Piezoelectric actuators (PEAs) are widely utilized in precision positioning and high-speed driving due to their exceptional performance. However, their hysteresis characteristics are significantly affected by temperature and driving voltage frequency variations. To address this, the Temperature-Rate Dependent Prandtl-Ishlinskii (TRDPI) model is proposed. This model enhances the Rate Dependent Prandtl-Ishlinskii (RDPI) framework by incorporating dynamic play operators, the Fermi–Dirac distribution function, and a corner compensation term to improve fitting accuracy. Model parameters are identified using an Improved Dung Beetle Optimizer (IDWO) algorithm, which employs Bernoulli chaotic mapping for population initialization, Levy flight strategy and T-distribution for position updates, and fractional-order dynamic boundary adjustment to avoid local optima and improve global search capability. Experiments were conducted to measure the temperature- and frequency-dependent behavior of the PEA under excitation within a frequency range of 1–200 Hz, across ambient temperatures ranging from 20 to 60 °C. Experimental results demonstrate that the TRDPI model performs better than the RDPI model in fitting the dynamic hysteresis of PEA. The relative maximum and mean square errors reduced to 6.11

This study introduces a traceable and quantitatively rigorous evaluation framework for digital-twin electrical capacitance tomography (ECT) system architectures applied to gas-liquid two-phase flow measurement. A threedimensional fluid-electric field coupling model (3D-FECM) is embedded into a digital-twin workflow that integrates multiphase-flow simulation, ECT forward modelling, virtual measurement generation, and image reconstruction, enabling systematic assessment of two-dimensional ECT (2D-ECT) and three-dimensional ECT (3D-ECT) sensors with 12-24 electrodes. Reconstruction performance and metrological stability are quantified using root mean square error (RMSE), mean absolute error (MAE), structural similarity index measure (SSIM), and standard uncertainty (SU) under controlled signal-to-noise ratio (SNR) conditions. Results show that 3D-ECT maintains reliable imaging quality for SNR > 50 dB and consistently outperforms 2D-ECT, which requires SNR >= 60 dB to reach comparable stability. Among the evaluated configurations, the 21-electrode 3D-ECT sensor demonstrates the most favorable overall performance. For liquid volume fraction (LVF) measurement, imagebased estimation achieves lower errors than capacitance-based methods (RMSE < 0.0152 for 3D-ECT and <0.0296 for 2D-ECT at 60 dB), with SU on the order of 10(-5), indicating excellent repeatability. High-speed experiments using a 21-electrode 3D-ECT system (100 fps) further validate the digital-twin predictions. These results establish clear performance benchmarks for ECT digital-twin architectures and provide a robust basis for future algorithm evaluation and intelligent ECT development.

The setting degree and interlayer strength of cement concrete are key indicators governing construction efficiency and structural safety, both of which are highly dependent on concrete moisture content. Therefore, this study developed a novel fringing electric field-based array sensor for depth-resolved moisture detection of cement concrete. A geometric model of the sensor array was simulated within COMSOL, and its structural configuration was optimized via orthogonal experimental design. With the optimal parameter configuration, the device achieved a penetration depth of 66.86 mm, a signal strength of 11.387pF, and a sensitivity of 0.267pF/

Reconstructing transient three-dimensional flame temperature fields in confined combustors from a single optical access is highly ill-posed due to wall emission and reflection and model mismatch, which can destabilize tomography and fixed-parameter filters. A physics-informed state-adaptive KalmanNet (SAKN) for online three-dimensional temperature reconstruction using a light field camera is proposed. SAKN constructs a radiation distribution factor based observation operator that incorporates wall bidirectional reflectance distribution function effects and window transmittance, and embeds two gated recurrent unit subnetworks into Kalman recursion to infer voxel-wise transition factors and time-varying noise covariances from innovations and state corrections. SAKN incorporates smoothing priors to suppress non-physical artifacts, maintaining stable reconstructions even under severe noise and model mismatch. In simulated leave-one-condition-out tests, SAKN achieves a mean normalized root mean square error (NRMSE) of 0.0138, compared with 0.041 for the smoothness-constrained updating recursive Kalman filter and 0.172 for the classical Kalman filter. The posterior covariance further yields physically interpretable uncertainty fields, revealing limited observability in low-temperature and near-wall regions.

