Testing helium turbo-expanders is often cost-prohibitive and technically demanding due to the extreme operating environments required. To address this, we propose a cross-fluid similarity framework that employs air as a viable surrogate for helium. Using the first-stage impeller of a radial-inflow turbo-expander as the benchmark, we derive similarity criteria rooted in compressible flow physics and the congruence of inlet velocity triangles. This framework adopts the tip Mach number (Mau) and flow coefficient (φ) as the foundational scaling parameters, supported by an iterative method to determine equivalent operating points for air. Comparative numerical analyses reveal that when Mau and φ are precisely matched, the internal aerothermodynamics—including temperature, pressure, and Mach number fields—show remarkable fidelity to the original helium conditions. Key flow phenomena, such as vortex evolution and irreversible entropy production, remain spatially and qualitatively consistent across both fluids. These findings provide mechanistic validation for the use of air in preliminary turbo-expander testing, offering a cost-effective and reliable pathway for the design optimization and performance characterization of cryogenic helium systems.

The end-winding insulation structure of stator windings in pumped-storage generator units is complex, with pronounced electric field concentration under out-of-phase conditions, making them critical concerns in insulation design and condition-based maintenance. Although the finite element method (FEM) offers reliable accuracy, the strong nonlinearity of the anti-corona layer results in a computation time exceeding 104 seconds per single solution, rendering it impractical for parameter optimization and rapid on-site assessment. This paper proposes a fast prediction method for end-region potential distribution based on a deep neural network (DNN). Taking a 334 MW unit as the research object, a three-dimensional electroquasistatic finite element model with six stator coils is established and validated through power-frequency withstand voltage and ultraviolet imaging experiments. Training samples are generated via design of experiments (DoE), and a multilayer DNN surrogate model with a 7-dimensional input (comprising 3D spatial coordinates and four physical parameters) and a 1-dimensional output is constructed to directly reconstruct the spatial potential field at the end region. The results demonstrate that the surrogate model achieves a maximum relative error of less than 2% along the entire path compared with the high-fidelity FEM solutions, with a single prediction time of approximately 38 s—representing a speedup factor of approximately 272—while also exhibiting good generalization capability. This method provides a feasible technical approach for rapid reconstruction of end-region field distribution and optimization of insulation structures.

[Objective]Liquid hydrogen,owing to its remarkably high specific energy and broad application prospects has attracted widespread attention in research and development,particularly in aerospace propulsion and sustainable clean energy storage and transport.However,the practical large-scale use of liquid hydrogen faces notable and demanding safety challenges.Liquid hydrogen has an exceedingly low boiling point;once a leak occurs,it rapidly vaporizes to form a hydrogen cloud.At flammable concentrations,this cloud can ignite or even explode upon encountering a flame or a static spark,posing serious safety hazards.Therefore,an in-depth investigation of liquid hydrogen leakage,diffusion,and vaporization behavior in open environments is crucial for devising effective safety protection strategies and emergency-response mechanisms.Owing to the inherently high risk,technical complexity,and substantial economic cost of liquid hydrogen experiments,publicly available data on liquid hydrogen release remain extremely scarce,severely limiting further progress in this field.[Methods]To address this research gap,a large-scale open-environment liquid hydrogen release experiment was executed,aiming to systematically obtain the diffusion characteristics and evolution process of the released liquid hydrogen.For this purpose,a comprehensive experimental platform was built,integrating a liquid hydrogen storage and release system,a sensor-array system,a remote monitoring and control system,and video and infrared observation systems.This platform collects key parameters in real time—including hydrogen volume fraction,temperature,wind speed,and wind direction—and records the hydrogen cloud evolution from multiple angles,ensuring thorough experimental data acquisition.By fully considering diverse environmental and operational parameters—such as wind speed,ambient temperature,release angle,ground-cover type,and release flow rate—the experiment completed 12 sets of valid release conditions.[Results]Given the large data volume,this study focused on presenting the detailed results and analysis of the second release experiment.During this experiment,multiple high-definition cameras were placed on the ground,and gimbaled aerial cameras captured the real-time spatial distribution of the hydrogen cloud from multiple perspectives,moreover,the sensor array recorded the dynamic changes in the hydrogen volume fraction at each measurement point.Based on these data,the liquid hydrogen release process was divided into four characteristic stages—ground spread,buoyant transition,stable plume,and cloud dissipation—and the evolution features of each stage were systematically summarized and analyzed.For the quantitative analysis,the sensor data were used to track the hydrogen cloud movement.Furthermore,this study employed the open-source computational fluid dynamics(CFD)code OpenFOAM to perform three-dimensional,transient numerical simulations under the second experimental condition.The deviations between the simulated results and experimental observations fell within acceptable tolerances.[Conclusions]This study not only provides large-scale experimental data on liquid hydrogen leakage in open environments—offering valuable foundational data for subsequent research—but also provides robust theoretical support through advanced CFD simulations coupled with empirical experiments.The integrated findings hold considerable reference value and practical importance for advancing the safe,reliable,and sustainable development of the hydrogen energy industry worldwide.

