Abstract:The effective acquisition of underwater acoustic information is the basis for conducting underwater acoustic science research. The underwater acoustic vector sensor is a new kind of underwater sound receiving transducer compared to traditional hydrophones, in which the vector sensor is the core carrier. It can obtain the complete ocean sound field information, including both sound pressure and particle velocity. In particular, the frequency-independent natural cosine directivity and the advantages of suppressing isotropic marine environmental noise make acoustic vector sensor (AVS) have great potential to be applied in underwater security defense, marine resource exploration, long-distance communication, and environmental monitoring. This article reviews the main developments of AVS over the past decade, including the exploration of novel sensing mechanisms, the application of new piezoelectric active materials, and the design of innovative structures. Furthermore, the comprehensive challenges to be addressed in future work, such as device design at very low frequencies or in deep ocean conditions, platform suitability considerations, and advanced calibration techniques are analyzed and discussed.

Abstract:The productization process of time-grating displacement sensors has revealed that the design of these sensors is primarily focused on the sensor structure itself, lacking a systematic design and optimization method. The design and experimental process are fragmented, making it impossible to fully utilize statistical knowledge to establish a model. The product design heavily relies on the experience of designers, which has significantly constrained the productization process of the sensors. To address the aforementioned issues, this article uses the magnetic field time-grating displacement sensor as the research subject. The experimental design is implemented using the uniform design method, and the experimental results are analyzed using stepwise regression analysis to construct the sensor model design method. Simulation tests and prototype tests are performed, and the results show that the sensor model, built based on this method, exhibits good prediction performance. The established accuracy model fits the simulation data by 89% with a confidence level of over 92% . It can effectively respond to the accuracy of the time-grating sensor, which is of great significance for the optimal design of the time-grating displacement sensor.

Abstract:Accurate measurement of contact force is the foundation and prerequisite for achieving active compliant assembly of components. This article proposes an assembly contact force measurement method of large components based on the distributed 3D force sensors for the precise and compliant control requirements of redundant drive parallel mechanisms. First, the kinematic model of assembly positioning mechanisms is formulated. An assembly contact force calculation method based on a dynamic model is proposed. Subsequently, to address the problem of gravity compensation during the compliant assembly process of components, a centroid selfcalibration method of the terminal based on multi-attitude is proposed. The centroid parameter is solved using the least squares method. Then, numerical simulation methods are used to analyze the measurement errors and installation angle errors of the sensor, as well as the influence of the centroid calibration strategy on the measurement error of assembly contact force. Finally, dynamic assembly contact force measurement experiments of large components are implemented in the laboratory. Compared with the six-dimensional force sensor, the experimental results show that the assembly contact force standard deviation measured by the distributed three-dimensional force sensors is reduced by 41. 6% , and the assembly contact torque standard deviation is reduced by 47. 1% .

Abstract:Terrain matching can provide position information to calibrate the inertial navigation system ( INS). However, underwater vehicles are often limited in their operational time within areas where the depth geographic information feature map can be matched. The position information provided by terrain matching is inherently uncertain. These factors lead to several challenges, including unknown segments of usable information and error characteristics that exhibit a variance-changing Gaussian distribution. To address these issues, this article proposes an adaptive Kalman filter based on performance monitoring of the measurement information (MMAKF) algorithm. First, a forward and backward adaptive Kalman filter utilizing data backtracking technology is designed. Then, the filtering results are employed to calibrate the INS. The results of sea tests show that the proposed MMAKF algorithm is effective for the comprehensive calibration of the INS during short-duration position information. This calibration method effectively corrects both position and attitude to values close to the reference, thereby achieving rapid and comprehensive calibration of the underwater INS.

Abstract:Currently, the types for fault diagnosis of three-phase inverters are mostly linear loads. As nonlinear loads are more and more widely used, the original methods make it difficult to solve the harmonic problems caused by nonlinear loads during faults. In addition, the existing fault diagnosis methods are mostly for switches or sensors, without comprehensive consideration of the above faults. To address the aforementioned problems, a simultaneous diagnostic method for open-circuit faults and current sensor faults in two-level three-phase inverters with nonlinear loads is proposed in this article. The inverter three-phase current state-space equation is formulated and the actual duty cycle function is derived. The continuous duty cycle function is used instead of the discrete switching function. Therefore, the state-space equation meets the observation requirements. A discrete sliding mode observer based on composite control is designed. It adopts the structure of zero-phase-shift repetitive control and proportional integral control in series, which can effectively track the harmonic current. Adjustable factor detection variables and adaptive thresholds are designed. The fast and accurate fault diagnosis can be realized by adjusting the adjustment factors. A test bench is established and experimentally evaluated that the diagnosis time for open-circuit faults is less than 3 ms, and the diagnosis time for sensor faults is less than 6 ms. The proposed fault diagnosis method can realize fast diagnosis of both open-circuit faults and sensor faults under nonlinear loads.

Abstract:To enable the robot avoid obstacles in real-time while performing complex tasks, an interactive learning approach for robot obstacle avoidance based on DP-KMP is proposed. First, the whole framework of this approach is constructed, which adopts the segmentation-generalization strategy to implement the learning of demonstrated trajectories with rapid segmentation and the planning of sub-trajectories for obstacle avoidance. During the learning phase, a trajectory segmentation strategy based on the DP algorithm is proposed to improve the efficiency of segmentation, while a Gaussian mixture model strategy is used to extract the reference database from each sub-trajectory. In the trajectory planning phase, the KMP model is used to implement the trajectory reproduction and generalization, while the reference database update strategy based on human-robot interaction feedback is introduced to enhance the success rate of obstacle avoidance through human-robot interaction. Aiming at the issue that this update strategy may fail to plan a successful obstacle avoidance trajectory, two available conditions are proposed for inspecting the sub-trajectories generated by segmentation. Finally, the effectiveness of mentioned available conditions is verified by simulation. Experimental results show that the proposed approach segments the demonstrated trajectories in just 0. 084 and 0. 107 s for two different experiments, respectively. Additionally, the KUKA cobot successfully avoids all static and suddenly changing obstacles through multiple interactions with user during the execution of the different lift-place tasks.

