This paper proposes an embedded digital twin (DT) framework tailored for the real-time condition monitoring of DC-DC converters, addressing the stringent requirements of aerospace power systems. The proposed approach employs a non-invasive, FPGA-based hierarchical DT architecture. Specifically, the low-layer executes the Physical BUCK control logic and high-frequency sampling tasks without modifying the physical circuitry. Simultaneously, the upper-layer manages the iterative updating of the DT model, the execution of the Hybrid Recursive Least Squares (Hybrid-RLS) parameter identification algorithm, and real-time data scheduling. Microsecond-level synchronization is achieved through efficient resource scheduling, ensuring the DT continuously monitors and dynamically evolves to accurately reflect the physical converter’s operating conditions. By algebraically decoupling nonlinear devices such as power switches and diodes, the proposed DT model ensures high universality, making it suitable for converters operating in both continuous conduction mode (CCM) and discontinuous conduction mode (DCM). Crucially, the Hybrid-RLS algorithm is employed to rapidly identify global circuit parameters, and these results serve as inputs to a Particle Swarm Optimization (PSO) algorithm. This strategy effectively reduces the heuristic search dimensionality, enabling stable and high-precision health monitoring of the switching devices. Finally, the clustering characteristics of the converter’s behavior under various test conditions are analyzed to verify the effectiveness of the proposed method in tracking degradation trends. These outcomes demonstrate a significant advancement toward resource-efficient, edge-deployed condition monitoring for power electronic converters.

Power electronic converters are widely used in electric propulsion aircraft, yet faults caused by the degradation of electrolytic capacitors, inductors, and MOSFETs greatly reduce system reliability. The digital twin (DT) can provide insights into its physical counterpart, enabling predictive maintenance to improve life-cycle management. However, limited onboard computational resources often hinder the accurate, real-time representation of converter behavior in existing DTs. A hierarchical embedded DT for the Buck-Boost converter is proposed in this paper based on a novel model decoupling strategy and a Hybrid Recursive Least Squares (Hybrid-RLS) algorithm. By decomposing high-order state equations into lower-order subsystems, this method significantly reduces computational complexity, allowing the parameters of the converter to be identified with hard real-time capability and the states between the digital and physical spaces to be synchronized. The proposed DT as well as the parameter identification method are verified by an experimental prototype, demonstrating superior convergence and robustness under dynamic load conditions compared to traditional algorithms.

The fuel cell temperature fluctuates with load variations, causing external characteristics such as the polarization curve and maximum power point to drift. Most existing methods treat fuel cell energy management and thermal management as independent objectives, ignoring their inherent electro-thermal coupling, which significantly degrades control accuracy and increases costs. To address this, this paper proposes a hierarchical electro-thermal control method for fuel cell hybrid vehicles. In the thermal management layer, the cooling system is regulated to track the optimal temperature reference, ensuring the stack operates at theoretical maximum output voltage and power capability. In the energy management layer, a temperature-driven dynamic power boundary is utilized, and the output power is regulated adaptively based on the real-time thermal state. Simulation results under the CWTVC driving cycle show that the proposed strategy effectively maintains the stack at its optimal thermal state and avoids damaging low-load and high-load zones. Compared to a conventional energy management strategy (EMS), the total operational cost is reduced from 55.68 USD to 55.24 USD, with a 7.5% reduction in degradation costs. These results demonstrate that the collaborative strategy significantly improves the safety, durability, and economic performance of the fuel cell system.

In this article, a deadbeat predictive control (DPC) method with low measurement effort and the capability to effectively suppress torque ripple is proposed for switched reluctance machine (SRM). First, to enhance prediction model accuracy, a flux-linkage model based on second-order B & eacute;zier curves combined with full-position interval interpolation is proposed. This method only requires determining three parameters-aligned-position incremental inductance, unaligned-position incremental inductance, and unaligned-position saturation inductance to accurately describe flux-linkage characteristics. Next, Varignon's principle is applied to analyze the mechanism by which phase current harmonics influence torque. On this basis, the beetle antennae search algorithm, which enables rapid optimization, is used to optimize the current waveform, resulting in a novel harmonic-optimized current profile. The developed enhanced DPC methodology incorporating pulsewidth modulation demonstrates significant torque ripple reduction under diverse operating conditions, minimal offline measurement requirements, and high estimation accuracy, thus providing an efficient and practical solution for achieving high-performance SRM control in industrial applications.