ABSTRACT This study centers on a 45# mineral oil and FR3 natural ester transformer oil blend. Under extremely cold conditions, it explores how varying moisture content impacts the breakdown characteristics of this mixture and examines the internal factors related to breakdown voltage. Moisture absorption experiments, breakdown tests, and computational simulations of moisture migration trajectories within oil matrices were conducted, with data analysis for temperatures below 0°C using the Weibull distribution. Results indicate that at a fixed oil mixture ratio, reducing moisture content greatly enhances the insulation oil's insulation performance. Around –30°C, increasing the natural ester proportion in the mix effectively raises the breakdown voltage. Under low—temperature conditions (0°C and below), when moisture content is low, increasing the natural ester share most markedly improves the insulation ability. However, as moisture content rises, this enhancement effect diminishes. Through simulation, the movement of moisture droplets in the transformer oil was analysed. It was found that they migrate to high‐electric‐field‐strength areas and gather, stretch, and connect the two poles. Further research shows that the movement speed of free water in mineral oil is about two orders of magnitude faster than in natural ester, and this speed difference grows with increasing voltage.

Testing helium turbo-expanders is often cost-prohibitive and technically demanding due to the extreme operating environments required. To address this, we propose a cross-fluid similarity framework that employs air as a viable surrogate for helium. Using the first-stage impeller of a radial-inflow turbo-expander as the benchmark, we derive similarity criteria rooted in compressible flow physics and the congruence of inlet velocity triangles. This framework adopts the tip Mach number (Mau) and flow coefficient (φ) as the foundational scaling parameters, supported by an iterative method to determine equivalent operating points for air. Comparative numerical analyses reveal that when Mau and φ are well matched, the internal aerodynamic and thermodynamic characteristics—including temperature, pressure, and Mach number fields—show remarkable fidelity to the original helium conditions. Key flow phenomena, such as vortex evolution and irreversible entropy production, remain spatially and qualitatively consistent across both fluids. These findings provide mechanistic validation for the use of air in preliminary turbo-expander testing, offering a cost-effective and reliable pathway for the design optimization and performance characterization of cryogenic helium systems.

