基于稀疏模型的曲线光顺算法

Sparity Optimization Based Curve Fairing

为了对曲线进行有效光顺,将曲线光顺问题建模成基于稀疏模型的优化问题,并提出了新的曲线光顺算法.通过分析非光顺曲线的例子,对现有的判别准则进行了总结;根据总结出的判别准则,曲线的光顺性与曲率图二阶差分的稀疏性与大小密切相关,将曲线光顺问题建模成一种基于稀疏模型的优化问题;并给出了一种高效的迭代调整算法求解该模型.数值实验结果表明,文中算法是实用和有效的;与其他光顺算法进行对比时,该算法也有十分突出的表现.

In this paper we propose a sparse modeling based optimization method to formulate the curvefairness problem. Based on an overall review on previous fairness criteria and our comprehensive workingexperience on ship design, our proposed fairness criteria not only covers all the previous cases but also considerscases which are never studied before. The curve fairness problem is further formulated as a sparsemodeling based optimization problem according to our proposed criteria. An effective and efficient iterativerefinement algorithm is proposed. Numerous experimental results demonstrate the practicability and effectivenessof our fairing algorithm, and the comparisons show that our fairing algorithm outperforms the stateof the art curve fairness methods.