This PDF file contains the front matter associated with SPIE Proceedings Volume 12059, including the Title Page, Copyright information and the Table of Contents

Based on the Villari effect of giant magnetostrictive material, a new giant magnetostrictive force sensor is proposed. The magnetic flux is measured by the Hall element in the structure, to realize the measurement of external load. Firstly, according to the structure and working principle of the giant magnetostrictive force sensor, its closed magnetic circuit is analyzed. Secondly, based on the basic theory of electromagnetic field, the magnetic circuit of the giant magnetostrictive force sensor is calculated by the finite element method. Finally, the specific factors affecting the output characteristics of the sensor are deeply discussed. The results show that the closed magnetic circuit can effectively avoid magnetic flux leakage, and it is reasonable to use the hall principle and Villari effect to measure the external load, which provides a technical way for the further research of giant magnetostrictive force sensors.

Electromechanical braking technology is an effective way to improve the braking response of mine hoist. Based on the analysis of mechanical disc brake of mine hoist, the control process is analyzed in detail, and then the control strategy is obtained. An automatic control system based on linear quadratic regulator and PI controller is proposed. The braking goal of self-adaptive adjustment of braking clearance and on-line adjustment of braking force is realized. The results of the experiment indicate that the brake gap adjustment time is less than 10s, stable within 1±0.2mm, and the steady-state error is less than 2%. The positive pressure of the brake has a linear relationship with the motor voltage, with a slope of 920.4 and an intercept of -1298.88. The problems of low automation and high pollution of the existing hydraulic brake have been solved, which provides a new way for the control system of electromechanical brake.

The vibrating tube densimeter measures the liquid density based on the relationship between the natural frequency of the resonator and the liquid density in the tube. However, the problem of constructing a complete mathematical model and optimizing the structure size and working mode of the resonator remains unsolved. The paper presents a complete set of working equations of the resonator constructed by Euler Bernoulli beam theory. In addition, the analytical formula of the resonator sensitivity and the Q-factor was deduced. According to the analytical formula, the variation rule of the resonator sensitivity and the Q-factor under different structure sizes and working modes is analyzed, which provides a theoretical basis for optimizing the structure size and working mode of the vibrating tube resonator.

Aiming at the problems of unsmooth and low efficiency in path planning of the indoor mobile robot, a path planning obstacle avoidance system is proposed. First, the global path is generated in the environment grid map by A* algorithm. Second, as the pose constraint, the Reeds-Shepp path set can generate the executable control instructions by extracting the key points of the path. Third, the static/dynamic obstacle is recognized by a binocular camera and avoided by a decision-making method. The experimental result shows that the generated path can meet the actual motion constraints of the mobile robot, the relative error of the obstacle depth distance is 0.03%, which meets the requirements of the robot path planning.

Non-orthogonal shaft laser theodolite(N-theodolite) measurement system is a new kind of measurement instrument utilized in large-scale metrology, which is made up of two same N-theodolites. To achieve the transformation of N-theodolites’ coordinate systems, a calibration method of extrinsic parameters is proposed. With the help of a scale-bar of a certain length, the calibration can be performed conveniently. The correctness of the calibration method is verified by the measurement experiments of point coordinates and distances. The experimental results show that the RMSE of spatial points and the distances of measured points are 0.319mm and 0.150mm respectively. We demonstrate that the proposed calibration method of extrinsic parameters is feasible for the N-theodolite measurement system.

Worm gears are used frequently in some heavy-duty, high-precision applications, and high-precision worm gear machining relies on high-precision gear hobbing machines. The imported high-precision hobbing machine is expensive, and the research group proposed to use the TE data analysis of the transmission chain to guide the precision improvement of the hobbing machine, thus obtaining a high-precision hobbing machine solution. Firstly, through the transmission error testing system named FMT of the research group, the precision measurement of the transmission error is realized, and the data is analyzed by spectrum to find the large error link. Then, different correction schemes are adopted for different error links to achieve the purpose of eliminating the original error by the reverse error. After many inspections and corrections, the transmission accuracy is greatly improved, and the hobbing machine TE is changed from the initial 26″ to 7.40″. This solution can improve the accuracy of machine tool transmission at low cost and can be applied to other types of machine tools.

An electronic inclinometer can measure the angle between faces or between a face and a horizontal plane. In recent years, electronic inclinometers have played an important role in bridge monitoring, medical devices calibration, and aerospace, with a wider measuring range and higher accuracy. However, there are few studies on the calibration methods and uncertainty budget of the electronic inclinometer. In this paper, the traceability chain, calibration environment, measurement standard, and calibration method are investigated to provide a direct measurement method to calibrate the indication error of the electronic inclinometer by using the dividing head as a standard angle generator and a special fixture. A mathematical model is established, and the uncertainty budget is performed for the calibration result of the indication error. The result shows that the ratio of the expanded uncertainty of the electronic inclinometer to the indication error is less than 1»3. Thus, a feasible calibration method for the indication error of the electronic inclinometer is obtained.

Fiber optical shape sensor has been widely used in industry because of its high stability and compatibility, and its shape reconstruction algorithm has always been the focus of scientific research. The article investigated shape measurement method of sensing optical fiber based on differential geometry, and discusses the measurement principle of Frenet frame and Rotation Minimizing Frame(RMF) in optical fiber shape sensor, compares the application effects of the above two methods. Analysing the applicability of the two methods in practical engineering application when the sensing fiber is in different deformation forms.

In order to eliminate the asymmetric projection error, this paper proposes a method to calculate projection point of circle center by use of concentric circles. Firstly, the basic principles and mathematical model are introduced briefly. Next, take arbitrary secants on a circle, the relationship between the secant midpoint and the projection point of the infinity point on the imaging plane is obtained based on the properties of projective geometry. The equations which contain parameters of secant midpoints are established according to the geometric constraints. Then the coordinate of projection point of circle center is derived from the coordinates of secant midpoints. Finally, in order to eliminate the algorithm error in the calculation process, the coordinates of all solved projection points are averaged to get the final coordinates. The simulation results show that the method can get the projection point of circle center accurately and stably, and does not have intrinsic projection error. Besides, the accuracy of the proposed method is 1.71nm higher than that of the direct ellipse fitting method.

Aiming at the microporous array structure, a measurement system combining vision and probe which can take into account the measurement accuracy and efficiency is proposed. The acquisition of visual features needs to ensure the accuracy and rapidity. Therefore, for the scanning acquisition of visual features, an image segmentation algorithm based on image gradient and a fast scanning method of part surface area are designed in this paper. The image segmentation algorithm based on image gradient combines the advantages of global binarization and local Binarization in threshold segmentation. Firstly, the local eigenvalues of each pixel are obtained, then the threshold segmentation is carried out according to the eigenvalues, and the region edge features are used to distinguish the non-part region, part surface and micropore feature region; The fast scanning method of part surface area realizes fast scanning through image segmentation results and certain scanning strategies; Using the space engine injection disk to verify, it shows that the region segmentation not only ensures the accuracy, but also ensures the real-time performance of the algorithm; The fast scanning method of part surface can avoid invalid scanning, effectively improve the scanning efficiency, and does not need the operator to provide a priori knowledge.

