微V槽超精密机床几何误差建模及补偿算法的研究

发布时间：2018-04-01 05:30

  本文选题：微V槽超精密机床　切入点：垂直度误差　出处：《广东工业大学》2015年博士论文

【摘要】：光电子、微电子器件中广泛应用微V槽阵列结构,如光纤连接器、波分复用器、LCD背光板等。这些元器件一般需采用专用超精密机床加工,使其达到亚微米级尺寸加工精度和纳米级表面粗糙度,以满足光学级功能要求。基于自主研发的微V槽超精密机床,本文围绕着通过几何误差补偿来提高其加工精度这个主题,全面分析机床的结构和其加工工艺特征,对机床垂直度误差模型、几何误差测量与辨识方法、单项运动副误差多项式拟合函数的优化、几何误差建模及补偿算法等几个关键问题开展深入的理论和实验研究。精密的误差模型是实现误差精密补偿的关键。本文比较分析传统基于小角度误差假设的垂直度误差变换矩阵的不足,推导出改进后的垂直度误差模型,基于现有的误差测量与辨识方法,对垂直度误差变换矩阵改进中涉及的垂直度误差角实施精算；针对微V槽超精密机床XFYZ型结构特征,基于多体系统理论和前述的垂直度误差改进模型,建立精密的几何误差模型；对比垂直度误差精算前后的几何误差模型以检测垂直度误差变换矩阵改进的效果。数控机床几何误差的测量与辨识是一项复杂且费时的工作。如何快速精密地辨识出各单项几何误差一直是几何误差补偿研究的重要课题。本文在分析目前常用的9线几何误差辨识法的基础上,推导高精度的测量效率更快的6线几何误差辨识法,并用实验检测该方法的可靠性。为便于几何误差补偿技术的实施就需要确定机床轴系运动到空间某坐标处各单项运动副误差的具体值,这就要求将各单项运动副误差拟合成关于机床运动坐标的函数。本文详细分析多项式拟合法的优点和不足,并基于统计学原理,推导出多项式拟合的优化算法,通过对机床几何误差测量数据的处理,验证该算法的优化效果。几何误差的精密补偿是垂直度误差的精算、单项几何误差的测量与辨识、单项运动副误差的多项式拟合及其优化等所有工作的最终目的。本文简单介绍数控修正指令的直接计算方法,并分析该方法的不足；推导出较优的数控修正指令的附加指令算法,并通过实例演算检测其修正效果。
[Abstract]:MicroV-slot arrays are widely used in optoelectronic and microelectronic devices, such as optical fiber connectors, wavelength division multiplexers (WDM) and LCD backlight panels.These components need to be machined with special ultra-precision machine tools to achieve sub-micron size machining accuracy and nanometer surface roughness to meet the requirements of optical function.Based on the micro-V-slot ultra-precision machine tool developed by ourselves, this paper focuses on the topic of geometric error compensation to improve the machining accuracy. The structure of the machine tool and its processing technology characteristics are analyzed in an all-round way, and the verticality error model of the machine tool is analyzed.Some key problems, such as geometric error measurement and identification method, optimization of polynomial fitting function of single motion pair error, geometric error modeling and compensation algorithm, are studied deeply in theory and experiment.Precision error model is the key to realize precision error compensation.This paper compares and analyzes the shortcomings of the traditional perpendicularity error transformation matrix based on the assumption of small angle error, and deduces the improved verticality error model, based on the existing methods of error measurement and identification.The perpendicularity error angle involved in the improvement of perpendicularity error transformation matrix is actualized, and according to the XFYZ structural characteristics of micro-V-slot ultra-precision machine tool, based on the theory of multi-body system and the improved model of perpendicularity error mentioned above,A precise geometric error model is established, and the geometric error model before and after the perpendicularity error actuarial is compared to detect the improvement effect of the perpendicularity error transformation matrix.The measurement and identification of geometric errors of NC machine tools is a complicated and time-consuming task.How to identify geometric errors quickly and accurately has always been an important topic in the research of geometric error compensation.Based on the analysis of the commonly used 9-line geometric error identification method, this paper deduces the 6-line geometric error identification method with high accuracy and faster measurement efficiency, and tests the reliability of the method by experiment.In order to facilitate the implementation of geometric error compensation technology, it is necessary to determine the specific value of each single motion pair error from the machine tool shafting to a space coordinate, which requires that each single motion pair error be synthesized into a function about the machine tool motion coordinate.In this paper, the advantages and disadvantages of polynomial fitting are analyzed in detail. Based on the principle of statistics, the optimization algorithm of polynomial fitting is derived. The optimization effect of the algorithm is verified by processing the measuring data of geometric error of machine tools.The precision compensation of geometric error is the ultimate purpose of all the works, such as the actuarial of perpendicularity error, the measurement and identification of single geometric error, the polynomial fitting and optimization of single motion pair error, etc.This paper briefly introduces the direct calculation method of numerical control correction instruction, and analyzes its shortcomings, deduces the better additional instruction algorithm of numerical control correction instruction, and detects its correction effect by example calculus.
【学位授予单位】：广东工业大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TG502
，

本文编号：1694205