Spatial error obtained is a crucial process for industrial robot accuracy calibration. This study proposes an error measurement system for robot tool center point, comprising a novel non-contact measurement device based on R-test device, and an automated rotary device equipped with several standard spheres across different ranges. The geometric accuracy of the proposed system was calibrated by Coordinate Measurement Machine and a self-designed calibration stage. Spatial error measurement is established based on geometric modelling and error vector mapping, followed by a novel error identification and compensation algorithm incorporating the Least Squares Method, Cartesian stiffness identification and Ridge Estimation. The calibration experiment is conducted on EPSON C4-A901 industrial robot, with a radius varying from 50 mm to 100 mm under different loading conditions. The results indicates that the proposed measurement system and method effectively reduced average errors by 52

Kinematic compensation is essential for improving the positioning accuracy of industrial robots. Conventional model-based calibration often lacks flexibility and scalability, while many learning-based approaches suffer from overfitting and limited deployability. This study develops a hybrid calibration framework that integrates a compact non-contact R-test measurement system with a tailored deep-learning-based error predictor, termed StarCalNet. StarCalNet is implemented as a compact multilayer perceptron with element-wise feature interaction to capture configuration-dependent error patterns while keeping the model size modest. A similarity-aware mechanism further exploits multi-scale configuration similarity to regularize learning of joint-space error distributions. Calibration experiments are conducted on an EPSON C4-A901 robot in nine workspace regions and three trajectory radii (50-100 mm). Measurements are obtained with a self-designed non-contact rotary-sphere (R-test) system that integrates three laser displacement sensors and a compact flange-mounted automated rotary module. This configuration enables fully automated multi-radius trajectory acquisition at substantially lower hardware cost than laser-tracker-based setups. The experimental results demonstrated that the proposed calibration method reduced average spatial errors by 41%-65% across 27 regions in robot working space. Compared to the Levenberg-Marquardt method under the identical conditions, the proposed approach achieved over 40% improvement in compensation performance.

Thermal errors arising from machine tool thermal deformation significantly affect the accuracy and stability of CNC machining. Real-time thermal error compensation, implemented via high-precision data-driven models, is a widely accepted method for ensuring consistent machining quality. However, because the model’s prediction accuracy relies entirely on the availability of a sufficient number of labeled samples, the practical application of thermal error compensation technology is limited by challenges including difficulties in acquiring modeling data and high modeling costs. To address thermal error modeling of CNC machine tools under small sample conditions and to enable multi-model expansion of the thermal error model, this study proposes a thermal error modeling method based on incremental transfer learning. Firstly, using support vector regression (SVR), an initial thermal error model for the source domain machine tool is established based on small sample error data; subsequently, an incremental learning approach combined with a sample screening mechanism founded on Karush–Kuhn–Tucker (KKT) conditions is employed to optimize the model’s boundaries, thereby enhancing accuracy and generalization. Secondly, a thermal error transfer model for various types of machine tools is constructed by integrating KKT screening with correlation alignment-based domain adaptation methods (CORAL). Finally, the proposed model (K-CA SVR) is compared with the traditional SVR model and a convolutional neural network (CNN) model, while real-time error compensation experiments are conducted on various types of machine tools. The results indicate that the actual thermal error compensation accuracy for the source domain machine tool exceeds 90

In vision-based measurement applications, black light-absorbing objects pose a challenge due to their poor reflection of the structured light emitted by the infrared module of an RGB-D camera. To address this, an algorithm based on reference environment information is proposed. Concurrently, a robotic depalletising system is established for randomly stacked single-layer volutes. During the depalletising process, the algorithm obtains the volutes position information within a depalletising framework from the vision system, which includes an RGB-D camera and an upper computer, and outputs the spatial positioning data and adsorption pose for end adsorption device connect with a six-axis industrial robot. Following this, for the method of positioning each individual volute, the position of the centre in XY-plane is determined by contours at the end of the protrusion and calculated by using the offset algorithm, while the volute's depth data is inferred from the depth of the cardboard using a linear transformation. Additionally, a regional hand-eye calibration method has been developed to improve twodimensional calibration precision, which divides the graspable area into quadrants, each with a unique coordinate transformation matrix. By conducting depalletising experiments with three different types of volutes, the maximum error in the image processing procedure is less than 9 mm. The average error is below 4 mm, and the standard deviation is less than 2 mm. After the hand-eye calibration phase, the maximum error observed for the volutes is within 8 mm, with an average deviation slightly less than 4.5 mm and a standard deviation below 2 mm.