Abstract:Due to the variety of cross-sectional areas in 10 kV distribution network lines, when the non-contact voltage device measures the voltage of different line diameters, the line-electrode coupling capacitance becomes time-varying. This variation complicates the determination of the voltage dividing ratio, making it difficult to accurately invert the line voltage. In this paper, under the condition that the outer copper electrode is not grounded, the self-calibration method of dual-probe coupling capacitance based on the capacitanceswitching array is proposed. This method eliminates the influence of line-diameter variation on the measurement results by casting the capacitance between the inner and outer copper electrodes of the dual-probe. Additionally, the sensitivity formula of each measured voltage is derived to quantify the effect of measurement errors on the inverted line voltage, and the optimized probe parameters are used to realize the adaptive measurement of the distribution network line diameter. Finally, the simulation circuit is designed and the experimental platform is constructed to complete the calibration of the sensor parameters and perform the adaptive testing of different line diameters. The experimental results show that when the cross-sectional area of the 10 kV distribution line changes from 70 mm 2 to 150 mm 2 , the maximum relative error between the line voltage inverted by the sensor and the real line voltage is 2. 685% , which meets the standard of electronic voltage transformer. This in turn can validates the effectiveness of the proposed method.

Abstract:Aiming at the control process of two-dimensional motion platform, this paper proposes a variable synchronization ratio circular interpolation method, specifically targeting the issue of large circular interpolation errors caused by the asynchronous motion of two physical axes. The proposed method uses the relationship between the change of interpolation point coordinates and the angle of the center of the circle. A virtual axis is constructed with the corresponding angle of the circle center at each interpolation point serving as the output. This virtual axis acts as the guide axis of synchronous motion, while the physical axes serves as the follower axes, establishing the synchronous motion relationship between the virtual axis and the physical axes. Based on the positional relationship between the guide axis and the follower axes, the synchronization ratio is obtained using the center angle as the intermediate parameter. The circular interpolation process is then transformed into the calculation of the synchronization ratio for each interpolation cycle. By adjusting the synchronization ratio during each interpolation cycle, the synthetic direction of the physical axes is changed, enabling precise circular interpolation. Experimental results show that the proposed method can control the two-dimensional motion platform to achieve circular interpolation at a relatively smooth speed. This not only improves the efficiency of the circular interpolation but also achieves high accuracy, making it highly suitable for applications requiring precise motion control. Keywords:two-dimensional motion platform; circular interpolation; synchronous motion; synchronizati

Abstract:The four-switch buck-boost converter has the advantage of multi-mode operation, which is suitable for a wide range of voltage regulation. To solve the problem of multi-mode efficiency reduction and multi-mode control, a loop-switch selective control strategy is proposed. Firstly, the Boost-T and Buck-T modes are introduced into the dead zone of the converter to form a four-mode control, which effectively eliminates the influence of the dead zone on the traditional two-mode control. Secondly, the switching boundary in the intermediate mode is optimized and the frequency conversion control in a small range is utilized to address the problem of reducing the efficiency of the converter after the introduction of the intermediate mode. Then, to solve the multi-mode control problem, a switchselective control strategy with input voltage feedforward is designed by exploring the law of the conjugate pole distribution of the system with the model parameters. Finally, an experimental prototype based on GaN is fabricated. Experimental results show that the output voltage ripples within 2. 5 V when the system load bursts, the output voltage ripples within 1 V when the mode is switched, and the maximum operating efficiency in the intermediate mode reaches 96. 3% when the load is full

Abstract:During the process of cutting coal seams with mining equipment, different characteristics of coal and rock can cause changes in the temperature of the cutting teeth. To explore the changing characteristics of the temperature field of the pick, the heat source distribution in the process of cutting the coal seam by the pick is analyzed. The theoretical calculation model of the temperature field of the pick in the process of cutting the coal seam is established using the heat source method. Considering the actual working conditions of the mining equipment, the calculation process of the convection heat transfer coefficient of the spray system is deduced. Taking the MG2 ×55 / 250-BWD shearer as the research object, a model of rotating cutting of coal and rock is formulated based on ABAQUS finite element software. The distribution and variation of the temperature field of the pick under different working conditions are studied. Based on Fluent fluid simulation software, the spray cooling model is established, and the convective heat transfer coefficient of the pick under different spray pressures is obtained. Then, the heat transfer coefficient is brought into the ABAQUS finite element model to simulate the temperature field of the pick during coal cutting under water spray cooling. Finally, a temperature collection experimental platform for drum-cutting coal and rock is established. A temperature collection system for cutting teeth is designed. Some experiments on drumcutting coal and rock under different working conditions are conducted, and the variation law of the temperature field of the cutting teeth is analyzed. The correctness of designed theoretical calculations and simulation models is evaluated. The results show that the temperature field of the pick has different distribution patterns under different working conditions, and the maximum temperature is positively correlated with the rotating speed of the cylinder, the traction speed, and the coefficient of coal rock firmness. Water spray pressure is inversely proportional to the cutting temperature, and the water spray cooling can greatly reduce the cutting temperature by over 20% , compared to dry cutting conditions. The research results provide new theories and methods for identifying coal-rock properties based on the temperature field of the cutting teeth.

Abstract:The transparent windshield is a key structural part of the aircraft, which requires outstanding performance. To reduce weight, the aircraft′s transparent windshield utilizes a heterogeneous multi-layer composite structure of “organic glass-transparent polycarbonate-organic glass”, where the performance is closely related to the thickness of each layer of the transparent windshield. Existing measurement methods cannot meet the comprehensive requirements for efficient and high-precision layer-by-layer thickness measurement of the transparent windshield. Therefore, a new method for spectral-domain low-coherence measurement of the multi-layer transparent windshield is proposed. Firstly, by using the orthogonal dispersion to enhance dispersion capability, the measurement range is increased. Then, a mapping model between the thickness of the transparent windshield and the spectral wavenumber is formulated, and the model parameters are calibrated with high-precision parallel flat crystal. Finally, a measurement system is established. Experimental results show that the proposed method has a maximum range exceeding 41 mm, with micrometer-level measurement accuracy, and a measurement speed less than 8 ms. It can meet the requirements for thickness measurement of the multilayer aircraft transparent windshield.

