Abstract:The inertial reference unit based visual axis stabilization is a main technique used to overcome external disturbance of the moving carrier photoelectric tracking system and realize micro-radian or even sub-micro-radian level tracking. The initial calibration of inertial reference unit requires the construction of a closed-loop control loop with the eddy current displacement sensor as the angular position feedback element. The angular position control accuracy is the main factor that affects the pointing accuracy of the photoelectric tracking system. To solve the problems such as base disturbance and sensor noise faced by inertial reference unit system position control, the varying gain PI control method based on frequency band estimation is proposed. The improved disturbance observer is designed to reduce the influence of middle and low frequency sensing noise on the control accuracy, and the frequency band estimator is designed to realize the subdivision of external disturbance frequency band and dynamically adjust the controller gain accordingly. Simulation and experimental results show that the VGPI-IDOB control algorithm can effectively improve the position control accuracy of the system. Under the condition of no input, the static output root mean square of the system is reduced by 67. 3% compared with that of the PI controller. Under the influence of 15 Hz, 1 mrad sinusoidal input, 20 Hz, 0. 097 mrad disturbance, the position control accuracy of VGPI-IDOB is improved by 72% compared with the PI structure.

Abstract:By analyzing the electro-optical effect of crystal and varied refractive index with electric field, a mathematical model of electrooptical modulation is established and a new half-wave voltage measurement method is proposed. Based on the mathematical model between the amplitude of voltage applied at both ends of the crystal and the output waveform of system, the half-wave voltage of crystal is deduced by determining the output distortion point. The characteristics of output waveform near the half-wave voltage and the relationship among the light source, power supply, time resolution and the system error are analyzed. Accordingly, two optimization schemes based on time resolution optimization and symmetric optimization are proposed, respectively. Specifically, the time resolution optimization amplifies individual waveform details to reduce data dispersion and error. While the symmetric optimization is to use the symmetry characteristics of corresponding waveform extreme points near the half-wave voltage and take the average or median value through multiple samples to reduce the error. The results show that the critical value method is a simple and effective method for measuring half-wave voltage, whose optimization scheme is efficient in reducing data fluctuation and error.

Abstract:To meet some special requirements of 2D displacement measurement in engineering or improve the efficiency of displacement measurement, a simple and compact dual-channel displacement measurement system utilizing laser self-mixing interference is established. Firstly, mathematical equations of the self-mixing interference system are given, which are based on the three-mirror Fabry-Perot cavity model. Then, under the condition of weak feedback, linear current modulation is applied. According to the linear relationship between the frequency of the self-mixing signal and the distance between the external object, when the distance between the two objects and the laser is different, the frequency domain will present two independent spectral peaks, and the phase is solved respectively. Therefore, the self-mixing dual-channel displacement measurement is realized. Then, the dual-channel self-mixing signal is generated by numerical simulation. The phase of two spectral peaks of the self-mixing signal is extracted according to the all-phase spectrum analysis technique, and the displacement curves of the two objects are reconstructed accordingly. Finally, an experiment system is established, the dual-channel displacement measurement experiments are carried out, and the measurement results are exhibited subsequently. The experimental results show that the system can completely distinguish two moving objects, and the relative error of displacement measurement is better than 8. 42% . The linear frequency modulated laser self-mixing interferometry can realize the dual-channel displacement measurement of arbitrary motion, and the number of measuring channels can be further increased by more beam splitting.

Abstract:Particle image velocimetry (PIV) technology is widely used in fluid flow measurement. The refractive index difference of the medium causes the light to deflect on the wall of the tube, which may result in image distortion and directly affect the accuracy of velocity measurement. In this article, the physical model of optical refraction is established, and the functional relationship between the object point and the image point in the circular pipe is achieved. Then, the pixel coordinates of the corrected image are obtained. The bilinear interpolation algorithm is used to obtain the pixel gray value to reconstruct the corrected particle image. Finally, the velocity field in the tube is calculated according to the multi-grid iterative algorithm. The static fluid grid image deviation and laminar flow velocity field measurement experiments are carried out respectively. The errors of optical correction box method, linear correction method and distortion correction method based on optical model were compared. The results show that the accuracy of the proposed image correction method based on geometric optics is better than those of the optical correction box method and the linear correction method. The accuracy and effectiveness of the established geometric optics model are fully evaluated by static and flow experiments.