The spatial error obtained is a crucial process for industrial robot accuracy calibration. However, the existing spatial error sensor systems are often limited by inflexibility, poor adaptability to varying conditions, or substantial cost. This study proposes a multisensor system for calibrating tool center point (TCP) of the robot, comprising a novel noncontact sensor system based on R-test, and an automated rotary device equipped with several standard spheres across different ranges. The geometric accuracy of the proposed system was validated by coordinate measuring machine (CMM). Spatial error measurement was established based on the geometric modeling and error vector mapping, followed by a novel error identification and compensation algorithm incorporating the Least Squares Method, Cartesian stiffness identification, and Ridge Estimation. A simulation was conducted to demonstrate the capability of the proposed model to identify and correct spatial errors down to the micrometer level. The calibration experiment was conducted on the EPSON C4-A901 industrial robot. The results showed that the proposed method reduced the average error modulus by 36.94%-43.86%, and the maximum error modulus by 35.90%-20.34% as the rotation radius increased from 50 to 100 mm. In addition, a laser tracker was employed for validation, and the spatial errors measured by both systems exhibited a similar trend and deviation. The calibration results obtained using the laser tracker showed a 49 % reduction in average error. This result substantiates the reliability of the proposed calibration algorithm and verifies the accuracy of the measurement system.

Accurate segmentation of densely stacked and weakly textured objects remains a core challenge in robotic depalletizing for industrial applications. To address this, we propose MaskNet, an instance segmentation network tailored for RGB-D input, designed to enhance recognition performance under occlusion and low-texture conditions. Built upon a Vision Transformer backbone, MaskNet adopts a dual-branch architecture for RGB and depth modalities and integrates multi-modal features using an attention-based fusion module. Further, spatial and channel attention mechanisms are employed to refine feature representation and improve instance-level discrimination. The segmentation outputs are used in conjunction with regional depth to optimize the grasping sequence. Experimental evaluations on camshaft depalletizing tasks demonstrate that MaskNet achieves a precision of 0.980, a recall of 0.971, and an F1-score of 0.975, outperforming a YOLO11-based baseline. In an actual scenario, with a self-designed flexible magnetic gripper, the system maintains a maximum grasping error of 9.85 mm and a 98% task success rate across multiple camshaft types. These results validate the effectiveness of MaskNet in enabling fine-grained perception for robotic manipulation in cluttered, real-world scenarios.

Fiber- reinforced plastics (FRP) are extensively employed in the fabrication of mannequins. However, the manual polishing process of FRP is characterized by a deficiency in efficiency and poses detrimental effects on the workforce. An industrial robot combined with an online measurement method is proposed for polishing the surface of the mannequin arm made of FPR material. In this paper, the factors affecting the removal depth of FRP materials are discussed by establishing a mathematical model. Based on the simulation analysis, the removal depth and the range of machining parameters are derived. The online measuring device manufactured by Keyence model LJ-X8200 is mounted on the flange of an ABB IRB6700 industrial robot, which obtains the 3D digital data from the arm surface before and after the polishing process. A sub-regional hand-eye calibration is applied to establish the spatial transformation matrix from the sensor coordinate to the workpiece coordinate to refine the calibration range and accuracy. Then, the processing data points are extracted by equal interval and coplanar method. In addition, the robot processing posture is obtained from the surface fitting curve based on the least squares method. Finally, the polishing processing experiments are performed to analyze robot intelligent location recognition and verify the surface removal effect based on the optimum polishing parameters. Comparing the 3D measuring data before and after the polishing process, it is shown that the average value of workpiece surface roughness has been reduced to 2.518 µm.

In the application of vision measurement, the black light-absorbing object is difficult to reflect the structured light from infrared emitter of the RGB-D camera. Therefore, an image recognition algorithm based on reference environment information is proposed to acquire the spatial positioning information of black volutes in the depalletizing system. The hardware system of the depalletizing system is mainly constructed of an upper computer, a six-axis industrial robot, an RGB-D camera and an end adsorption device. Firstly, the horizontal position information of each volute placed on the cardboard is obtained by the depth differences between the cardboard and the volute. Then, the depth information of the volute is obtained by the upper cardboard depth through collecting the position of the end vacuum suction cup triggered by feedback signal from vacuum generator. Secondly, a regional planar hand-eye calibration method is developed to improve the calibration accuracy in two-dimensional coordinates. The regional calibration method divides the robot working area into four regions: upper left, lower left, upper right, and lower right. The transformation matrix of each region is calculated separately. Finally, the depalletizing experiment is conducted on the three types of volutes. It is concluded that the average positioning error of the grasping center point of each volute obtained by our method is 3.795 mm, and its standard deviation is 1.769 mm. The average value of regional planar hand-eye calibration error is 4.044 mm, and its standard deviation is 1.501 mm. Under a stack of materials with dimensions of 1350 mm x 1350 mm x 1500 mm, the maximum error is controlled within 15 mm. Additionally, when combined with the end feedback compensation mechanism, the success rate for grasping all three volutes reaches 100%.