Aircraft pose is of great importance to the monitor during flight test. Commonly, traditional pose measurement is based on various sensors, which has the inconvenience of repair and replacement because of damage. Two-dimension (2D) single image processing method alleviate the inconvenience, but it has the ambiguity of single image three-dimension (3D) reconstruction. To address these problems, we accomplish 3D reconstruction of the aircraft’s structures via 2D multi-view images. Structures are obtained from 2D multi-view images of aircraft by a convolutional neural network (CNN) and then used to accomplish reconstruction. Structures typically represent the topological relationship between components of aircraft, reducing the self-occlusion of point features. To more precise evaluation of the experimental results, we propose a new Mean Per Frame Position Error (MPFPE) calculation for the structures position. Compared with the Mean Per Joint Position Error (MPJPE), the MPFPE takes the length of structures into account and mixes the multi-view images. Experiments show the mean error of our method is 1.47%, which shows great potential for aircraft pose estimation.

To real time obtain the pose of a line laser sensor in a manipulator with a high accuracy, in this paper, a line laser three-dimensional(3D) measurement system is designed based on visual positioning method. First, the position and posture of the target in the camera coordinate system is obtained by the coplanar PNP model under isometric constraint and the absolute orientation problem. Second, the pose change matrices of the target relative to the initial position at different times are calculated by the pose of the target in the camera coordinate system to achieve the visual positioning. Third, the relative position and posture relationship matrix between the target and the line laser sensor is solved by the standard small ball method to achieve hand-eye calibration. Fourth, the measurement data of the line laser sensor are unified into the world coordinate system by the pose change matrices and the hand-eye calibration matrix to achieve the point cloud splicing. The experimental result shows that the displacement accuracy of the visual positioning is 0.039 mm, and the rotation accuracy is 4.2×10^-3 rad. The contour measurement accuracy of the line laser 3D measurement system based on the vision positioning technology is 0.55 mm. It can be seen that the system meets the general industrial measurement requirements.

For structures with damping and gyroscopic effects, an iterative dynamic condensation method of finite element model is proposed in this paper, which is able to simultaneously reduce the mass, damping, gyroscopic and stiffness matrices of the rotor-bearing system. The numerical example of a simple rotor-bearing system is used to verify the present method. The critical speed and steady-state displacement response of the rotor-bearing system are calculated and compared using the proposed reduced model and full model, and the results show that the reduced model can accurately calculate the critical speed. When the speed is lower than the second-order critical speed, the reducd model proposed in this paper can accurately predict the steady-state response of the rotor system(except for the critical speed), and it can be used for the unbalance fault diagnosis of the rotor-bearing system.

Light strip extraction is important in a structured light measurement system. However, the extraction of light strip is incomplete, losing the information, in the detection of highly reflective surface. This paper presents an iterative threshold segmentation algorithm based on information entropy. Firstly, the image is initially segmented, then using the eightneighborhood detection to remove the noise and retain the light strip, after which the iterative segmentation is carried out with the inflection point of the information entropy of the extracted image as the termination condition to improve the integrity of the light strip segmentation and avoid the excessive segmentation of the background. Compared with the results of Otsu algorithm, the proposed method can retain more complete light strip information, with fewer breaking points.

The paper introduces an experimental scheme for measuring the phase of double laser pulses based on the combination of near-field time-stretched dispersion Fourier transform(TS-DFT) and Gerchberg-Saxton(GS) algorithm. The scheme uses dispersive fibers to perform near-field dispersion Fourier transform on the double pulse signal, and then uses the GS algorithm of time domain simulation to complete the phase recovery of the double laser pulses. Finally, the algorithm simulation of this scheme successfully recovered the phase information of the double pulses, which verified the feasibility of the scheme.

Road traffic sign recognition includes two parts: traffic sign recognition and lane line recognition. Based on the Hilens deep learning platform provided by Huawei cloud, this paper simulates the whole traffic scene and realizes the function of traffic sign recognition. In the aspect of detection and recognition of landmarks and vehicles, based on yolov3, this paper uses two kinds of backbone networks, Darknet53 and Resnet18, adds interference data features, and uses soft NMS, which improves the robustness and accuracy of the algorithm. The experimental results show that the detection accuracy of traffic signs in this paper can reach 97.09%. In the aspect of lane line detection, aiming at the problem of low accuracy of Hough transform, this paper optimizes it combined with least square method to improve its recognition efficiency.

We propose a method of measuring the optical thickness of stacked glass using an absolute distance measurement based on the principle optical balanced cross correlation. The optical balanced cross correlation system is mainly composed of a nonlinear crystal, which is used to generate two second-harmonic sub-pulses. The balanced cross correlation signal will be generated while the measurement pulses and the reference pulses are overlap by scanning the repetition rate. The signal is used as an error signal to lock the repetition rate of the laser by controlling the cavity length. In this study, multiple glasses are pasted one by one with different thickness and the thickness of each glass is measured with sub millimeter precision.

Scheimpflug principle is a description of the geometric relationship between the orientation of the plane of sharp focus, the lens plane, and the image plane of an optical system, when the three planes intersect along a common line, which is widely applied in single line laser scanning measurement. We extend Scheimpflug principle to multi-line laser scanning measurement to enlarge the depth of field in a specific orientation and leverage its strengths of high measurement efficiency in the meanwhile. The two classic scenarios of multi-line laser are theoretically analyzed with the construction of measurement model. Our experimental results verify the measurement model and prove the increase of depth of field in the measurement direction.

In this paper, a resonant piezoelectric pump is driven by a new type of piezoelectric vibrator is presented. A square piezoelectric vibrator with an arch cantilever beam structure is designed. Four arched cantilever beams are symmetrically distributed at the output end. Through the arch structure design, the stiffness of the cantilever beam is reduced under a certain size. The output displacement of the vibrator is improved and the output flow of the piezoelectric pump is improved. In this paper, the square piezoelectric vibrator is modeled and simulated. The experimental results show that the piezoelectric pump with square vibrator has good output performance. The flow rate reaches a peak of 49.8 ml/min at 505Hz when the driving voltage is 400 V_pp.

Loop closure detection (LCD) is an important and challenging task in simultaneous localization and mapping (SLAM). To improve the efficiency and accuracy of an indoor LiDAR LCD, a geometric feature-based method is proposed. First, the rotation invariance of the keyframe pair to be detected is realized by the Fourier transform. Second, the main features of the keyframe are extracted and fitted into geometric features, which are stored in a ring-shaped semantic image. Third, the similarity score of the two keyframes is obtained by calculating the distance between two corresponding ring semantic images, which is verified by the ICP algorithm. The experimental result shows that the time cost of each detection is reduced to less than 30ms without sacrificing the detection precision by using our suggested method.

Stereo deflectometry can specify the position of a workpiece and reduce the difficulty of geometrical calibration. But the measurement scope is limited, and this issue is especially severe for the measurement of complex surfaces. A method is proposed to extend the measuring scope of stereo deflectometry. The nominal model of the surface under test is aligned with the overlapped measurement area of the stereo vision system, and the other areas outside overlapped region are measured using the monoscopic SCOTS approach with each camera, respectively. This method effectively combines the advantages of stereo and monoscopic deflectometry, and the measuring accuracy and flexibility can be improved.