Abstract Magnetic target positioning (MP) technology has emerged as a pivotal positioning technique owing to its inherent advantages of non-contact operation, high precision, and excellent portability. However, conventional methods (e.g. magnetic dipoles, standalone neural networks) lack target classification capability, suffer low accuracy in complex regions, and adapt poorly to multi-scenarios, limiting their practical application. To address this, we propose a hybrid random forest-multi-layer perceptron (RF-MLP) method for simultaneous MP classification and localization. The framework uses RF for efficient target classification and MLP for high-precision regression positioning, enabling automatic extraction of discriminative features while simplifying the network structure. A space multi-plane magnetic sensor array is deployed to capture rich spatial target information, further boosting the model’s performance. Experiments show it achieves 100% classification accuracy, tenfold faster computation than standalone MLP, and triaxial RMSEs of 0.474, 2.787, 5.540 mm, outperforming classical algorithms. This work resolves key limitations of conventional MP, providing a reliable solution for high-precision, multi-functional positioning in complex scenarios.

Bubbly flow measurement is essential in both industrial and biomedical applications, with electrical impedance tomography (EIT) emerging as a promising non-invasive tool for real-time imaging. This study explores the effect of electrode numbers on EIT performance in bubbly flow and investigates the potential of principal component analysis-regression (PCAR) for accurate gas void fraction (GVF) measurement. Using a 3D fluid-electric coupling model, time-resolved simulations and impedance measurements were conducted across varying electrode configurations. Results indicate that an 8-electrode EIT, combined with PCAR, achieves high accuracy and stability in GVF estimation. Experimental validation supports these findings, showing that PCAR-based GVF estimations align closely with ground truth and outperform traditional pixel averaging. While increased electrode numbers reduced image artifacts, they did not enhance GVF measurement accuracy. These findings advance multiphase flow assessment, offering valuable insights for enhanced monitoring in both industrial and biomedical applications.

Enhanced infrared remote sensing calibration accuracy requires higher precision in spaceborne blackbody emissivity measurement. Nonthermal equilibrium conditions caused by blackbody temperature gradient lead to spectral emissivity deviation and distortion. This work proposes a temperature-dependent source term decoupling based on the radiation distribution factor (SD-RDF) method, which records multidimensional correlation information during the energy transfer process through RDF, decoupling the temperature-dependent source term from geometric and material properties. Compared with the standard Monte Carlo method (MCM), relative errors (REs) in directional emissivity are controlled within +/- 0.025%. The emissivity measurement deviation is positively correlated with temperature nonuniformity. The emissivity correction framework is established using SD-RDF for the nonisothermal target, determining optimal reference temperature through spectral measurement deviation minimization. It achieves adaptive correction strategies for different temperature distributions and spectral bands. The effects of nonuniformity and sensor arrangement on correction accuracy are investigated. Experimental results show that emissivity curves calculated using SD-RDF-based correction temperature exhibit high consistency with a perfect isothermal blackbody in the 8-16 & micro;m. The SD-RDF-based correction method achieves four to eight times accuracy improvement with measurement error controlled within 0.04%, outperforming existing reference temperature strategies. The method provides essential theoretical support for enhancing spaceborne blackbody precision and achieving high-accuracy infrared calibration

Mosquitoes of the genera Aedes and Culex are major vectors of mosquito-borne diseases, posing serious threats to public health. Accurate detection of these species is therefore crucial for disease prevention and vector control. Traditional identification methods are time-consuming, labor-intensive, and prone to human error. With the rapid development of deep learning, automated mosquito detection has become feasible; however, existing object detection models still struggle with small-object recognition and high computational complexity. To address these limitations, this study constructs a self-developed dataset and proposes a lightweight mosquito detection model based on YOLOv8, termed LW-YOLO. The model integrates HGNetv2, Rep-Ghost, and SCDH modules into the backbone, neck, and head, respectively, enhancing both detection accuracy and computational efficiency. Experimental results show that LW-YOLO achieves a precision of 0.978, recall of 0.972, and mAP50 of 0.987, improving by 1.6%, 1.25%, and 0.7% over the baseline YOLOv8. Meanwhile, its parameter count and computational cost are reduced from 3.0 M and 8.1 GFLOPs to 1.2 M and 4.4 GFLOPs, corresponding to decreases of 60% and 45.7%, respectively. The proposed LW-YOLO model not only achieves accurate detection of Aedes and Culex mosquitoes, providing technical support for mosquito-borne disease prevention, but also offers a promising lightweight solution for deployment on resource-constrained embedded or edge devices.

