High-speed maglev trains represent a new intelligent transportation mode that combines environmental friendliness with high speed. Under high-speed operations of 600 km/h, the maglev train levitation system faces numerous challenges, such as track irregularity and response delay. Most existing levitation control methods are problem specific and lack sufficient adaptability. While reinforcement learning-based methods offer improved adaptability, they do not guarantee theoretical stability. This article presents a safe reinforcement learning (RL) control method for the levitation system based on the Lyapunov stability theory and control barrier function (CBF). First, the system model is developed, considering various disturbances and time delays. Subsequently, stability constraints were constructed based on the Lyapunov theory, and a control law was designed by integrating a deep RL algorithm. Considering physical constraints, a cost function is constructed, and a CBF is introduced to guide the system away from the constrained boundary. Through the rigorous mathematical analysis, it is proven that the proposed method guarantees the stability and safety of the levitation system. Experimental results show that under track irregularity and actuator time-delay disturbances, the root-mean-square error (RMSE) of the air gap controlled by the proposed method is reduced by 35.7% and 24.3%, respectively, compared with the adaptive sliding mode method, and it exhibits stronger robustness and better time-delay resistance.

To address the attenuation of levitation force between the electromagnets and rails in maglev vehicles caused by high-speed operation, this study first analyzes the generation mechanism of eddy current effects (ECEs) based on electromagnetic field theory. A high-precision surrogate model for air-gap magnetic field calculation is established, incorporating the actual topological structure of the levitation electromagnets. Building on this foundation, an analytical electromagnetic force model accounting for ECE in both rails and electromagnets is developed. This model enables efficient parametric analysis of eddy current influences and is directly applicable to inverse problems in electromagnet topology optimization. To address the challenges posed by ECE on the engineering application of speed increase in medium- and low-speed maglev vehicles, this article analyzes the limitations of existing end-mounted levitation electromagnet structures. Based on the current balance optimization principle between front and rear suspension points, a nondominated solution set for structural optimization parameters is derived, aiming to mitigate uneven load distribution and compensate for ECE. Dynamic simulation results demonstrate that the optimized electromagnet structure effectively compensates for ECE while simultaneously enhancing track compatibility at levitation points. Experimental results indicate that at 140 km/h, the currents at both front and rear levitation points of the end-mounted electromagnets reach a balanced state, remaining below the rated value of 35 A. The outcomes of this study provide theoretical guidance for optimizing levitation systems in medium- and low-speed maglev vehicles at higher speed levels.

Electro-magnetic Suspension (EMS) high-speed maglev is currently the fastest ground transportation system, and its dynamic issues remain a research hotspot, though experimental studies are scarce. Focusing on the EMS highspeed maglev vehicle developed in China, vehicle-track-guideway coupled vibration tests were conducted. Dynamic data from the vehicle, levitation/guidance systems, track, and guideway under variable-speed conditions were collected, followed by data processing and dynamic performance evaluation. The research results indicate: the vibration acceleration of the vehicle's multi-body system decays progressively, and the car-body Sperling index reaches an excellent level; the dynamics response of the vehicle after entering the R400 m curve from the straight line increase significantly, with the fluctuation range of the levitation gap exceeds +/- 4 mm during curve negotiation, and the guidance gap shifts by 5-8 mm; the maximum dynamic deflections at mid-span of the guideway, along with the vertical/lateral vibration accelerations of the track and guideway, are all within safe limits, and the vibration frequencies correspond to their modal frequencies.

Onboard Internet of Things (IoT) edge computing provides a real-time and efficient platform for levitation control of the high-speed maglev train. Aiming at the stability control problem of levitation systems, a closed-loop control framework integrating Transformer networks with model predictive control (MPC) to enhance levitation stability and dynamic response performance is proposed. To handle time-varying dynamics and disturbances, an edge-deployed temporal Transformer network predicts short-horizon system states, improving MPC optimization within finite prediction horizons. Simultaneously, safety constraints are introduced with the combination of the control barrier function (CBF) method, ensuring critical variables, including levitation gap, remain within safe bounds. The framework covers nonlinear dynamics, linearized system equations, Transformer-based forecasting, MPC design with constraints, and CBF-embedded safety control. Experiments on the hardware-in-the-loop test platform demonstrate significant performance improvements over traditional methods: maximum deviation reduced by 50%, recovery time shortened by 60%, and response speed increased by 50%, while maintaining safety during high-speed operation. However, these results were obtained in a simulated environment rather than under actual operating conditions at 600 km/h, and the effects of aerodynamic disturbances at 600 km/h still require further investigation in subsequent research.

