Long-stator linear induction motor (LSLIM) with unequal-length stator segment (ULSS) can significantly reduce system power supply capacity and construction costs, with crucial application value in fields like aerospace high-speed equipment testing. However, the ULSS configuration exacerbates the nonlinear and fast time-varying characteristics of motor parameters, causing severe current control instability during operation, which bottlenecks the performance enhancement of advanced electromagnetic launch equipment. To address this, a sixth-order nonlinear state-space model incorporating the coverage ratio factor an and stator length coefficient bn is established. Research shows that while the traditional Jacobian-based local linearization criterion indicates a stable equilibrium point, time-domain simulations reveal intense instability at the full coupling point. To uncover the mechanism, bifurcation theory is introduced; via bifurcation diagrams and maximum Lyapunov exponent (MLE), the system’s chaotic evolution under full coupling is confirmed. A stability boundary heatmap is then constructed via two-dimensional parameter scanning to define critical control parameters. Simulink simulations demonstrate that selecting parameters within this boundary effectively suppresses current oscillations. This validates the instability mechanism interpretation and the effectiveness of the defined parameters, providing theoretical guidance for controller parameter selection.

Magnetically suspended rotor (MSR) systems have gained widespread industrial adoption owing to their frictionless operation and exceptional reliability. However, harmonic current generated by unbalanced mass and sensor runout threatens the system stability. Repetitive control (RC) effectively suppresses harmonic current, but its parameter design relies on an accurate decoupling model of the system. The decoupling model for the MSR system is often simplified to a second-order linear system. Such a simplification, however, necessitates explicit consideration of system uncertainties caused by unmodeled nonlinearities during the RC design process. Especially under strong gyroscopic effects, the parameter uncertainty is further increased. In this article, an active disturbance rejection controller (ADRC) based on phase compensation (PC) is used to suppress coupling disturbances and improve the control performance of harmonic suppression. Firstly, the dynamic model of the MSR system is established, and both internal and external disturbances are thoroughly analyzed. Then, the RC-PCADRC scheme is designed, integrating the complementary strengths of RC and ADRC, with a particular emphasis on PC to improve stability margins. A comprehensive stability analysis is conducted, along with parameter optimization guidelines. Finally, the effectiveness and superiority of the proposed scheme are validated through both simulations and experiments.

This paper addresses the challenge encountered by segmented-power-supply supersonic linear induction motors (LIMs, >340m/s) under supply-section commutation and severe current-impact conditions, where key parameters (e.g., mutual inductance and resistance) undergo nonlinear and abrupt variations within an extremely short time and are difficult to observe in real time. Conventional model-based observers typically exhibit slow convergence during fast transients, whereas data-driven machine-learning approaches suffer from high-dimensional data, strong feature coupling among parameters, and limited interpretability. To overcome these issues, a physics-guided feature extraction method is proposed based on parameter-sensitivity analysis of proportional-integral (PI) controller residuals. First, a dynamic mathematical model of a segmented-supply LIM is established using a virtual-mover formulation, and a steady-state voltage feedforward equation is derived. Second, under a small-parameter-error assumption, an analytical mapping between parameter perturbations and the residual of the current-loop PI controller output is formulated, based on which a sensitivity matrix is constructed and its coupling mechanism is analyzed. Furthermore, a feature orthogonalization transformation is proposed to generate a reactive residual and an active compensation residual, thereby achieving bidirectional decoupling between the identification results of stator resistance and mutual inductance. Finally, an online parameter identification scheme is developed using a damped Newton iterative law based on the inverse sensitivity model, and learning-rate scheduling and low-pass filtering are introduced to suppress measurement noise and higher-order nonlinear errors. Full-system MATLAB/Simulink simulations demonstrate that, under high-speed mover entry/exit and abrupt parameter-change conditions, the proposed method can separate aliased residual signals into clear trajectories of parameter-mismatch features and accurately track the trend of actual parameter variations. Without requiring additional sensors or constructing extra observer models, the method provides highly interpretable feature inputs for subsequent low-cost, small-sample machine-learning-based parameter identification, and thus has significant engineering value.

The segmented power supply scheme for long-stator linear motor facilitates reducing power capacity and achieving a high power factor. However, the segment-switching process leads to overcurrent under high-speed conditions. This paper proposes a novel segment-switching strategy based on the time-optimal control theory. It employs time-optimal feedforward voltage and planned current trajectory during the switching transient process. Thus, it ensures rapid disconnection of the exiting segment and rapid establishment of the current in the incoming segment, while suppressing transient current overshoot. The mathematical model of the long-stator linear motor is established in the process of segment-switching. It derives the minimum times required to force the exiting segment current to zero and to establish the incoming segment current to the reference value by time-optimal control theory. Furthermore, the time-optimal voltages and current trajectories are calculated. The time-optimal current trajectories are used as the reference command for the current loop. The time-optimal feedforward voltages are introduced into the current loop control. Hence, it achieves rapid disconnection of the exiting segment and fast, accurate establishment of the incoming segment current. Experimental and simulation results collectively validate the effectiveness of the proposed segment-switching strategy.