Abstract:Worm wheel grinding is the finishing process of face gear. The dressing accuracy of grinding wheel directly affects the grinding accuracy of face gear. In this article, the influence of dressing process error on grinding tooth surface error is analyzed. The modeling and compensation method for forming dressing errors of the worm wheel for face gear grinding is proposed. Firstly, the mathematical model of the worm grinding wheel is formulated, the forming and dressing principle of face gear worm grinding wheel is analyzed, and a multi-axis coupling linkage method for dressing the worm grinding wheel of the general cylindrical grinding machine is proposed. Then, the dressing process error is decomposed into axial position error and radial position error. The influence of axial position error and radial position error on grinding tooth surface error are analyzed, and the error compensation method of forming dressing process is proposed. Finally, the grinding wheel compensation dressing, face gear grinding and measurement experiments are carried out. The experiments show that the tooth profile error of the left tooth surface is decreased from 51. 9 μm before compensation to 7. 9 μm after compensation. The tooth profile error of the right tooth surface is decreased from 35. 3 μm before compensation to 17. 6 μm after compensation. The effectiveness of the error compensation method is evaluated.

Abstract:To address challenges faced by most volume measurement methods, such as difficulties in transferring, poor real-time performance, and sparse, noisy, and perturbed Lidar point cloud data, this article presents a vision-based correction method for Lidar volume measurement for small ‘L’-shaped objects. The proposed method first aligns camera and Lidar data through joint calibration and time-stamped nearest-neighbor matching. Subsequently, it leverages target detection algorithms to extract information from images while simultaneously performing ground segmentation on point cloud data to distinguish ground and non-ground points. By employing visual projection and point cloud clustering, the method segments target point clouds and utilizes KDtree to identify ground points in proximity to the target point cloud. Finally, a 3D box fitting algorithm is proposed to provide initial rough estimation of the point cloud target′s 3D box. A visual correction model is established to refine the target′s 3D box, and enable accurate volume calculation. Experimental results show that for ‘L’-shaped objects like weapon crates, medical boxes, and barrels, the proposed algorithm achieves promising results within a certain range. The average relative error in volume measurement is less than 4. 44% , with a maximum error below 6. 12% and a maximum repeatability error of 5. 61% . In addition, the integration of the visual correction model significantly enhances the algorithm′ s accuracy and stability. The processing of frame on an embedded platform takes 55 ms, demonstrating the capability to achieve realtime, high-precision volume measurement. This method holds great promise for practical engineering applications.

Abstract:The cell culture apparatus can be used for the research on cell biology, space biotechnology and space medicine. In this article, a flow-loading automatic cell culture device is designed and manufactured, which can be used for automatic vitro cell culture in underground and space microgravity environment. It can provide periodic and quantitative input of different reagents to the cell culture plate regularly, and meet the requirements of miniaturization, convenient assembly and high reliability required by space cell culture. The whole device adopts the modular design method, and uses Solid Works software to establish 3D models of basic components, reagent supply components and culture components, respectively. The Flow Simulation module in Solid Works software is used to conduct finite element analysis, fluid simulation, and structural optimization of the pin plate, the key component connected between the cell culture device and the cell culture plate. The velocity field and pressure field of the channel inside the pin plate are observed, and the channel is optimized by replacing right angles with rounded corners. The results show that the device is feasible and can meet the requirements of cell culture device. The fluid stability in the optimized pin-plate channel is stronger, and the flow rate and pressure distribution are more uniform. In the ground environment, the optimized needle plate has increased the speed uniformity of fluid in the transverse channel by 95% , 96% , and 94% , respectively under low, medium, and high flow rates. Similarly, in the longitudinal channel, the speed uniformity has been increased by 68% , 70% , and 72% , respectively under low, medium, and high flow rates. The experimental results show that the optimized flow at the outlet of the pin-plate channel is closer to the set output flow at the inlet pump. The designed device can be used in space biotechnology research.

Abstract:To address the problems of cumbersome operation and low efficiency of manual annotation in current traditional sitting posture annotation methods, this article proposes an automatic annotation method for pressure data based on three-dimensional sitting posture. Real time synchronous collection of binocular visual data and pressure data based on dual pointer timestamp matching. The normalization and adaptive median filtering are utilized to process pressure data, and remove the dimensional influence and peak noise of pressure data. By using 2D pose estimation and matching optimization, coordinate transformation, and multi-point triangulation to process visual data, 6 key point information of the 3D human body are extracted. A skeleton image based on sitting posture features and a 3D projection angle feature between adjacent nodes are constructed, and a 3D sitting posture based annotation information generation model is formulated. The annotation information is utilized to label the pressure data and create an annotated sitting pressure dataset. In practical applications, the average time difference between real-time synchronous collection of single sample data is only 21 ms, and the accuracy of label generation is 98. 98% . The average time for automatic labeling is 0. 199 s, and the labeling speed is 13. 3 times faster than manual labeling.

