| Author: | Lu, Yang |
| Title: | Adaptive model predictive control and dynamic analysis for maglev systems |
| Advisors: | Ni, Yiqing (CEE) Wang, Sumei (CEE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Civil and Environmental Engineering |
| Pages: | xxii, 252 pages : color illustrations |
| Language: | English |
| Abstract: | Maglev technology has emerged as a revolutionary solution for modern transportation, offering high-speed, low-friction, and energy-efficient alternatives to conventional rail systems. Countries like Japan, Germany, and China have led the way in deploying operational maglev trains, showcasing advancements in dynamic modeling, levitation control, and system integration. Despite these achievements, several critical challenges persist in maglev suspension control. From a performance perspective, maintaining precise control of the levitation gap to ensure passenger comfort and operational safety remains difficult, particularly under poor track conditions. Nonlinear electromagnetic force characteristics inherently complicate controller design, leading to challenges in ensuring stability and accuracy. Additionally, variations in system parameters, such as mass changes due to passenger load variations and external environmental disturbances, pose significant difficulties for conventional controllers. These issues, combined with measurement noise and disturbances caused by track irregularities, further degrade control performance and system robustness. Addressing these multifaceted challenges comprehensively constitutes the primary objective of this research. A detailed dynamic model of the electromagnetic suspension (EMS) system was developed to analyze its nonlinear characteristics and constraints. Conventional Proportional-integral-derivative (PID) control methods were initially implemented, revealing significant limitations in scenarios involving high-frequency disturbances, such as rapid levitation gap variations and measurement noise. To overcome these issues, Model Predictive Control (MPC) was designed and validated, demonstrating superior performance in trajectory tracking, stability, and constraint handling compared to PID control. The integration of Kalman filters further enhanced the robustness of MPC under noisy measurements. To address model inaccuracies and dynamic parameter variations, an Adaptive Model Predictive Control (AMPC) strategy was proposed. The AMPC framework, utilizing ARX-based adaptive models, showed remarkable improvements in handling real-time system changes. Simulation and experimental results validated its capability to minimize tracking errors, maintain stability, and outperform traditional MPC under dynamic conditions, such as mass variations and external disturbances. The thesis further extended the analysis to a two-point suspension system and three-dimensional dynamic models that incorporate train-track-bridge interactions. Simulations demonstrated the AMPC controller's effectiveness in maintaining levitation stability and reducing oscillations under track irregularities and high-speed operations, highlighting its robustness and practical applicability. The findings of this research contribute to the advancement of maglev suspension control systems by offering improved performance, adaptability, and stability. Future work will focus on integrating detailed control circuit simulations, exploring machine learning-based adaptive control algorithms to further enhance real-time adaptability, mitigate reliance on accurate modeling, and address residual limitations observed in AMPC strategies, as well as optimizing energy efficiency for multi-bogie suspension systems. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13981