基于核函数的三坐标空间测量误差估计

Error Estimation of Coordinate Measuring Machine Based on Kennel Function

由于三坐标测量过程中涉及多种测量误差因素,因此在快速测量过程中对误差特性进行详细分析就比较复杂。本文以移动桥式三坐标测量机为研究对象,运用基于非参数回归模型中的核函数估计法对手动三坐标测量机的空间测量误差进行了多元非线性建模和预测。通过对采集数据进行回归拟合,并与参数建模的拟合结果进行对比分析后可知,非参数回归模型中的核函数估计法具有更好的预测效果。

Since the error is affected by various factors in the process of CMM measuring,the detailed error characteristic analysis in fast measurement can be complicated. Taking a moving bridge CMM as the research object,and based on the kennel function estimation in non-parametric regression,the multivariable nonlinear modeling and estimation of manual CMM measuring errors are carried out. By comparing the fitting and regression results from the data acquisition and the parametric modeling,the kennel function estimation method shows a better result.