Aiming at the problems of limited projection information, high noise sensitivity, and the inverse problem's ill-condition in the real-time temperature reconstruction of the three-dimensional (3D) dynamic flames, this work proposes a dynamic laminar reconstruction method that integrates light-field imaging with smoothness-constrained updating recursive Kalman filter (SURKF). This work innovatively integrates flame-radiation tensor decomposition and Tikhonov regularization into the Kalman filter (KF), and introduces the recursive least-squares method for online optimization of the state transition matrix. Analysis shows that the average relative errors of SURKF are 3.60 % and 2.60 % under the dynamic triangular and sinusoidal wave forms, respectively, which is 85.7 % more precise than the classical KF. The parameter optimization analysis reveals that the synergistic modulation of the process noise covariance (Q = 10) and the forgetting factor (r = 0.3-0.7) significantly balances the dynamic response and stability. In the low-temperature region (< 1200 K), reconstruction error is two times higher than that in the high-temperature region, but still maintains a relative error within 3 % through the adaptive regularization constraint. The work verifies the effectiveness of the SURKF algorithm in real-time 3D temperature tomography using a light-field camera. It provides a high-precision solution for dynamic combustion monitoring in aero-engine combustion chambers, gas turbines, and other scenarios.

To address the current lack of standardized evaluation systems and complete metrological traceability chains for flatness assessment algorithms, this study proposes a flatness digital measuring instrument model based on the minimum zone (MZ) method. By combining triangle and cross criteria, along with point cloud geometric feature analysis, barycentric coordinate methods, and projection techniques, a constraint framework satisfying the four fundamental sampling points is established. Based on rigorous mathematical and geometric derivations, a unified standard for constructing the flatness digital measuring instrument model is developed, and sampling procedures as well as reference model examples under different criteria are provided. Utilizing this model set and its implementation methodology, a series of validation experiments were conducted to assess the feasibility and applicability of various flatness evaluation algorithms and measurement software. Experimental results demonstrate that the proposed model is effective for verifying and evaluating flatness assessment algorithms, supporting accuracy validation down to 0.1 mu m. This research provides a reproducible and traceable technical pathway for the standardized verification of flatness algorithms, supporting quality control in ultra-precision manufacturing.

To address the lack of traceability mechanisms and standardized evaluation methods in current flatness assessment algorithms, this paper proposes a flatness digital measuring instrument (DMI) model based on the least squares method (LSM). Leveraging the inherent stability and measurement accuracy of LSM, an adaptive iterative equilibrium (AIE) theory and a range inclusion constraint principle are introduced to regulate the position of the fitted plane, thereby enabling effective control of the standard evaluation value. During model construction, a cross-shaped division method and a quadrant-based point distribution strategy are applied to generate the preliminary points under Gaussian noise interference. A Gaussian-Newton iterative algorithm, coupled with a constraint-based termination criterion, is employed to compute the AIE point and finalize the models. Considering both general and specialized algorithm validations requirements, two sets of standard model collections are designed. Experiments using representative flatness evaluation algorithms. Results demonstrate that the proposed model effectively facilitates algorithm comparison and accuracy verification, supporting evaluation precision down to the 0.1 mu m level. This model provides a repeatable technical approach for flatness algorithm validation, establishes a foundation for their traceability, and offers great potential for the standardized development of standardized development of DMI models for geometric quantities.

Four-dimensional electrical impedance tomography (4D-EIT) offers a non-intrusive sensing approach for industrial process monitoring, particularly for visualizing dynamic multiphase flow behaviors in complex reactors. However, its practical deployment is often constrained by limited spatial resolution and insufficient quantitative accuracy caused by the ill-posed nature of the inverse problem. To address these challenges, this study proposes a digital-twin-driven 4D-EIT framework that integrates a fully coupled fluid-electric simulation model with a lightweight deep learning reconstruction strategy. An Intelligent Bubbling Flow Measurement Model is developed to construct a high-fidelity virtual bubble column and corresponding EIT measurement system, generating a large-scale 4D synthetic dataset of 40,404 samples. A Lite-UNet network is trained to reconstruct three-dimensional bubble distributions directly from boundary voltage measurements. Virtual experimental results show that the proposed method achieves a CC of 0.7647–0.9343, an RMSE of 0.1317–0.1937 and a PSNR of 14.3827–19.6898 dB, significantly outperforming Tikhonov regularization. Physical experiments further validate accurate reconstruction of spatial structures and consistent gas void fraction trends with camera-based measurements. These results demonstrate improved quantitative accuracy, robustness, and dynamic imaging capability, providing an effective artificial intelligence-enabled solution for industrial multiphase flow monitoring.