With the advantages of low control complexity and high efficiency, series resonant dc-dc converters (SRCs) have been widely used in application scenarios that require galvanic isolation. To improve power density, it is desirable to use as small dc capacitors as possible. However, the operating characteristics under both normal and faults conditions of SRCs exhibit substantial discrepancies if the dc capacitors decrease significantly. To address this issue, the characteristics of the SRC operated with small dc capacitors is analyzed under semiconductor devices open circuit and short circuit faults conditions. On this basis, a fault tolerant strategy is proposed, with which soft switching can still be achieved in case of semiconductor devices faults. Both of the efficiency and the reliability of the SRC have been enhanced. The correctness of the theoretical analysis and the performance of the proposed fault tolerant strategy are validated through experimental results.

To address the issue that the mover position of the long primary segmented linear induction machines (SLPLIMs) cannot be directly estimated using flux linkages and back electromotive forces (EMFs) due to the slip ratio. A novel sensorless control strategy is proposed for SLPLIMs, utilizing the amplitude variation of coupled mover flux linkages. The method is built upon a proposed fused flux-amplitude space vector (FFASV), which is synthesized from multiple estimated flux linkages. The phase of this vector directly indicates the mover position. The system architecture comprises flux observers based on the model reference adaptive system (MRAS), the FFASV observer, and a dedicated position observer, whose design is derived from phase-locked loop (PLL) principles. To address the issue of inconsistent parameters among different stator segments in the SLPLIM, flux observers based on MRAS identify motor parameters before the mover enters the powered stator segment. To address the complexity of the inductance matrix caused by the asymmetric distribution of the motor windings, the form of the inductance matrix is simplified by optimizing the selection method of the alpha beta-axis in the Clarke transformation, thereby reducing the number of parameters to be identified. Only the alpha-axis or beta-axis- flux linkages are used to construct the orthogonal coupled flux linkage independently, thereby decreasing the number of signals that require processing and lowering the overall computational burden. The effectiveness of the proposed sensorless control strategy is verified on an experimental platform with a SLPLIM powered by three inverters.

The second harmonic current (SHC) is significant in the cascaded H-bridge converter with dc-dc stage and supercapacitor under high current and variable output frequency, increasing the risk of overcurrent in the dc-dc. Conventional suppression strategies designed for fixed-frequency SHC constrain the dc-dc controller bandwidth to the SHC frequency, inevitably degrading dynamic performance. To address this issue, this article proposes a feedforward-based SHC suppression strategy. By analyzing the transmission path of SHC, an accurate feed-forward compensation is introduced at the dc voltage reference point, blocking the second harmonic signal from entering the dc-dc control loop. This decouples the SHC suppression strategy from the controller, eliminating the bandwidth constraint imposed by the SHC frequency. Therefore, the proposed strategy ensures both effective SHC suppression and enhanced dynamic performance under variable frequency conditions. Furthermore, the dc-dc controller design process is significantly simplified. Finally, the experimental results validate the effectiveness of the proposed SHC suppression strategy.

To address the problems that ultra-high-speed segmented power-supply long-stator linear induction motors are prone to switch open-circuit faults during power-supply transitions, leading to phase current distortion, instantaneous thrust fluctuations, and possible protection-triggered shutdowns, a spatiotemporal precision fault ride-through method based on the principle of "space as the primary factor and time as the secondary factor" is proposed. This method first establishes a spatial coupling relationship using the real-time position of the mover and the coverage ratio of stator segments, enabling rapid localization of the faulty segment and faulty switch, and implements selective temporary blocking of the faulty segment. Subsequently, the power-supply switching instant is taken as a key temporal node for secondary fault discrimination. Specifically, when the motor operates to the power-supply switching moment of a stator segment, a secondary state assessment of the faulty segment is triggered. If the fault persists and the accumulated number of occurrences reaches a preset threshold, it is identified as a permanent fault and permanent blocking is executed. If the fault is cleared, power supply is restored. Verification on a Matlab/Simulink simulation platform demonstrates that, compared with traditional fixed-delay reclosing strategies, the proposed method significantly shortens the outage duration of faulty segments, reduces the impact on the system’s continuous traction capability, and effectively avoids blind reclosing on permanent faults. As a result, the method achieves rapid fault localization, precise blocking and unblocking, and accurate fault-type identification for open-circuit faults, thereby enhancing the continuous operation capability, stability, and reliability of ultra-high-speed linear motor drive systems under fault disturbances.