Abstract:Considering the redundancy of localization matching between reference point and access point due to the spatial structure obstruction and weak signal penetration, a spatial dual-refinement localization algorithm named “horizontal refinement of reference points and vertical refinement of access points” is proposed. Firstly, the traditional mean value is replaced with the high-order statistical information of the strongest received signal to characterize each reference point, and the processing methods of small-area fusion and boundary reference point sharing are combined to achieve the fuzzy clustering in the target space, so as to weaken the adverse effect of absolute discrimination at the edges; secondly, based on the dimensionality-reduced subspace, the spatial differentiation and coverage reliability of each access point are measured comprehensively to screen out the subspace with high recognition value and high stability. Finally, the first-level area discrimination is performed by judging the strongest received signal source, and the second-level location estimation is achieved with the WKNN algorithm. The proposed horizontal streamlining strategy possesses the regular clustering and complies with the structural constraints of the scene more precisely according to the realistic roadshow test. It′s found that the vertical streamlining strategy improves the average positioning accuracy by at least 17% compared to the traditional access point selection algorithm, and filters out the large positioning error of more than 4. 5 m under the condition of distribution density of 1 m×1 m reference points.

Abstract:Controlling a brushless direct current motor requires the position information of rotor, and deviation in position recognition can reduce motor efficiency. This study focuses on brushless direct current motor with switch-type Hall sensors and, and estimates the relative installation deviation of Hall sensors based on the minimum deviation principle. It clarifies the ideal commutation position recognition method based on LBEMF to eliminate the installation deviation between Hall sensors, and a pre-calibration method is used to calculate the delay introduced by the signal conditioning circuit. Additionally, a rotor position information correction strategy is proposed. An adaptive position information prediction algorithm based on the Sage-Husa method is designed to filter out the deviation of position information that pre-calibration cannot eliminate. The results of hydrogen circulation pump experimental platform show that, the MAPE of speed fluctuation and the average phase current are decreased by 72. 4% and 62. 8% with the pre-calibration method respectively, and it provides a significant improvement in system efficiency. Compared with the traditional KF prediction algorithm, the proposed algorithm reduces speed fluctuation, speed curve overshoot and commutation time fluctuation by 16. 0% , 19. 4% , 42. 1% , and 35. 0% , respectively, which demonstrates the higher disturbance resistance and more accurate and stable prediction of commutation timing.

Abstract:High-temperature vibration monitoring is essential for failure diagnosis and equipment maintenance. This paper presents a La3Ga5 SiO14(LGS) surface acoustic wave (SAW) high-temperature vibration sensor based on the symmetrical structure of four-end fixed beam, capable of withstanding temperatures up to 800℃ and providing high sensitivity and wide band by using an optimized algorithm. Then the mathematical model of SAW vibration sensor with a four-end fixed beam was constructed, and the sensitivity and natural frequency of the sensor were analyzed. Subsequently, the structural model of the sensor was established, and the mechanical and electrical simulations were carried out in the range of 20℃ ~800℃, which reveals the mechanical and electrical properties of the sensor at high temperature. The results show that the sensitivity of the proposed sensor in this study is about 7. 047 times higher than that of the sensor with single SAW resonator, and the sensitivity increases progressively with temperature at 20℃ ~800℃, reaching 9. 879 1 ×10 -6 / g at 800℃. The natural frequency decreases with temperature, showing 3 018. 4 Hz at 800℃ . Finally, the feasibility of the design was preliminarily verified by experiments, which provides a new idea for the optimization design and high-temperature application of SAW vibration sensor.

Abstract:The DAMAS algorithm has high spatial resolution. But, the unsatisfactory dynamic range leads to false acoustic sources in the imaging results. The F-DAMAS algorithm is proposed to solve this problem, which exploits the functional beamforming ( FB) algorithm to improve the imaging performance. By powering the PSF in the DAMAS algorithm, the linear relationship among the output power of the FB algorithm, the powered PSF and the acoustic source distribution is established. And the equation group of the F-DAMAS is formed. Furthermore, the Gauss-Seidel iterative method with positive constraints is utilized to derive the acoustic source distribution. Compared with FB and DAMAS algorithms, simulations and experiments on mono and incoherent acoustic sources show that the proposed algorithm can effectively improve resolution and dynamic performance of the image. According to the relationship between the coverage area of the acoustic center and the change of the power index, it recommends that the value range of the power index in the algorithm is 6 ~ 14.