The deviation between the actual and the theoretical kinematics parameters of the articulated arm coordinate measuring machine (AACMM) will lead to a decrease in the measurement accuracy. To improve the measurement accuracy, a carbon fiber standard gauge is designed and used to calibrate the AACMM based on space distance and single point. The coordinate system of the AACMM is established, based on which the homogeneous transformation matrixes from the probe to the base are derived. By regarding the parameter identification problem as fitting a nonlinear regression model, a search algorithm identification model is built. The data acquisition system based on the carbon fiber standard gauge is developed. And Particle Swarm Optimization (PSO) is used to identify kinematics parameters. Experimental results show that the accuracy of single point cone measurement and the measurement accuracy of space distance are all improved greatly after calibration.

The modern industry has higher precision demand for micro-angle measurement, the measurement methods and measurement techniques are constantly progressing. A micro-angle measurement method based on the interference imaging principle of F-P etalon was proposed, and the micro-angle measurement can be achieved in conjunction with an effective data processing technique. This paper performs error analysis and experimental study on the measurement system. Theory analyzes the effect of temperature change, nonlinearity of focal length calculation model, focusing error, and F-P etalon interval on micro-angle measurement results, and proposes possible solutions. A micro-angle measurement experiment with different F-P etalon intervals (2 mm, 3 mm, 5 mm) were performed using existing experimental devices. The experimental results show that the uncertainty of the micro-angle measurement results at interval d = 3 mm is optimal; under the current test conditions, the angle measurement uncertainty in the 300" measurement range is not greater than 0.08".

Robot joint is the main actuator for the attitude control of robot system, which has an important effect on the performance of the whole machine. Lost motion is the key indices to characterize robot joint transmission accuracy and performance, and its measurement and evaluation is an important foundation for robot joint performance improvement. The measurement level of robot joint lost motion is still limited to the measurement of single transmission device, such as reducer and motor. The researches on measuring robot joint lost motion is still small. This paper analyzes the measuring principle and methods of robot joint lost motion, and designs a testing machine for robot joint, which provides a strong support for establishing the overall quality evaluation system for robot joint. The correctness, effectiveness and practicability of the proposed method and theory have been verified by targeted experiments. It shows that the testing machine can meet the requirements of measuring lost motion of different types of robot joints.

Due to its anisotropic shrinkage characteristics and the instability of forming process, small module plastic gears are easy to produce complex appearance defects after forming. The traditional manual detection can not accurately identify all kinds of defects, let alone complete the full inspection and statistical analysis. Therefore, a prototype of small module plastic gear on-line detection and sorting system based on machine vision is developed. The system has the function of on-line classification and identification of plastic gear surface defects, and can realize full inspection. The watershed algorithm based on local extreme value and Canny edge extraction algorithm can accurately extract the gear defect area, mark the defect location and count the defect characteristics, complete the determination of defects such as size out of tolerance, shrinkage, burr, surface foreign matter and incompletely filled part in 0.3s, and display the defect location through the human-computer interaction interface, with a discrimination accuracy of 95%.

Triboelectric nanogenerators (TENGs) have broad application prospects in energy havesting and self-powered sensing devices. In this paper, a wind-driven TENG based on Ti₃C₂Tx MXene/amino-functionalized multi-walled carbon nanotubes (MXene/NH₂-MWCNTs) was developed for formaldehyde gas monitoring. The power supply and gas sensor were successfully integrated into one device. The MXene/NH2-MWCNTs with formaldehyde sensing property work both as a frictional layer and electrode in the TENG. The peak-to-peak value of open-circuit voltage from the wind-driven TENG can reach up to 136 V. The output voltage is obviously reduced in varying degrees after being exposed to formaldehyde with different concentrations, resulting from the change of electroconductivity of MXene/NH₂-MWCNTs, which is the design principle of the self-powered formaldehyde gas sensor. Meanwhile, the self-powered formaldehyde gas sensor exhibits high response and low detection limit of 10 ppb at room temperature. This work proposes a new thought to design the self-powered formaldehyde sensor, which has the widespread application prospect to harvest ambient energy for detecting harmful gas without any external power sources.

In the practical application of speckle pattern interference, it is often necessary to measure the deformation or defects of some specimens placed in the box. Transparent glass observation windows are usually used on one side of these boxes. In the measurement of laser speckle pattern interference, glass windows often cause overexposure in a certain area of the field of view, which has a great influence on the measurement. In this paper, a method of recognizing and eliminating high reflective region based on polaroid is proposed. By adjusting the polaroid to filter the stray and chaotic obtrusive light, the overexposed area in the field of view is greatly reduced, and then the exposure area caused by glass windows in the field of view is identified and extracted by feature extraction technology. Finally, the overexposed area is interpolated and filled according to the surrounding region information to obtain the measurement map without overexposure. Experimental results show that the proposed method can effectively extract and eliminate the overexposed area, and good measurement results are obtained.

Aiming at the problems of overshoot, weak disturbance resistance ability and poor adaptive ability of permanent magnet synchronous motor using PI control algorithm. Firstly, the field oriented control (FOC) model of permanent magnet synchronous motor is analyzed. Secondly, based on the active disturbance rejection control and fuzzy control principle, the speed loop controller is improved, and the fuzzy active disturbance rejection control algorithm is designed. The control algorithm preserves the advantages of active disturbance rejection control, and realizes the online tuning function of active disturbance rejection control parameters, which improves the adaptive ability of the system. Finally, the simulation experiment is carried out on the MATLAB / Simulink platform, and the results show that the fuzzy active disturbance rejection control algorithm has the characteristics of no overshoot, high control accuracy and strong adaptive ability.

In order to measure the full-field thermal deformation of materials in high temperature environment, a full-field thermal deformation measurement method based on high temperature digital image correlation is proposed in this paper.In this method, monochromatic polarized light is used as the lighting source, and the polarizing camera combined with narrowband filter is used as the image acquisition system to collect images under high temperature environment. Through optical imaging and optical filtering system, visible radiation can be initially filtered to minimize the thermal radiation on the specimen surface. And the image gray average method is also used in this paper to eliminate the thermal disturbance error and vibration error in the measurement process to improve the measurement accuracy.Finally, a set of high temperature digital image correlation measurement system was built, and the thermal deformation test was carried out on the ceramic materials. The experimental results show that the system can realize the clear image with high-contrast acquisition under high temperature conditions, and can effectively eliminate the interference of vibration and thermal disturbance noise in the measurement under high temperature .

There are many redundant parameters in Convolutional Neural Network(CNN) when it is used for target recognition in a specific scene, which will greatly occupy the calculation amount and affect the operating efficiency of software and hardware, and cannot meet the real-time target detection requirements of the algorithm in a specific scene. In this paper, channel pruning, layer pruning and their hybrid pruning experiments were carried out on you only look once version 3(YOLOv3), a typical CNN target recognition model. The result of the hybrid pruning can greatly reduce the model parameters and the amount of calculation, and can reduce the model of resource utilization through the comparative analysis. And the volume of the model after hybrid pruning was reduced 94.4% when mAP only loss 0.9%. The model inference time was reduced by 36.6%. This study could provide references for the optimization of object recognition model in structured scenes such as road and workshop.

