A study on automatic on-machine inspection system for 3D modeling and measurement of cutting tools

A 3D model of the maximum rotating envelope of a milling cutter with tool holder is required for Computer Aided Manufacturing (CAM) process design and machining simulation. The user may define the 3D model of the whole tool assembly in the tool library of CAM software. However, it is not convenient and reliable. Considering these problems, a new method based on single view 3D reconstruction algorithm has been proposed in previous research work, which is able to quickly reconstruct the 3D model of a cutter with tool holder while they are installed onto the spindle. As the extension of this work, this paper focuses on the recent progresses in order to improve the automation, accuracy, efficiency and reliability of tool modeling system. First, an improved flexible on-machine camera calibration procedure is proposed. The accurate motion of machine tool axis is used to calibrate the camera on machine tool instead of a physical calibration board. The whole procedure of calibration can be conducted automatically by running NC code. Therefore, the automation of vision system can be guaranteed. Second, the contour extraction module is improved by using a method of silhouette image composition. This method is applied to solve the problem of translucent and fuzzy cutter profile induced by motion blur. Third, the new algorithm for contour partitioning and classification are proposed, which is more reliable and robust. The reliability and accuracy of the vision system can be guaranteed. Finally, the vision system with an 8 mm lens and 1 mm extensions has been tested on different type of machine tool with smaller cutters. The average measurement accuracy is about 35 microns verified by comparison with a commercial tool setting system.     

数字图象离散线性变换的点辖域方法

数字图象的线性变换在计算机视觉图象处理，模式识别和计算机图形学等领域有着广泛的应用。在实数域中这种变换可以提供连续的一一映射。然而，由于图象象素是离散的，常导致变换结果出现不如意的孔洞，破坏了变换图象的质量，给进一步的处理和识别过程带来困难。     

Efficient local transformation estimation using Lie operators

Conventional translation-only motion estimation algorithms cannot cope with transformations of objects such as scaling, rotations and deformations. Motion models characterizing non-translation motions are thus beneficial as they offer more accurate motion estimation and compensation. In this paper, we introduce low-complexity transformation estimation methods with four motion models based on Lie operators, which are linear operators that have found applications in optical character recognitions. We show that individual Lie operators are capable of capturing small degrees of object transformations. We propose an efficient local transformation estimation algorithm in order to further improve the accuracy of the translation-only estimation by integrating all four motion models. Simulations with an MPEG-2 video codec on two video sequences show that the proposed transformation estimation approach can noticeably improve the motion compensation performance of the translation-only method by achieving higher PSNR (peak signal-to-noise ratio) values for the predicted frames, with only a small fraction of the complexity required by the translation motion search.     

Design and implementation of a linear jerk filter for a computerized numerical controller

The proposed filter is placed in front a buffered digital differential analyzer, and is formed by combining three modified moving average filters. It can export the position commands to ensure smooth and accurate motion of a tool with a linear jerk change. These output commands can guarantee contour accuracy despite the error in the chord height. The acceleration and jerk can be designed simply by specifying the number of registers. The filter can be implemented using three circular buffers to simplify the arithmetic and reduce the computation time. The high precision motion commands are confirmed by installing the filtering algorithm in a digital signal processor of a computerized numerical controller. The radius of the command trajectory does not become distorted at high speed 30 m/min. The commands filtered by the linear jerk filter stabilize the beginning and end of the actual motion of the machine table.     

Structural design,numerical simulation and control system of a machine tool for stranded wire helical springs

A stranded wire helical spring (SWHS) is a unique cylindrically helical spring, which is reeled by a strand that is formed of 2∼16 wires. In this paper, a parametric modeling method and the corresponding 3D model of a closed-end SWHS are presented based on the forming principle of the spring. By utilizing a PC + PLC based model as the motion control system, a prototype machine tool is designed and constructed, which improves the manufacturing of the SWHS. Via the commercial CAD package Pro/Engineering, numerical simulation is carried out to test the validity of the parametric modeling method and the performance of the machine tool. The scheme of the tension control system is analyzed and the control mechanism is set up, which have achieved the constant tension of each wire. A human-machine interface is also proposed to achieve the motion control and the tension control. Experimental results show that the tension control system is well-qualified with high control precision.     

An application of Bertrand curves and surfaces to CADCAM

A method is proposed for generating tool centre paths with an accuracy equal to the resolution of the machine tool. The method exploits the properties of Bertrand curves and surfaces, to generate tool motion which secures the desired degree of continuity in the machined contour or surface. It is shown that for cubic splines or bicubic patches, accurate tool motion can be generated by a fast DDA-like (Digital Differential Analyser) digital integration scheme, which reduces integration to successive additions. Accuracy is achieved by eliminating the integration error inherent in DDA.     