Abstract:The accurate positioning of mobile agricultural equipment in the facilities is the key to the development of intelligent and unmanned agriculture. In view of the positioning problem of agricultural machinery in the facility in a GPS-denied environment, the ultra-wide band (UWB) system is proposed to establish the navigation and positioning system of agricultural machinery in the facility. To improve the positioning accuracy of agricultural machinery equipment, the variational Bayesian Kalman filtering (VBKF) is utilized to smooth the four measured distances of the UWB system to enhance the estimation accuracy of the measured distance, and the centroid positioning algorithm (CPA) is used to calculate the position coordinates of the target node ( TN) . To further improve the positioning accuracy, an improved Taylor series algorithm (TSA) is implemented to optimize the localization results of the VBKF-CPA method. With the mobile robot as the experimental platform, the dynamic and static simulation experiments are conducted indoors by using the UWB positioning system to evaluate the effectiveness of the proposed method. The experimental results show that the VBKFCPA-TSA algorithm can improve the positioning accuracy of the TN and obtain more stable localization results. The mean error on the x-, y-, and z - axis are reduced from 0. 085, 0. 071, and 0. 064 m to 0. 034, 0. 032, and 0. 028 m, and the average estimation accuracies are increased by 60% , 54. 9% , and 56. 3% , respectively. The dynamic localization trajectory of the VBKF-CPA-TSA algorithm is closer to the real movement trajectory, which verifies that the proposed positioning algorithm is able to ameliorate the positioning accuracy of the UWB system in the agricultural facilities and provide a novel method for agricultural mechanical positioning in GPS-denied environment

Abstract:The pseudo capacitive flexible pressure sensor based on reversible redox reaction has high sensitivity and can be used for weak pressure detection. However, currently, the linearity of the pseudo capacitive flexible pressure sensor is poor and can only maintain high sensitivity within a limited pressure range. For this reason, this paper uses MXene material as the electrode to design a double scale random microstructure ionic gel film with pores inside and rough surface, which increases the buffer space during the compression process, makes the stress deformation of the gel film more uniform, and ensures that the sensitivity remains stable during the compression process. Experimental data shows that the sensor has ultra-high linearity ( correlation coefficient ~ 0. 994) in the range of 0 ~ 1 MPa, excellent sensitivity ( ~ 2 133. 7 kPa -1 ), fast response and recovery time ( ~ 15 and~ 23 ms respectively), low detection limit ( ~ 2. 5 Pa), and excellent mechanical stability. The sensor can be used underwater to detect water depth with high linearity, and at the same time, the sensor can detect weak water flow changes caused by propeller disturbances at different water depths with high sensitivity.

Abstract:The dual-redundant acceleration sensor of the maglev train is externally mounted on the electromagnet. It is affected by vibration, electromagnetic interference, temperature and humidity change during operation. Its measurement characteristics are unstable, which are reflected by dynamic measurement noise. At present, the comparison test method is often used to detect redundant sensors in engineering practice. The problem of false alarm and miss alarm is easy to occur under dynamic noise, and the detection accuracy is low. Thus, in this article, an adaptive multi point generalized likelihood ratio test algorithm is proposed for sensor fault detection. The multi point decision form is used to enhance the robustness to outliers. In addition, the sliding window variance estimation is introduced to recursively estimate the variance of parity vectors, and the decision function is adjusted to realize the adaptiveness to dynamic noise. The effectiveness of the proposed algorithm is evaluated by experiments on a small-scale suspension test-rig. Compared with similar traditional algorithms, the detection accuracy of the proposed algorithm increased by 15% in static noise experiments, and increased by 13% in dynamic noise experiments. The false alarm rate and missing alarm rate are significantly reduced with good robustness to dynamic noise.

Abstract:Faced with the limited labeled sample problem in practical engineering, particularly in extreme labeling scenarios where only one labeled sample is available for each fault type, the existing semi-supervised diagnosis methods suffer from a significant deficiency in the fault identification ability. To address this issue, a novel semi-supervised fault diagnosis method based on the decoupled feature pseudo-label propagation (DFPP) algorithm is proposed. Firstly, the locally selective combination in the parallel outlier ensembles (LSCP) method is introduced to separate fault samples. Subsequently, the DFPP method is proposed. In DFPP, the adversarial decoupled auto-encoder (ADAE) is applied to extract the enhanced fault features, and the incorporation of fault feature dimension reduction, pseudo-centroid calibration of feature distribution, and distance measurement are adopted to efficiently achieve pseudo-label propagation in situations. Finally, a fault classifier is trained by using pseudo-labeled fault samples, and the combination of anomaly detection ensures accurate fault diagnosis with high precision. Experimental results conducted on two datasets of rotating components demonstrate that the proposed method can achieve average diagnostic accuracies exceeding 97% and 90% in the same working condition and cross working condition with extremely limited labeled samples, respectively, which is significantly superior to the comparison methods.

