Top

Noise Resistant Feature Fusion for Digital Twin Queue Control of Battery Swapping Stations Using DLinear and Kolmogorov Arnold Networks

Authors

Files

Abstract

Battery swapping stations coordinate vehicle queues and charged battery availability under missing records, sensor errors, and delayed communication. This paper presents an engineering research design that links noise-resistant feature fusion to forecasting and digital twin queue control. A decomposition-based linear predictor represents persistent temporal variation, while a Kolmogorov-Arnold network adjusts its residual forecast using inventory, service availability, and observation quality. Reliability-weighted feature groups condition the predictor before a constrained rolling-horizon controller evaluates feasible charging and service actions. The numerical reference comparison examines forecast error, waiting time, observation degradation, and operating sensitivity. The reported arrival mean absolute error is 2.8 vehicles per five-minute interval for the full configuration and 4.1 for the decomposition-only configuration. Mean waiting time is 7.2 minutes, compared with 10.8 minutes for reactive control, while the ninety-fifth percentile decreases from 24.6 to 16.8 minutes. At the highest evaluated noise level, forecast error is 4.3 vehicles per interval for the full configuration and 7.0 for the decomposition-only configuration. These comparisons describe the numerical case presented in the manuscript; independent empirical interpretation requires raw event records, documented training runs, and controller execution logs. The methodological contribution is an explicit interface between observation reliability, bounded forecast correction, and physical action feasibility. The design provides a basis for reproducible implementation, paired operational evaluation, and subsequent field validation.

References

Renga, D., Spoturno, F., & Meo, M. (2024). Reinforcement learning for charging scheduling in a renewable powered battery swapping station. IEEE Transactions on Vehicular Technology, 73(10), 14382–14398. https://doi.org/10.1109/tvt.2024.3404108

Saha, M., Thakur, S. S., & Bhattacharya, A. (2024). Bilevel planning of electric vehicle charging station and battery swapping station considering real-time uncertainty. International Journal of Green Energy, 21(12), 2724–2752. https://doi.org/10.1080/15435075.2024.2326063

Mao, S., Jin, J., & Xu, Y. (2024). Routing and charging scheduling for EV battery swapping systems: Hypergraph-based heterogeneous multiagent deep reinforcement learning. IEEE Transactions on Smart Grid, 15(5), 4903–4916. https://doi.org/10.1109/tsg.2024.3386609

Li, X., Cao, Y., Hu, Z., Zhang, X., Liu, Z., & Lei, X. (2025). A battery swapping service management system reliability-based design concerning station information delay. IEEE Transactions on Consumer Electronics, 71(1), 1063–1078. https://doi.org/10.1109/tce.2025.3531384

Gu, Z., & Liu, W. (2026). Robust time‐varying estimators using weighted fusion for multisensor networked systems with missing observations and two‐step random delays. International Journal of Adaptive Control and Signal Processing, 40(3), 579–596. https://doi.org/10.1002/acs.4105

Zhang, Y., Kong, X., Zhou, W., Liu, J., Fu, Y., & Shen, G. (2024). A comprehensive survey on traffic missing data imputation. IEEE Transactions on Intelligent Transportation Systems, 25(12), 19252–19275. https://doi.org/10.1109/tits.2024.3478816

Zhang, J., Cao, G., Liu, Z., Ding, Y., & Feng, J. (2025). DecMamba: Mamba utilizing series decomposition for multivariate time series forecasting. Computers, Materials & Continua, 82(1), 1049–1068. https://doi.org/10.32604/cmc.2024.058374

Sára Pusztaházi, L., Eigner, G., & Csiszár, O. (2024). Parametric activation functions for neural networks: A tutorial survey. IEEE Access, 12, 168626–168644. https://doi.org/10.1109/access.2024.3474574

Wang, Y., Han, Y., Hu, X., & Geng, Z. (2025). Long-term series forecasting for industrial processes based on multiscale hybrid decomposition feature extraction network. IEEE Transactions on Industrial Informatics, 21(7), 5401–5410. https://doi.org/10.1109/tii.2025.3555980

Mandi, J., Kotary, J., Berden, S., Mulamba, M., Bucarey, V., Guns, T., & Fioretto, F. (2024). Decision-focused learning: Foundations, state of the art, benchmark and future opportunities. Journal of Artificial Intelligence Research, 80, 1623–1701. https://doi.org/10.1613/jair.1.15320

Alghamdi, W. (2024). Synchronization patterns for digital twin systems. Journal of Applied Data Sciences, 5(3), 1026–1037. https://doi.org/10.47738/jads.v5i3.267

Boonraksa, T., Boonraksa, P., Pinthurat, W., & Marungsri, B. (2024). Optimal battery charging schedule for a battery swapping station of an electric bus with a PV integration considering energy costs and peak-to-average ratio. IEEE Access, 12, 36280–36295. https://doi.org/10.1109/access.2024.3374224

Gahlaut, T., & Dwivedi, G. (2026). Obtaining a balance between charging stations and battery swapping stations. Journal of the Operational Research Society, 77(1), 313–325. https://doi.org/10.1080/01605682.2025.2515152

Xiao, Y., Liu, H., Zhang, Q., & Chen, J. (2024). Reinforcement learning based automatic synchronization method for nuclear power digital twin model. IEEE Access, 12, 87625–87632. https://doi.org/10.1109/access.2024.3416999