Brushless DC motors (BLDCMs) have been widely adopted across various fields. However, implementing advanced algorithms for high-performance torque control typically relies on accurate mathematical models, requiring precise parameters including inductance, resistance, and back electromotive force (EMF) coefficients. Unlike permanent magnet synchronous motors (PMSMs) that achieve unified mathematical models through coordinate transformation, BLDCMs exhibit distinct mathematical models during commutation and non-commutation phases due to their "three-phase six-step" driving mode. Addressing this challenge, this paper conducts a comprehensive investigation into the mathematical models of BLDCMs across different operational phases and modulation methods, ultimately establishing a unified mathematical framework. Building upon this foundation, a model reference adaptive algorithm-based parameter identification method is proposed. This innovative approach enables offline identification of inductance and resistance parameters, while simultaneously performing online identification of inductance and back EMF coefficients. Experimental results have verified that this method can achieve high identification accuracy for stator inductance, resistance, and back EMF coefficients. Particularly suited for hoist motors requiring frequent start-stop operations, this methodology shows significant practical value in applications demanding robust parameter identification under dynamic operating conditions.

The multi-phase converter has been widely used for fuel cell (FC) applications. However, despite the continuous operation of the converter, the switch fault would deteriorate the control performance of the closed-loop controller, and cause increased input current ripple. To this end, in this paper, a robust fault-tolerant control based on the framework of active disturbance rejection control (ADRC) is proposed for a four-phase interleaved converter. In particular, the system control gain is updated based on the duty cycles, so that the control performance under switch fault can be maintained. On the other hand, the method of phase shift reconfiguration is proposed to suppress the input current ripple of multi-phase converters under switch fault, in which the optimal phase shift is calculated by optimizing the object function derived from the input current ripple model. The experimental results are provided to validate the effectiveness of the proposed method.

Abstract Fuel cell truck (FCT) is expected to become the first commercial fuel cell vehicle. However, the high capital cost and operation cost are still the huge challenges preventing FCT commercialization process. The multi-stack fuel cell system (MFCS) is applied in FCT to improve the vehicle’s economy and reliability. The paper proposes a hierarchical design scheme to size the rated power of fuel cell stacks. Firstly, the upper layer aims to obtain the rated power of FCS and batter energy capacity with the objective of minimizing the vehicle’s operation cost. Then, the lower layer determines each stack’s rated power to decrease hydrogen consumption. The results show that the proposed scheme can save a lot of calculation time compared to single-step size design scheme. In addition, the triple stack FCS can save hydrogen cost 25.6% and 27.5% compared to double stack FCS under two driving cycles.

Fuel cell hybrid vehicles (FCHVs) have drawn tremendous attention due to the advantages of zero emissions. Existing energy management strategies (EMSs) typically fail to adequately address the coupled relationship between power allocation and thermal dynamics in the powertrain system, which is a gap that affects the economic and durability performance of FCHVs. To address this, this research proposes a hierarchical EMS that innovatively integrates: 1) a fuzzy-encode Markov chain (FMC) speed predictor for speed forecasting at the upper level; and 2) a multiobjective model predictive control (MPC) that optimizes system operating cost, durability, and thermal safety of battery and fuel cell systems at the lower level. In the validation phase, the impacts of different membership functions and state numbers on speed prediction accuracy are explored firstly. Then, the effects of weighting factors in multiobjective function are studied. Furthermore, an effectiveness evaluation method is set up to score each strategy with dynamic programming (DP) as the upper benchmark. The comparison with the other three benchmark strategies proves that the suggested strategy is more cost-effective, thermally safe, and life-extended, bringing an overall performance improvement of at least 20.61% and a score improvement of at least 14.02 points in the [0,100] range.

This paper proposes a novel vehicle speed prediction method based on a fuzzy Markov chain model. To enhance prediction accuracy, a triangular membership function is employed to map the acceleration into fuzzy states, enabling smooth transitions between adjacent states. The proposed method is evaluated under three driving cycles, demonstrating superior performance with average prediction error reductions of 4.28

In this article, the impact of stator resistance on initial rotor position detection for permanent magnet synchronous motor (PMSM) using a high-frequency (HF) voltage injection method is analyzed comprehensively. Theoretical derivation shows that the stator resistance appears as a phase shift term in the error signal for position estimation using the rotating HF voltage injection (RHFVI) method. The accuracy of initial rotor position detection is, therefore, degraded, and an obvious average error occurs when $R/L_{d}$ is greater than 200. On the contrary, using the pulsating HF voltage injection (PHFVI) method, the stator resistance appears as amplitude term in the position error signal, the initial rotor position can be estimated accurately, and there is nearly no dc bias error at a steady state. Finally, experimental data validates the effectiveness of the theoretical analysis in the article.