Abstract:The clinically used middle-ear implant for round-window stimulation is not designed specifically for this excitation method, and there are problems such as large individual differences in postoperative performance, unstable performance, and gain value that cannot reach the theoretical value. To address these issues, a round-window stimulating type piezoelectric actuator that matches the size of the round-window membrane and can monitor the preload acting on the round-window membrane is designed. Firstly, combined with the anatomical structure of the human ear, the mechanical structure and macroscopic dimensions of the actuator are designed. Then, the force analysis of the flextensional amplifier is carried out, and the particle swarm algorithm is used to optimize the design of the flextensional amplifier to obtain the optimal output displacement magnification. After that, a piezoelectric actuator-human ear coupling mechanics model is formulated, the influence of the cross-sectional area of the support spring and the number of piezoelectric stack layers on the performance of the actuator′s hearing loss compensation is analyzed, and the cross-sectional area of the support spring and the number of piezoelectric stack layers are ascertained. Finally, the prototype of the piezoelectric actuator is manufactured according to the determined design size. A laser vibration test bench is established to test the dynamic characteristics of the piezoelectric actuator. The experimental results show that the designed round-window stimulating type piezoelectric actuator can monitor the preload acting on the round-window membrane. It has the advantages of wide operating frequency band (100~ 10 000 Hz) and low harmonic distortion (THD ≤0. 29% ). The designed piezoelectric actuator is suitable for the hearing loss compensation of the round-window stimulation, which provides a new solution for the clinical treatment of mixed hearing loss.

Abstract:Pneumatic control valves act as typical terminal actuators in the process industry, which suffer from difficulties in fault identification and severe consequences of faults due to the high incidence of faults and diverse fault types. Therefore, intelligent fault detection and diagnosis of pneumatic control valves have crucial practical significance. In this paper, an adaptive multi-scale features fusion network is proposed for the pneumatic control valve fault diagnosis. Firstly, a multi-scale feature extraction network with fusion self-attention mechanism is constructed to automatically extract spatial and detail features of signals. Then a weighted adaptive feature fusion network is designed to perform the weighted fusion of multi-scale features to improve the fault feature characterization capability of model. Finally, the feature identification and fault classification are performed by the Long short-term memory neural network with SoftMax function. The experimental results show that the model achieves an average accuracy of 96. 82% on the DAMADICS valve benchmark experimental platform, which are higher than other comparative models. Comparison with the detection results in the latest literature reveals that the model developed in this paper also has certain advantages such as the number of detectable faults and detection accuracy, and the detection performance of model is experimentally verified.

Abstract:The problems of strong linear indistinguishability of different fault features and insufficient labeling of fault data exist in bearing compound fault classification, which seriously affect the classification accuracy. This paper proposes a semi-supervised compound fault classification method based on the two-stage SVM with wavelet kernel diffusion. To address the strong linear inseparability of fault features, the wavelet kernel function is used to transform them in high-dimensional space, and the maximal overlap discrete wavelet packet transform is applied to obtain the energy distribution of the signal in different frequency bands as the fault features. Aiming at the insufficiency of the fault data labeling, a two-stage SVM classification model with incremental kernel space label diffusion is proposed. Based on the kernel difference distance in the wavelet kernel space, we expand the neighboring samples and boundary samples in the coarse partition stage using incremental kernel space label diffusion, and the training of models is completed based on expanded samples at the segmentation phase. Three sets of bearing composite fault data validate the effectiveness of the proposed method, and the experimental study shows that under the condition of a single class of training samples of 5, the proposed method improves the classification accuracy by 7. 5% on average than SVM, and outperforms other popular algorithms.

Abstract:Power towers in unmanned aerial vehicle (UAV) inspection images have the characteristics of multi-attitude and large-aspect ratios, which are difficult to accurately locate and distinguish towers with different degrees of inclination according to unique prior knowledge. To improve the incline detection accuracy, this article proposes a method for processing UAV images based on the improved refined rotation RetinaNet (Multi-Head-KF-R 3 det). Firstly, the incline detection head is added to the original R 3 det model to achieve the classification of power towers and their inclination degrees, as well as the accurate location. Then, the Kalman-filter intersection over union loss is introduced into the regression loss function to effectively improve the overall detection accuracy and recall rates of incline detection without additional hyperparameters. Finally, the improved model is reasonably compressed based on the design principles of ghost lightweight network, which lays a foundation for the deployment of the model in embedded devices. The experimental results show that the mAP and recall rates of Multi-HeadKF-R 3 det on multi-scale and multi-attitude power tower datasets can reach 94. 5% and 94. 9%, respectively.