A new calibration device using a double-ball rotary structure and three contact displacement sensors is proposed to measure the spatial errors of six-axis serial robot. The error model is divided into kinematics part and static stiffness part, established by Modified DH (MDH) model and virtual joint method. The MDH parameter errors are compensated by modifying the angles of each axis with Newton Raphson compensation method and ridge estimation. The stiffness parameters are also identified by the least square method. The self-designed double-ball rotary device is calibrated by coordinate measuring machine (CMM), while the three-point measuring device is calibrated by gauge block. Then, the kinematic and static stiffness calibration and compensation experiments are carried out on EPSON C4 A901 robot in turn. The spatial error under none load and 1.0 kg, 1.5 kg, 2.0 kg gravity load are measured separately. The experiment results show that the motion accuracy of the robot is improved to the range from 0.0163 mm to 0.0236 mm under the gravity load conditions between weights of 1.0 kg and 2.0 kg. Finally, by comparing the measuring results from the laser displacement sensor and laser interferometer, the measurement accuracy of each axis of the self-designed device is equivalent to the laser displacement sensor and the laser interferometer in micron range.

The sphericity error is one of the most important tolerances to evaluate the dimensional accuracy of bearing balls. A continuous dynamic measurement system based on an industrial camera and a Z-shaped transparent track is developed to measure the sphericity error of bearing balls. The Z-shaped track, which is considered for the parameter setting of length, width and inclination, is designed to achieve a pure rolling of the ball around three perpendicular axes. During the whole measurement process, 15 images are dynamically captured by an industrial camera and divided into three groups according to the three mutually perpendicular directions. Then, a sub-pixel edge contour of the bearing ball from each image is extracted by an image-processing step, such as image de-noising, contrast enhancement, region-of-interest extraction and edge detection. Finally, the roundness error and sphericity error, as defined by ISO 12181-1 and ISO 3290-1, are acquired from the extracted contour of each image. Three different diameter bearing balls with a precision grade of G1000, representing a sphericity error of approximately 25 μm, are applied as experimental objects. The two times standard deviation roundness error obtained from the contour of each image are all less than 7 μm after six repeated measurements. In addition, the two measurement results respectively acquired from the proposed method and a commercial projection measuring instrument have good consistency.

航空发动机叶片三维面形数据重建是评价发动机叶片加工精度的重要手段.提出一种基于增强特征信息的双目视觉三维重建方法:首先在发动机叶片表面张贴圆形标记以增强叶片表面特征信息;其次通过相机的多角度拍摄获得能够覆盖叶片全貌的图片,并利用圆心特征匹配算法实现左、右图片中对应的圆形标记点的匹配;最后利用双目视觉三维重建原理计算得到三维点云数据,从而重构发动机叶片面形.对重构后的发动机叶片三维数据与白光扫描设备(精度0.05 mm)所得的扫描数据进行对比可得,发动机单叶片的叶背、叶盆的偏差平均值分别为0.1032和-0.1014 mm,标准偏差分别为0.0966和0.0571mm.

Aluminium/magnesium hybrid structure has been paid increasing attention recently due to its excellent lightweight. However, aluminium/magnesium is easy to form pores, cracks, and massive intermetallic compounds when fusion welding is used. Friction stir welding is as solid-state method, which avoids the above-mentioned problems caused by fusion welding. This article systematically reviews recent progress in friction stir welding of aluminium/magnesium from temperature field, metal flow and weld composition. Moreover, the effects of control parameters on the temperature field, metal flow and weld composition are discussed objectively, respectively. Finally, this work puts forward some problems that need to be solved urgently in the future based on summarising the recent progress of friction stir welding of aluminium/magnesium.

Small-scale aspheric optical elements have an urgent market demand and a wide range of applications. With the development of science and technology and the increasing requirements on product quality, the polishing technology of small size aspheric optical elements is becoming much more important in the field of ultra-precision machining. This paper first gave a brief introduction of the commonly used material for small size aspherical optical lenses and molds. Then, the applicable polishing technologies and their development status were introduced in detail, which included the computer controlled optical surface (CCOS), abrasive jet polishing (AJP), magnetorheological finishing (MF), ion beam polishing (IBP), bonnet polishing (BP), chemical mechanical polishing (CMP), shear-thickening polishing (STP), laser polishing (LP), and several kinds of compound polishing technologies. Finally, the development of polishing technology for small size aspheric optical components was summarized and prospected.

