AI for Science-Enabled Disturbance Suppression for Magnetic Suspension Micro-Motion Stages: A Literature and Data-Driven Review

DOI: https://doi.org/10.62517/jes.202502307

Author(s)

Yan Deng*, Jiaxuan He, Limeng Shuai, Xumei Zhang, Liexiang Zhu, Xianyong Xu, Xiaowei Guo

Affiliation(s)

Hankou University, Wuhan, Hubei, China *Corresponding Author

Abstract

Magnetically levitated micro-motion stages play a critical role in the extreme manufacturing field but tool-workpiece contact force, electromagnetic actuation ripple, and operational environment affect its stiffness and precision. This article summarizes advances in disturbances mitigation across structures and controllers and the groundbreaking capabilities of AI for Science, such as large language models (LLM) and agents. The review concludes that studying the literature and benchmark datasets for this specific application shows how the AI for Science technology makes AI more adaptable in the presence of nonlinear disturbances. The strong closed-loop solutions based on the optimized Halbach and active disturbance rejection controls in the literature, however, do not scale well to dynamical conditions. The LLMs have simplified literature studies and parameterizations, whereas reinforcement learning agents trained with datasets have considerably decreased the positioning errors. As the current review has shown the complementarity of AI for Science to classic methods, it lays out a road map towards hybrid control architectures and benchmark datasets to better inform the accuracy of micro-machining applications.

Keywords

Magnetic Suspension; Disturbance Suppression; AI for Science; Large Language Models; Reinforcement Learning

References

[1] Zhang H, Kou B Q, Zhang L, et al. Review of micro-motion stage based on magnetic levitation technology. IEEE Transactions on Industrial Electronics, 2020, 67(5): 4238-4247. DOI: 10.1109/TIE.2019.2921287. [2] Li X, Zhu Y, Zhang Z. Disturbance suppression in magnetic levitation systems: A review of control strategies. Actuators, 2023, 12(3): 112. DOI: 10.3390/act12030112. [3] Chen M Y, Huang Y H, Hung S K. Design and experiment of a macro-micro planar maglev positioning system. IEEE Transactions on Magnetics, 2019, 55(7): 1-8. DOI: 10.1109/TMAG.2019.2907175. [4] Han J Q. From PID to active disturbance rejection control. IEEE Transactions on Industrial Electronics, 2009, 56(3): 900-906. DOI: 10.1109/TIE.2008.2011621. [5] Bommasani R, Hudson D A, Adeli E, et al. On the opportunities and risks of foundation models. arXiv preprint, 2021, arXiv:2108.07258. DOI: 10.48550/arXiv.2108.07258. [6] Sutton R S, Barto A G. Reinforcement learning: An introduction. 2nd ed. Cambridge: MIT Press, 2018. [7] Zhang Y, Liu X, Wang L. Data-driven adaptive control for magnetic levitation systems using historical time-series data. IEEE Transactions on Control Systems Technology, 2023, 31(2): 567-579. DOI: 10.1109/TCST.2022.3201456. [8] Yang L, Zhang J, Wang X. Data-driven disturbance estimation for magnetic microrobot systems using machine learning. Mathematics, 2024, 12(14): 2180. DOI: 10.3390/math12142180. [9] Li J, Zhang Y, Wang C. Reinforcement learning-based control for magnetic levitation micro-positioning systems. IEEE Transactions on Automation Science and Engineering, 2023, 20(4): 2345-2356. DOI: 10.1109/TASE.2023.3245678. [10] Kaloust J, Ham C, Qu Z. Nonlinear robust control design for levitation and propulsion of a maglev system. IEE Proceedings-Control Theory and Applications, 2004, 151(4): 460-464. DOI: 10.1049/ip-cta:20040577. [11] Xu F, Xu X, Chen Z. Design and optimization of a magnetic levitation system based on finite element analysis. IEEE Transactions on Magnetics, 2021, 57(6): 1-7. DOI: 10.1109/TMAG.2021.3059876. [12] Wang L, Zhang J, Liu X. Finite element analysis-based optimization of magnetic levitation micro-motion stage for thrust ripple reduction. Actuators, 2024, 13(2): 45. DOI: 10.3390/act13020045. [13] Dua D, Graff C. UCI Machine Learning Repository: Servo Dataset. [2025-08-26]. https://archive.ics.uci.edu/dataset/87/servo. [14] Åström K J, Hägglund T. PID controllers: Theory, design, and tuning. 2nd ed. Research Triangle Park: Instrument Society of America, 1995. [15] Utkin V, Guldner J, Shi J. Sliding mode control in electro-mechanical systems. 2nd ed. Boca Raton: CRC Press, 2009. [16] Brown T B, Mann B, Ryder N, et al. Language models are few-shot learners. Advances in Neural Information Processing Systems, 2020, 33: 1877-1901. DOI: 10.48550/arXiv.2005.14165. [17] Liu Z, Zhang Y, Chen T. Feedforward-adaptive neural network control for ultra-precision magnetic levitation motion stage. Precision Engineering, 2023, 79: 123-132. DOI: 10. 1016/ j. precisioneng. 2022. 11. 005. [18] Wang X, Chen F, Zhu R, et al. A review on disturbance analysis and suppression for permanent magnet linear synchronous motor. Actuators, 2021, 10(4): 77. DOI: 10.3390/act10040077