operation data is acquired from the National Monitoring and Management Center for New Energy Vehicles in China to examine its robustness, reliability, and superiority. The verification results show that the proposed method can achieve accurate fault diagnosis for potential battery cell failure and precise locating of thermal runaway cells.

engineering, electrical & electronic

What problem does this paper attempt to address?

This paper proposes a new method for diagnosing faults in electric vehicle (EV) batteries by combining long short-term memory recurrent neural networks (LSTM) and an equivalent circuit model (ECM). The study utilizes an improved adaptive boosting method to enhance diagnostic accuracy and incorporates a predictive model to reduce computation time and improve diagnostic reliability. Taking into account the influence of driver behavior on the battery system, this method is able to assess potential fault risks and issue thermal runaway warnings in advance.

The main contributions of this paper are as follows:
1. A novel battery fault diagnosis model that combines LSTM and ECM is proposed, which accurately localizes potential faulty battery cells by coupling the advantages of modified adaptive boosting (MAB) coupling module.
2. Extensive training and validation using actual operational data sourced from China's National Monitoring and Management Center for New Energy Vehicles are conducted, improving the diagnostic capability of the model for vehicles with the same specifications.
3. An approximate optimization method (AOM) is proposed to optimize the hyperparameters of the model and enhance diagnosis performance.
4. Real-time fault diagnosis strategies and fault alarm protocols are designed and their effectiveness is validated using real-world EV data, particularly in the context of fault occurrences and thermal runaway accidents.

The structure of the paper includes an introduction to the battery monitoring center and actual data acquisition, an overview of voltage prediction methods including LSTM, ECM, MAB, and the predictive module, a detailed discussion of voltage prediction results, and an analysis of the actual fault diagnosis strategy and its robustness, reliability, and superiority. Finally, key conclusions are summarized.