Fuel cell hybrid electric vehicles (FCHEVs) express the huge potential in realizing green and low-carbon transportation, but high total cost in vehicle's purchase and operation stages still prevents their rapid and large-scale commercialization. Beside to satisfying the propulsion power demand, energy management strategies (EMS) attempt to save hydrogen, to extend fuel cell lifetime and to mitigate vehicular operating costs through optimizing the outputs of energy sources. Nevertheless, the discrepancies between the standard speed-time dataset for strategy design and the real-world ones for testing may seriously compromise EMS performance. In addition, powertrain component size (e.g. the capacity of battery, the maximal output power of fuel cell, etc.) directly influences FCHEVs' manufacture cost and also put constraints on output feature of energy sources, thereby affecting the power-allocation performance. In this case, focusing on either one of aforementioned three aspects would not be effective in bringing down FCHEV's operating costs, thus indicating the necessity of developing a combined optimization framework that consider vehicular EMS, speed planning and component sizing concurrently. This paper starts by analyzing FCHEVs' powertrain structure and models and follows by a comprehensive review on state-of-the-art approaches and optimization frameworks for EMS, speed planning and sizing design. Then, a case study on combined optimization of sizing-EMS combination and speed planning-EMS combination is conducted. Finally, the major challenges and future prospective of combined optimization are summarized. This paper not only provides a literature review for readers but also a guideline for prospect designers in the field of FCHEV powertrain design and control.

This paper devises a generalized two-layer predictive energy management strategy with a comprehensive operating cost analysis for fuel cell logistic vehicles under different application scenarios. In the upper layer, an improved speed predictor based on long-and-short-term memory neural network and fuzzy C-means clustering is proposed, which can recognize driving states in real time and select corresponding sub-models for speed forecasting. In the lower layer, a multi-objective cost function including hydrogen consumption cost and powersource degradation cost is established and the optimal control action is derived within each receding horizon using sequential quadratic programming. Moreover, the performance discrepancies caused by various factors such as optimization weighting coefficients, prediction horizon length, velocity prediction methods and solution method are analyzed. Compared with benchmark strategies, the proposed strategy could reduce vehicular total operating cost by 0.76 %-32.83 % and fuel cell aging cost by 0.75 %-16.04 % across all the cycles. In addition, the operating cost distribution law with respect to different logistic vehicle types and different component sizes are analyzed via a comparative study, which could be used as a guideline for prospective designers in control strategy development.

To address the limitations of conventional Hall sensor interpolation methods in PMSM vector control, including dynamic response lag and discontinuous estimation, this study proposes a rotor position and velocity estimation method based on Hall vector filter phase-locked loop. The approach converts three-phase Hall signals into rotating vectors through coordinate transformation, extracts fundamental sinusoidal components using a synchronous frequency tracking filter (SFTF), and achieves closed-loop observation of rotor position/speed through phase-locked loop (PLL) synchronization, effectively suppressing the Hall signal deviation and improving the estimation accuracy. Furthermore, an improved PLL architecture with open-loop feedforward compensation is developed to resolve the direct-current (DC) offset error in conventional quadrature PLL under variable-speed conditions. Experimental verification demonstrates the proposed method's superior performance over conventional interpolation approaches in dynamic response speed, position estimation accuracy, and anti-interference capability, confirming its engineering viability.

Hydrogen fuel cell unmanned aerial vehicles (UAV) have become a research hotspot at home and abroad due to their features of zero emission, high energy density, high energy conversion efficiency and long fly range. Multi-physical domain coupling exists among the subsystems of hydrogen fuel cell UAV, which requires multi-disciplinary collaborative design and optimization in the development process of hydrogen fuel cell UAV, and reduces the R&D efficiency. Simulation has become an indispensable method for the R&D of various complex systems. Using computer to realize the simulation of the system is not only convenient and flexible, but also can bring huge social and economic benefits. Therefore, in this paper, a high-accuracy model implemented in Simulink is developed, which contains detailed subsystem modeling and numerical solution methods. Then using Simulink to iteratively calculate and simulate the designed system model by numerical method. The establishment of the model is of great significance to improve the R&D efficiency and shorten the R&D cycle.

The rotating high frequency voltage injection (RHFVI) method is widely used for sensorless control of permanent magnet synchronous motor (PMSM) at low and zero speed. The advantage of this method is that parameters tuning is easy and the convergence is good. However, the accuracy of rotor position estimation is degraded due to time delay in the high frequency (HF) current demodulation process, especially with the increasing of the rotor speed. Aimed at this issue, dual heterodyne method based RHFVI strategy is proposed to improve the sensorless performance of PMSM in this paper. Compared with the conventional method, the HF voltage is still injected into the stationary reference frame, however, the rotor position is obtained based on the HF current in the estimated synchronous rotating reference frame. Both the positive and negative sequence component of the HF current are utilized for rotor position estimation by the dual heterodyne method. Position error due to the time delay is compensated and the accuracy of rotor position estimation is improved. In the experimental part, one PMSM is tested to verify the effectiveness of the proposed method.

