提高计算几何算法健壮性方法的研究  

Methods of Improve the Robustness of Computational Geometry

作　　者：谭建领[1] 张圣敏[1] 李海生 

机构地区：[1]黄河水利职业技术学院,河南开封475001 [2]北京亿维讯科技有限公司,北京100020

出　　处：《机电产品开发与创新》2007年第3期26-27,37,共3页Development & Innovation of Machinery & Electrical Products

基　　金：河南省自然科学基金(0411053100)

摘　　要：讨论了计算几何领域中影响算法健壮性的一些因素,根据影响算法健壮性的退化现象,提出了增强算法健壮性的策略,包括长精度数据的使用、对剖分域中的部分点进行合理赋权、在几何运算时采取误差控制等技术。经程序测试,这些方法不仅增加了算法的健壮性,而且可以提高系统处理问题的能力。The robustness is the key point of the computational geometry. The factors that influence the robustness of algorithm are analyzed and several methods that enhance the robustness of algorithm are put forward. A novel algorithm is proposed which gives the vertex on the sharp regions a reasonable weight in advance for Delaunay triangulation and controls the error effectively. So it can ensure the boundary consistency. The algorithm is developed and proved to be convergent.

关 键 词：计算几何 算法 健壮性 误差 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]