【作者】 沈金华；

【导师】 杨建国；

【作者基本信息】 上海交通大学 ， 机械制造及其自动化， 2008， 博士

【摘要】 数控机床误差补偿技术是提高机床精度的一种经济而有效的方法和手段。基于高等学校全国优秀博士学位论文专项资金资助项目、云南省省院省校科技合作计划项目、《上海市引进技术的吸收与创新计划》等项目,本文研究了数控机床误差测量、建模和实时补偿等数控机床误差补偿的关键技术及其应用。本文的主要研究内容包括:(1)分析了数控机床存在的空间定位误差,利用齐次坐标变换方法,通过在床身基座上建立基坐标系(参考坐标系),在各轴(刚体)上分别建立子坐标系,根据刀具、主轴到机架之间的“刀具-机架”运动链和工件到机架之间的“工件-机架”运动链之间的变换关系,建立了四种不同类型三轴数控机床的空间定位误差综合模型,为机床误差测量和补偿提供了理论基础。(2)研究和分析了激光测量仪在直线、平面对角线和空间体对角线进行数控机床误差测量时,激光测量仪所获得的测量值同机床空间定位误差值之间的关系,着重介绍了机床工作空间体对角线测量方法的优点与不足。通过对三个坐标轴方向的误差进行矢量分解,在现有的体对角线方法的基础上,提出了一种改进的分步体对角线测量方法,这种方法通过分步测量机床工作空间的四条体对角线可以快速获得分解后得到的9项位置误差。进一步提出直接利用这9项位置误差进行误差补偿的思想,并通过实验进行了验证。这种方法既节省了测量和辨识的时间,又提高了补偿的效率。(3)研究和分析了机床空间定位误差同机床温度变化之间的对应关系。通过使用可以快速测量空间定位误差的分步体对角线测量方法,在不同的温度条件下对数控机床进行了误差的测量,得到了机床9项位置误差随温度变化的曲线。提出了对不同温度条件下的空间定位误差进行有效预测的直接插补、曲线拟合和神经网络等方法,并对这些方法进行了对比。对空间定位误差不同温度下的热变化分析和预测,使得空间定位误差的建模和补偿不但考虑了机床的位置变化而且考虑了温度变化,通过补偿实验证明这种方法有助于进一步提高机床补偿的精度。(4)研究和分析了机床热误差建模方法,提出了一种基于偏最小二乘回归神经网络进行热误差建模的方法,该方法通过引入神经网络的学习原理对传统偏最小二乘回归方法中的线性方法进行了改进,用以解决机床热误差建模中温度变量共线性和热误差非线性的问题。文中论述了这种方法的基本算法并在一台数控机床上通过实验数据对模型的预测和鲁棒性能进行了验证。(5)研究和分析了机床误差实时补偿实施方法,提出了一种基于数控机床外部坐标系偏移功能,可以对数控机床热误差和定位误差等进行实时补偿的装置。通过硬件和软件的设计,可以根据机床的温度和位置信息进行误差的实时计算和补偿。进一步,提出了一种针对热误差补偿的简化装置。误差补偿装置的研制为误差的实时补偿提供了有效的实施方法。(6)介绍了在与云南机床厂合作研究项目中对其生产的CY-K360型数控车床进行误差补偿实施以提高产品精度的应用实例。通过对机床定位误差和主轴热误差进行检测,并使用基于外部机床坐标系偏移的误差实时补偿装置对数控机床实施误差补偿后,机床精度得到了明显提高。更多还原

【Abstract】 The error compensation for CNC machine tools is an economical and effective method to improve the machine tool precision. In this dissertation, key technologies and their applications for error compensation such as error modeling, measurement and real time error compensation implementation are studied, based on the projects supported by“The Foundation for the Author of National Excellent Doctoral Dissertation of P.R. China”,“The cooperative plan on science and technology of Yunnan province”and“The assimilation and innovation plan for introduced technology of Shanghai”.The main contents of this dissertation can be described as follows:(1) The volumetric positioning error of CNC machine tools is analyzed. A reference coordinate system is created on the machine body,some sub-coordinate systems are set on each saddle, so the relationship between tool and workpiece can be expressed as the relationship between“tool-machine body”chain which is from tool to machine body and“workpiece-machine body”chain which is from workpiece to machine body, then the synthesis models of four types of 3-axis machine tool are derived by the homogeneous coordinate transformation method which provides a theory base for the error measurement and compensation.(2) The relationship between the values measured with laser interferometer along the line, face and body diagonal directions and positioning errors are studied and analyzed. The advantages and shortcomings of body diagonal direction in accuracy calibration are discussed. A new sequential step diagonal measurement method has been proposed based on the conventional diagonal measurement. In this method, the error in three axis directions can be divided 9 synthesis errors and they can be identified based on the values measured along four body diagonals. The idea of compensating the volumetric error directly based on these 9 synthesis errors are proposed and verified by the measurement and compensation experiment. With this method, the time for error measurement and identification is saved and the efficiency for error compensation is improved.(3) The relationship between the volumetric positioning errors and the temperature fields are studied and analyzed. Volumetric errors under different temperature conditions are measured by the sequential step diagonal measurement method, and the variation curves of the 9 synthesis errors are obtained. The methods for the prediction of volumetric errors under certain temperature condition such as simple interpolation, curve fitting and neural networks are proposed and compared. The error compensation effect is improved by analyzing and predicting the volumetric positioning errors considering not only the position variation but also the temperature variation. The compensation results show that this method can help to further improve the precision of the machine tool.(4) The modeling methods for thermal error modeling is studied and analyzed. A partial least squares neural network modeling method for CNC machine tool thermal errors is proposed. This method uses the neural network learning rule to obtain the PLS parameters instead of the traditional linear method in partial least squares regression so as to overcome the multicollinearity and nonlinearity problem in thermal error modeling. The basic principle of PLSNN is described and the prediction performance and robustness of the new method is validated by the thermal error modeling application on a CNC turning center.(5) The implementation method for real time error compensation is studied and analyzed. The function of the CNC system for external coordinate system shift is analyzed and a real time compensation system has been developed. With hardware and software design, the machine tool error can be calculated and compensated based on the external thermal and position signals. Furthermore, a simplified thermal error compensation device is also proposed. The development of this device provides an effective implementation method for real time error compensation.(6) The real time error compensation technique is implemented on a CY-K360 CNC lathe as a real case in the cooperative research project with CY Group to improve their manufactured machine tools.The positioning errors considering temperature variation and the spindle thermal drifts are measured, predicted and finally compensated by the real time compensation system based on the external coordinate system shift and the precision of the machine is improved significantly.更多还原