A sphere precessions polishing (SPP) method is proposed to perform the precessions polishing function like the bonnet polishing technique. A compliant inflated hollow ball is adopted as the sphere polishing tool, and three motors are used to collaboratively drive the sphere tool rotating around a desired axis tilting with a precessions angle. A prototype is developed to validate the technical feasibility and demonstrate the polishing removal functions. The polishing spots of vertical polishing show a typical W shape-like profile, and tilted polishing results reveal a D shape-like profile. This proposed SPP method can be regarded as a potential candidate technique to realize the precision precessions polishing in the optical precision engineering. (C) 2021 Society of Photo-Optical Instrumentation Engineers (SPIE)

针对大口径光学元件的超精密加工需求,在自主开发的MK7160和2MK1760两台平面磨床的研制基础上,进一步研发了高精度平面磨床UPG80.与现有大多数大口径平面磨床及MK7160、2MK1760磨床相比,UPG80应用了更先进的关键配套工艺技术,包含液体静压支承技术、砂轮修整技术及智能监测系统等.同时,对磨床UPG80的静态精度进行了检测,建立了三坐标轴误差模型对磨床进行误差补偿.在此基础上,采用金刚石圆弧砂轮对530mm×530mm口径的光学非球面元件进行加工,获得了良好的加工效果,验证了磨床的可靠性.

A new calibration method is proposed to improve the circular plane kinematic accuracy of industrial robot by using dynamic measurement of double ball bar (DBB). The kinematic model of robot is established by the MDH (Modified Denavit-Hartenberg) method. The error mapping relationship between the motion error of end-effector and the kinematic parameter error of each axis is calculated through the Jacobian iterative method. In order to identify the validity of the MDH parameter errors, distance errors and angle errors of each joint axis were simulated by three orders of magnitude respectively. After multiple iterations, the average value of kinematic error modulus of end-effector was reduced to nanometer range. Experiments were conducted on an industrial robot (EPSON C4 A901) in the working space of 180 mm x 490 mm. Due to the measuring radius of DBB, the working space was divided into 30 sub-planes to measure the roundness error before and after compensation. The average roundness error calibrated by the proposed method at multi-planes decreased about 21.4%, from 0.4637 mm to 0.3644 mm, while the standard deviation of roundness error was reduced from 0.0720 mm to 0.0656 mm. In addition, by comparing the results of positioning error measured by the laser interferometer before and after calibration, the range values of motion errors of end-effector were decreasing by 0.1033 mm and 0.0730 mm on the X and Y axes, respectively.

Grinding is a typical machining process which utilizes a grinding wheel consisting of abrasive grains held together with a binder material. The material removal action is done by the masses of tiny and sharp abrasives directly. Therefore, the good condition of grinding abrasives and wheel is very important to achieve good grinding performance. Thus, it is necessary to perform the conditioning (including both the dressing and truing processes) of grinding wheel. Generally, it is relatively simple to achieve the conditioning of conventional grinding wheel. But it is not the case for the superabrasive and wheels, which some novel dressing and truing methods and equipment are needed and decisive for the high-quality conditioning of grinding wheel. In this chapter, the common abrasives and bonds used in a grinding process are roughly introduced to give a general background firstly. Then some typical conditioning methods and novel techniques for precision machining are presented, which are mechanical dressing (cup-wheel dressing), laser dressing (LD), electrical discharge dressing (EDD), and electrolytic in-process dressing (ELID), as this can provide an impressive basis for comprehension of the conditioning and preparation of a superabrasive grinding wheel in precision machining.

以大口径光学元件精密与超精密磨削为监控对象,搭建磨削机床智能监测系统,探索智能磨削监控共性技术.监测系统以NI-PXI为主要运行硬件平台,建立与数控系统的通信,以机床内部、内置和外置传感器相结合的方式,自动获取机床运行过程和磨削加工过程的重要动态过程信号以及其他相关数据,通过对磨削机床全生命过程延伸数据体系的管理与分析,实现机床热平衡监测与热误差补偿、磨削液循环监控以及砂轮磨削性能退化在线评估等目标,以智能监控方式确保磨削机床长期平稳运行和加工质量稳定.

文章介绍了取力器圆锥滚子轴承结构及轴承间隙及传统的调整方法,针对现有取力器锥轴承间隙调整的不足之处,根据三点检测法设计了一种新的取力器锥轴承间隙调整检测量表.锥轴承间隙调整检测量具的使用保证了取力器输出轴轴承间隙量化,不仅满足设计、质量要求,提高了检测结果重复性、稳定性、精确度,还解决了传统轴承间隙测量方法不能满足装配节拍的问题.