This article presents a novel position and speed estimation strategy based on moving horizon estimation (MHE) utilized in the linear synchronous motor (LSM) drives for electromagnetic suspension (EMS) high-speed maglev train. The mathematical models for LSM considering lumped disturbances are established. In light of this, an optimization driven MHE method that excels in performance, is put to use for estimating speed and position. Given lenient assumptions, an optimal solution to a quadratic program with equality constraints is computed in each cycle. Besides, the lumped disturbances observer was derived and constructed, characterizing residual perturbation at each iteration of the moving window estimation. Moreover, a proportional scaling LSM test platform based on EMS maglev train is established to validate the proposed method. To finalize, this article compares an extended Kalman filter (EKF), a modified MHE (mMHE), and standard MHE (sMHE)-based sensorless control methods. These proposal's steady state and transient state are tested with respect to different traction conditions, and estimation error and computational cost are analyzed. The results prove that the proposed scheme achieves superior in estimation precision while running in a real-time control environment, even in low-speed range.

To address the speed regulation challenges stemming from external running resistance variations and internal suspension-induced parameter fluctuations in electromagnetic suspension (EMS) maglev train linear synchronous motor (LSM) drives, this article presents a novel prediction-error-based active disturbance rejection speed control (PADRC). The mathematical model of train motion is established, incorporating lumped process disturbances. Based on this model, a predictive speed error model driven extended state observer (PESO) method is proposed to reject external resistance disturbance. To extract errors dynamics while mitigating noise effects, a recursive least squares (RLS) filter is integrated to proactively predict the variation trends of these disturbances. Besides, an online excitation flux linkage identification scheme is designed to minimize internal suspension-induced disturbances, characterizing the magnetic field variations influenced by excitation current and suspension air-gap. Moreover, stability analysis of the closed-loop system with the proposed scheme is conducted using Lyapunov stability theorem. To finalize, a scaled EMS maglev train LSM test platform is constructed to validate the proposed method. The results demonstrate that the proposed scheme achieves superior speed tracking precision in a real-time control environment, even in low-speed range.

The superconductor/guideway relationship is similar to the wheel/rail relationship of the highspeed railway, which is of paramount importance in maglev system. In the superconductor/guideway relationship, the levitation and guidance characteristics of high-temperature superconducting (HTS) materials directly impact the dynamics of HTS maglev vehicles. In light of this, this paper investigates the levitation and guidance characteristics of the superconductor/guideway relationship affected by different external factors. First, establish a two-dimensional superconductor/guideway relationship model considering multi field coupling. Second, propose a calculation method considering the interaction between superconductor and guideway to simulate the forced lateral and vertical vibrations as well as the longitudinal motion of dewar. Then, an equivalent three-dimensional mechanical model of superconductor/guideway relationship that takes into account the multi field coupling effects of electricity, magnetism, thermality, and force is established. Finally, focusing on the impacts of dewar’s working height, magnetic field’s irregularity, and HTS bulks’ working temperature, the dynamic characteristics of superconductor/guideway relationship in motion can be studied. This paper employs the mechanical model of equivalent three-dimensional superconductor/guideway relationship established to analyze the impacts of three external factors on levitation and guidance characteristics under multi-field coupling, providing effective references for the dynamic response of HTS bulks’ levitation and guidance characteristics.

The sectionalized traction power supply system (TPSS) of high-speed maglev lines severely limits the cross-sectional utilization of regenerative braking energy (RBE), causing most RBE to be dissipated by braking resistors. This paper proposes a DC-bus-side architecture that combines a hybrid energy-storage system (HESS) with a dual active bridge (DAB) converter to enable both local storage and bidirectional cross-sectional transfer of RBE. A rule-based energy-management strategy is developed to coordinate the grid, HESS, DAB, and braking resistors, and a life-cycle co-optimization model is established in which equipment ratings and key control thresholds are jointly optimized using an NPV-based penalized fitness function. The proposed framework is evaluated through a case study by comparing the baseline, single energy-storage system (SESS), HESS, DAB-integrated SESS, and DAB-integrated HESS architectures. Results show that, compared with capacity-only optimization under fixed energy-management parameters, the proposed capacity-control co-optimization improves the penalized fitness values of all four improved architectures. The DAB-integrated HESS architecture achieves the best overall techno-economic performance: Relative to the baseline, maximum demand is reduced by 25.37

This article presents a data-driven quantitative multiobjective finite-control-set model predictive control (FCS-MPC) for three-phase three-level neutral point clamped (NPC) inverters, aiming to achieve precise control of both the switching frequency and the neutral point voltage. Considering the switching weighting factor and the neutral point voltage weighting factor as inputs, and the average switching frequency within a fixed sliding window time and the maximum value of the neutral point voltage within another fixed sliding window time as outputs, a compact form dynamic linearization (CFDL) method is employed to construct an equivalent dynamic linearized data model at each dynamic operating point of the closed-loop system. Building upon this data-driven model, a meticulously crafted controller is devised to precisely regulate both the switching frequency and the peak deviation of the neutral-point potential. The simulation and experimental results validate the correctness and demonstrate the superiority of the proposed control approach.