To facilitate the in-depth study of the tribological behavior of texture friction pairs in the process of reciprocating friction. The micro-displacement deformation output model of compliant mechanism is deduced using a new method for indirectly measuring friction coefficient through micro-deformation of double parallel four-bar compliant mechanism. Then, the whole structure of the testing machine is designed, the data acquisition hardware platform of the testing machine is built, and the measurement and control and signal processing system based on virtual instrument technology is developed. To realize the real-time acquisition and synchronous display of tribological measurement data such as friction coefficient, load and speed. Finally, texture friction pairs with diameters of 800 μm, 900 μm, 1000 μm and 1200 μm are processed, and texture friction tests under different loads (50N-250N) and rotational speeds (10r/min-50r/min) are carried out on the testing machine. The test results show that the friction coefficient increases with the increase of rotational speed and decreases with the increase of load, The test results accord with the tribological laws, which are consistent with the law of klageriski test curve under small load. Stable performance of the tester with good credibility, which lays the foundation for the subsequent development of real-time dynamic measurement of the friction coefficient during the reciprocal cycle.

The production process of robot castings mainly includes processes such as robot core taking, core assembly, core setting, dipping, pouring, sand casting, polishing, and cleaning. The handling process involves every process of the casting production line and is one of the key processes of the casting production line. For the 6-axis robot used for handling, the kinematics model of the robot is carried out according to the improved DH parameter method, and the dynamic model of the robot is established by using Newton-Euler equation and Lagrangian equation respectively. Secondly, according to the requirements of casting handling technology, the motion trajectory is planned according to the established kinematics model and dynamics model, and the dynamic simulation is carried out with MATLAB. The joint torques of the robot dynamic models established by the two methods are very close, which verifies the correctness of the dynamic model.

Gears are important mechanical parts that transmit motion and power by interaction of meshed tooth surfaces. The processing quality of tooth surfaces has a direct influence on the performance of gear transmission. In order to verify whether the machining accuracy of the gear tooth surface meets the design and use requirements, it is necessary to measure its accuracy with related instruments. The geometric characteristics of the gear tooth surface include tooth profile, helix, tooth pitch and surface roughness, etc. For the measurement of tooth profile, commonly used instruments can not obtain the values of tooth profile shape, waviness, and surface roughness in one measurement. The cylindrical gear tooth profile measurement scheme based on the roughness profiler can realize the simultaneous measurement of tooth profile deviation, waviness and surface roughness. This is a relatively novel method, and there is a lack of relevant study on the uncertainty evaluation of the measurement results. This article introduces the measurement principle of cylindrical gear tooth profile based on roughness profiler, and analyzes the source of uncertainty in the measurement. The feasibility of the method is verified through experiments, and the uncertainty evaluation result is given.

The signal of the inductive type time-grating angular displacement sensor is essentially a triangular wave which is similar to a sine wave and contains various harmonic components. How to suppress the harmonic components of each frequency and improve the sinusoidal properties of the sensor signal is an important way to improve the measurement accuracy of the sensor. This paper proposes a modulation method for the sine regular change of the air-gap permeance of the sensor, and the basic model of the sensor is established. Two sensor models with different structural parameters are proposed, and the corresponding simulation and error analysis are carried out. The results show that the measurement error of the f2(θ) type sensor is ±25" in the range of 0°-6°. The sensor designed based on the method presented in this paper has an excellent effect on the sinusoidal enhancement of the signal and the suppression of the error component. It has an important significance for the design of the time-grating angular displacement sensor.

We demonstrate a method for simultaneous measurement of thickness and refractive index of transparent materials based on spectrally-resolved interferometry. By extracting the phase from the interference spectrum, the measured optical path difference can be obtained directly without phase shift device in the measurable range. This unique advantage makes it capable of measuring physical thickness and refractive index by introducing transparent medium and processing three useful optical path differences. In the long-term stability test, the repeatability experiment of spectral resolution interferometric ranging technology using femtosecond laser is carried out within 60 min, and the standard deviation is 341nm.

In recent years, with the development of wind power industry, the installed capacity is increasing rapidly, and the installation sites are developing towards the ocean and remote mountainous areas, which makes the maintenance of wind turbines difficult and the cost increases. In order to reduce the maintenance cost and improve the maintenance efficiency, and ensure the economic and reliable operation of the equipment, many state recognition and fault early warning methods have been proposed one after another, and the intelligent fault diagnosis method based on supervisory control and data acquisition(SCADA) data and various kinds of machine learning has gradually become a research hotspot. In this paper, the generator system of a wind turbine in a wind farm is taken as the research object. By using the massive SCADA data recorded in operation, a multivariate state estimation technique (MSET) is established to predict specific operation parameters with relevant operation parameters as input. This method uses clustering algorithm to clean up the selected SCADA data, then uses MSET to establish the prediction model, and calculates the prediction residual by sliding window method to realize the fault diagnosis. Finally, the effectiveness of the method is verified by actual SCADA data

Due to the components of DD motor exist errors in the process of machining and assembly, which will generate error motions in rotary axis when the DD motor starts running. To solve the above problems, this paper presents a measurement method for three degrees of freedom error motions of rotary axis based on auto-collimation principle. The proposed system consists of two collimated laser, two beam splitters, two quadrant photodiode detectors (QPDs), four lenses and a high precision steel ball mounted on the top of DD motor. The error motions of the rotary axis result in the changes of the position of high precision steel ball. Thus, the shift values of laser beams reflected from the high precision steel ball are detected by the QPDs. The experimental results show that the radial and axial error motions of DD motors is less than 3.2μm and range from -1.5μm to 2μm respectively

With the development of ultra-precision equipment, the roughness of contact surface can reach the nanometer level, and the surface morphology has a significant impact on surface contact, friction, wear and lubrication. At present, the surface morphology description is mainly based on the measurement, which is scale dependent, and the statistical parameters obtained by different sampling length and measurement resolution are different, so it is impossible to realize the accurate characterization of nanoscale rough surface. Because the rough surface is self-affine, another method can be introduced to characterize the rough morphology, fractal theory. The simulated rough surface has the advantage that it is not limited by the sampling length, and can realize the unique characterization of the rough surface. In this study, a nanoscale anisotropic three-dimensional fractal surface is established based on W-M model, and the relationships between fractal dimension D, roughness coefficient G, contour arithmetic mean deviation Sa and contour height standard deviation Sq are studied based on statistical principle. Finally, it is determined that the key parameter for the characterization of nanoscale rough morphology is the roughness coefficient G

The asynchronous optical sampling (ASOPS) ranging method using dual optical frequency combs (OFCs) has high measurement accuracy and fast measurement speed, is has been widely used in precision manufacturing, aerospace and scientific research. In recent years, with the development of photonic integration technology, the on-chip soliton microcomb technology has gradually matured. Microcomb has the advantages of small size, high repetition rate, great integration, and system stability, providing an ideal light source for ranging. This paper simulates on the key parameters of dual soliton microcombs ranging system with high repetition rate. The ranging model and numerical simulation method based on soliton microcomb are proposed, the effects of different repetition rate and repetition rate different on ranging accuracy are compared. The optimal parameters of soliton microcomb are obtained.

