基于点云数据的自由曲面加工误差评定

Machining Error Evaluation of Free-Form Surface Based on Point Cloudy Data

针对自由曲面加工误差检测困难的问题,提出了一种自由曲面加工误差评定方法。该方法以高精度蓝光扫描仪获取到的点云作为实测数据,以自由曲面的立体印刷模型作为理论模型。首先设置动态偏差阈值去除点云的远距离噪声,利用基于曲面变化度的局部离群系数方法去除点云的近距离噪声;然后采用最近点迭代算法分离点云位置误差的旋转量,以自适应粒子群算法依据最小区域原则分离位置误差的平移量;最后对点云进行基准变换后,计算点云与模型之间的偏差。实验结果表明,相对于最近点迭代算法,所提方法计算的位置误差向量均方根误差减小了66%,轮廓度减小了18%。提出方法的位置误差分离精度更高,轮廓度误差评定更准确,适用于扫描点云对自由曲面的加工误差评价,可以有效判断出超差区域。

Aiming at the difficulty in detecting the machining error of free-form surface blade,a method for evaluating the machining error of free-form surfaces is proposed,where the point cloud is obtained by the high-precision blue-light scanner as the measured data,and the STL model of the free-form surface is taken as the theoretical one.The dynamic deviation threshold is set to remove the far outliers of point cloud,and the near outliers of the point cloud are removed with the SVLOF method.Then the iterative closest points algorithm is used to separate the rotation of the point cloud position error,and the adaptive particle swarm algorithm is used to separate the translation amount according to the minimum region principle.The deviation between the point cloud and the model is calculated after the reference transformation of the point cloud.The experimental results show that the root mean square error of the position error vector evaluated with the proposed method is reduced by 66%and the profile error is reduced by 18%compared with the iterative closest points algorithm.The position error separation accuracy of this method is higher,and the contour error evaluation is more accurate.The proposed method is suitable for the processing error evaluation of the free-surface by the scanning point cloud,and can effectively judge the out-of-tolerance area.