To address the issues of excessive current ripple and poor dynamic response in conventional angle position control (APC) for high-speed switched reluctance generator (SRG), this paper proposes an online parameter identification-based model-free predictive control (MFPC) strategy. First, the system dynamics are represented as an ultra-local model (ULM), enabling the design of an extended state observer (ESO) for two-step current prediction to compensate for control delays. Second, an improved Recursive Least Squares (RLS) algorithm with covariance resetting and error clearance is implemented to accurately identify dynamic inductance online, thereby enhancing the prediction accuracy of the ESO. Third, a bus current estimation-based adaptive feedforward compensation (AFC) technique is introduced to accelerate DC-bus voltage regulation and system dynamic response. Finally, simulations conducted on a 250 kW SRG platform demonstrate that the proposed method achieves superior dynamic performance and significantly reduced current ripple compared to conventional APC method.

Considering the poor torque ripple suppression performance resulting from inefficient torque allocation in conventional offline torque sharing functions (TSFs), this article proposes an online adaptive TSF based on deadbeat predictive control (DPC). First, the proposed method accounts for the current tracking capability by dynamically distributing the optimal current reference profiles for the incoming and outgoing phases. The rate of change of current (RCC) in the adjacent phases is also considered during the formulation of the TSF expression. Second, to accommodate various operating conditions, a control angle self-adjustment algorithm is proposed. The optimal turn-on angle is determined by minimizing torque ripple through an iterative calculation method. The optimal turn-off angle is identified by comparing the estimated value with the maximum flux linkage at the ( $k +1$ ) control period. Compared to the optimal angle calculation strategies in the existing literature, the proposed method is not only computationally efficient but also mitigates the angle oscillation phenomenon during the iterative calculation process. Finally, simulation and experimental validation of the proposed method are conducted on a three-phase 12/8 switched reluctance motor (SRM). The results demonstrate that the proposed method offers superior torque ripple suppression and higher torque-per-ampere ratios across different operating conditions compared to the traditional torque control method.

This paper proposes a sensorless velocity observation method for ultra-high-velocity linear induction motors (LIMs) based on multi-voltage data fusion. The study addresses the critical challenges in velocity measurement encountered during high-velocity LIM operation, where rapid mechanical stress, environmental factors, and transonic unsteady aerodynamic characteristics induce severe disturbances in mover position/velocity measurements and electromagnetic mechanisms. These disturbances may further damage costly velocity measurement equipment. To overcome these limitations, a mathematical model correlating multi-voltage data with mover velocity is established, and a data fusion algorithm is introduced to enhance the precision of velocity observation. Compared to conventional single-voltage-based velocity observation methods, the proposed multi-voltage fusion approach significantly reduces the number of required parameters and computational complexity while achieving higher accuracy. Simulation results validate that the average error between the multi-voltage fusion-based velocity observation method and the actual velocity is less than 5‰. Experimental tests on a prototype further demonstrate that the maximum average error of the synchronous voltage-based velocity observation method remains below 9‰, confirming its practical applicability in engineering scenarios.

Linear induction motors (LIMs) with segmented stators are commonly used in long-stroke electromagnetic propulsion platforms, Due to the thermal effect of high current, the resistance of the motor’s mover deviates from the measured value, making it difficult to stably output the maximum thrust. Therefore, accurate online identification of the mover resistance is crucial for maximizing the thrust output of LIM. The motor parameters estimated by traditional adaptive identification methods fluctuate significantly between segments. This paper proposes an adaptive parameter identification method based on stator voltage and current reconstruction. In this method, the coupled parts of each stator segment covered by the mover are equivalently summed up. A mathematical model of a virtual LIM is then reconstructed, which is immune to the effects of stator segmentation and switch switching. Subsequently, a voltage-model-based flux observer and a current-model-based flux observer are established to observe the mover’s magnetic flux, enabling online identification of the mover resistance. Simulation results show that the proposed identification method can quickly and accurately identify the mover resistance, ensuring that the motor operates at the maximum thrust point.

When the segmented long primary linear induction motor (LPLIM) is powered by segments series-fed method, the motor parameters do not vary with the rotor movement even during the mover crossing the two stator segments. However, the segments series-fed method may results in the switch layout complexities, especially during the switching process. To address this, the transient switching process of the dual three-phase linear induction motor is analyzed and a corresponding phase-domain model is developed by introducing a coupling matrix and current constraint matrix. The model and control effectiveness are validated through simulation results based on a test platform of LP-DSLIM in our Lab.