An Analytical Method for Decoupled Local Smoothing of Linear Paths in Industrial Robots

The linear-format path is widely adopted to approximate the continuous contour in robot controllers. The tangential discontinuity of the linear paths usually causes the discontinuity of the joint velocity. To comply with the joint kinematics limits, the robots have to stop at each corner point, resulting in a great loss of efficiency. To achieve a smooth motion, this paper presents an analytical decoupled C³ continuous local path smoothing method for industrial robots. The tool position path is smoothed in the reference frame while the tool orientation is smoothed in the rotation parametric space based on the exponential coordinates of rotations. The quintic B-splines are inserted at the corners of the linear segments to achieve the G³ continuity of the tool position path and tool orientation path. The orientation smoothing error is constrained analytically. By reparameterization of the remaining linear segments using specially constructed B-splines, the C³ continuity of the tool position path and tool orientation path is achieved. Then, the synchronization of the tool orientation path and tool position path can be guaranteed by sharing the same curve parameter. Besides, to improve the smoothness of the angular motion on the remaining linear segments during parameter synchronization, the transition lengths of the inserted B-splines are optimized. The proposed local smoothing method guarantees that the generated smooth orientation path in the rotation space is invariant with the selection of the reference frame and the orientation of the tool frame, and ensures the jerk-continuous motion with a smoother angular motion on the remaining linear segments, which could improve the motion smoothness of the tool path and tracking accuracy. The effectiveness of the proposed method is validated through simulation and experiments.     

Matrix representations,linear transformations,and kernels for disambiguation in natural language

In the application of machine learning methods with natural language inputs, the words and their positions in the input text are some of the most important features. In this article, we introduce a framework based on a word-position matrix representation of text, linear feature transformations of the word-position matrices, and kernel functions constructed from the transformations. We consider two categories of transformations, one based on word similarities and the second on their positions, which can be applied simultaneously in the framework in an elegant way. We show how word and positional similarities obtained by applying previously proposed techniques, such as latent semantic analysis, can be incorporated as transformations in the framework. We also introduce novel ways to determine word and positional similarities. We further present efficient algorithms for computing kernel functions incorporating the transformations on the word-position matrices, and, more importantly, introduce a highly efficient method for prediction. The framework is particularly suitable to natural language disambiguation tasks where the aim is to select for a single word a particular property from a set of candidates based on the context of the word. We demonstrate the applicability of the framework to this type of tasks using context-sensitive spelling error correction on the Reuters News corpus as a model problem.     

Systematic geometric rigid body error identification of 5-axis milling machines

A 5-axis milling machine has 39 independent geometric error components when the machine tool is considered as a set of five rigid bodies. The identification of the deterministic component of the systematic error is very important. It permits one to improve the accuracy close to the repeatability of the machine tool. This paper gives a new way to identify and compensate all the systematic angular errors separately and then use them further to identify the systematic translational error.Identification based on a new mathematical method and a stable numerical solution method is proposed. The model explains from first principles why some error components have no effect in a first order model. The identification of the total angular systematic errors can be done independently from the translation errors. However, the total translation error depends on the angular errors and the translation errors of each machine tool slide. The main problems solved are to find enough linear independent equations and avoid numerical instability in the computation. It is important to separate numerical problems and linear dependence. The very complex equations are first analyzed in symbolic form to eliminate the linear dependencies. The total of linear independent components in the model is reduced from 30 to 26 for the position dependent errors and from 9 to 3 for the position independent components. Secondly, the large system of linear equations is broken down in many smaller systems. The model is tested first with simulated errors modeled as cubic polynomials. An artifact-based identification is proposed and implemented based on drilling holes in various locations and orientations. New ways to measure the volumetric error directly are proposed. Direct measurement of the total volumetric error requires considerably less measurement than measuring all 6 components of each machine slide especially in the case of a 5-axis machine.     