With the rapid development of maglev transportation technology, maglev trains have entered the commercial operation stage worldwide. At present, the maintenance of maglev trains is shifting from traditional planned maintenance to preventive maintenance. It is necessary to study how to scientifically and efficiently achieve the maintenance of maglev vehicle equipment to ensure the safe and reliable operation of maglev trains and improve economic benefits. This article aims to establish a scientific magnetic levitation train maintenance management information system using a Reliability-Centered Maintenance (RCM) strategy, which balances both economy and practicality. Firstly, based on Failure Mode and Effects Analysis (FMECA), an RCM logical decision-making method is proposed for identifying reasonable maintenance modes. By categorizing failure levels to assess their impact, an optimized maintenance strategy for magnetic levitation trains is proposed. By establishing an economic benefit evaluation model and considering the characteristics of remote maintenance and RCM analysis data requirements for magnetic levitation trains, a data warehouse is used as the basic tool for data storage, sharing, analysis, and application. The constructed remote maintenance data warehouse system for magnetic levitation trains can realize functions such as monitoring data query, visualization, and real-time warning of service status, providing decision-making references for the maintenance of the magnetic levitation train system.

In the long-term operation of high-speed maglev trains, deterioration of levitation electromagnets leads to variations in key system parameters and threatens operational stability. This study proposes a condition monitoring framework based on parameter identification using Bayesian physics-informed neural networks (B-PINNs) under incomplete data. The method identifies time-varying parameters with an accuracy of the order of 10 −2 , and remains effective under sparse measurements with a low sampling frequency of 10 Hz. Using the dual strengths of B-PINNs in parameter identification and uncertainty quantification, a hierarchical monitoring strategy is established: thresholds and instantaneous rates of change (ROC) capture anomalies and abrupt events, ROC trends reveal gradual electromagnet degradation, and uncertainty-based indicators highlight potential risks associated with parameter instability. By integrating these indicators, the framework improves the robustness of condition monitoring, enabling early fault detection and supporting the long-term safe operation of high-speed maglev trains.

The electrodynamic suspension (EDS) system with cross-connected null-flux coil (NFC) configurations shows great potential for high-speed Maglev trains and electromagnetic launch systems. However, existing test platforms cannot experimentally simulate cross-connected NFC tracks. In this article, we propose a novel experimental approach to investigate the influence of crossconnection cables on dynamic performance in a laboratory environment. Firstly, a new rotating double-disk experimental EDS device equipped with cross-connected NFCs is introduced, with its structure, functions, and experimental procedures described in detail. Secondly, an efficient numerical model of the NFC-type EDS system, capable of accommodating general magnet arrays and arbitrary magnetization angles, is developed to support the electrical parameter design of the NFCs. Thirdly, comparisons of transient electromagnetic forces between numerical simulations (for both linear and rotary motion) and rotational experimental measurements demonstrate a high degree of consistency, validating the effectiveness of the proposed test platform. Finally, the transient performance of the EDS system with and without cross-connection cables is experimentally evaluated and compared. The results indicate that the cross-connected NFC configuration enhances the guidance force by 250% without compromising suspension performance. This experimental method, presented here for the first time, enables laboratory simulation of the dynamic behavior of cross-connected EDS systems and demonstrates considerable scalability. Moreover, the platform significantly enhances the capabilities of EDS testing by allowing flexible simulation of various coil and magnet configurations. As such, it provides an important reference framework for the design and experimental investigation of advanced EDS systems.

