重构模型精度的评定和产品改进

Precision Evaluation of Reconstruction Model and Product Improvement

以叶片CT扫描后的数据点云为研究对象,对其进行数据预处理、曲线拟合以及重构模型。针对重构模型采用了两种方式评定重构模型质量,第一种方式是以叶片重构模型与其CAD模型的均方误差、平均误差以及最大误差来评定总体重构精度。另一种方式就是以重构模型上的截面曲线与CAD模型上的对应理论曲线的均方误差、平均误差、最大误差、截面曲线周长和截面面积来评定重构模型局部重构质量。此外,在重构模型的过程中,对拟合的曲线进行等比例缩放,从而改进产品模型,得到一系列形状相同,尺寸不同的产品族。

The data of point cloud after CT scanning was preprocessed, section curves were fitted and the model was reconstructed. Two kinds of modes were used to assess the quality of reconstruction model. One is to evaluate the total reconstruction precision by mean square error, average error and maximal error between the blade＇s reconstruction model and its CAD model. The other is to evaluate the local reconstruction quality by mean square error, average error, maximal error, the perimeter and the area between the section curve on reconstruction model and the corresponding theory curve on CAD model. In addition, in the process of reconstructing model, fitted curves were zoomed according to the same scale, this can improve product model. A series of productions which have similar shape and different dimension were got.