This study develops a full-organ, system-level three-dimensional (3D) modeling and dynamic impedance simulation framework for four-dimensional (4D) respiratory monitoring using electrical impedance tomography (EIT). The proposed framework was used to evaluate representative single-layer, dual-layer, and triple-layer electrode configurations under three representative respiratory scenarios, including healthy ventilation and rep-resentative inflamed lung conditions. A patient-specific thoracic model incorporating realistic anatomy, regional lung deformation, and time-varying conductivity was constructed to assess 16-, 32-, and 48-electrode configurations under controlled respiratory conditions. Simulation results showed that multi-layer configurations improved volumetric sensitivity and reconstruction accuracy compared with single-layer sensing, with SSIM increasing from ap-proximately 0.54–0.72 to 0.61–0.75 and RMSE decreasing from 0.27–0.35 to 0.20–0.31. Among the representative electrode configurations evaluated in this study, the dual-layer configuration provided a favorable balance between reconstruction performance and practical implementation requirements for 4D respiratory monitoring. Clinical data acquired using the dual-layer setup further demonstrated spatially and temporally consistent 4D ventilation patterns that aligned closely with physiological expectations, supporting the potential of this workflow for individualized bed-side respiratory assessment. The proposed framework provides a basis for evaluating representative multi-layer EIT configurations under controlled respiratory conditions and can be extended in future studies to investigate broader anatomical variability, pathological conditions, and alternative electrode-layer arrangements.

This paper proposes an innovative dual-path differential method for waveguide temperature measurement through the time-of-flight difference of waves, adopting a transmitted wave propagation approach. Two geometric structures, namely Dual-Path and Spiral Dual-Path are designed. Tube diameter ratio and acoustic energy distribution are discussed to obtain optimal received waveforms. In finite element simulations and practical experiments, 6061 aluminum alloy is selected as the experimental material, and experiments are conducted within the temperature range of 0-400℃. The two new structures are compared with three traditional waveguide structures, including L-Form, Notched, and Perforated, improving the average resolution from 4℃/10 ns to 2℃/10 ns and 1℃/10 ns, reducing the absolute value of average error from 3.6℃ to 1.9℃ and 1.6℃, and decreasing the standard deviation by 43% and 57% respectively. The results indicate that optimizing the waveguide geometry is an effective approach to enhancing performance.

To address the issues of insufficient localization accuracy, low matching efficiency, and high cost of fingerprint library construction in complex indoor environments, a multi-source fusion localization method based on mutated signal segmentation fitting (MSSF) and an improved adaptive unscented Kalman filter is proposed. Firstly, a segmentation partitioning strategy integrating Delaunay Triangulation, flood-like lighting algorithm, and localized fitting algorithm for small-area segmentation is designed to construct a double-layer fingerprint library (DLFL) based on a virtual access point signal propagation attenuation model. This achieves high-precision signal fitting and rapid zone matching under sparse sampling conditions. Secondly, an improved adaptive unscented Kalman Filter algorithm incorporating Sage-Husa filtering is proposed. It dynamically estimates the observation noise covariance through an innovation augmentation mechanism, enhancing its robustness in non-linear noise environments characterized by Wi-Fi signal fluctuations and PDR drift. Combined with the DLFL, this further realizes a localization algorithm that fuses Wi-Fi signals and PDR information. Experimental results show that the proposed MSSF algorithm achieves an average error of merely 1.508 dB across 12 test points; the DLFL based on MSSF reduces the online matching time from 0.024 s to 0.004 s; and the multi-source fusion localization algorithm achieves a per-step average error of 0.326 m, representing a 56.7% improvement compared to traditional fusion localization methods. These results validate the comprehensive advantages of the proposed method in terms of accuracy, efficiency, and adaptability.

