逆向工程中约束驱动数据点云曲面特征优化

Constraint Driven Optimization of Surface Features from Point Cloud in Reverse Engineering

为了获得产品原始设计意图,提高重构模型的整体质量,提出一种实用的逆向工程中约束驱动数据点云曲面特征优化方法,其中包括约束分解和有效的数值求解.在约束分解部分,通过设计结构矩阵分割算法消除几何约束系统中曲面特征间的耦合约束,提出了基于多尺度特征的凝聚算法来实现几何约束系统的简化和分解;在数值求解部分,基于罚函数法建立了约束优化的数学模型,采用BFGS法进行了数值求解.对优化后的逼近误差与约束满足误差进行分析的结果表明,采用文中方法可以低数量级的逼近误差的放大,实现约束满足误差的减小,获得一种全局优化的结果.

In order to capture the original design intent and improve the whole quality of reconstructed models,a practical technique is presented for geometric constraint driven optimization of surface features from point cloud in reverse engineering,including constraint decomposition and numerical solution.In the stage of constraint decomposition,coupled constraints of complex surface features are eliminated by Design Structure Matrix partitioning algorithm.A new clustering method based on multi-scale surface feature analysis is proposed to reduce the geometric constraint system and decompose it into constraint subset.In the stage of numerical solution,mathematical models of optimization are built with exponential penalty.The BFGS method is studied for stable numerical solution of surface feature model optimization.Approximate error and constraint satisfaction error is analyzed.And the results show that the proposed method can achieve a global optimization result by increasing the approximate error at low level and decreasing the constraint satisfaction error.