直线伺服6 位置环虚拟轴机床的运动学解耦

使用直线电机驱动虚拟轴机床 ,可满足高速时动态响应和提高精度的要求。如何解决 6位置环虚拟轴机床的强耦合 ,一直是虚拟轴机床伺服控制中的一个难题。为此 ,将局部结构化方法与解耦理论相结合 ,提出一种新的解耦算法 ,从理论上实现了对 6位置环虚拟轴机床的平动和转动之间的解耦     

Toward Machine Learning Through Genetic Code-like Transformations

The gene expression process in nature involves several representation transformations of the genome. Translation is one among them; it constructs the amino acid sequence in proteins from the nucleic acid-based mRNA sequence. Translation is defined by a code book, known as the universal genetic code. This paper explores the role of genetic code and similar representation transformations for enhancing the performance of inductive machine learning algorithms. It considers an abstract model of genetic code-like transformations (GCTs) introduced elsewhere [21] and develops the notion of randomized GCTs. It shows that randomized GCTs can construct a representation of the learning problem where the mean-square-error surface is almost convex quadratic and therefore easier to minimize. It considers the functionally complete Fourier representation of Boolean functions to analyze this effect of such representation transformations. It offers experimental results to substantiate this claim. It shows that a linear classifier like the Perceptron [38] can learn non-linear XOR and DNF functions using a gradient-descent algorithm in a representation constructed by randomized GCTs. The paper also discusses the immediate challenges that must be solved before the proposed technique can be used as a viable approach for representation construction in machine learning.     

A configuration independent error model of machine tools: Hyperpatch model and metrology pallet

A machine tool error model based on a three-dimensional hyperpatch error map was developed to describe the machine tool error field over a workspace. The error map was generated by a set of measurement points without using a detailed relationship among joints and linkages of a machine tool. The modeling is based on the motion that an ideal workspace is distorted to an actual workspace through a prescribed parametric representation. The concept was implemented on a three-axis CNC machining center. A metrology pallet was designed to serve as a reference coordinate frame of the workspace. A touch trigger probe was used to collect the required data. Experimental results show that this error model is capable of modelling the warm machine tool distortion behavior. The proposed model is based on the observed phenomena rather than the structural configuration of machine tools. Therefore, the advantage of this approach is that only a few measurements are required to construct the prediction model for a variety of machine tool configurations and for working environments.     

Design and implementation of look-ahead linear jerk filter for a computerized numerical controlled machine

In this study, an open programmable logic controller (PLC) from Fuji electric prescribed in the Structured Text program was applied to develop a look-ahead linear jerk filter (LALJF) for a computerized numerically controlled (CNC) machine. To ensure the smooth and accurate motion of a tool with a linear change in jerk during real-time machining, the proposed filter was formed by combining a look-ahead algorithm with three modified moving average filters (3MMAF). The look-ahead algorithm performed a single look-ahead step-change in the speed of the speed curve. Based on a step-changing speed profile, given maximal acceleration/deceleration and maximum jerk, the speed curve was modified before it was passed through a linear jerk filter to reduce machining time. The speed commands filtered by the proposed filter stabilize machine table at the beginning and end of its motion, and at any point at which its speed exhibits a step-change. The theoretical and computational aspects of the LALJF are presented together with experimental results of its implementation on an X–Y table. The multiple-step-changing speed curve of a CNC machine and the speed curve of a high-speed measurement system were constructed in order to verify the feasibility and precision of the proposed method.     

基于性能计数事件的计算机功耗估算模型

精准快速获取计算机系统的实时功耗是功耗优化研究的基础,因此提出并建立了一种高精度的计算机功耗估算模型。通过分析统计系统运行时代表性的性能计数事件,应用机器学习理论分析性能事件与功耗的关系,建立多核计算机系统实时功耗估算模型。模型构建时使用多元线性回归(Multiple Linear Regression,MLR)方法以及支持向量回归(Support Vector Regression,SVR)方法分析二者关系,并对两种方法建立的功耗估算模型进行了对比分析。实验结果表明,基于性能事件的功耗估算模型可准确估计计算机实时功耗,估算误差不高于3%。同已有模型相比较,该估算模型精度更高、通用性更好。     

Digital twin modeling for loaded contact pattern-based grinding of spiral bevel gears

To distinguish with the conventional tooth flank grinding only considering geometric accuracy, an innovative digital twin modeling is proposed for loaded contact pattern based grinding of spiral bevel gears. Where, data-driven grinding simulation, sensitivity analysis strategy, adaptive decision and control are developed. Focusing on loaded contact pattern optimization, numerical loaded tooth contact analysis (NLTCA) considering noncentrosymmetric problem and tooth flank roughness is developed for data-driven relationship establishment. Then, an adaptive data-driven tooth flank grinding decision and control model is established. Where, the universal motion concept (UMC) machine settings is selected as the optimal design variable. It is actually an infinite approximation to the target tooth flank in form of an adaptive control system. Moreover, with point-to-point material removal distribution, the different optimization strategies are proposed for accurate tooth flank grinding. In particular, the overcutting problem on the tooth flank grinding programming is investigated. Finally, Levenberg-Marquardt method is applied to solve the established nonlinear lease square model for the accurate machine tool settings having modification variations. Thus, this accurate data-driven digital twin modeling can achieve loaded contact pattern-based grinding. The provided numerical and test instances can verify the proposed digital twin modeling.     