Abstract:Connectors are essential components of electronic devices, and the cleanliness of their working contacts is a necessary condition for the normal operation of electronic equipment. To address the frequent issues of false positives and false negatives caused by the diverse types and styles of connectors, as well as the limited and variable foreign object samples, this article proposes a novel zeroshot anomaly detection method. By synthesizing random anomalies on unrelated background images, it constructs pairs of normal-anomaly sample images. Through network prediction, a discrepancy score map representing the pixel-level similarity between sample pairs is obtained, enabling anomaly detection and localization. By employing anomaly region mask supervision, the network focuses on the pixel differences between normal and anomaly samples, reducing the network′s attention to the semantic information of the images themselves and minimizing the need for real samples. Thus, the generalization ability of the detector is enhanced. To evaluate the effectiveness of the algorithm, the network is trained solely on synthesized data and evaluated on the DeepPCB dataset, achieving a mAP (mean average precision) of 88. 2% . After transfer learning, the mAP increases to 99. 1% , which is the best performance on this dataset to date. Experimental results demonstrate the strong generalization ability of the proposed zero-shot anomaly detection method.

Abstract:Scoliosis is a spinal disorder that is highly prevalent in the adolescent population. To address the radiation hazards associated with X-ray assessment of scoliosis, this article designs and implements a personalized modeling and scoliosis assessment system for the spine based on morphological recognition. First, the feature points related to the spine are extracted, and the additional feature points are generated according to the relative positional relationship between the feature points. Secondly, the feature point correction algorithm and the filtering algorithm are designed and applied to improve the positional accuracy of the feature points. Finally, the vertebral bone model is aligned to the spine line fitted by the feature points in Unity to obtain a personalized 3D spine model, and the Cobb angle, the thoracic kyphosis angle, and the lumbar lordosis to assess scoliosis. An experiment is implemented on 28 subjects to compare and analyze the results of the systematic assessment with those of the X-ray assessment. The Pearson correlation coefficient between the Cobb angle and the actual Cobb angle is 0. 82, with a mean absolute error of 3. 4° and a root-mean-square error of 4. 2°. The Pearson correlation coefficient between the thoracic kyphosis angle and the actual thoracic kyphosis angle is 0. 80, with a mean absolute error of 3. 4° and a root-mean-square error of 3. 8°. The Pearson correlation coefficient between the lumbar lordosis and the actual lumbar lordosis is 0. 78, the mean absolute error is 3. 2°, and the root mean square error is 3. 7°. The experimental results show that the scoliosis assessment system is highly accurate, which is easy to use. It can be applied to scoliosis screening in a large adolescent population.

Abstract:The error caused by thermal deformation is one of the main factors affecting the accuracy of CNC machine tools. Correspondingly a thermal error prediction method based on CNN-GRU combined neural network is proposed to reduce the impact of thermal error on the accuracy of CNC machine tools. By conducting thermal error experiments, the temperature rise data and thermal error data of the linear feed system of a specialized CNC machine tool are collected for spiral surfaces. Then the fuzzy c-means clustering and grey relation analysis are carried out to screen temperature sensitive points in the feed system, and a CNN-GRU thermal error prediction model is established using temperature rise data of temperature sensitive points and thermal error of feed system as data samples. To verify the accuracy and practicality of model, a comparative analysis is conducted with traditional thermal error prediction models based on CNN-LSTM and LSTM. The results showed that the CNN-GRU model possesses the high prediction accuracy and robustness, whose average absolute error, root mean square error, and determination coefficient of the prediction results are better than those of the CNN-LSTM model and LSTM model. The proposed thermal error model can lay the foundation for subsequent error compensation and provide ideas for predicting thermal errors in CNC machine tools.

Abstract:Proposes a method to improve the sensitivity and consistency of stress measurement for steel plates based on magnetoelastic effects. From the framework of the theoretical model of magnetoelastic, it analyzes the roles of strong magnetization and weak AC magnetization:Strong magnetization can improve the sensitivity and consistency of stress measurement by reducing the initial magnetic permeability and unifying the parameters of the magnetoelastic model. Weak AC magnetization provides a minor oscillating magnetic field, which can effectively eliminate the influence of interfering magnetic fields. A finite element simulation study on stress measurement of steel plates based on the magnetoelastic method is conducted, and the results demonstrate that when the initial relative permeability is low, the change rate of normal magnetic induction intensity of the steel plate under stress is relatively high, which indicates a high sensitivity of stress measurement. Tensile experiments are conducted on steel plates with different thicknesses under different strong magnetization conditions to measure the changes in normal magnetic induction intensity. The experimental results show that compared with unmagnetized case, the sensitivity of stress magnetic measurement has been improved by strong magnetization by tens of times, and the consistency has been improved by several to hundreds of times. The stress measurement sensitivity and consistency parameter of 3mm thick steel plates using single magnet moving magnetization method have been improved to 1. 665 mGs/ MPa and 2% . The sensitivity of 5 mm thick steel plates using dual magnets moving magnetization method has been enhanced to 1. 41 mGs/ MPa, and the consistency parameter of 5 mm thick steel plates using single magnetic moving magnetization method has been increased to 0. 2%. The sensitivity and consistency parameter of 7 mm thick steel plates using dual magnets moving magnetization method have been improved to 1. 2 mG/ s/ MPa and 1. 8%.

