空间统计方法在自由曲面加工误差分解中的应用

为了便于对自由曲面的加工误差进行分析，应用空间统计方法分析加工误差数据，将构成加工误差的系统误差和随机误差这两部分分解出来。该方法首先对加工误差进行空间统计分析，判断加工误差的空间自相关性，然后构造基于B样条曲面的确定性曲面回归模型，计算各个测点的残差后对该回归模型进行充分性分析，将服从空间独立分布的残差作为分解后的随机误差，进而得到系统误差。针对一个自由曲面进行仿真验证。结果表明，该误差分解方法精确、有效。再对上述自由曲面进行数控加工，并获得该工艺系统的系统误差，根据该系统误差进行补偿加工，显著提高了零件的加工精度。 

Application of Spatial Statistical Analysis in Decomposition of Machining Errors for Free-Form Surfaces

To analyze machining errors of free-form surfaces, a spatial statistical analysis method was presented for the machining errors decomposition, forming systematic errors and random errors. Firstly, the machining errors were analyzed with the method to judge the spatial autocorrelation of these errors. Then a regression model was built based on B-spline surfaces for the deterministic surface. And the adequacy of the regression model was analyzed through calculating the residual errors of all the measured points. Once the residual errors were of spatial independence, the residual errors could be treated as random errors, and the systematic errors were easily obtained. A free-form surface simulation was taken to validate the method. Results show that the error decomposition is accurate and effective. According to the systematic errors of the process system obtained in the surface machining on a NC machining center, the compensation made for the machining error has significantly improved the machining accuracy of free-form surface.