To address the common problems in conventional drying technologies (long drying time, quality deterioration and nutrient loss) as well as the specific problems of high moisture content and easy spoilage during long-term storage of lettuce, electrohydrodynamic (EHD) drying was explored as an efficient non-thermal drying technology. The effects of EHD drying, hot air drying (HAD), and natural air drying (AD) on the drying characteristics and nutritional quality of lettuce were investigated. EHD drying was conducted at a fixed 6 cm needle array-plate air gap distance while input alternating current (AC) voltage (24 kV, 28 kV, 32 kV), and the frequency is 50 Hz. Electrical characteristics included voltage, current, discharge power, and drying energy consumption were investigated. Results showed that the synergistic effects of ion wind and non-uniform electric field mass transfer during EHD drying reduced its specific energy consumption to only 10%-24% of HAD. The EHD-32 kV sample achieved the highest rehydration rate, with total chlorophyll content of 0.53 mg/g-1.55 and 2.2 times higher than HAD and AD, respectively. In contrast to AD, which exhibited excessive hardness and poor palatability, EHD drying promoted the conversion of free water to bound and immobilized water, thereby improving texture and palatability. Meanwhile, EHD drying better retained the diversity and abundance of volatile compounds. Overall, EHD is a promising non-thermal drying technology that provides an effective strategy to mitigate typical drawbacks of conventional drying while enhancing the drying efficiency and quality of lettuce.

Improving the efficiency of hydrogen liquefaction cycles is essential for reducing costs and promoting clean hydrogen energy. As the core refrigeration component, the helium turbo-expander (HTE) experiences significant efficiency losses within the impeller passage. However, current design methodologies often involve complex manual iterations that can limit systematic 3D optimization. This paper explores a parametric approach for 3D impeller design based on cylindrical projection (Cylindrical Projection based Parametric Impeller Design). This method allows for the definition of blade profiles at the hub, mid span, and shroud sections using projection parameters (Tu, Tr), facilitating smooth geometric transitions. To evaluate the approach, a final stage HTE impeller for a 5 TPD hydrogen liquefier was analyzed. Numerical simulations indicate that adjusting the flow path geometry using these parameters can lead to significant performance variations: Case C1 showed a calculated 16.58% reduction in helium mass flow at a constant refrigeration power of 30,551.4 W, while Case C2 yielded a predicted isentropic efficiency of 92.33% (a 2.81% absolute increase) with a 3.5% reduction in required inlet pressure. Flow field analysis using the Omega vortex identification method suggests that these improvements are associated with the suppression of high loss vortex structures. Specifically, the concave blade profiles appear to mitigate transverse pressure differences, reducing the intensity of passage vortices. These results demonstrate that the parametric projection method offers a useful alternative for the geometric optimization of cryogenic turbo-expander impellers.

Electrohydrodynamic (EHD) drying is a promising non-thermal technology, but the quantitative relationship among the non-uniform electric field, ionic wind, drying kinetics, and product quality remains insufficiently understood. In this study, the drying characteristics of yellow carrots were investigated under passive ambient-air drying and EHD drying at 26, 30, and 34 kV, with a needle-to-plate electrode spacing of 10 cm., and complemented by a multiphysics-coupled simulation of electric potential, electric field, space-charge transport, and airflow. Significant differences were observed among the drying treatments for the key drying-performance and quality indicators (P < 0.05). Compared with ambient-air drying, the average drying rate increased by 1.65-, 2.12-, and 2.39-fold at 26, 30, and 34 kV, respectively, while the drying time decreased by 38.9%, 50.0%, and 55.6%. At 34 kV, drying was completed within approximately 8 h, the effective moisture diffusivity reached 1.05 × 10⁻⁵ m²/s, the rehydration ratio approached 3.7, and the measured extractable carotenoid content reached approximately 0.14 mg/g. The simulation reproduced the voltage-dependent increase and periodic spatial variation in ionic-wind velocity, although the absolute airflow velocity was overestimated, respectively. Mechanistically, increasing voltage strengthened the electric field, space-charge density, Coulomb force, and ionic wind, thereby enhancing both external mass transfer and internal moisture migration. From an engineering perspective, 34 kV achieved the best drying performance under the tested conditions, indicating that voltage should be optimized rather than simply maximized. Overall, these findings support the efficient and quality-oriented design of EHD drying systems.