Pixel-wise video stabilization

In this paper, we present a novel video stabilization method with a pixel-wise motion model. In order to avoid distortion introduced by traditional feature points based motion models, we focus on constructing a more accurate model to capture the motion in videos. By taking advantage of dense optical flow, we can obtain the dense motion field between adjacent frames and set up a pixel-wise motion model which is accurate enough. Our method first estimates dense motion field between adjacent frames. A PatchMatch based dense motion field estimation algorithm is proposed. This algorithm is specially designed for similar video frames rather than arbitrary images to reach higher speed and better performance. Then, a simple and fast smoothing algorithm is performed to make the jittered motion stabilized. After that, we warp input frames using a weighted average algorithm to construct the output frames. Some pixels in output frames may be still empty after the warping step, so in the last step, these empty pixels are filled using a patch based image completion algorithm. We test our method on many challenging videos and demonstrate the accuracy of our model and the effectiveness of our method.     

Transfer matrix method for linear multibody system

A new method for linear hybrid multibody system dynamics is proposed in this paper. This method, named as transfer matrix method of linear multibody system (MSTMM), expands the advantages of the traditional transfer matrix method (TMM). The concepts of augmented eigenvector and equation of motion of linear hybrid multibody system are presented at first to find the orthogonality and to analyze the responses of the hybrid multibody system using modal method. If using this method, the global dynamics equation is not needed in the study of linear hybrid multibody system dynamics. The MSTMM has a small size of matrix and higher computational speed, and can be applied to linear multi-rigid-body system dynamics, linear multi-flexible-body system dynamics and linear hybrid multibody system dynamics. This method is simple, straightforward, practical, and provides a powerful tool for the study on linear hybrid multibody system dynamics. This method can be used in the following: (1) Solve the eigenvalue problem of linear hybrid multibody systems. (2) Obtain the orthogonality of eigenvectors of linear hybrid multibody systems. (3) Realize the accurate analysis of the dynamics response of linear hybrid multibody systems. (4) Find the connected parameters between bodies used in the computation of linear hybrid multibody systems. A practical engineering system is taken as an example of linear multi-rigid-flexible-body system, the dynamics model, the transfer equations and transfer matrices of various bodies and hinges; the overall transfer equation and overall transfer matrix of the system are developed. Numerical example shows that the results of the vibration characteristics and the response of the hybrid multibody system received by MSTMM and by experiment have good agreements. These validate the proposed method.     

基于运动能量模型的人体运动检索算法

提出一种新的借助于运动中产生的能量对运动进行描述的模型。引入运动协调性的概念,利用支持向量机和索引序列对运动进行粗分类。采用基于Keogh下界的线性索引算法精确地度量输入运动和候选动作之间的相似性。实验结果表明,该检索方法具有较好的速度和准确性。     

New fuzzy control model and dynamic output feedback parallel distributed compensation

A new fuzzy modeling based on fuzzy linear fractional transformations model is introduced. This new representation is shown to be a flexible tool for handling complicated nonlinear models. Particularly, the new fuzzy model provides an efficient and tractable way to handle the output feedback parallel distributed compensation problem. We demonstrate that this problem can be given a linear matrix inequality characterization and hence is immediately solvable through available semidefinite programming codes. The capabilities of the new fuzzy modeling is illustrated through numerical examples.     

An update method for digital twin multi-dimension models

Digital twin, as an effective means to realize the fusion between physical and virtual spaces, has attracted more and more attention in the past few years. Based on ultra-fidelity models, more accurate service, e.g. real-time monitoring and failure prediction, can be reached. Against the background, some scholars studied the related theories and methods on modeling to depict various features of physical objects. Some scholars studied how to use Internet of Things to realize the connections and interactions, thereby keeping the consistency between the virtual and physical spaces. During this process, a new question arises that how to update the models once digital twin models are inconsistent with the practical situations. To solve the problem, this paper proposed a general digital twin model update framework at first. Then, the update methods for multi-dimension models are further explored. The cutting tool is the core component of machine tools which are the key equipment in industry. The precise cutting tool models are essential for realizing the digitalization and servitization of machine tools. Therefore, this paper takes a cutting tool as the application object to discuss how to conduct physics model update based on the proposed framework and methods. Through model update, a more accurate and updated tool wear model could be obtained, which contributes to the prognostics and health management for machine tools.