Ping Huang

The thicknesses of pipe walls and anti-corrosion coatings, together with local void defects, are important parameters affecting the safe operation of long-distance oil and gas pipelines. To enable non-contact quantitative inspection, this study proposes an air-coupled ultrasonic thickness-resonance method. COMSOL finite-element models were established for single-layer steel, single-layer Plexiglass, Plexiglass–steel bilayer structures, and structures containing void defects with different thicknesses. Fast Fourier transform, combined moving-average–Savitzky–Golay preprocessing, and Welch power spectral density estimation were employed to extract resonance frequencies for thickness and defect-depth inversion. The numerical results show that Welch PSD estimation reduces spectral-estimation fluctuations through segment averaging and improves the stability of resonance-frequency extraction and thickness inversion. The relative thickness-inversion errors for single-layer steel and Plexiglass were within 2% and 2%, respectively, while those for the individual layers of the bilayer structures were within 1%. For void-defect models, resonance components associated with the intact thickness and the residual thickness of defective regions could be distinguished. The defect-depth errors for single-layer steel and Plexiglass were within 2% and 1%, respectively, while those for the bilayer structures were within 2%. Air-coupled ultrasonic through-transmission experiments were further conducted on 6.5 and 8.2 mm single-layer steel plates. The inverted thicknesses were 6.358 and 8.33 mm, corresponding to relative errors of 2.1% and 1.6%, respectively. The experimental results provide preliminary support for the thickness-resonance relationship and frequency-extraction method for single-layer steel plates, whereas the performance of Plexiglass–steel bilayer structures and defect-depth inversion requires further experimental validation.

To address key technical challenges in detecting radial circular runout in large rotating components—including difficult electromechanical runout acquisition, insufficient error separation for eccentric coupled signals, and poor environmental adaptability, This paper proposes a high-precision detection method integrating differential eddy current sensing with a Multi-Source Fusion Adaptive Robust Levenberg–Marquardt (MFAR-LM) algorithm. A coaxial symmetric differential sensor operating at 20 kHz was developed, featuring hardware-level common-mode interference suppression, signal amplification, and epoxy encapsulation for oil and dust resistance, enabling non-contact electromechanical runout detection. The MFAR-LM algorithm introduces a multi-source adaptive damping strategy based on residual improvement, gradient variation, and local curvature. Combined with Hessian matrix regularization and Huber robust weighting, it effectively evades local optima, suppresses outliers, and accurately extracts true runout from complex signals. Experimental results show that the sensor output fits the theoretical exponential model with an R ² of 0.9975, a 16 h static drift below 70 ppm, a compensated thermal error under 2 μm(± 5 °C), and a dynamic measurement deviation under 2 μm at 300 r min ⁻¹ . The MFAR-LM algorithm enhances detection accuracy from 130–160 μm to 9 μm ( R ² > 0.997), speeds up convergence by 30%, and achieves an expanded uncertainty of 2.5 μm ( k = 2, 95% confidence level). This integrated method provides an effective solution for high-precision in-situ geometric measurement in high-end major equipment.

The continuous growth of IIoT systems has significantly increased the number of connected devices and message interactions, creating higher requirements for communication mechanisms in terms of scalability and adaptability under dynamic network environments. Although MQTT is widely used for its lightweight communication, its traditional centralized broker architecture limits scalability and fault tolerance in large-scale data transmission, reducing system scalability and fault tolerance. Additionally, static QoS configuration is difficult to adapt to dynamic environmental changes, resulting in high end-to-end latency and limited system throughput. To address these issues, this paper proposes a distributed MQTT communication method based on intelligent QoS routing and hierarchical collaboration (DAIS-MQTT). This method designs a network routing algorithm based on a hierarchical tree structure (LCN), which effectively addresses the scalability limitation of centralized proxies by enabling multi-level proxy collaboration and self-recovery from faults. At the same time, it proposes a QoS routing algorithm based on intelligent decision trees (IQR), which jointly optimizes proxy selection and QoS levels to dynamically adapt to changes in the network environment, thereby solving the problem of insufficient adaptability in static QoS configurations. Experimental results show that compared with the traditional MQTT-based communication method, the DAIS-MQTT method reduces the average message delay by 29.9%, increases system throughput by 28.2%, and maintains a reliable transmission rate of 98.7% in unreliable network environments, making it suitable for high-dynamic and large-scale IIoT communication scenarios.