Abstract:Sensor arrays are commonly used for condition monitoring in mechanical equipment. However, traditional wireless communication faces challenges in closed metal enclosures due to electromagnetic shielding, making effective transmission difficult. Ultrasonic waves, capable of penetrating metal structures, show significant potential as carriers for information transmission. While multichannel synchronous transmission can enhance communication efficiency, signal crosstalk remains a pressing issue. To address this, this paper proposes a crosstalk suppression technique based on acoustic metamaterials. By utilizing the bandgap effect of Lamb wave dispersion curves, the method blocks the propagation of sound waves between channels. The metamaterial structure is designed through numerical simulation to determine the bandgap range, and experiments are conducted to verify the effectiveness of crosstalk suppression. The study successfully achieves independent communication across three channels, significantly reduces inter-channel crosstalk, and ensures high decoding accuracy, there by meeting the communication requirements in closed metal structures. This research provides a novel and effective solution to the crosstalk issue in multi-channel ultrasonic communication.

Abstract:Gas-liquid two-phase flow is prevalent in various fields such as chemical engineering, petroleum and natural gas transportation, and the nuclear industry. As a crucial parameter of gas-liquid two-phase flows, pressure drop is essential for the accurate measurement and precise prediction of two-phase flow rates. In this paper, a pressure drop measurement system is designed, and experiments are carried out under the pressure range of 0. 1 ~ 0. 5 MPa. Pressure drop is measured for horizontal gas-liquid two-phase flow in the pipe with an inner diameter of 50 mm at different pressures. The influence of experimental characteristic parameters on pressure drop is analyzed and discussed. A pressure drop experimental database is established and existing pressure drop prediction correlations are compared and verified. The results indicate that the existing pressure drop correlations are not applicable to high pressures and large-diameter pipes. Therefore, a new pressure drop correlation is proposed based on the analysis of the experimental parameters. When compared with the experimental database, over 93% of experimental data exhibit a relative error within ±30% , and the mean absolute percentage error is 13. 57% . This verifies the extrapolative capability and applicability of the new pressure drop prediction correlation.

Abstract:Polarization metasurfaces are innovative optical devices that can precisely control the polarization state, phase, and amplitude of light by designing micro-nano structures of different materials. Most existing polarizers have limited functionality and larger sizes due to their fixed physical structures and material properties, making it difficult to meet the demands for miniaturization and multifunctionality in modern optical systems. This article investigates a tunable phase-change polarization metasurface based on an all-dielectric GaAs/ Sb2 Se3 structure operating in the near-infrared wavelength range ( 780 ~ 1 100 nm). By integrating polarization conversion meta-atoms and phase-switching meta-atoms, a linear polarization device with high transmittance (99. 625% ), high extinction ratio (32. 8 dB), and controllable polarization angle is designed. Notably, the introduction of the phase-change material Sb2 Se3 enables dynamic tuning of the polarizer under different phase states, offering significant advantages for controlling the polarization state across various wavelength ranges. The effectiveness of the design is evaluated through theoretical analysis and finite - difference time - domain simulations, demonstrating the broad application potential of this metasurface in infrared detection, imaging, and communication. This research provides an innovative solution for the miniaturization and multifunctionality of optical systems and lays the foundation for future material optimization and exploration of new applications.

Abstract:The robotic precise peg-in-hole assembly in the unstructured environment is a problem of non-cooperative games. The position uncertainty of the peg brings challenges to the subsequent hole search and insertion. Hence, the robot needs to adjust the position of the peg to eliminate the peg-in-hole posture deviation. In this article, the peg adjustment is divided into two stages, including rough adjustment and fine adjustment. First, in the rough adjustment phase, the force-angle samples of the peg are collected when the peg does not contact the hole. They are used as the input of the MLP model for training. In this way, the robot arm for rough compensation is guided. Next, in the fine adjustment phase, the RLVAC model is formulated to estimate the peg-in-hole contact state and accurately adjust the position of the peg. A peg-hole contact fuzzy inference model is established to estimate the peg-in-hole contact state. Based on the peg-in-hole contact state, the optimal admittance control parameters are found by the reinforcement learning algorithm, which incorporates the fuzzy reward mechanism to realize the tight fit of the peg-in-hole surface. Finally, the comprehensive experiment is implemented on the peg with an unknown posture. Comparison with other conventional methods is analyzed in terms of adjustment accuracy, running time, and success rate.

Abstract:To study the effect of reference plane inclination on micro-thrust measurement, a high-precision ground inclination measurement method based on an inverted pendulum structure is proposed. The effect on system stiffness and measurement accuracy is explored by precisely controlling the inclination angle of the reference. Calibration experiments using a high-precision electromagnetic calibrated force revealed that the gravitational moment of the inverted pendulum was about 162. 80 mNm and the initial inclination angle of the installation reference was about -1. 78 mrad. Next, an experimental study on the effect of reference plane inclination is conducted by piezo deflector stage, followed by an uncertainty analysis. The experimental results indicate that the reference plane inclination has a significant amplifying of pendulum deflection angle when the pendulum gravity moment stiffness is approaching the flexible pivot stiffness. The amplification is about 5. 73 when the gravitational moment is 162. 80 mNm and K0 = 191. 46 mNm/ rad, which is 0. 25% different from the simulation results. Due to the amplification effect of the inverted pendulum, the stiffness of the inverted pendulum will be deviated by 57% at a reference plane inclination angle of 10 μrad. The deviation of the micro-thrust measurements was better than 0. 98% and the deviation of the system stiffness was better than 2% with the model correction. Considering the effects of the experimental environment and instrumentation, the relative uncertainty of the experimental measurements was 1. 40% . The results of the study provide support for improving micro-thrust measurement performance. Keywords:micro-newton thrust measurement; electromagnetic f

Abstract:Photoelectric insolation meter may experience significant errors when measuring scattered radiation under sunny and cloudy weather conditions due to their shading design. To reduce the measurement errors, the reflection direct radiation error and diffuse distribution error of the four-, six-, and eight-divided shading schemes under different weather conditions are analyzed using TracePro simulation. The analysis concludes that although the six- and eight-divided shading scheme can capture the dome of the sky more effectively, increasing the number of the shading holes makes it more difficult to correct the diffuse distribution error. Consequently, this paper establishes a diffuse distribution error correction model for the four-divided shading scheme and verifies its validity through simulation and actual measurement. After the correction, the correlation between the scattered radiation received by the total radiation sensor and the radiation measured by the scattering sensor is significantly improved, with the Pearson correlation coefficients above 0. 95, and the average relative error reduced to below 4% . In the actual test, the monthly cumulative error of the photoelectric heliometer was reduced from 12. 25% to 2. 66% , significantly enhancing measurement accuracy. Keywords:photoelectric insolation meter; shading scheme; reflected direct radiati