This paper is concerned with Inerter-Spring-Damper(ISD) suspension design with semi-active inerter and semi-active damper configuration. First, a quarter vehicle model with semi-active suspension is established, and three control laws， named Sky-Hook (SH) control, Acceleration-Driven-Damping (ADD) control and Power-Driven-Damper (PDD) control， are derived to control the inertance and damper in the semi-active part. Hybrid control of semi-active inerter and damper is carried out, in which, switch and continuous control are used in the process of SH control, the combinations of different control algorithms are simulated as well. The simulation results show that the RMS of the car body acceleration with the proposed semi-active suspension can be greatly improved. And in the frequency domain, body vibration can also be restrained in wide frequency range, the proposed semi-active suspension has a good ability to attenuate the body vibration and improve the ride comfort. The proposed control strategy can compensate the inaction of inerter, and even benefit the suspension performance at a little higher frequency range.

The self-driving articulated arm coordinate measuring machine (AACMM) is a new type of flexible coordinate measuring equipment. The integrated joint module is introduced to the AACMM joint for self-driving control and measurement. The error source of the joint angle of the self-driving AACMM is analysed based on the internal components of the joint module, and single and comprehensive models of the single-joint module angle error are established. Numerical simulation and calibration experiment of torsion deformation angle error model of the single joint module are carried out. Results show that the angle error of each joint module includes the joint torsion deformation error, the angle error of the harmonic reducer, the angle measurement error of the magnetic encoder, and the error of the servo motor control system. The simulation error of the joint rotation angle shows a cosine variation trend under different loads, and the error fluctuation range is positively correlated with the load. The joint rotation angle calibration error shows a periodic variation, and the overall error fluctuation range is [-80.7 arcsec, 61.31arcsec]. The analysis and modelling calibration of the angle error of the AACMM single joint module offer an important theoretical foundation for the further research of the AACMM joint angle error compensation to improve the positioning and measurement accuracy of the measuring machine.

Laser-induced breakdown spectroscopy (LIBS) is an effective technology to analyze the content of the target elements. The surface morphology of the target will affect the coupling between the laser and the target, which will change the plasma spectrum and lead to inaccurate results. The surface relief and surface roughness are taken as the research parameters of the target surface morphology, the influence of which on LIBS spectrum are researched. The LIBS spectra are acquired on a set of ferroalloy targets, whose included angles θ with horizontal direction changed from -10° to 10°, or surface roughness are different. On the basis of theoretical derivation, we explore the variation trends of line intensity, line integral area, line intensity ratio of different main elements, and line intensity ratio of the same main element with surface morphology parameters. The experimental results have an increasing trend with the increase of θ and a decreasing trend with the decrease of surface roughness. The line intensity ratios are closely related to the change of surface morphology. The line integral area of Cr Ⅰ 429.3438nm has a large variation amplitude and higher correlation coefficient R, which is suitable for characterizing the change of LIBS spectrum with the target surface morphology. The results can provide a valuable reference for reducing the influence of target surface morphology on LIBS detection.

For improving the positioning accuracy of the robot, the kinematic parameters need to be calibrated. A modified differential evolution algorithm is proposed to identify the kinematic parameters in this paper. The control parameters of differential evolution algorithm are adjusted adaptively for a better convergence. Herein, the kinematic model is established based on the product of exponentials formula. To verify the efficiency of the proposed algorithm, experiments are performed using a six-degree-of-freedom robot and laser tracker. The maximum positioning error is reduced from 6.094mm to 0.4935mm and the average positioning error is reduced from 3.076mm to 0.08911mm by using the proposed algorithm.

The low-frequency information of seismic records can enhance the recognition ability of lithological bodies and make the inversion results clear and reliable. Affected by conventional acquisition technology, low-frequency information is usually missing in the imaging profile. Therefore, extrapolating low-frequency seismic data from band-limited seismic data is an important research topic in full-waveform inversion (FWI). Most of the existing methods directly use machine learning to extrapolate the low-frequency, but the amplitude of the seismic records will be greatly attenuated with the increase of offset and time. The energy gap of seismic records is wide, and the contribution of high-energy seismic records to the network weight is far greater than that of low-energy data. Therefore, it is difficult to extrapolate deep low-energy data. To solve this problem, we propose a method of layered low-frequency extrapolation with deep learning. The seismic records are divided into several layers according to the change of depth and the similarity of energy, and the convolutional neural network is used for training. Experimental results show that this method can accurately extrapolate low-frequency data, and the extrapolation data in the deep layer are close to true data in both the time domain and the frequency domain. In addition, this method occupies lesser computing resources and has the potential for field data application. We verify the effectiveness of the method through two datasets obtained from the Marmousi model and the overthrust model.

The influence function of the actuators can be expressed as a Gaussian function. The exponent of the Gaussian function influences the reconstruction precision of the deformable mirror (DM). A finite element model of the PZT stacked array DM with 55 actuators was created, and each actuator was separately controlled. The influence of the structural parameters of the DM on the Gaussian index and coupling coefficient was investigated. Based on the simulation results, the design of the DM was optimized. In addition, using simulation, the fatigue life and working bandwidth of the new model were studied. Specific aberration surfaces were reconstructed. The results showed that the optimized DM had a smaller reconstruction error compared with the pre-optimized model. This study can be used as a reference for the design of DMs.

Aircraft images captured by a third-party camera during take-off and landing can be used for monitoring and aircraft pose measurement. Hazy weather would severely affect the aircraft image quality and incur the worse visual perception. Haze removal from the aircraft image has become an important task for practical industrial applications. Existing deep learning algorithms need the hazy image and corresponding hazy-free ground-truth image simultaneously for the same scene and time, to learn the dehazing process. However, the ground-truth aircraft images are difficult to obtain, which hinders those approaches from addressing the actual aircraft image dehazing problem. In this paper, we present an endto- end ground-truth information agnostic deep dehazing network for single C919 aircraft image dehazing problem. Instead of the requirement of ground-truth image, we train the network only by utilizing the pair of hazy and predehazed images. The pre-dehazed image can be easily obtained by the conventional dehazing manner without deep learning, and the Natural Image Quality Evaluator (NIQE) is introduced to find the best dehazing model. Compared to existing dehazing algorithms, the proposed algorithm can be capable of addressing real-world hazy C919 aircraft images effectively and achieve the best dehazed performance on our collected aircraft dataset.