Abstract:This article addresses the issues of low accuracy and time-consuming nature associated with traditional density detection methods in ore grinding classification by proposing an intelligent pulp density detection method. Through mechanistic analysis of the pulp fluid, linear known terms and nonlinear unknown terms are identified. A holistic recognition of pulp density is performed by combining Gaussian process regression with a regularized stochastic configuration algorithm. Additionally, the variance estimated by the mechanistic model is set as the training objective for the data-driven model, enhancing the model's capacity to acquire data information. Meanwhile, a collaborative computing method is employed to apply the adaptive intelligent detection method in the industrial domain, ensuring realtime detection and adaptability of the pulp density detection model. Based on industrial data experimental analysis, the proposed method shows an average absolute error of 7. 13, a root mean square error of 9. 31, a determination coefficient of 99. 51% , and a sample quantity proportion of relative error δ < 1. 0% at 83. 58%. These results are better than those of other comparative algorithms. The effectiveness of the pulp density detection model is enhanced.

Abstract:Aiming at the problems of excessive residual current in the station, low operation rate of residual current device (RCD) and frequent leakage accidents caused by erroneous wiring faults of user neutral and ground wires, an erroneous wiring leakage user location method based on timing correlation characteristics is proposed. There is a causal relationship between the load current of connected users and the residual current of station. It’ s noted that the influence of normal users is limited, while the load current of abnormal users dominates the change of residual current in the station. First, the Apriori algorithm is used to infer the strong correlation between the residual current in the station area and the load current of faulty when there are erroneous connections; then, the adaptive Lasso regression model is constructed to characterize the relation between residual current of the station area and the load current of each user, correspondingly the suspicious user variables in different fault scenarios can be screened out; By utilizing the additional absolute value of standardized regression coefficient of suspicious users, the erroneous wiring leakage users with strong correlation characteristics with obviously abnormal changes in the residual current in the station can be quickly identified. Finally, the effectiveness of proposed method is verified with the laboratory data of realistic distribution network.

Abstract:The high-precision positioning in the whole workspace domain is the key to realizing the full-range precision operation of large workspace robots. To improve the reliability and spatial adaptability of parameter calibration in the full workspace domain of the robot, this article studies the non-probabilistic quantification methods of many uncertainties in the robot itself, analyzes the difference of the influence of uncertain parameters on the end positioning accuracy of the robot in different workspace domains, and partitions the whole workspace domain according to the non-probabilistic reliability index of positioning accuracy. A non-probabilistic reliability calibration method for robots under the framework of workspace partition is proposed. The example shows that, after partition calibration compensation, the average lower and upper bounds of error intervals in x, y, and z directions decrease by 40. 16% , 59. 36% , and 59. 08% , 40. 87% , as well as 54. 24% , 33. 98% , respectively. Moreover, the compensated robot has a fast response speed and small fluctuation during movement. It is proved that the proposed method is effective in reducing the end error range in the whole working domain, improving the absolute positioning accuracy of the robot and the spatial adaptability of the calibration.

Abstract:To solve the problem of delay interference in the execution of a multi-robot path planning system, a k-robust safe interval path planning algorithm based on safety interval is proposed in this article. Firstly, the global time interval is introduced, and the time occupied by robots and the k-robustness factor are added together as the obstacle time interval of nodes. Then, the global time interval is updated, and the safe interval constraint of the global time interval is used to avoid conflicts between robots. Secondly, the A ∗ method with k time expansion is proposed as the core algorithm of multi-robot low-level path planning, where k is the set robust factor. This method can deal with the robust planning problem in the spatio-temporal relationship of multi-robot. Finally, all robot path planning is completed under the constraint of a global time interval. Simulation results show that the solution success rate of the proposed algorithm is 37% higher than that of the existing improved k-robust conflict-based search algorithm, and the solution time is reduced by two orders of magnitude. It provides a more efficient and robust scheme with a delay margin for multi-robot path planning, which provides a more efficient and robust scheme with delay margin for multi-robot path planning, and further verifies the effectiveness and feasibility of the proposed algorithm in real robots.