Abstract:An improved particle swarm optimization-simulated annealing algorithm (IPSO-SAA) is proposed to enhance the measurement accuracy of the coordinate measuring machine (CMM) by identifying the optimal measurement area for the measured object. First, the distribution pattern of volumetric errors within the CMM measurement space is analyzed. Individual geometric error models are fitted using the least squares method, and an optimization model for the point errors in the CMM space is established. The proposed IPSO-SAA method, which combines adaptive weighting, adaptive disturbance force and simulated annealing algorithm, outperforms conventional particle swarm optimization (PSO) and adaptive particle swarm optimization (APSO) algorithms. Comparative experiments show that IPSO-SAA is superior to PSO and APSO algorithms in terms of the best, worst, mean, and standard deviation values. Additionally, the optimization speed is increased by 45. 1% and 29. 2% , respectively. The results obtained from the IPSO-SAA algorithm identification indicate that, for a planning optimization space of 30 mm×30 mm×30 mm, the optimal measurement area in the CMM identified by the IPSO-SAA algorithm is 206 mm ≤ X ≤ 236 mm, 350 mm ≤ Y ≤ 380 mm, and - 262 mm ≤ Z ≤ - 232 mm. Comparative experiments using a high-precision standard ball, with a diameter of 15. 874 7 mm and a sphericity of 50 nm, demonstrate that when placed within the optimal measurement area in the CMM, the minimum diameter measurement error of the standard ball is 1. 7 μm, validating the correctness of the proposed method. The method presented in this study is general and can be used to determine the optimal measurement area of CMM for other measured objects.

Abstract:In recent years, the development of global navigation satellite system has matured. However, in complex environments such as urban canyons, forest shading and other scenes, issues such as insufficient visible stars and serious satellite signal attenuation, leading to positioning blind areas. Different from the conventional strategy of finding positioning optimization methods from the satellite end and the receiver end, the GNSS positioning method based on metasurface assistance changes the radio environment from the propagation path. It realizes beamforming of satellite signals at different reflection angles, effectively establishing a virtual line-of-sight link. Firstly, the positioning model is built and a metasurface that can reflect the 15° incident Beidou B1 signal to 30° is designed. Secondly, the relationship between the quality of its reflected signal, receiving angle, and receiving distance is analyzed through darkroom measurement experiments, verifying the beamforming capability. Finally, the pseudo-range single point positioning accuracy of Beidou signal based on metasurface reflection is evaluated. The root mean square error in the east, north and zenith directions were 2. 138, 4. 456 and 8. 268 m, respectively, which verifies the feasibility of metasurface assisted GNSS localization and provides a reference for the subsequent metassural dynamic positioning.

Abstract:To address the problem that the existing reflective grating and reading head cannot set more grid lines to further improve the stability and accuracy when the volume is small, a method for improving the displacement measurement accuracy of the reflective grating is proposed. The method corrects the grating Moire signal quality by analyzing the error of three main factors that affect the grating Moire signal, including DC, amplitude, and orthogonality deviation. The arctangent subdivision method is used to interpolate and subdivide the corrected grating Moire signal to obtain the measurement result. Meanwhile, it improves the CORDIC algorithm used in the orthogonality error correction and arctangent subdivision. Experimental results show that the subdivision error is reduced from 45″ before correction to 6″ after correction, which verifies the repeatability of the system, the consistency and hysteresis of the forward and reverse displacement. The measurement accuracy reaches ± 1 μm under the same grid logarithm, which is significantly better than that of the commercial grating displacement senso Key r.

Abstract:Infrared and visible images exhibit complementary characteristics, making their fusion highly suitable for achieving high accuracy and robustness in target detection for applications such as autonomous driving. However, existing multimodal object detection algorithms often feature large models and long inference times, making them unsuitable for deployment on edge devices. Additionally, direct fusion methods fail to fully leverage the strengths of different modalities. To address these challenges, we propose a fusion object detection algorithm that integrates a gradient operator and an attention mechanism. A gradient operator is employed to design a customized convolutional layer for capturing image texture. In the infrared branch, coordinate attention is incorporated to enhance target localization capabilities. Additionally, a weight generation network is introduced to adaptively balance the features of both modalities. The algorithm is modular and lightweight, making it ideal for edge device deployment. Experiments on benchmark datasets demonstrate that the proposed method achieves mAP@ 0. 50 and mAP@ 0. 5 ∶ 0. 95 scores that are 6. 3% and 7. 2% higher, respectively, than singlemodal detection using visible images, and 11. 3% and 9. 8% higher than infrared detection. The inference frame rate reaches 22. 7 FPS, meeting real-time processing requirements. Keywords:object detection; dual-modal; f

Abstract:Segmenting the inner cavity regions of precision parts from industrial CT images is crucial for accurate dimensional measurements. However, inner cavities are often non-closed and exhibit CT grayscale values similar to the background, making accurate segmentation challenging for existing algorithms. To address this, this paper introduces a local multi-scale convex hull algorithm that integrates convex hull concepts with mathematical morphology. Starting from an initial segmentation, the algorithm fills the inner cavity regions, followed by closing and Boolean operations to achieve complete segmentation. Comparative experiments with various segmentation methods demonstrate the effectiveness of the proposed approach, achieving an F1 score of 0. 973 5 on CT images of automotive parts. The results indicate that the proposed algorithm offers high accuracy and efficiency, enabling the precise and rapid segmentation of non-closed inner cavity regions in diverse industrial CT image applications.