Automatic history matching is a process of using an optimization algorithm to adjust the parameters of the reservoir model. The reservoir model can reproduce the historical performance of the reservoir and realize the prediction for future production. Accurate prediction of oil well performance guarantees to establish a reliable reservoir model, which is traditionally realized by ESMDA and ensemble Kalman filter. We design and implement history matching using a 3D-pix2pix generative adversarial network(3D-pix2pix GAN) structure for the first time, which can correct the parameters of the complex heterogeneous reservoir based on dynamic response. The adversarial generative network includes generator and discriminator. The generator attempts to use the fast feedforward operation of historical production data (input) to reconstruct the calibrated model, while the discriminator attempts to distinguish the pseudo output and the prior (real data) so that 3D-pix2pixGAN finally learns an infinitely close to the real reservoir model. The most significant contribution of this work is to train a 3D-pix2pixGAN model to correct reservoir model parameters. Compared with traditional work ow, 3D-pix2pixGAN has several advantages. First, the reservoir parameters estimated from history matching help to improve 3D reservoir characterization. Second, the reservoir obtained by history matching can accurately predict the future production of water and oil. Third, 3D-pix2pixGAN is used as a proxy model instead of using a numerical simulator in the training process to reduce the amount of computation and achieve end-to-end offline processing.

Deep sort algorithm is a multi-object tracking algorithm with high tracking accuracy and speed. However, due to the lack of detection filter and the association stage of a single frame, the accuracy of multi-object tracking is remaining enhancement. In this paper, we propose a DO-Adaptive NMS algorithm to filter the detections, and combine the K nearest neighbor algorithm with the intersection of union algorithm to sharpen features of the trajectories. Besides, we put forward a weighted algorithm of motion information and appearance information, which takes the disappear time of trajectories into consideration. Experiments show that the methods mentioned above all perform better than the original algorithm.

In the process of imaging, atmospheric turbulence will lead to image degradation, such as noise, blur, geometric distortion, thus reducing the quality of feature point extraction. In order to solve this problem, we analyze images with atmospheric turbulence degradation and find that image blur and geometric distortion have great influence on feature extraction. Image blur is a representation of high-frequency information loss, so detectors based on gray gradient will extract fewer points. On the other hand, geometric distortion is reflected by the movement of pixels in the image patch, which will also cause the movement of feature points, especially when they are extracted according to their neighborhoods. In this paper, we propose Wiener Filter and Linear Minimum Variance Unbiased Estimation (WFLMVUE) strategy to deal with image blur and geometric distortion respectively. A simplified filter based on Wiener’s method is used to remove noise and ambiguity. Then the base frame and auxiliary frames are used to estimate the position of feature points by linear minimum variance unbiased estimation. Experimental results show that WF-LMUVE has great advantages in increasing the number of feature points and improving their location accuracy.

In fields such as wind power, shipping and so on, there are great demands for linear and (or) rotary precision positioning of heavy or supersized mechanicals; and the areas where required high reliability, like aerospace and weaponry, there are also urgent needs to perform synchronous position measurement through a multipoint redundant structure in a narrow space to deal with the impact of strong shock vibration and electromagnetic interference. As grating encoders and magnetic grating encoders are difficult to meet the requirements of the extreme space size and working environment above, this paper proposes an embedded measurement method that integrates magnetic sensing modules (MSMs) and the measured transmission part. Its principle is that place several small MSMs next to the measured transmission part to form a specific time-varying magnetic field, and then make use of the periodic equipartition structure of the transmission part itself to realize the permeance modulation to get the induced voltage that reflects the position change, finally, calculate the displacement. Based on the magnetic equivalent circuit (MEC), this paper builds an embedded time-grating (ETG) sensor model applied to gear parts. According to that, take a spur gear as the object for rapid optimization design, which is verified by three-dimensional (3-D) time stepping finite element analysis (TSFEA). Finally, an experimental prototype is made by the optimized sensor parameters, and experimentally verify the results. The results show that the high-accuracy measurement is realized, and the measurement error achieves ±5″ within the entire range.

Turntable bearings are widely used in the fields of national defense, industrial manufacturing and precision machining, and the precise measurement of their displacement is very important. A cylindrical embedded magnetic field angular displacement sensor based on PCB process is presented in this paper. In this design, the coil required by the electromagnetic induction sensor is distributed in the flexible PCB , and there is no need to provide a specific shape of the iron core to place the metal coil, which is easy to achieve embedded measurement. In this paper, the magnetic field distribution of cylindrical coil is modeled according to Biot-savar's law, and the mathematical relationship between magnetic induction intensity and coil structure is obtained. The coil structure is designed according to the mathematical relationship obtained. The feasibility and accuracy of the theoretical model were verified by finite element simulation software. Finally, a prototype was made for measuring angular displacement of bearing with inspection. Experiments show that the resolution of the sensor can reach 0.5 "and the original error is ±80" in the range of 0~360°.

Highway tunnel lighting equipment, as an important equipment to ensure traffic safety in highway tunnel, has some problems in its production, installation, commissioning and operation, such as low intelligence degree and weak man machine interaction operation. On the basis of studying the existing test methods of highway tunnel lighting equipment, this paper designs an improved test process and method of highway tunnel lighting equipment, builds a cloud management platform for test data, realizes the automatic test of highway tunnel lighting equipment by using remote client combined with Ethernet, realizes the intellectualization of test process and the standardization of test data. The database and background management software deployed on the remote server can realize the classified management of many kinds of test data. The system can improve the testing efficiency of lighting equipment for the highway tunnel lighting equipment production enterprises, and improve the management efficiency and quality for the highway tunnel lighting management department.

Aiming at the problems of cumbersome, time-consuming, inefficient and unable to create an accurate and controllable light environment for the input of the dimming system in the existing road tunnel lighting dimming quality inspection process, this paper designed a set of road tunnel lighting brightness intelligent detection system. The system fixes 5 illuminance sensors in a vehicle-mounted manner to achieve continuous measurement of longitudinal measurement points on the tunnel road surface and synchronous measurement of lateral measurement points; a programmable light source installed outside the tunnel provides a controllable brightness input for the dimming system, and uses the LoRa wireless technology completes the long-distance wireless communication with the inspection vehicle, which can realize the analysis and management of the measurement data. The whole system is convenient, efficient and intelligent, which can be applied to highway tunnel lighting quality monitoring and operation management.

Nowadays, the number of assembled wind turbines in the world is growing more rapidly, which brings an urgent need for intelligent operation and maintenance of wind turbines. The intelligence of wind turbine operation and maintenance is based on the high-precision classification and recognition of SCADA system data. In response to this demand, this paper establishes a wind turbine normal data discrimination model that combines SCADA system data preprocessing and random forest integrated learner. First, obtain a determinable sample dataset according to the principles of statistics and the NearMiss under-processing method. Then build a decision tree, use the features in a variety of SCADA datasets to train and learn the sample dataset, and form a random forest to determine the normal data model of wind turbines. The results show that the model can effectively classify whether the SCADA data of wind turbines is normal, achieve a higher accuracy rate, and improve the reliability of discrimination, which is of great significance to the subsequent research on intelligent operation and maintenance of wind turbines.