Combining Si-IGBT and SiC-MOSFET in a hybrid parallel configuration integrates the lower conduction loss characteristics of Si-IGBT with the fast switching and lower losses of SiC-MOSFET. This hybrid approach offers significant advantages in improving the efficiency of power switching devices. However, the effects of different drive delays on switching losses between Si-IGBT and SiC-MOSFET are not well understood. Improper designing results in higher switching losses instead of lower switching losses, which can seriously affect the safety and reliability of the switching device. This work establishes a switching loss analysis model that takes into account the effects of drive delay. Simulation analysis of the switching losses in a hybrid parallel double-pulse circuit with different gate drive time delays for turn-on and turn-off was conducted to determine the optimal delay time for minimizing the switching losses. The simulation results show that under the proposed optimal delay timing, the turn-on losses are reduced by 32

The hybrid modulation strategy (HMS) offers the advantages of simple voltage balancing and low current harmonic distortion, rendering it suitable for the cascaded H-bridge (CHB) inverter with supercapacitor (SC) and dc–dc stage. The rounding function is essential in the HMS-based voltage balancing method, as it determines the number of the inserted submodules (SMs) within each control cycle. However, the voltage fluctuations in the dc-link among the SMs may either increases or decreases, affected by the rounding function employed. To explore this, three common rounding functions are considered, including fix(x), round(x), and ceil(x). Among these dc-link voltage fluctuations under the three functions, fix(x) yields the lowest, followed by round(x), and ceil(x) results in the highest fluctuation. Simulation and experimental results demonstrates the correctness of the theoretical analysis in this paper.

In order to study the characteristics of the long stator section on high-speed electromagnetic suspension (EMS) Maglev trains, a lumped-parameter model was derived based on the equivalent circuit theory in this paper for calculating the performances of the iron-core synchronous linear motor within one stator section. The German Transrapid 08 type Maglev train was selected as a typical example for studying, the section parameters in the equivalent model were obtained by the analytical method and the electromotive force, as well as propulsion force were solved by a 3-D finite element method. Via the proposed lumped-parameter model, the characteristics of driven current, terminal voltage and acceleration distance, etc. of the stator section could be solved, also the potential of applying the current configuration to a higher train speed was discussed. Finally, taking the key parameters of stator sections and the characteristics of acceleration/ power factor as dominant variables and design targets, respectively. Based on the derived model, Elitist Non-dominated Sorting Genetic Algorithm (NSGA-II) was employed for solving its multi-objective optimization problem, and some typical Pareto solutions were chosen for investigating the optimal targets under different configurations.

Low switching frequency of power semiconductor devices is expected to decrease losses in the cascaded H-bridge (CHB) inverter with supercapacitor and dc–dc stage. This paper proposes a switching frequency reduction method accompanied by a hybrid modulation strategy (HMS) for the CHB inverter. By optimizing the comparison output between the reference wave and the triangular carrier, unnecessary state jumps in the switching signals for the H-bridge can be avoided when rotating the switching mode of the H-bridge in the HMS. The proposed method can reduce the average switching frequency among the H-bridges without side effects on capacitor voltage balance and ac outputs. Finally, its effectiveness is demonstrated through simulation results of five SMs based on MATLAB/Simulink.

This paper explores the challenges and design considerations of multilevel energy storage converters (MESC) tailored for high-power applications, with a particular focus on high-power linear motor drives. The research evaluates different converter topologies including cascaded H-bridge (CHB) and modular multilevel converter (MMC) to identify configurations that optimize power delivery, energy management, and operational efficiency. The selection and analysis of these topologies are driven by the need to address increasing demands for energy efficiency and system reliability in high-power settings. The paper aims to guide the design choices by highlighting the critical factors that influence converter performance and system integration.

Linear motors with stator segmented powered are widely used in high-speed electromagnetic propulsion applications. And fast and smooth segment switching is required to achive higer propulsion speed. In this paper, the dynamic characteristics of the current during the segment switching process are analyzed. And the upper and lower limits of the electrical angle required to complete the segment switching at any speed are obtained by calculation. Then, a novel segment switching method is proposed, which achieves fast switching without overcurrent at any speed. Simulation results verify the effectiveness of the proposed segment switching method.

In inductively coupled power transfer (ICPT) systems, the mutual inductance and load are the key parameters to realize high performance control, which is difficult to measure directly in rail transit. Thus, a method realizing mutual inductance and load identification using only the output voltage and current of a high-frequency inverter is proposed in this paper. Dual-frequency modulation, where one is the resonant frequency and the other is the non-resonant frequency, is used in the high-frequency inverter. Then the amplitudes and phases of the output voltage and current of the high-frequency inverter at the resonant frequency and the non-resonant frequency can be obtained based on the fast Fourier transform (FFT) method. The circuit is decomposed according to different frequencies. Then a mathematical model of the ICPT system is established. Therefore, the mutual inductance and load can be identified. Finally, the effectiveness of the proposed method is verified based on an ICPT prototype. Experimental results show that the identification errors of mutual inductance and load are less than 5