Fault diagnosis of polar buoy energy systems using an Arrhenius-corrected second-order RC model and deep learning

Abstract

Polar ice–sea amphibious ecological observation buoys are prone to energy system faults under harsh operating conditions, including low temperatures, low-irradiance conditions, and long-term unattended operation. Faults such as ohmic resistance (R₀) increase, capacity degradation, leakage fault, and polarization-resistance increase can reduce sensor uptime, interrupt data transmission, and cause missing observation data. In this study, we develop a temperature-dependent second-order resistor–capacitor (RC) model for a 4S40P LiFePO₄ battery pack using Arrhenius-based temperature correction to address the scarcity of field fault samples and labeled fault data. Representative load profiles are then introduced into the developed model, and fault injection is used to generate time-series data under different operating states. On this basis, a 1D-CNN-BiGRU fault diagnosis model is constructed using terminal voltage (V), terminal current (I) and their first-order differences (dV) and (dI) as four-channel time-series inputs. Simulation results from five independent training runs show that the 1D-CNN-BiGRU achieves an accuracy of 0.9593 ± 0.0076, a Macro-F1 score of 0.9594 ± 0.0076, and a weighted-F1 score of 0.9593 ± 0.0076. These results indicate consistently high classification performance across the five operating states—normal, R₀ increase, capacity degradation, leakage fault, and polarization resistance increase. Robustness analysis further shows that uncertainty-aware training with bounded temperature-dependent voltage perturbations substantially improves classification robustness under moderate battery-model uncertainty, although severe low-temperature uncertainty remains challenging. Overall, this study provides a simulation-based framework for fault diagnosis and robustness evaluation in unattended polar buoy energy systems.

Keywords

Polar buoy energy system / Arrhenius-based correction / Second-order RC model / 1D-CNN-BiGRU / Fault diagnosis

References

[1] Belgibayeva A, Rakhmetova A, Rakhatkyzy M, Kairova M, Mukushev I, Issatayev N, et al.Lithium-ion batteries for low-temperature applications: limiting factors and solutions. J Power Sources, 2023, 557232550

[2] Gao ZC, Chin CS, Woo WL, Jia JBIntegrated equivalent circuit and thermal model for simulation of temperature-dependent LiFePO₄ battery in actual embedded application. Energies, 2017, 10: 85.

[3] He GH, Xue JF, Zhao CK, Cui T, Liu CGDeep learning based short-term motion prediction of floating wind turbine under shutdown condition. Appl Ocean Res, 2024, 151104147

[4] Ji ZQ, Hao Q, Lan MS, Lou LW, Zuo GY, Zhang TZ, et al.Automated under-ice monitoring system for fish detection in Central Arctic Ocean. Intell Mar Technol Syst, 2026, 47.

[5] Kawakita de Souza A (2021) Lithium-ion battery OCV and dynamic test data of a LiFePO4 cylindrical cell. Mendeley Data, V1. https://doi.org/10.17632/p8kf893yv3.1

[6] Lee CM, DeGrandpre M, Guthrie J, Hill V, Kwok R, Morison J, et al.Emerging technologies and approaches for in situ, autonomous observing in the Arctic. Oceanography, 2022, 35210-221.

[7] Luo XY, Zhang SX, Wang XY, Guan XPIntegrated-energy ship-power optimization scheduling considering the uncertainty of photovoltaic output. Intell Mar Technol Syst, 2026, 4: 6.

[8] Mei P, Zhang Y, Zhang WLow-temperature lithium-ion batteries: challenges and progress of surface/interface modifications for advanced performance. Nanoscale, 2023, 15987-997.

[9] Senemmar S, Zhang J (2021) Deep learning-based fault detection, classification, and locating in shipboard power systems. In: 2021 IEEE Electric Ship Technologies Symposium. IEEE, pp 1–6. https://doi.org/10.1109/ESTS49166.2021.9512342

[10] Sun B, Qi X, Song D, Ruan HReview of low-temperature performance, modeling and heating for lithium-ion batteries. Energies, 2023, 167142

[11] Tsai CM, Wang CS, Chung YJ, Sun YD, Peng JWMultisensor fusion time-frequency analysis of thruster blade fault diagnosis based on deep learning. IEEE Sens J, 2022, 22: 19761-19771.

[12] Wang QY, Wen Q, Wei ZQ, Mao HY, Tao RT, Zhang HEnergy-efficient reconfigurable FPGA accelerator for marine convolutional neural network-long short-term memory applications. Intell Mar Technol Syst, 2026, 4: 9.

[13] Wen L, Li XY, Gao L, Zhang YYA new convolutional neural network-based data-driven fault diagnosis method. IEEE Trans Ind Electron, 2018, 655990-5998.

[14] Xu YD, Hu MH, Fu CY, Cao KB, Su Z, Yang ZState of charge estimation for lithium-ion batteries based on temperature-dependent second-order RC model. Electronics, 2019, 81012

[15] Yu JB, Zhang YChallenges and opportunities of deep learning-based process fault detection and diagnosis: a review. Neural Comput Appl, 2023, 35: 211-252.

[16] Zhao Y, Yin B, Sang GF, Jiang ZX, Li HS, Wu BL, et al.Applications of optical fiber sensors in marine observation: a review. Intell Mar Technol Syst, 2025, 338.

Funding

National Key Research and Development Program of China(2022YFC2807603)

National Natural Science Foundation of China(42306260)

Shanghai Frontiers Science Center of Polar Science(SOO2025-04)

Shanxi Province Water Conservancy Science and Technology R&D Service Project(2025GM22)

Rights & permissions

The Author(s)