基于遗传算法的三维空间柱面拟合  

Fitting of Spatial Cylindrical Surface Based on Genetic Algorithm

作　　者：秦世伟[1,2] 潘国荣[1] 谷川[1,3] 施贵刚[1] 

机构地区：[1]同济大学测量与国土信息工程系,上海200092 [2]上海大学土木工程系,上海200072 [3]上海市政工程设计研究总院,上海200092

出　　处：《同济大学学报（自然科学版）》2010年第4期604-607,618,共5页Journal of Tongji University:Natural Science

基　　金："十一五"国家科技支撑计划资助项目(2006BAJ03A07)

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.The definition of spatial cylindrical surfaces is introduced to the definition of corresponding planar curves. Fitting of the spatial cylindrical surface with genetic algorithm （GA） is proposed. 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 have to be identified in this method. Non- independent parameters are constrained. Objective function is built based on least squares method,and optimized by GA. The effectiveness and its advantages in fitting is testified with an engineering example.

关 键 词：空间柱面拟合 遗传算法 平面曲线 卷板辊 圆度 

分 类 号：P258[天文地球—测绘科学与技术]