Low-carbon steel is widely used in construction, bridges, automobiles, shipbuilding, and pressure vessels industries as a fundamental material in China’s industrial manufacturing sector. Stress detection techniques for low-carbon steel are not only a current research focus but a key enabler for shifting maintenance of large-scale carbon steel structural components from post-failure repair to pre-failure prevention. To address the challenge of high-precision stress detection, we propose an eddy-current stress detection method using a gradiometer coil configuration. By incorporating the magneto-mechanical coupling effect in ferromagnetic materials and Maxwell’s equations, an eddy current stress detection model for low-carbon steel is established. This model is used to determine the influence of stress on eddy current signals in the differential gradiometer coil under an alternating magnetic field. A low-carbon steel stress detection platform was established to systematically analyze the influence of excitation frequency and lift-off distance on the detection signal. The results indicate that the detection signal increases with excitation frequency and exhibits a strong functional relationship with stress intensity. At an excitation frequency of 5 kHz, the stress detection sensitivity is 0.07242 mV/MPa, with a linearity coefficient R2 = 0.993 and excellent stability. Meanwhile, the lift-off distance significantly affects detection results; the detection signal decreases exponentially as the lift-off distance increases. After fitting the results, the coefficient of determination is R2 = 0.932. Overall, this study presents an eddy current stress detection method employing a gradiometer-structured coil, offering a new technical pathway for the nondestructive characterization of stress in low-carbon steel structures.

Accurate identification of tool wear conditions is of great significance for extending tool life, ensuring processing quality, and improving production efficiency. Current research shows that signals collected by a single sensor have limited dimensions and cannot comprehensively capture the degradation process of tool wear, while multi-sensor fusion recognition methods cannot effectively handle the complementarity and redundancy between heterogeneous sensor data in feature extraction and fusion. To address these issues, this paper proposes Hi-MDTCN (Hierarchical Multi-scale Dilated Temporal Convolutional Network). In the network, we propose a hierarchical signal analysis framework that processes the signal in segments. When processing intra-segment signals, we design a Multi-channel one-dimensional convolutional network with attention mechanism to capture local wear features at different time scales and fuse them into a unified representation. When processing signal segments, we design a Bi-TCN module to further capture long-term dependencies in wear evolution, mining the overall trend of tool wear over time. Hi-MDTCN adopts a dilated convolution mechanism, which can achieve an extremely large receptive field without building an overly deep network structure, effectively solving problems faced by recurrent neural networks in long sequence modeling such as gradient vanishing, low training efficiency, and poor parallel computing capability, achieving efficient parallel capture of long-range dependencies in time series. Finally, the proposed method is applied to the PHM2010 milling data. Experimental results show that the model’s tool condition recognition accuracy is higher than traditional methods, demonstrating its effectiveness for practical applications.

Existing magnetic tracer detection systems primarily rely on fundamental wave signal acquisition using non-differential sensor configurations. These sensors are highly susceptible to external interference and lack tomographic localization capabilities, hindering their clinical application. To address these limitations, this paper presents a novel method for achieving the deep spatial localization of tracers. The method exploits second harmonic signal detection at non-zero field points. By considering the combined nonlinear characteristics of the coil’s axial spatial magnetic field distribution and the Langevin function, a correlation model linking the signal peak and bias field is established. This model enables the determination of the tracer’s precise spatial location. Building on this framework, a handheld device for localizing magnetic nanoparticle tracers was developed. The device harnesses the second harmonic response generated by coupling an AC excitation field with a DC bias field. Our findings demonstrate that under conditions of reduced coil turns and weak excitation fields, the DC bias field exhibits exclusive dependence on the axial distance of the detection point, independent of particle concentration. This implies that the saturated DC bias field corresponding to the second harmonic signal can be used to determine the magnetic nanoparticle sample detection depth. The experimental results validated the method’s high accuracy, with axial detection distance and concentration reduction errors of only 4.8% and 4.1%, respectively. This research paves the way for handheld probes capable of tomographic tracer detection, offering a novel approach for advancing magnetically sensitive biomedical detection technologies.