Abstract:Traditional LiDAR systems encounter significant challenges with pose estimation accumulation errors during SLAM and require improvements in algorithm efficiency and real-time performance. This paper introduces a tightly coupled LiDAR-IMU integration method with enhancements in three key areas: Point cloud undistortion, feature extraction and registration, and tightly coupled pose estimation. To mitigate point cloud distortion caused by high-speed motion, a continuous time-domain motion correction method is proposed, along with a pre-fitting plane mechanism for feature extraction to balance accuracy and computational efficiency. To address the high computational cost of KNN search during feature matching, a tracking mechanism is introduced to reduce complexity. For improving state estimation accuracy, the method employs a framework optimized using a nonlinear geometric observer. Evaluation on public datasets demonstrates that the proposed method reduces trajectory APE by 30. 52% and 21. 36% compared to LIO-SAM and Fast-LIO, respectively. Additionally, computational efficiency improves by 59. 9% and 43. 7% , making the method highly effective for real-time applications.

Abstract:A human-machine cooperative health assessment method is proposed for mechanical equipment to evaluate its health condition and support hierarchical maintenance decisions. First, symptom parameters are extracted from collected vibration, pressure, and torque signals. A novel fuzzy residual shrinkage network is then developed to establish the status membership function of the mechanical equipment, forming the individual assessment model based on the extracted parameters. Next, the status memberships from each model are integrated into a collective hesitation fuzzy health assessment matrix. The Best-Worst Method is applied to calculate the priority of each assessment model, while TOPSIS with linguistic Z-numbers is employed to analyse the impact of different operational states on the equipment′s behaviour. Finally, a hesitation fuzzy weighted average operator is used to define the health index of the mechanical equipment, and health levels are identified using the versatile k-means clustering method to support hierarchical maintenance decisions. Validation results demonstrate that the proposed method excels in adaptability to different conditions and stability in performance.

Abstract:Aiming at the problem that the accuracy rate and false alarm rate are difficult to balance during the maintenance of key components of wind turbines, a wind turbine gearbox fault early warning method based on two-stage multi-dimensional data generation and real-time health index is proposed. To account for the impact of wind speed on the unit′s operational state, a Gaussian Mixture Hidden Markov Model leveraging historical wind speed data is proposed to forecast short-term wind speed. Next, to enhance early warning accuracy, a real-time dynamic threshold-setting method utilizing a two-stage multi-dimensional data generation process is introduced. Based on the predicted wind speed sequence and generator data, the threshold interval for the current oil temperature is established. Finally, the actual gearbox oil temperature values and the output of the health status discriminator are integrated to assess whether the wind turbine gearbox is in an abnormal state. Simulation results using real-world data demonstrate that the proposed method significantly reduces the false alarm rate and provides a potential fault warning for the wind turbine gearbox up to 17 hours in advance.

Abstract:To mitigate the impact of thermal errors on the positioning accuracy of the CNC machine tool feed system and improve the consistency of processed products, a thermal error prediction model based on the RIME-optimized BP neural network is introduced. Temperature sensors and a laser interferometer are deployed under various operating conditions to collect temperature and lead screw thermal error data. Fuzzy C-means clustering and grey relational analysis are applied to select features from temperature samples, identifying key temperature feature points. The RIME-BP thermal error prediction model is constructed using temperature and screw position coordinates as inputs and screw thermal error as the output. For the H650GA gear grinding machine, the K-fold cross-validation method is used to validate the model′s prediction accuracy, which is compared with GA-BP, BP, and SVM models. The results show that the proposed model achieves an average coefficient of determination (R 2 ) of 0. 995, which is 3. 54% , 9. 58% , and 17. 75% higher than the GA-BP, BP, and SVM models, respectively. The proposed method provides theoretical and technical guidance for thermal error compensation and holds significant engineering application potential. Keywords:thermal error prediction; feed system; feature selection; ri

Abstract:The influence of non-uniform microstructure, uneven roughness and small thickness change of thermal barrier coating on terahertz waves is highly overlapping and reduce the accuracy of thickness measurement. To address this, a thickness measurement method based on a residual cross-domain network driven by an abnormal dominant loss function is proposed to minimize prediction outliers. First, an analytical model of terahertz signals is developed, demonstrating that two key parameters for thickness measurement— time of flight and refractive index—can be derived from time-domain and frequency-domain data. A Fast Fourier Transform layer is introduced to extract frequency-domain features, while a gated recurrent layer captures changes in the time-domain. Additionally, a division layer is designed based on transmission rules, enabling the construction of an interpretable cross-domain residual network. Given the overlapping effects of roughness and microstructural changes on terahertz signal peaks, an abnormal dominant loss function is established to assign greater weight to outliers. Finally, thermal barrier coating samples were prepared, and terahertz experiments were conducted. The results demonstrate that the proposed method achieves a maximum relative thickness measurement error of less than 2. 5% .

Abstract:The influence of non-uniform microstructure, uneven roughness and small thickness change of thermal barrier coating on terahertz waves is highly overlapping and reduce the accuracy of thickness measurement. To address this, a thickness measurement method based on a residual cross-domain network driven by an abnormal dominant loss function is proposed to minimize prediction outliers. First, an analytical model of terahertz signals is developed, demonstrating that two key parameters for thickness measurement— time of flight and refractive index—can be derived from time-domain and frequency-domain data. A Fast Fourier Transform layer is introduced to extract frequency-domain features, while a gated recurrent layer captures changes in the time-domain. Additionally, a division layer is designed based on transmission rules, enabling the construction of an interpretable cross-domain residual network. Given the overlapping effects of roughness and microstructural changes on terahertz signal peaks, an abnormal dominant loss function is established to assign greater weight to outliers. Finally, thermal barrier coating samples were prepared, and terahertz experiments were conducted. The results demonstrate that the proposed method achieves a maximum relative thickness measurement error of less than 2. 5% .

Abstract:The influence of non-uniform microstructure, uneven roughness and small thickness change of thermal barrier coating on terahertz waves is highly overlapping and reduce the accuracy of thickness measurement. To address this, a thickness measurement method based on a residual cross-domain network driven by an abnormal dominant loss function is proposed to minimize prediction outliers. First, an analytical model of terahertz signals is developed, demonstrating that two key parameters for thickness measurement— time of flight and refractive index—can be derived from time-domain and frequency-domain data. A Fast Fourier Transform layer is introduced to extract frequency-domain features, while a gated recurrent layer captures changes in the time-domain. Additionally, a division layer is designed based on transmission rules, enabling the construction of an interpretable cross-domain residual network. Given the overlapping effects of roughness and microstructural changes on terahertz signal peaks, an abnormal dominant loss function is established to assign greater weight to outliers. Finally, thermal barrier coating samples were prepared, and terahertz experiments were conducted. The results demonstrate that the proposed method achieves a maximum relative thickness measurement error of less than 2. 5% .