In this article, dual second-order generalized integrator (DSOGI)-based rotating high-frequency voltage injection (RHFVI) method is proposed for sensorless control of surface-mounted permanent magnet synchronous motor (SPMSM) at zero- and low-speed regions. Compared with the conventional bandpass filter (BPF) and high-pass filter (HPF), DSOGIs are proposed to suppress the position estimation error due to the phase shift in the high-frequency (HF) demodulation process. The first SOGI serves as an adaptive BPF and it is used to separate the HF current and fundamental current in the stationary reference frame. Compared with the conventional BPF, the resonance frequency of the SOGI can be adaptively adjusted according to the rotor speed. Therefore, the phase shift of BPF is eliminated, especially with the increase in rotor speed. The second SOGI serves as a band stop filter (BSF) and it is used to extract the negative sequence component of HF current that contains the rotor position information. Finally, the conventional BPF and HPF are removed, and the accuracy of rotor position estimation is improved. The experimental data validate the effectiveness of the proposed method.

Real-time simulation (RTS) is a crucial technology for the early-stage development of power electronics converters in electrified transportation. However, RTS faces challenges in accurately and rapidly identifying the switching statuses, particularly for uncontrolled devices, which limits the simulation capabilities of high-switching-frequency converters, such as resonant converters. In this article, a surrogate model assisted switch identification method is proposed for field programmable gate array (FPGA)-based RTS of resonant converters. This method allows for precise detection of the switch statuses with minimal computational effort, enabling accurate and fast RTS while avoiding zero-crossing oscillations. The RTS of the LLC resonant converter is implemented in an FPGA with a 50 ns time-step and demonstrates high accuracy compared to the Simulink reference model. Furthermore, the developed RTS is successfully applied to the controller hardware-in-the-loop (CHIL) testing, of which the results are validated through experiments on the LLC prototype.

In this article, the sensorless fast bidirectional start-up and speed regulation issues of permanent magnet synchronous motors (PMSMs) are addressed in the framework of indirect field-oriented control (IFOC). The proposed sensorless control scheme consists of a reduced-order adaptive speed observer and an alignment controller. As the input of the speed observer and the corrective term of the alignment controller, a hybrid terminal sliding-mode flux observer (HTSMO) is developed. Different from conventional flux observers (CFOs), the proposed one removes the assumptions about flux measurement or pure integration of the stator flux model, and a prior knowledge of load torque is also not required. Under the fast flux estimation and tracking, the alignment error can be corrected, and the rotor speed can synchronize with the electrical one. The stability analysis and the parameter tuning principles are also presented in detail. A set of PMSM control experiments driven by the ST digital signal processor (STM32F429) confirms the quick response and robustness of the proposed scheme.

An improved pulsating high frequency voltage injection method using resistance saliency reflected by eddy current loss is proposed for sensorless control of PMSM at zero and low speed regions. Firstly, compared with the conventional method that rotor position is estimated based on the saturation induced inductance saliency, in this article, by solving the high frequency (HF) voltage equation at dq-axis, it shows that the resistance saliency reflected by the eddy current loss is a good candidate for sensorless control of SPMSM, especially for the motor with ultra-low or even no inductance saliency. Secondly, using the proposed strategy to extract the HF current containing the resistance-based saliency information and construct the position error signal, rotor position is estimated at steady state and during the dynamic process. Finally, the effectiveness of the proposed method is proved by the experiment results.

Rotating high frequency voltage injection is a good candidate for sensorless control of permanent magnet synchronous motor (PMSM) at zero and low speed regions. However, satisfied sensorless performance is still challenging for surface-mounted PMSM (SPMSM) with ultra low or no inductance saliency. Aiming at this issue, eddy current loss reflected resistance saliency based sensorless control for SPMSM is researched in this article. Firstly, an improved mathematics model of PMSM considering resistance saliency is established, including both inductance and eddy current loss reflected resistance effect. Secondly, by solving the dq-axis HF (high frequency) voltage equation, it shows that resistance-based saliency can be used to estimate rotor position for SPMSM, which is key important for motor with ultra low or no saturation induced inductance saliency. Thirdly, a new heterodyne strategy is proposed to construct the position error signal for the PLL based observer, then, rotor position estimation can be achieved at steady state. Finally, the experiment results verify the effectiveness of the proposed method.