The semiconductor industry’s device dimensions continue shrinking and device architectures increase in 3D complexity, while incorporating new materials. To keep pace with these changes, new critical in-line metrology accurately and efficiently evaluating the structural profiles will be needed. Small angle X-ray scatterometry shows promise to be considered for critical dimension (CD) metrology for future nodes. In this paper, we report simulation results of the transmission small angle X-ray scattering (T-SAXS) metrology to evaluate its measurement capability for 3D periodic architectures. Based on measurability analysis for various 3D structural models, T-SAXS shows a good potential solution to the future 3D architectures measurement.

A high-precision two-dimensional (2D) coplanar micro-vibration generator with a flexure decoupling mechanism is proposed in this study. The generator is composed of a 2D symmetrical coplanar platform and a slider pair. The 2D coplanar platform structure is fully symmetric to achieve decoupling of motion in the X and Y directions. The vertical support with a slider pair is used to guarantee strict planar motion with heavy loads. Under open-loop control, the developed generator has the capability of translational motion stroke of 48.71 μm and 49.10 μm in the X and Y directions, respectively, and the maximum coupling error is approximately 0.78%. A PID control system was designed to realize the closed-loop control of the vibration generator. The control system can effectively reduce the inertial vibration and settling time of the generator. In addition, the closed-loop controlled vibration generator also has good motion tracking characteristics.

With the development of semiconductor manufacturing processes, critical dimension small-angle x-ray scattering (CDSAXS) has been identified as a potential solution for measurement. It is worthy of exploring how to achieve fast parameter extraction. In this paper, we propose a XSCNN model based on deep learning to reconstruct the parameters related to structure and measurement conditions. Simulation experiments performed on a trapezoidal grating have demonstrated that XSCNN can produce satisfactory results. It is expected that deep learning will provide a practical solution in CD-SAXS.

To meet the requirement of high precision measurement for high aspect ratio microstructures, a contact trigger probe based on optical lever is designed in this paper. An elastic mechanism, which is composed of a tungsten stylus, a cross shaped leaf springs, a floating plate and a mirror is used to transmitted displacement of the probe tip. An optical lever principle is used to amplify the angle offset to improve the sensitivity. The four quadrant detector is be adopted to detection light spot displacement, And the micro adjustment mechanism adjusts the position of the four quadrant sensor to make the measurement in the linear range. When the probe is triggered in different directions, the light spot will pass through multiple mirrors and produce double angle offset amplification, Through the plane mirror group, the light beam passes through the floating plate twice to realize the secondary amplification of the angle, it causes the light spot on the four quadrant detector to shift. The measurement accuracy is greatly improved. The experimental results show that the resolution of the probe in all directions is better than 1nm, and the maximum resolution in x-axis direction is 0.1nm.

The precision and accuracy of interferometres are limited by three typical errors, including phase shift errors, gain differences, and zero offset. A sinusoidal signal processing techniques for interferometers to correct the errors is proposed in this paper. The errors are compensated by the orthogonalized circuits based on vector operations, the regularized circuits based on automatic gain control (AGC) technology and the DC offset compensation circuits based on low-pass filter, respectively. Miniature semiconductor polarization Michelson interferometer system has been developed using the proposed sinusoidal signal processing techniques and compared with the SIOS commercial interferometer. Experimental results show that the errors can be compensated effectively, and the measuring standard deviation of the developed interferometer is 57 nm. The proposed system can be used to correct the typical errors in sinusoidal signal processing fields.

REVO head is a high efficiency and accuracy five-axis measuring system designed and applied on coordinate measuring machine. It is now widely used in the measurement of precision parts of aircraft engine. As a high precision and high efficiency measurement system, the accuracy of the measurement result can only be guaranteed, in the case of ensuing the correctness of the measurement methods. This papers presents two measurement methods of REVO five-axis head based on coordinate measuring machine, using an NPL designed and manufactured freeform material standard consisting of an aluminium block with geometric shaped blended to form a pseudo freeform surface as a measurement example. By comparing the result of REVO five-axis head CMM and the result given by NPL. The accuracy of the measurement method can be verified. And then combined with the results, this paper explores how to better ensure the accuracy of measurement results. This has profound effect on future precision-parts measurement.

Publisher’s Note: This paper, originally published on 19 November 2021 was withdrawn on 30 November 2021 per author request.

The laser interferometric particle imaging technology is used to irradiate the particle spray field with flaky laser beams and collect the scattered light images of the particles on the defocus surface of the imaging lens with a high-speed CCD camera. Based on the existing laser interference particle imaging measurement methods, this paper proposes an automatic recognition algorithm of interference fringe pattern based on morphology Hough transform and a fringe frequency extraction algorithm based on Fourier transform. The overlapping particles are accurately recognized through the improved particle interference fringe pattern recognition algorithm, so as to improve the particle measurement accuracy. Two algorithms are used to simulate and analyze the basic circle and interference fringes, the correct center coordinates and fringe frequency are obtained, and the interference particle images in particle fields with different concentrations are simulated and verified.

Focused ultrasound transducer is widely used in medical treatment instruments and equipment, which can achieve the purpose of tumor ablation through thermal, cavitation and mechanical effects under high power and high sound intensity conditions, and precisely destroy the target tissues. The optical method based on laser interference technology has the characteristics of high precision, high resolution and high measurement efficiency for the measurement of acoustic field characteristics of the focusing transducer, while avoiding the disadvantages of hydrophone measurement under high sound intensity conditions. In this paper, the influence of the transmissive reflective film on the acoustic transmission is theoretically deduced, and the influence of the transmissive reflective film on the focal position of the focusing transducer is confirmed. Experiments are conducted to investigate the change of acoustic pressure of the transmissive reflective film on the focus of the focusing transducer at different power conditions and different positions of the focusing transducer

Compared with the traditional hydrophone method, the optical method has the advantages of high precision and high measurement speed. In this paper, a variety of optical detection methods of sound field are introduced. Taking optical interferometry as an example, the velocity of acoustic particles in sound field is measured by measuring the vibration of pellicle with laser vibrometer. Acoustic reflective pellicle is the key medium, and its main installation methods are placed in water and on water. The hydrophone method is used to explore the effect of sound transmitting and reflective pellicle placed in water on the performance of ultrasonic sound field. The experimental results show that the existence of surface sound transmitting pellicle has no effect on the ultrasonic sound field of focused transducer.

Accuracy measurement is an essential front step to error compensation of industrial robots. Different measurement methods have been tried all the way over the past years. In this paper, we propose a novel method to measure the positioning distance accuracy based on the multi-station technology of the laser tracer device. Compared with the laser tracker device, this method is more precise in measuring the coordinate values of the robot end-effector. Besides, the laser tracer is more portable than the laser tracker. At the last part, a practical example is presented in which we have carried out in an domestic industrial robot.

With the continuous development of science and technology, cylindrical lithium batteries, as new energy batteries, are widely used in many fields. In the production process of lithium batteries, various defects may occur. To detect the defects of lithium batteries, a detection algorithm based on convolutional neural networks is proposed in this paper. Firstly, image preprocessing is introduced on the collected lithium battery dataset. Secondly, the K-means clustering algorithm is used on the processed dataset to generate anchor boxes for lithium battery defect detection. Then the detection network YOLOv3 is trained with the given dataset. Finally, the detection network YOLOv3 is applied to output the type and location information of the defect. The experimental results show that the mean average precision (mAP) value of the detection algorithm on the lithium battery validation dataset reaches 94% and the detection speed is 25 frames per second. The proposed algorithm can effectively locate and classify the bottom defects of the lithium battery.