Abstract:A visual detection method of relative liquid level without camera calibration and liquid level gauge and closed-loop control system are proposed for the robotic pouring task related to multiple types of transparent containers and liquids. Firstly, the characteristics of the service robots′ liquid pouring tasks are analyzed and the target detection method in computer vision field is deployed to detect liquids and containers simultaneously. The relative liquid level height is obtained by calculating the height-proportion between the detected container and detected liquid, which avoids the complicated hand-eye calibration processes during the measurement of liquid level′s absolute height. Secondly, the proposed liquid level detection method is geometrically modeled and analyzed by applying the pinhole imaging model, and the laws on measurement errors of relative liquid level height are deduced in typical cases. Thirdly, the images of various liquids and containers are collected to train YOLOv5s, which is deployed to detect target object and obtain the relative height of liquid level. The test results verify the effectiveness of proposed method. The paired average precision of new method is 86. 7% for the new types of liquids in new shaped containers that do not appear in the training set. Finally, to avoid solving or estimating corresponding Jacobian matrices in visual servo theory, the detected liquid level height is combined with PD control to get a closed-loop control system. Several kinds of liquid pouring tasks on both manipulator platforms are successful with the same parameters in PD controller, which prove the effectiveness and robustness of proposed closed-loop control method.

Abstract:For large area data collection and environmental monitoring in remote areas with no mobile network coverage, this article first designs a LoRa communication protocol between the UAV mobile gateway and the ground nodes. Based on this, a spreading factor prediction model based on the improved extreme learning machine ( PG-ELM) is proposed to achieve dynamic optimization and adjustment of the spreading factor. To improve the prediction accuracy and efficiency, the model uses signal strength, signal-to-noise ratio, distance, packet loss rate, temperature and relative humidity as inputs. The particle swarm optimization algorithm and the grey wolf optimization algorithm are fused to optimize the ELM model. The LoRa communication data sample sets are obtained through the UAV mobile communication experiment, which are then used to train and optimize the PG-ELM model. The results show that, with a data size of 20 kB, the proposed scheme reduces the data collection time by about 78% and 26% compared with single SF12 and SF7. It also lowers the average communication energy consumption by more than 70% compared with single SF12, achieves a packet delivery rate of 98% , and has significant advantages in energy efficiency and prediction real-time performance.

Abstract:In the real environment, most intelligent wheelchairs work in complex scenarios, and their autonomous navigation requires high requirements for path safety. The asymptotic optimal random search tree RRT star algorithm basically meets the optimal path planning of mobile robots. However, due to the large size of the intelligent wheelchair itself, it is easy to come into close contact with the environment. Therefore, the environmental model can be expanded and different search steps can be defined to keep the planned path away from obstacles. Secondly, to ensure that users can achieve higher comfort and more efficient destination when using intelligent wheelchair navigation, the heuristic constraint sampling idea and the gravitational field idea in the artificial potential field are used to prune the redundant nodes in the planning process of this algorithm. Therefore, the operating memory of the system is reduced. Subsequently, combined with the minimum turning radius of the wheelchair, a minimum path curvature constraint strategy and a cubic Bspline curve algorithm are proposed to smooth the path, which make it more suitable for wheelchair driving. Finally, a comparative experiment is conducted on the improved algorithms on MATLAB and Gazebo simulation platforms, and the proposed algorithm is applied to intelligent wheelchair entities. The experimental results show that the algorithm can effectively solve the global path planning problem of intelligent wheelchairs, significantly improve the efficiency of global path planning, and have a certain degree of security. It provides an effective reference for the mobile robot field.

Abstract:To address the problem of cooperative perception for multiple intelligent and connected vehicles (ICVs) in road scenarios, this article proposes a cooperative perception registration algorithm for ICVs, which is based on sparse semantic features. The proposed algorithm aims to extend the perception range of ICVs by point cloud ensemble registration. Therefore, the cooperative perception for ICVs is achieved. Firstly, the sparse semantic features are obtained by geometric feature extraction based on road semantic features. Secondly, the angle deviation among the road semantic features is calculated to provide the initial registration value. The point cloud semantic information is used as the registration constraint condition to realize the global semantic ensemble registration. The experiments show that the proposed algorithm effectively expands the cooperative sensing range of multi-ICVs. The accuracy and robustness of multipoint cloud ensemble registration are enhanced. Compared with the current mainstream algorithm JRMPC, the registration accuracy of the proposed algorithm is improved by 2. 45% .