Abstract:In robot-assisted bone milling surgery, the generation of a large amount of heat can easily lead to thermal necrosis. The irrigation fluid is commonly used to reduce the milling temperature. However, traditional research on bone milling temperature does not consider the influence of irrigation fluid. To address this, this study proposes a bone milling temperature model that comprehensively considers the effect of irrigation fluid. First, based on the finite element method, a temperature field model of cortical bone milling with a ball-end milling cutter is formulated. Then, the convective heat transfer coefficient of the irrigation fluid at different flow rates is calibrated using the point heat source temperature field theory. The results indicate that, at flow rates of 31. 6, 43. 3 and 65 μm 3 / s, the convective heat transfer coefficients are 400, 800 and 1 100 W/ ( m 2·℃ ), respectively. Finally, response surface methodology is utilized to analyze the effects of various milling parameters on the maximum milling temperature. The analysis shows that milling depth has the most significant impact on the maximum temperature, followed by irrigation fluid flow rate, feed speed, and milling angle. This temperature model and the accompanying analysis provide valuable references for selecting preoperative milling parameters and help reduce the risk of thermal necrosis during bone milling. Keywords:bone milling temperature; finite element meth

Abstract:In robot-assisted bone milling surgery, the generation of a large amount of heat can easily lead to thermal necrosis. The irrigation fluid is commonly used to reduce the milling temperature. However, traditional research on bone milling temperature does not consider the influence of irrigation fluid. To address this, this study proposes a bone milling temperature model that comprehensively considers the effect of irrigation fluid. First, based on the finite element method, a temperature field model of cortical bone milling with a ball-end milling cutter is formulated. Then, the convective heat transfer coefficient of the irrigation fluid at different flow rates is calibrated using the point heat source temperature field theory. The results indicate that, at flow rates of 31. 6, 43. 3 and 65 μm 3 / s, the convective heat transfer coefficients are 400, 800 and 1 100 W/ ( m 2·℃ ), respectively. Finally, response surface methodology is utilized to analyze the effects of various milling parameters on the maximum milling temperature. The analysis shows that milling depth has the most significant impact on the maximum temperature, followed by irrigation fluid flow rate, feed speed, and milling angle. This temperature model and the accompanying analysis provide valuable references for selecting preoperative milling parameters and help reduce the risk of thermal necrosis during bone milling. Keywords:bone milling temperature; finite element meth

Abstract:In robot-assisted bone milling surgery, the generation of a large amount of heat can easily lead to thermal necrosis. The irrigation fluid is commonly used to reduce the milling temperature. However, traditional research on bone milling temperature does not consider the influence of irrigation fluid. To address this, this study proposes a bone milling temperature model that comprehensively considers the effect of irrigation fluid. First, based on the finite element method, a temperature field model of cortical bone milling with a ball-end milling cutter is formulated. Then, the convective heat transfer coefficient of the irrigation fluid at different flow rates is calibrated using the point heat source temperature field theory. The results indicate that, at flow rates of 31. 6, 43. 3 and 65 μm 3 / s, the convective heat transfer coefficients are 400, 800 and 1 100 W/ ( m 2·℃ ), respectively. Finally, response surface methodology is utilized to analyze the effects of various milling parameters on the maximum milling temperature. The analysis shows that milling depth has the most significant impact on the maximum temperature, followed by irrigation fluid flow rate, feed speed, and milling angle. This temperature model and the accompanying analysis provide valuable references for selecting preoperative milling parameters and help reduce the risk of thermal necrosis during bone milling. Keywords:bone milling temperature; finite element meth

Abstract:In robot-assisted bone milling surgery, the generation of a large amount of heat can easily lead to thermal necrosis. The irrigation fluid is commonly used to reduce the milling temperature. However, traditional research on bone milling temperature does not consider the influence of irrigation fluid. To address this, this study proposes a bone milling temperature model that comprehensively considers the effect of irrigation fluid. First, based on the finite element method, a temperature field model of cortical bone milling with a ball-end milling cutter is formulated. Then, the convective heat transfer coefficient of the irrigation fluid at different flow rates is calibrated using the point heat source temperature field theory. The results indicate that, at flow rates of 31. 6, 43. 3 and 65 μm 3 / s, the convective heat transfer coefficients are 400, 800 and 1 100 W/ ( m 2·℃ ), respectively. Finally, response surface methodology is utilized to analyze the effects of various milling parameters on the maximum milling temperature. The analysis shows that milling depth has the most significant impact on the maximum temperature, followed by irrigation fluid flow rate, feed speed, and milling angle. This temperature model and the accompanying analysis provide valuable references for selecting preoperative milling parameters and help reduce the risk of thermal necrosis during bone milling. Keywords:bone milling temperature; finite element meth

Abstract:In robot-assisted bone milling surgery, the generation of a large amount of heat can easily lead to thermal necrosis. The irrigation fluid is commonly used to reduce the milling temperature. However, traditional research on bone milling temperature does not consider the influence of irrigation fluid. To address this, this study proposes a bone milling temperature model that comprehensively considers the effect of irrigation fluid. First, based on the finite element method, a temperature field model of cortical bone milling with a ball-end milling cutter is formulated. Then, the convective heat transfer coefficient of the irrigation fluid at different flow rates is calibrated using the point heat source temperature field theory. The results indicate that, at flow rates of 31. 6, 43. 3 and 65 μm 3 / s, the convective heat transfer coefficients are 400, 800 and 1 100 W/ ( m 2·℃ ), respectively. Finally, response surface methodology is utilized to analyze the effects of various milling parameters on the maximum milling temperature. The analysis shows that milling depth has the most significant impact on the maximum temperature, followed by irrigation fluid flow rate, feed speed, and milling angle. This temperature model and the accompanying analysis provide valuable references for selecting preoperative milling parameters and help reduce the risk of thermal necrosis during bone milling. Keywords:bone milling temperature; finite element meth