To investigate the application of electrohydrodynamic drying in yellow carrots and clarify the role of ionic wind during drying, natural air drying was used as the control. Corona discharge drying experiments were conducted at 26, 30, and 34 kV, combined with multiphysics simulations to analyze ionic wind characteristics. Experiments evaluated drying characteristics of yellow carrots, while simulations focused on ionic wind velocity, electric field strength, and electric potential.Results showed that EHD drying significantly accelerated moisture removal compared with AD, and the drying efficiency increased with voltage. At 34 kV, drying was completed within 8 h, nearly half the time required for AD. Effective moisture diffusivity increased with voltage, reaching 1.05 × 10⁻⁵ m²/s at 34 kV. The rehydration ratio also improved with voltage, reaching 3.7 at 34 kV. Carotenoid retention increased with voltage; the 34 kV group achieved the highest content (about 0.14 mg/g).Simulation results demonstrated that ionic wind intensity and charge concentration increased with voltage. At 34 kV, the peak ionic wind velocity near the plate reached 0.2 m/s with a more pronounced directional airflow. The enhanced ionic wind and non-uniform electric field synergistically promoted mass transfer, which was identified as the key mechanism improving EHD drying efficiency.

This study uses molecular simulations to examine the adsorptive separation of helium-light hydrocarbon mixtures by the metal-organic framework (MOF) UPC-66 and its activated form UPC-66a. Grand canonical Monte Carlo (GCMC) simulations were conducted to obtain single-component adsorption isotherms of CH4, C2H2, C2H4, C2H6, and He at 298 K and 220 K. The ideal adsorbed solution theory (IAST) was applied to predict selective adsorption coefficients, and its accuracy was evaluated by comparison with direct GCMC-calculated selectivities for UPC-66a. Multicomponent GCMC simulations were further performed for a five-component gas mixture. To clarify competitive adsorption mechanisms, isosteric heats of adsorption were determined using the Clausius-Clapeyron relation and GCMC simulations, while adsorption energy distributions, adsorbate probability density distributions, and binding energies were analyzed. The results show that both MOFs can effectively extract helium from light hydrocarbon mixtures, with separation performance significantly enhanced at lower temperatures. The observed selectivity mainly arises from the stronger binding interactions between the adsorption sites and light hydrocarbon molecules compared with helium.

This paper proposes a visualization monitoring method based on Bi-kernel t-distributed Stochastic Neighbor Embedding (Bikernel t-SNE). The method introduces the input data and feature kernel matrices to construct a dual kernel mapping from low-dimensional to high-dimensional spaces, improving the identification accuracy of anomalous outliers. Subsequently, Kernel Entropy Component Analysis (KECA) is applied to transform the feature kernel matrix, so the algorithm can distinguish between normal and abnormal values even when the feature kernel matrix values of the new sample data approach zero and thereby achieve accurate fault warnings. Finally, a confidence boundary combining the fault detection Composite Index (CI) and Kernel Density Estimation (KDE) is designed to perform fault detection in a two-dimensional (2D) scatter plot, with fault monitoring and warning realized by evaluating whether the characteristic data exceed the limits. The algorithm's effectiveness is validated through case studies using the faulty and healthy Supervisory Control And Data Acquisition (SCADA) data from a wind farm. The experimental results demonstrate that the proposed method can accurately distinguish faulty data points and significantly reduce the number of alarms. The fault sensitivity reaches 91.3%, representing an 11.89% improvement in monitoring accuracy over conventional t-SNE.

In response to the challenges of frequent viewpoint changes, appearance variations, and occlusions in UAV tracking scenarios, as well as the issue of excessive computational redundancy in simple scenes, we propose an Adaptive Multi-View Complementary Tracker (AMV-Tracker) with a dynamic activation mechanism. First, we introduce a dual-template tracking strategy, where a static template preserves the original target information while a dynamic template provides short-term variation information, thereby enhancing robustness against appearance changes and occlusions. Second, we design a dynamic activation module to reformulate the Transformer-based backbone network, enabling a feature extraction network that dynamically activates relevant modules according to different backgrounds. Third, we employ a multi-view image mutual information module to model the target, ensuring stable feature representation under viewpoint variations. Additionally, to mitigate localization errors caused by scale variations, we introduce a target boundary regression branch and propose a more precise head prediction network. The proposed method is validated on the LaSOT and UAV123 datasets, demonstrating its effectiveness and superiority. Furthermore, the algorithm is deployed on an embedded device, Jetson Orin NX, for performance evaluation, achieving a processing speed of 46 FPS. Compared to classical tracking algorithms, the proposed method exhibits higher accuracy across various challenging scenarios and effectively addresses issues related to viewpoint changes, rapid motion, and scale variations.

