基于遗传算法和最速下降法的Bézier曲线拟合  

Bézier curve fitting based on genetic algorithm and steepest descent algorithm

作　　者：白向军[1] 彭国华[1] 陈晓[1] 

机构地区：[1]西北工业大学数学系,陕西西安710129

出　　处：《计算机工程与设计》2009年第1期194-196,共3页Computer Engineering and Design

基　　金：国家自然科学基金项目(60672135)

摘　　要：Bézier曲线最小二乘拟合,最终可转化为求解给定数据点的参数优化问题,遗传算法可以求解该优化问题,但易陷入局部早敛。为了防止局部早敛,提出将最速下降法和遗传算法相结合的混合算法用在曲线拟合上,该算法有效的解决了曲线拟合中遗传算法的局部早敛问题。最后通过实例验证了算法的可行性和有效性。Bezier curve fitting is solved then became a parameter optimization with given date knot, the genetic algorithms may solve this problem, but it is easy to fall into partial solution. In order to prevent the local early convergence, the mix algorithm based on genetic algorithm and steepest descent algorithm using in the curve fitting is suggested, which can solve the problem that it is easy to fall into partial solution of genetic algorithm. Finally, examples of data fitting are given to show feasibility and validity of our method.

关 键 词：曲线拟合 遗传算法 最速下降法 BÉZIER曲线 最小二乘拟合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]