Abstract:Aircraft telemetry data are the only source for ground-based assessment of satellite in-orbit status. Anomaly detection facilitates condition-based dynamic decision-making during aircraft operations and effectively reduces failures. However, existing methods primaril f y ocus on short-term variations, making it difficult to identify collective anomaly patterns effectively. To address this issue, this article proposes a collective anomaly detection method for aircraft telemetry data based on long time-scale characteristic modeling optimization. First, a temporal correlation model is formulated to extract high-dimensional patterns from telemetry data segments and generate prediction results. Then, using the residuals between the prediction results and observed data, a statistical model is developed to extract distribution characteristics and establish anomaly detection criteria. Finally, iterative prediction is employed to automatically adjust model inputs, enhancing the robustness of collective anomaly detection. Validation using actual aircraft attitude angle telemetry data shows that, compared with the VAE-LSTM model, the proposed method improves the detection rate of anomaly segments by 0. 041 and the F1 score by 0. 039. These results show the method′s advantages in improving detection accuracy and reducing missed detections, providing reliable data support for condition-based satellite operations and maintenan

Abstract:Aircraft telemetry data are the only source for ground-based assessment of satellite in-orbit status. Anomaly detection facilitates condition-based dynamic decision-making during aircraft operations and effectively reduces failures. However, existing methods primaril f y ocus on short-term variations, making it difficult to identify collective anomaly patterns effectively. To address this issue, this article proposes a collective anomaly detection method for aircraft telemetry data based on long time-scale characteristic modeling optimization. First, a temporal correlation model is formulated to extract high-dimensional patterns from telemetry data segments and generate prediction results. Then, using the residuals between the prediction results and observed data, a statistical model is developed to extract distribution characteristics and establish anomaly detection criteria. Finally, iterative prediction is employed to automatically adjust model inputs, enhancing the robustness of collective anomaly detection. Validation using actual aircraft attitude angle telemetry data shows that, compared with the VAE-LSTM model, the proposed method improves the detection rate of anomaly segments by 0. 041 and the F1 score by 0. 039. These results show the method′s advantages in improving detection accuracy and reducing missed detections, providing reliable data support for condition-based satellite operations and maintenan

Abstract:Aircraft telemetry data are the only source for ground-based assessment of satellite in-orbit status. Anomaly detection facilitates condition-based dynamic decision-making during aircraft operations and effectively reduces failures. However, existing methods primaril f y ocus on short-term variations, making it difficult to identify collective anomaly patterns effectively. To address this issue, this article proposes a collective anomaly detection method for aircraft telemetry data based on long time-scale characteristic modeling optimization. First, a temporal correlation model is formulated to extract high-dimensional patterns from telemetry data segments and generate prediction results. Then, using the residuals between the prediction results and observed data, a statistical model is developed to extract distribution characteristics and establish anomaly detection criteria. Finally, iterative prediction is employed to automatically adjust model inputs, enhancing the robustness of collective anomaly detection. Validation using actual aircraft attitude angle telemetry data shows that, compared with the VAE-LSTM model, the proposed method improves the detection rate of anomaly segments by 0. 041 and the F1 score by 0. 039. These results show the method′s advantages in improving detection accuracy and reducing missed detections, providing reliable data support for condition-based satellite operations and maintenan

Abstract:Aircraft telemetry data are the only source for ground-based assessment of satellite in-orbit status. Anomaly detection facilitates condition-based dynamic decision-making during aircraft operations and effectively reduces failures. However, existing methods primaril f y ocus on short-term variations, making it difficult to identify collective anomaly patterns effectively. To address this issue, this article proposes a collective anomaly detection method for aircraft telemetry data based on long time-scale characteristic modeling optimization. First, a temporal correlation model is formulated to extract high-dimensional patterns from telemetry data segments and generate prediction results. Then, using the residuals between the prediction results and observed data, a statistical model is developed to extract distribution characteristics and establish anomaly detection criteria. Finally, iterative prediction is employed to automatically adjust model inputs, enhancing the robustness of collective anomaly detection. Validation using actual aircraft attitude angle telemetry data shows that, compared with the VAE-LSTM model, the proposed method improves the detection rate of anomaly segments by 0. 041 and the F1 score by 0. 039. These results show the method′s advantages in improving detection accuracy and reducing missed detections, providing reliable data support for condition-based satellite operations and maintenan

Abstract:Aircraft telemetry data are the only source for ground-based assessment of satellite in-orbit status. Anomaly detection facilitates condition-based dynamic decision-making during aircraft operations and effectively reduces failures. However, existing methods primaril f y ocus on short-term variations, making it difficult to identify collective anomaly patterns effectively. To address this issue, this article proposes a collective anomaly detection method for aircraft telemetry data based on long time-scale characteristic modeling optimization. First, a temporal correlation model is formulated to extract high-dimensional patterns from telemetry data segments and generate prediction results. Then, using the residuals between the prediction results and observed data, a statistical model is developed to extract distribution characteristics and establish anomaly detection criteria. Finally, iterative prediction is employed to automatically adjust model inputs, enhancing the robustness of collective anomaly detection. Validation using actual aircraft attitude angle telemetry data shows that, compared with the VAE-LSTM model, the proposed method improves the detection rate of anomaly segments by 0. 041 and the F1 score by 0. 039. These results show the method′s advantages in improving detection accuracy and reducing missed detections, providing reliable data support for condition-based satellite operations and maintenan

Abstract:Artificial intelligence algorithms are widely used in fault diagnosis of modular multilevel converter ( MMC). However, the existing algorithms require a large number of target domain samples to train the model. To address the problem that it is difficult to diagnose accurately under small samples, a MMC small sample discrete fault diagnosis method based on a multi-source fusion graph and SE-BiGRU-ResNet model is proposed. Firstly, according to the characteristics of an open-circuit fault, the output phase current and bridge arm voltage is selected as the key fault parameters. Secondly, the 1D fault parameters are mapped into the corresponding 2D feature images by using the recurrence plot, Markov transition field, and the Gramian angular field algorithm. To comprehensively strengthen the feature saliency of the image, the multi-source fusion graph is obtained by adding each graph according to the channel dimension. Finally, based on the residual network (ResNet), to improve the ability of the model to capture key spatiotemporal features, the squeeze-excitation (SE) module and the bidirectional gated recurrent unit (BiGRU) module are introduced. The SE-BiGRU-ResNet model is formulated to train and test the multi-source fusion graph. Compared with other methods, the experimental results show that the accuracy of fault diagnosis of IGBT in the fault bridge arm and positioning sub-module reaches 98. 10% and 99. 13% in the case of small samples, and the diagnostic accuracy is high. The test process has a second-level response time. It still has good diagnostic performance and strong generalization ability under extreme conditions. Keywords:modular multilevel converter; fault diagnosis; sm