Hydrogen turbo-expanders are core components in large scale hydrogen liquefaction systems. However, due to the high risks and costs associated with hydrogen experiments, direct experimental investigation of their internal flow characteristics is challenging. This paper proposes a comparative research method based on similarity theory, enabling the study of different working fluids (helium and hydrogen) within an impeller of identical geometry. By establishing appropriate similarity criteria-specifically, maintaining equal inlet Mach number (Mu) and flow coefficient (q)-flow similarity across different fluids can be achieved. Corresponding hydrogen operating conditions were derived based on these criteria. Using the first stage helium turbo-expander impeller from a 5 tons/day hydrogen liquefier as the baseline, two hydrogen operating conditions were determined by either fixing the inlet temperature or the inlet pressure. Numerical simulations were conducted to compare the flow field structure, loss distribution, and vortex evolution. The results demonstrate that under the proposed similarity criteria, both the macroscopic flow fields and microscopic loss mechanisms of hydrogen within the helium impeller exhibit a high degree of consistency with the original helium operating condition, validating the effectiveness of this method. The study further indicates that under high Reynolds number conditions, viscous effects are minor, making strict equality of the Reynolds number unnecessary. This similarity based approach not only provides theoretical support for the aerodynamic design of hydrogen turbines but also lays the groundwork for future substitute experimental studies using conventional working fluids, such as air at ambient temperatures.

In the field of photovoltaic (PV) system monitoring, fault detection faces two critical challenges: data imbalance and fault diversity, as well as incomplete complex fault information. To tackle these issues, this paper proposes a dual-mechanism anomaly detection with generative adversarial network (DMAD-GAN) and an integrated fault diagnosis with fine-grained information fusion (IFD-FGIF). DMAD-GAN utilizes GAN to integrate dual mechanisms for anomaly detection in PV datasets, with coordinate-space attention enhancing the perception of subtle features and differences in PV panels. The anomaly scoring mechanism utilizes an improved loss function to compute anomaly scores, assessing the degree of anomaly for each sample. In the IFD-FGIF method, t-SNE is used to visualize features for fault pre-classification to determine the presence of new faults. A fine-grained information fusion module is designed, leveraging ResNet50 to extract features from fault key areas and original images. This module integrates fine-grained features, original features, and fine-grained attributes. Fault attributes and categories are determined using an attribute classifier and Euclidean distance. If a new fault is identified during pre-classification, the network undergoes transfer learning to recognize and adapt to the new fault. The experimental results demonstrate that the proposed method outperforms other networks, achieving an anomaly detection accuracy of 95.86%. The fault fine-grained recognition accuracy is 95.62%. The accuracy of fine-grained information fusion has improved by 4%, and unsupervised learning of new faults has been successfully achieved. The proposed method can enhance the intelligent operation and maintenance capability of PV power plants, reduce false alarm rates in fault detection, and minimize operational risks caused by potential faults, thus effectively shortening downtime and lowering maintenance costs.

Due to their low friction and long lifespan, aerostatic bearings are widely used in high-speed cryogenic turbo expanders. Variations in bearing gas viscosity and density lead to differences in the static and dynamic characteristics of these bearings. In this study, a fluid-structure coupled model of the bearing-rotor system is established. The bearing lubricant equation and rotor motion equation are solved simultaneously to obtain the nonlinear dynamic response of the system. The findings reveal that the system exhibits rich nonlinear phenomena, indicating that the load capacity of helium bearings is greater than that of hydrogen and air bearings, while the stability of air bearings is superior to that of hydrogen and helium bearings.

