Computational Technology Resources - CCC

P. Han^1,2,3,4, J. Xu^1,2,3,4, Y. Sun^1,2,3,4, F. Ni^1,2,3,4, R. Mai^1,2,3,4, X. Liu⁵ and G. Lin^1,2,3,4

¹College of Transportation, Tongji University, Shanghai, China
², National Maglev Transportation Engineering R&D Center, Shanghai, China
³, State Key Laboratory of High-speed Maglev Transportation Technology, China
⁴, Key Laboratory of Maglev Technology in Railway Industry, China
⁵College of Automotive and Energy Engineering, Tongji University, Shanghai, China

doi:10.4203/ccc.15.11.8

P. Han, J. Xu, Y. Sun, F. Ni, R. Mai, X. Liu, G. Lin, "A Koopman-Theory-Based Data-Driven Modelling and Control Method of High-Speed Maglev Trains", in J. Pombo, (Editor), "Proceedings of the Seventh International Conference on Railway Technology: Research, Development and Maintenance ", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 15, Paper 11.8, 2026, doi:10.4203/ccc.15.11.8

Keywords: maglev train, suspension control, data-driven, Koopman operator, dynamic mode decomposition, model predictive control.

Abstract

The suspension system is one of the key systems of high-speed maglev trains, and its strong nonlinear characteristics and system parameter perturbations under persistent service conditions lead to errors between ordinary linearization control models and actual systems. This paper proposes an online data-driven modelling and model predictive control method based on Koopman operator theory. First, based on the Extended Dynamic Mode Decomposition method, this paper designs offline and online data-driven modelling methods for suspension systems, which achieve global high-dimensional linearization of nonlinear systems without relying on internal mechanisms, and can dynamically capture the time-varying characteristics of system parameters. Second, the online data-driven model was combined with model predictive control to design an online adaptive suspension predictive control strategy with multiple constraint handling capabilities. Finally, under simulated parameter perturbation conditions, the proposed method is validated. The results show that the online Koopman MPC method is more adaptable to changes in system parameters, effectively suppresses system state fluctuations, and significantly reduces the steady-state error in the suspension gap. The proposed method provides a practical and feasible technical solution for the contactless smooth operation and suspension shakedown test of high-speed maglev trains.

download the full-text of this paper (PDF, 6231 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 15 PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON RAILWAY TECHNOLOGY: RESEARCH, DEVELOPMENT AND MAINTENANCE Edited by: J. Pombo Paper 11.8 A Koopman-Theory-Based Data-Driven Modelling and Control Method of High-Speed Maglev Trains P. Han^1,2,3,4, J. Xu^1,2,3,4, Y. Sun^1,2,3,4, F. Ni^1,2,3,4, R. Mai^1,2,3,4, X. Liu⁵ and G. Lin^1,2,3,4 ¹College of Transportation, Tongji University, Shanghai, China ², National Maglev Transportation Engineering R&D Center, Shanghai, China ³, State Key Laboratory of High-speed Maglev Transportation Technology, China ⁴, Key Laboratory of Maglev Technology in Railway Industry, China ⁵College of Automotive and Energy Engineering, Tongji University, Shanghai, China doi:10.4203/ccc.15.11.8 Full Bibliographic Reference for this paper P. Han, J. Xu, Y. Sun, F. Ni, R. Mai, X. Liu, G. Lin, "A Koopman-Theory-Based Data-Driven Modelling and Control Method of High-Speed Maglev Trains", in J. Pombo, (Editor), "Proceedings of the Seventh International Conference on Railway Technology: Research, Development and Maintenance ", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 15, Paper 11.8, 2026, doi:10.4203/ccc.15.11.8 Keywords: maglev train, suspension control, data-driven, Koopman operator, dynamic mode decomposition, model predictive control. Abstract The suspension system is one of the key systems of high-speed maglev trains, and its strong nonlinear characteristics and system parameter perturbations under persistent service conditions lead to errors between ordinary linearization control models and actual systems. This paper proposes an online data-driven modelling and model predictive control method based on Koopman operator theory. First, based on the Extended Dynamic Mode Decomposition method, this paper designs offline and online data-driven modelling methods for suspension systems, which achieve global high-dimensional linearization of nonlinear systems without relying on internal mechanisms, and can dynamically capture the time-varying characteristics of system parameters. Second, the online data-driven model was combined with model predictive control to design an online adaptive suspension predictive control strategy with multiple constraint handling capabilities. Finally, under simulated parameter perturbation conditions, the proposed method is validated. The results show that the online Koopman MPC method is more adaptable to changes in system parameters, effectively suppresses system state fluctuations, and significantly reduces the steady-state error in the suspension gap. The proposed method provides a practical and feasible technical solution for the contactless smooth operation and suspension shakedown test of high-speed maglev trains. download the full-text of this paper (PDF, 6231 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions