【摘要】 航空发动机叶片点云数据三维重建是焊接修复模型构建的关键步骤,其重建精度直接决定了焊接修复精度。鉴于航空发动机叶片是空间自由曲面,其高度扭曲性以及区域分段函数的特点使得拟合困难,现有的最小二乘法存在精度差、光顺性差以及拟合的曲面不易调整等缺点。为此提出了一种基于移动最小二乘法(MLS)的曲面重建方法。由于移动最小二乘"紧支性"的存在,能够更加适合航空发动机叶片是空间自由曲面的特点;而基函数的阶数与权函数的提出,使得拟合曲面能有很高的光顺性,也使得拟合得到的曲面更容易调整。仿真结果表明:与传统的最小二乘以及正交最小二乘法相比,航空发动机叶片拟合精度提高。故移动最小二乘法更加适合航空发动机叶片三维重建。更多还原

【Abstract】 Three-dimentional reconstruction of the point cloud data is the key step in the construction of the welding repair model,and the precision of the reconstruction directly determines the precision of welding repair. In view of the fact that the aero engine blade is a free-form surface,the distortion of the height and the characteristic of the regional subsection function make the fitting difficult. The existing least squares method has many disadvantages,such as poor accuracy,poor smoothness and difficult to adjust the surface. A method of surface reconstruction based on moving least squares( MLS) is proposed. Because it is the characteristic of space free surface that moving least squares "compactness"can be more suitable for aero engine blade,and the order of the basis function and weight function is proposed,which have a very high fitting surface smoothness that makes the fitting easier to adjust. The simulation results show that compared with the traditional least squares method and the orthogonal least square method,the accuracy of aeroengine blade fitting is improved. The moving least squares method is more suitable for 3 D reconstruction of aeroengine blade.更多还原