Composites are gradually replacing traditional metal materials in aeronautical manufacturing. Despite of excellent properties, composite materials suffer from barely visible impact damage which can be caused by low velocity objects. In this paper, an impact localization algorithm is proposed based on dynamic characteristic calibration and reference database to improve the accuracy and robustness. The installation parameters of fiber Bragg grating (FBG) sensor are determined by the dynamic characteristic calibration. The impact positions are obtained by calculating the similarity degree between the random impact signal and the signals in the reference database. The algorithm is tested over a carbon fiber reinforced polymer (CFRP) plate attached with FBG sensors. Impacts are stimulated by using a hammer. Noise signals are added to part of the FBG sensor signals during experiments. The results show the proposed algorithm can accurately locate the impacts and deal with the signal interference.

This paper presents the characterization of step structures at multiple scales using a metrological AFM rig based on nano measuring machine. The AFM system achieves a measurement range of 25 mm × 25 mm × 5 mm with sub-nanometer resolution. An AFM signal amplifier is designed to match the signals of the AFM scanner and the NMM controller, and to increase noise attenuation. Step-height standards of 500 nm, 10 µm and 2 mm have been measured and evaluated using the presented AFM system. Methods to characterize ultra-high steps and sidewalls are also introduced.

The reconstruction of 3D point cloud is the core of reverse engineering and widely applied in industrial field. Focused on the problem of data redundancy and calculation, the reconstruction procedure is realized and the influence of various triangulation methods on geometric features is analyzed. The point cloud data pre-processing is implemented first based on C++ Point Cloud Library (PCL) in the paper, including filtering and smoothing, outlier removal, valid points extraction, simplification, and hole filling. Then the Greedy Projection Triangulation and the Poisson Reconstruction methods are applied separately to reconstruct the mesh models. The spherical center distance and diameter of calibration board are selected as the geometric characteristic parameters to assess the reconstruction quality. The relative error is calculated according to the true value and the average of multiple measurements on the parameters. For the distance feature, the results show that the two methods have similar accuracy. For the diameter feature, the Greedy Projection Triangulation is further suitable than the Poisson reconstruction, and the relative error of which is less than 0.18%.

Stability detection of melt pool has become a challenging task in laser cladding, due to the high temperature and brightness during laser cladding. In this paper, an enhanced mask R-CNN for object instance segmentation of melt pool is proposed to boost the performance of detection. In order to enrich the dataset and improve the generalization of the neural network, the data enhancement method of elastic deformations is used to simulate the irregular deformation of melt pool topography caused by the interference of the external environment. Meanwhile, the MobilenetV2 structure is introduced into mask R-CNN to solve the problem of a large number of parameters and slow running speed of network model, and transformer model was used to replace the classifier of the original network. Experimental results show that the proposed method can improve testing speed by 14.7% without decreasing the segmentation accuracy. Finally, a dynamic stability detection method of melt pool is proposed in this paper.

The variousness as well as the inaccessibility of defect characteristics make the defect detection of monocrystalline silicon solar cells more challenging. To address these problems, a novel domain adaptive target detection algorithm based on pseudo-label learning, which is an efficient and feasible weakly supervised learning method, is proposed in this paper. Firstly, in the early stage of the model, the loss function is improved to solve the problem that the model is not easy to converge due to a large number of false labels in the pseudo labels. Then, to classify the sample space better, entropy regularization is applied to the sample boundary data, and the unlabels of the target domain images are labeled with pseudo labels. Finally, the source domain data and the target domain data are used for training together, and the generalization performance of the model obtained is greatly improved. The results show that in solar cell image detection, the accuracy of the domain adaptive method based on pseudo-label learning can reach over 90%, which is better than the target detection accuracy of using only the source domain dataset.

The image quality and accuracy of the atomic force microscope are seriously affected due to drift and hysteresis of Piezoelectric positioning platform. Currently, the distortion correction methods had been confined in tracking a certain area or image correction after images are acquired, which makes it difficult for the atomic force microscope (AFM) to obtain low distortion images. In this study, the scanning path of the AFM is redesigned. The scanning route is spiral in a whole with many blocks which are scanned once at a time, and the drift of blocks is corrected in real-time during the scanning process. This method is suitable for real-time correction of drift during long-time scanning, compared with the traditional scanning method.

Electric vehicle charging pile fault diagnosis (CPFD) technology has achieved rapid development and successfully implemented in the field of electric vehicle charging piles. However, in real life, failure data is very difficult to obtain, as a result, it will cause data samples to be imbalanced seriously and make CPFD more and more challenging. To solve this problem, a novel Borderline-SMOTE-based imbalance correction for CPFD is proposed in this paper. With regard to the imbalance correction, Borderline-SMOTE over-sampling technology is utilized to solve the problem of unbalanced samples. For CPFD implementation, the LightGBM ensemble learning combined with a grid search cross-validation algorithm is designed to build a fault detection model. Related experiments have proven the proposed methods can achieve the highest diagnostic accuracy, which is superior to other popular methods.

Defect detection methods for photovoltaic (PV) devices based on electroluminescence (EL) imaging technology are crucial to maintain productivity and prolong components’ life. However, due to the lack of feature extraction capability for morphologically complex defects and some small defects, the traditional object detection algorithms perform not well in EL defect detection. Therefore, an improved algorithm based on the Faster Region-based Convolutional Network algorithm (Faster R-CNN) is proposed to improve the detection performance of multi-scale defects. Specifically, an improved residual module based on deformable convolution and attention module is proposed to improve the detection rate of morphologically complex defects. And a Feature Pyramid Network (FPN) is utilized in the proposed algorithm to improve the detection performance of small defects. In addition, the GIOU loss, instead of the original smooth L1 loss, is utilized to improve the boundary box regression accuracy. Experimental results on the detection of the EL defects show the high efficiency of the proposed algorithm. The method is expected to provide more guiding feedback in both practical design and reliable diagnosis of the PV industry.

The treatment force and moment on the teeth play an important role in orthodontics. However, there are many difficulties to investigate the biomechanical mechanism of tooth movement in vivo. Orthodontic simulation system becomes an acceptable method to reproduce the orthodontic process and measure the mechanical parameters. In this study, an orthodontic mechanics test platform based on a six-axial force/moment sensor is developed to provide a quantitative evaluation of orthodontic forces and moments exerted by the invisible braces. First, the mechanical design and working principle of the test platform are explained. Then, the hardware design and data processing of the six-axial force/moment sensor are illuminated. The calibration of the sensor is described. Finally, the maxillary model of a central incisor with specific displacement are tested and discussed. The experimental results show that the proposed test platform can simulate the position change of the concerned tooth and reflect the magnitude change of their mechanical parameters during the orthodontic treatment. This study provides an effective technical solution for the investigation of the biomechanical mechanism of tooth movement during the orthodontic process.