Coherent noise in digital holographic microscopy (DHM) seriously degrades the accuracy of quantitative phase imaging, limiting its applications in fields such as nondestructive testing. However, traditional numerical denoising methods struggle to achieve an ideal balance between noise suppression, detail preservation, and computational efficiency. To address this challenge, we propose a multi-scale attention efficient network (MAENet). This network employs a dual-encoder architecture to achieve complementary extraction of multi-scale features. To efficiently integrate the features from these two branches, a dual-branch dense attention fusion (DDAF) module is designed. It performs a weighted fusion of features from the dual branches via an adaptive attention mechanism and enhances feature representation via dense residual connections, significantly boosting the model’s denoising performance. Furthermore, a hierarchical fusion strategy is adopted to preserve high-frequency details in the shallow layers of the network while performing feature fusion in the deeper layers, thereby maximizing protection of image textures while effectively suppressing noise. To address the lack of paired training data in real-world scenarios, a DHM simulation system capable of simulating the key physical characteristics of coherent noise was constructed. Extensive experiments on the simulated dataset show that MAENet achieves a PSNR of 33.25 dB and an SSIM of 0.93042, outperforming various mainstream denoising algorithms and demonstrating its excellent performance in suppressing coherent noise, providing an effective solution for denoising in coherent imaging systems.

Particle image velocimetry (PIV) measurements in complex flows are fundamentally limited by systematic errors induced by particle trajectory curvature, which are inadequately modeled by conventional straight-line displacement assumptions. To address this limitation, we propose a Runge-Kutta forward diffeomorphic deformation interrogation (RK-FDDI) framework that explicitly incorporates curved particle trajectories into the image warping process, in which displacement compensation is performed along particle trajectories rather than being approximated by instantaneous velocity vectors. A hybrid-norm adaptive step-size strategy is further introduced to balance global flow smoothness and local curvature intensity. The deformation-compensated images are progressively refined via optical-flow iterations, allowing subpixel in-plane motion compensation and enhanced recovery of fine-scale flow structures. Extensive evaluations were conducted on synthetic particle image pairs generated from multiple flow fields, spanning various curvature intensities, particle concentrations, and particle diameters. Compared with forward-difference interrogation with post-correction (FDI2CDI) and forward diffeomorphic deformation interrogation (FDDI), RK-FDDI demonstrates improved robustness and higher accuracy. In the sea surface flow, the root mean square error is reduced by 21.89% and 20.71% relative to FDI2CDI and FDDI, respectively. As the curvature increases, RK-FDDI suppresses the error growth observed in FDDI, achieving a maximum reduction of 25.7%. RK-FDDI maintains consistently low errors across different particle concentrations and particle diameters, with particularly improved performance at a concentration of six particles per interrogation window and particle diameters around three pixels. Furthermore, experimental validation based on a 2D PIV system shows that RK-FDDI reliably recovers flow features in real measurements, confirming its applicability to achieve high-accuracy velocity reconstruction for complex flows.

Light field imaging based on spontaneous radiation presents challenges in achieving accurate three-dimensional (3D) flame measurements within confined spaces due to complex radiative transfer, limited optical access, and wall radiation. This work introduces an integrated framework comprising two novel approaches: a null-collision backward Monte Carlo light field imaging (NCBMC-LFI) for forward simulation and a kernel-based maximum a posteriori (K-MAP) method for inverse problems. The NCBMC-LFI for confined space can mitigate the effect of wall radiation on reconstruction accuracy, while K-MAP integrates reduced-order modeling with Bayesian inference. The proposed methods achieve average deviations of 1.9 K at a high spatial resolution scenario, surpassing LSQR and regularization methods while maintaining robust performance under 3 % 10 % noise. Experimental validation using confined C2H4/air-premixed flames demonstrates the effectiveness of reconstructing 3D temperature distributions (900 K 1550 K), consistent with other methods and kinetic simulation at the same equivalence ratio, with computation times reduced to 10 % of classical methods.

In this work, a novel bipedal stepping piezoelectric actuator based on stick–slip principle is proposed. An integrated stator with symmetrically positioned dual driving feet has been developed, characterized by its element analysis. Key structural parameters were optimized, and a prototype was fabricated for a series of experiments. The experimental results demonstrate that the actuator attains a maximum output speed of 4550 μm/s at a voltage of 150 V and a frequency of 350 Hz. The stepping efficiency is measured at 0.886, with a maximum horizontal load capacity of 170 g. Furthermore, the actuator exhibits a displacement resolution of 180 nm, making it highly suitable for precision actuation and applications in fields such as biomedical engineering.