Abstract:Artificial intelligence algorithms are widely used in fault diagnosis of modular multilevel converter ( MMC). However, the existing algorithms require a large number of target domain samples to train the model. To address the problem that it is difficult to diagnose accurately under small samples, a MMC small sample discrete fault diagnosis method based on a multi-source fusion graph and SE-BiGRU-ResNet model is proposed. Firstly, according to the characteristics of an open-circuit fault, the output phase current and bridge arm voltage is selected as the key fault parameters. Secondly, the 1D fault parameters are mapped into the corresponding 2D feature images by using the recurrence plot, Markov transition field, and the Gramian angular field algorithm. To comprehensively strengthen the feature saliency of the image, the multi-source fusion graph is obtained by adding each graph according to the channel dimension. Finally, based on the residual network (ResNet), to improve the ability of the model to capture key spatiotemporal features, the squeeze-excitation (SE) module and the bidirectional gated recurrent unit (BiGRU) module are introduced. The SE-BiGRU-ResNet model is formulated to train and test the multi-source fusion graph. Compared with other methods, the experimental results show that the accuracy of fault diagnosis of IGBT in the fault bridge arm and positioning sub-module reaches 98. 10% and 99. 13% in the case of small samples, and the diagnostic accuracy is high. The test process has a second-level response time. It still has good diagnostic performance and strong generalization ability under extreme conditions. Keywords:modular multilevel converter; fault diagnosis; sm

Abstract:Artificial intelligence algorithms are widely used in fault diagnosis of modular multilevel converter ( MMC). However, the existing algorithms require a large number of target domain samples to train the model. To address the problem that it is difficult to diagnose accurately under small samples, a MMC small sample discrete fault diagnosis method based on a multi-source fusion graph and SE-BiGRU-ResNet model is proposed. Firstly, according to the characteristics of an open-circuit fault, the output phase current and bridge arm voltage is selected as the key fault parameters. Secondly, the 1D fault parameters are mapped into the corresponding 2D feature images by using the recurrence plot, Markov transition field, and the Gramian angular field algorithm. To comprehensively strengthen the feature saliency of the image, the multi-source fusion graph is obtained by adding each graph according to the channel dimension. Finally, based on the residual network (ResNet), to improve the ability of the model to capture key spatiotemporal features, the squeeze-excitation (SE) module and the bidirectional gated recurrent unit (BiGRU) module are introduced. The SE-BiGRU-ResNet model is formulated to train and test the multi-source fusion graph. Compared with other methods, the experimental results show that the accuracy of fault diagnosis of IGBT in the fault bridge arm and positioning sub-module reaches 98. 10% and 99. 13% in the case of small samples, and the diagnostic accuracy is high. The test process has a second-level response time. It still has good diagnostic performance and strong generalization ability under extreme conditions. Keywords:modular multilevel converter; fault diagnosis; sm

Abstract:Artificial intelligence algorithms are widely used in fault diagnosis of modular multilevel converter ( MMC). However, the existing algorithms require a large number of target domain samples to train the model. To address the problem that it is difficult to diagnose accurately under small samples, a MMC small sample discrete fault diagnosis method based on a multi-source fusion graph and SE-BiGRU-ResNet model is proposed. Firstly, according to the characteristics of an open-circuit fault, the output phase current and bridge arm voltage is selected as the key fault parameters. Secondly, the 1D fault parameters are mapped into the corresponding 2D feature images by using the recurrence plot, Markov transition field, and the Gramian angular field algorithm. To comprehensively strengthen the feature saliency of the image, the multi-source fusion graph is obtained by adding each graph according to the channel dimension. Finally, based on the residual network (ResNet), to improve the ability of the model to capture key spatiotemporal features, the squeeze-excitation (SE) module and the bidirectional gated recurrent unit (BiGRU) module are introduced. The SE-BiGRU-ResNet model is formulated to train and test the multi-source fusion graph. Compared with other methods, the experimental results show that the accuracy of fault diagnosis of IGBT in the fault bridge arm and positioning sub-module reaches 98. 10% and 99. 13% in the case of small samples, and the diagnostic accuracy is high. The test process has a second-level response time. It still has good diagnostic performance and strong generalization ability under extreme conditions. Keywords:modular multilevel converter; fault diagnosis; sm

Abstract:Artificial intelligence algorithms are widely used in fault diagnosis of modular multilevel converter ( MMC). However, the existing algorithms require a large number of target domain samples to train the model. To address the problem that it is difficult to diagnose accurately under small samples, a MMC small sample discrete fault diagnosis method based on a multi-source fusion graph and SE-BiGRU-ResNet model is proposed. Firstly, according to the characteristics of an open-circuit fault, the output phase current and bridge arm voltage is selected as the key fault parameters. Secondly, the 1D fault parameters are mapped into the corresponding 2D feature images by using the recurrence plot, Markov transition field, and the Gramian angular field algorithm. To comprehensively strengthen the feature saliency of the image, the multi-source fusion graph is obtained by adding each graph according to the channel dimension. Finally, based on the residual network (ResNet), to improve the ability of the model to capture key spatiotemporal features, the squeeze-excitation (SE) module and the bidirectional gated recurrent unit (BiGRU) module are introduced. The SE-BiGRU-ResNet model is formulated to train and test the multi-source fusion graph. Compared with other methods, the experimental results show that the accuracy of fault diagnosis of IGBT in the fault bridge arm and positioning sub-module reaches 98. 10% and 99. 13% in the case of small samples, and the diagnostic accuracy is high. The test process has a second-level response time. It still has good diagnostic performance and strong generalization ability under extreme conditions. Keywords:modular multilevel converter; fault diagnosis; sm