基于遗传算法的三维空间柱面拟合

Fitting of Spatial Cylindrical Surface Based on Genetic Algorithm

摘要:

从二维平面曲线的定义出发，进而扩展至三维空间柱面的定义。提出采用遗传算法拟合得出空间柱面的参数，将空间柱面唯一确定。以空间圆柱面为例，说明采用遗传算法进行空间柱面拟合的原理以及方法。以空间曲面的表达参数作为需辨识的参数值，并且对相关的参数值加以约束，基于最小二乘原理建立优化目标函数，采用遗传算法进行进化寻优。实例研究表明本文方法的有效性与拟合精度优势。

Abstract:

The definition of spatial cylindrical surfaces is extended, amplifying on the definition of corresponding planar curves. Fitting of the spatial cylindrical surface using genetic algorithm is suggested in this paper. Taking the fitting of cylinder surface for example, the principle and method of spatial cylindrical surface fitting based on genetic algorithm is described. Express parameters are to be identified in this method. Non-independent parameters are constrained. Objective function is built based on least squares method. Optimization is achieved using GA. The effectiveness and advantage in fitting precision of the method proposed in this paper is testified using an engineering example.