In response to the issues of low accuracy, perception degradation, and poor reliability of single-sensor simultaneous localization and mapping (SLAM) technologies in complex environments, this study presents a novel Inertial Measurement Unit (IMU)-centered multi-sensor fusion SLAM algorithm (IFAL-SLAM) integrating Light Detection and Ranging (LiDAR), vision, and IMU, based on factor graph elimination optimization (IMU-centered multi-sensor Fusion, Adaptive Lagrangian methods). The proposed system leverages a multi-factor graph model, centering on the IMU, and applies a covariance matrix to fuse visual-inertial and LiDAR-inertial odometries for bias correction, using loop closure factors for global adjustments. To minimize the optimization costs post-fusion, a sliding window mechanism is incorporated, coupled with a QR decomposition elimination method based on Householder transformation to convert the factor graph into a Bayesian network. Finally, an adaptive Lagrangian relaxation method is proposed, employing matrix-form penalty parameters and adaptive strategies to enhance convergence speed and robustness under high rotational dynamics. Experimental results indicate that the proposed algorithm achieves absolute trajectory errors of approximately 0.58 m and 0.24 m in large and small complex scenes, respectively, surpassing classic algorithms in terms of accuracy and reliability.

The aerostatic bearing-rotor system plays a critical role in ensuring the stable operation of highspeed, high-precision rotating machinery. Despite its importance, the system is often affected by nonlinear sub-synchronous vibration instabilities, which limit its performance and development. To address this issue, this study proposes a novel tangentially supplied (TS) bearing. Modified gas lubrication Reynolds equations are derived, and a fluid-structure interaction model for the gas bearing-rotor system is established. By predicting gas film thickness distribution, the bearing lubrication equations and rotor motion equations are solved simultaneously. The nonlinear dynamics of the system are analyzed using bifurcation diagrams, rotor center orbits, and frequency spectrum plots, with comparisons to conventional radially supplied (RS) bearing. The results reveal differences in gas film pressure distribution, where the tangentially supplied bearing demonstrates slightly reduced load capacity but significantly suppresses sub-synchronous vibrations. Additionally, it raises both the bifurcation onset speed and the instability threshold speed, thereby improving system stability. Finally, theoretical predictions are validated through speedup and coast-down experiments.

Timely and accurate failure analysis of photovoltaic (PV) systems is crucial forensuring the stable operation of power grids. However, existing failure analysis and diagnosis algorithms based on deep neural networks excessively rely on high-quality failure state data collected by sensors. This is extremely difficult to achieve in real photovoltaic power plants that are commonly equipped with self-protection mechanisms. To address this issue, we propose a Digital Multi-Twin integrating Theory, Features, and Vision (TFV-DMT) for failure analysis of PV strings in PV systems. This method first constructs theoretical simulation twins, feature twins, and visual twins based on the concept of digital twins, specifically tailored for actual PV systems, and designs a multi-twin collaborative model for model updating and failure diagnosis. Secondly, to better construct the visual twin, we introduce a Two-Dimensional Gram Angle Field Transformation Algorithm (TDGAF) to achieve targeted two-dimensional mapping of PV feature data, facilitating a more direct expression of failure characteristics. Furthermore, by constructing a Swish-activated Deep Convolutional Generative Adversarial Network (SDCGAN) to achieve balanced augmentation of mapping data, the model bias of theoretical simulation twins can be reduced. Finally, we propose a Swin-LT network that incorporates a Lightweight Dual-Channel Attention Module (LDAM) to better analyze the features of the visual twin, enabling more precise fault diagnosis. The algorithm has been validated on a real 250 kW grid-connected PV system, with results indicating that the proposed digital twin model is effective, achieving a diagnosis accuracy rate of 98.8 % for string failures.