基于微粒群算法的叶片曲面形状误差评定

Form Error Evaluation of Cured Blade Surface Based on Particle Swarm Optimization

以涡轮机叶片型面的形状误差评定为例,利用NURBS曲线插值构造出截面设计曲线,提出一种四控制点法构造与测量点最近的NURBS截面设计曲线,建立了计算曲面形状误差的数学模型,并应用微粒群算法计算测量点到曲面的最短距离,实现了曲面形状误差的评定。通过与传统的BFGS和DFP优化方法的计算结果进行比较,表明该方法能快速准确地计算叶片曲面的形状误差。

The form error evaluation of cured surface of turbine blade was investigated. The section curves were constructed as NURBS by interpolating design points. A new four control points algorithm for finding the least distance between measure points and design curve was presented. The mathematical model of surface form error was built. The particle swarm optimization （PSO） was applied for solving the minimum distance between measure points and design surface. Comparing with BFGS and DFP, the experiment results show the proposed method can rapidly and accurately calculate the surface form error.