Superconducting electrodynamic suspension (EDS) trains promise very high-speed transport, yet electromechanical coupling can induce vibrations because of low or even negative damping, aggravated by track irregularities and high-frequency harmonics from discrete levitation and guidance coils (LGCs). An electromechanical coupling framework is developed together with an electromagnetic shunt damper (EMSD) for vibration suppression. An EDS train model integrates dynamic circuit representations of LGCs and damping coils with vehicle motion equations. The electromagnetic force module is validated against finite element simulations and experiments. A speed-based stepping window algorithm enables long-track simulation with reduced computational burden for mutual inductance and circuit matrices. Using fixed point theory, geometric and electrical parameters of the electromagnetic damper (EMD) and EMSD are optimized, and parameter sensitivity analysis examines how deviations in electromagnetic quantities affect attenuation across frequency bands. Simulations of suspension bogie system under random track irregularities show clear suppression by both an EMD and the EMSD; relative to EMD, EMSD reduces lateral acceleration RMS by 27.3% and vertical acceleration RMS by 17.5%. For a full train system at 600 km/h, EMSD reduces lateral acceleration RMS by 28.9% to 38.0% and vertical acceleration RMS by 2.4% to 7.4%, and lowers the vehicle body Sperling ride comfort index. By aligning electromagnetic and mechanical resonances, the optimized EMSD provides strong damping, most of all in the lateral direction. The framework offers a practical and scalable route to improve the dynamic stability of superconducting EDS trains, and future work will explore adaptive parameter tuning and intelligent control to coordinate damping across multiple vibration modes.

Traditional high-speed maglev train control is based on a simplified single-point levitation system, ignoring the coupled disturbance interactions between levitation points. At higher operating speeds, the competitive effects between levitation points intensify, significantly reducing levitation control performance, and requiring higher precision tracks to maintain stability. This paper proposes a cooperative model predictive control method with planning trajectory communication (CMPC-PTC) for maglev levitation system, which enhances robustness under track irregularities compared to single-point control methods. Specifically, by introducing planning trajectory communication, we establish a levitation system equation that considers the coupling between levitation points. A robust model predictive control (RMPC) method is then applied to handle disturbances in levitation system, leading to the design of a novel levitation system control method. Through mathematical analysis, we prove the recursive feasibility and stability of this method. Numerical simulation results demonstrate that the proposed cooperative control method significantly reduces airgap tracking errors under track irregularity excitation compared to single-point control methods. Experimental results on a high-speed maglev track coupling test platform validate the effectiveness of this method. By enhancing the robustness of levitation control, the proposed method reduces the precision requirements for track construction, which is of great significance for improving economic efficiency of maglev transportation systems.

The superconducting electrodynamic suspension (EDS) train utilizes superconducting coils (SCs), levitation and guidance coils (LGCs), propulsion coils (PCs), and damping coils (DCs) for operation, which has the potential for high-speed train. The impact of propulsion systems on vehicle dynamics is extremely important and little studied. Therefore, it is crucial to conduct a comparative study of the dynamic characteristics with different propulsion configurations. This article establishes an equivalent circuit model considering the four electromagnetic systems, including PC, LGC, DC, and SC interactions. This model also incorporates air resistance and vehicle dynamics to capture the electromagnetic-dynamics coupling effects. Secondly, based on the numerical model, the static thrust fluctuation characteristics of both double-layer and single-layer propulsion systems and the vehicle's vibration behavior are investigated. Finally, the damping effects of the DCs on vibration suppression in the propulsion system are discussed. The results show that, after incorporating the PCs into the numerical model, the guidance stiffness of the suspension bogie decreased and levitation stiffness increased. The comparison reveals that the propulsion system has a large impact on the electromagnetic force fluctuation of the system, but a little effect of the dynamics. From the perspective of engineering manufacturing and long-term operational economy, it is therefore suggested that high-speed EDS trains may opt for a single-layer propulsion system.

In high-speed railways, the electric current is transmitted from the power supply to the high-speed train via the pantograph mounted on the train and the catenary system erected along the track. The current collection efficiency of high-speed trains is directly determined by the sliding contact stability between the pantograph and the catenary. Previous studies widely assume the pantograph to be a lumped mass model, which is regarded to efficiently represent the low-frequency behavior (below 20 Hz). But this approach fails to accurately capture the flexibility of the pantograph head's strip, which has important implication on sliding contact at high speeds. A hybrid-pantograph model, incorporating a lumped mass framework and a flexible dual-strip, is validated through pantograph head modal testing. The model's accuracy is verified against experimental results. The comparative analysis with respect to the traditional lumped mass model reveals that the hybrid model can describe the impact of the pantograph's low-frequency pitch motion in contact force. Furthermore, the high-frequency fluctuations in contact force and loss-of-contact phenomena are related to the pantograph head elasticity, which is neglected by the lumped mass model, can also be captured by the hybrid model. To satisfy the demand of numerical accuracy at higher speeds, the impact of pantograph pitch motion and the pantograph head flexibility must be incorporated in simulations.