Yan, H., Sun, C., Liao, H., & Tang, X. (2024). Optimal pricing and charging strategy design for non-cooperative battery swapping stations. IEEE Transactions on Mobile Computing, 23(12), 13573–13588. https://doi.org/10.1109/tmc.2024.3427784

Zhao, H., Zhang, H., Tian, C., Wei, Z., & Zhou, A. (2026). SRLinear: Lightweight long-term time series forecasting via symbolic regression. IEEE Transactions on Artificial Intelligence, 7(1), 210–224. https://doi.org/10.1109/tai.2025.3571377

Wang, G., Tan, X., Lin, Z., Luo, B., Zhong, S., & Xu, K. (2026). FreqEvo: Enhancing time series forecasting with multi-level frequency domain feature extraction. IEEE Transactions on Knowledge and Data Engineering, 38(2), 1099–1113. https://doi.org/10.1109/tkde.2025.3632365

Shuai, H., & Li, F. (2025). Physics-informed Kolmogorov-Arnold networks for power system dynamics. IEEE Open Access Journal of Power and Energy, 12, 46–58. https://doi.org/10.1109/oajpe.2025.3529928

De Giosa, M., Falini, A., Iavernaro, F., Losito, S., Mazzia, F., & Rubino, G. (2026). A robust variant of cubic smoothing spline approximation. Numerical Algorithms, 101(1), 221–241. https://doi.org/10.1007/s11075-024-02003-7

Liu, H., Huang, K., & Li, Y. (2026). Event-triggered robust fusion estimation for multisensor nonlinear cyber-physical systems under denial-of-service attacks. IEEE Transactions on Industrial Informatics, 22(6), 5462–5471. https://doi.org/10.1109/tii.2026.3660924

Wu, K., & Politis, D. N. (2024). Bootstrap prediction inference of nonlinear autoregressive models. Journal of Time Series Analysis, 45(5), 800–822. https://doi.org/10.1111/jtsa.12739

Ren, K., Chen, C., Sung, H., Ahn, H., Mitchell, I. M., & Kamgarpour, M. (2025). Recursively feasible chance-constrained model predictive control under Gaussian mixture model uncertainty. IEEE Transactions on Control Systems Technology, 33(4), 1193–1206. https://doi.org/10.1109/tcst.2024.3477089

Adamek, J., Heinlein, M., Lüken, L., & Lucia, S. (2024). Deterministic safety guarantees for learning-based control of monotone nonlinear systems under uncertainty. IEEE Control Systems Letters, 8, 1030–1035. https://doi.org/10.1109/lcsys.2024.3407635

Lai, M., Wulff, S. S., Cao, Y., Robinson, T. J., & Rajapaksha, R. (2025). Temporal cross‐validation in forecasting: A case study of COVID-19 incidence using wastewater data. Quality and Reliability Engineering International, 41(2), 672–688. https://doi.org/10.1002/qre.3686

Bitencourt, J., Wooley, A., & Harris, G. (2025). Verification and validation of digital twins: A systematic literature review for manufacturing applications. International Journal of Production Research, 63(1), 342–370. https://doi.org/10.1080/00207543.2024.2357741

Liu, S., & Zhou, D. J. (2024). Using cross‐validation methods to select time series models: Promises and pitfalls. British Journal of Mathematical and Statistical Psychology, 77(2), 337–355. https://doi.org/10.1111/bmsp.12330

Wu, C., Shi, Q., Gao, S., & Wang, Y. (2024). Service capability evaluation of electric vehicle charging and battery swapping stations: A game theory‐based combination weighting method. IET Energy Systems Integration, 6(4), 581–592. https://doi.org/10.1049/esi2.12155

Brennan, R. W., & Lesage, J. (2025). Evaluating the use of grey-box system identification for digital twins in manufacturing automation. International Journal of Computer Integrated Manufacturing, 38(7), 893–914. https://doi.org/10.1080/0951192x.2024.2386980

Li, X.-Y., Xu, Y., Zhu, Q.-X., & He, Y.-L. (2025). Industrial data imputation based on multiscale spatiotemporal information embedding with asymmetrical Transformer. IEEE Transactions on Neural Networks and Learning Systems, 36(8), 14937–14948. https://doi.org/10.1109/tnnls.2025.3527581

Zhang, K., Wen, Q., Zhang, C., Cai, R., Jin, M., Liu, Y., Zhang, J. Y., Liang, Y., Pang, G., Song, D., & Pan, S. (2024). Self-supervised learning for time series analysis: Taxonomy, progress, and prospects. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(10), 6775–6794. https://doi.org/10.1109/tpami.2024.3387317

Kyriazi, F., & Thomakos, D. (2026). Simplicity, robustness and speed in time series forecasting: LEAF-learning by exponentially adaptive forecasting, a novel univariate predictive analytics benchmark. Journal of the Operational Research Society, 77(4), 951–962. https://doi.org/10.1080/01605682.2025.2510530

Rombouts, J., Ternes, M., & Wilms, I. (2025). Cross-temporal forecast reconciliation at digital platforms with machine learning. International Journal of Forecasting, 41(1), 321–344. https://doi.org/10.1016/j.ijforecast.2024.05.008

Cao, B., & Chu, C. (2025). A review: Location of charging and/or battery swapping stations for electrical vehicles. International Journal of Production Research, 63(24), 10633–10676. https://doi.org/10.1080/00207543.2025.2554949

Details