This study examines pantograph aerodynamic lift at 400 km/h, and uncovers the dynamic behaviors and mechanisms that influence pantograph–catenary performance. Using computational fluid dynamics (CFD) with a compressible fluid model and an SST k-ω turbulence model, aerodynamic characteristics were analyzed. Simulation data at 300, 350, and 400 km/h showed lift fluctuation amplitude increases with speed, peaking near 50 N at 400 km/h. Power spectral density (PSD) energy, dominated by low frequencies, peaked around 10 dB/Hz in the low-frequency band, highlighting exacerbated lift instability. Component analysis revealed the smallest lift-to-drag ratio and most significant fluctuations at the head, primarily due to boundary-layer separation and vortex shedding from its non-streamlined design. Turbulence energy analysis identified the head and base as main turbulence sources; however, base vibrations are absorbed by the vehicle body, while the head causes pantograph–catenary vibrations due to direct contact. These findings confirm that aerodynamic instability at the head is the main cause of contact force fluctuations. Optimizing head design is necessary to suppress fluctuations, ensuring safe operation at 400 km/h and above. Results provide a theoretical foundation for aerodynamic optimization and improved dynamic performance of high-speed pantographs.

To address the suspension airgap fluctuations and vertical instability caused by rotor vibration in magnetically suspended flywheel energy storage systems (MS-FESS) under high-speed operating conditions of maglev trains, this paper proposes a high-precision stable control method for rotor axis trajectory. In complex operational scenarios (e.g., high-speed cruising, emergency braking), unbalanced disturbances and sensor harmonic noise in the magnetically suspended flywheel rotor dynamically couple through the mechanical base and suspension system, degrading airgap control accuracy. Existing methods exhibit limitations in coupling modeling, dynamic adaptability, and collaborative suppression of multi-source disturbances, failing to meet high-speed operational demands. To resolve these issues, this study first establishes a multi-degree-of-freedom coupled dynamic model of the flywheel-maglev system, quantifying the vibration energy transfer mechanism from the base to suspension electromagnets and revealing the nonlinear correlation between airgap fluctuations and compensation currents. Subsequently, an adaptive tracking filter with dynamic parameter adjustment is designed to synchronously suppress synchronous and harmonic disturbances through real-time speed tracking and phase correction, addressing the phase lag issue of traditional notch filters across wide speed ranges. Finally, experimental validation is conducted on a high-speed maglev flywheel prototype and a simulated train platform. Results demonstrate that the proposed method reduces base vibration intensity by 8–10 times, decreases suspension airgap fluctuation amplitude by 40%, stabilizes rotor axis trajectory amplitude within 10 μm, and limits vibration acceleration below 0.05 m/s 2 , significantly enhancing system stability and energy efficiency. This research dynamically adjusts the filter parameters through real-time rotational speed tracking and phase correction, overcoming the phase lag problem of traditional notch filters. It provides theoretical and practical guidance for vibration suppression and trajectory control in the flywheel energy storage system of maglev trains.

Maglev vehicles are emerging as a promising solution for next-generation transportation, offering advantages such as high speed, low energy consumption, and strong gradient-climbing capabilities. However, the maglev vehices levitation systems are inherently nonlinear, operate within narrow airgap, and are highly sensitive to external disturbances. Additionally, track irregularities, sudden load variations, electromagnetic coupling, and actuator faults can cause significant fluctuations in system dynamics and may even lead to instability. To address the challenge of achieving high-precision and stable levitation control under multi-source uncertain disturbances, this paper presents a physics-informed neural network learning-based nonlinear model predictive control (PINMPC) scheme featuring a variable terminal constraint set. In particular, a deep physics-informed neural network (PINN) is developed to perform identification of system unknown parameters, enabling the controller to adapt to dynamic changes. Based on the updated model, a customized nonlinear model predictive control (NMPC) strategy is formulated to handle system nonlinearities, external disturbances, and input/output constraints. Furthermore, a variable terminal constraint set is designed to mitigate constraint violations caused by input delays and to ensure recursive feasibility. Finally, simulation results validate the effectiveness of the proposed method under input delay and external disturbances.

The long stator linear synchronous motor (LSLSM) is employed to drive the electromagnetic suspension (EMS) maglev train. Accurate motor parameters are essential for achieving high-performance traction control. However, the parameters of LSLSM vary with the complex operating conditions of the maglev train. To ensure precise real-time estimation of motor parameters, an online parameter identification algorithm based on a dual extended Kalman filter (DEKF) is proposed, taking into account the influence of the voltage-source inverter (VSI) nonlinearity. The proposed DEKF approach enables the simultaneous estimation of $d q$-axis stator inductances and stator resistance within one control cycle, which is not affected by the rank-deficient problem. The comparative simulation results validate the accuracy of the proposed algorithm.