| 摘要: |
| 针对大尺寸设备装配误差检测过程复杂、耗时较长以及测量结果不准确等问题，提出了一种基于局部特征配准的模型装配误差分析方法。首先通过激光扫描设备获取零部件实体外表面点云数据；对点云数据进行去噪、自适应曲率精简等预处理操作，降低运算时间复杂度；导入IGES理论设计模型，并将其拆分成若干面片形式；通过三点法将点云数据模型与理论模型进行粗配准，提高精确配准的准确性，防止局部最佳拟合的情况发生；然后选中特征明显的1～2个面片与点云数据进行ICP精确配准；最后结合最小包围盒、法向计算等方法对点云模型的误差进行整体计算显示。经过铸造件实验验证，结果表明，该方法能够精确地计算分析出模型的误差分布情况。 |
| 关键词: 点云精简 IGES面片拆分 三点法配准 ICP 误差分析 |
| 基金项目:国家自然科学基金项目（No. 51805071），大连市人才专项计划项目（2019RQ026、No. 2018RD05），中央高校基本科研业务费（DUT18RC（3）073） |
|
| Model Assembly Error Analysis Method Based on Local Feature Registration |
|
Wang Guohua1, Liu Tao2, Tang Zengwu2, Li Te1, Liu Haibo1, Wang Yongqing1
|
|
1. Key Laboratory for Precision & Non-traditional Machining of Ministry of Education, Dalian University of Technology, Dalian 116024;2.1. Institute of Shandong Non-Metallic Materials, Jinan 250031
|
| Abstract: |
| Aiming at the problems of complexity, long time consuming and inaccurate measurement results of large-size equipment assembly error detection process, a model assembly error analysis method based on local feature registration is proposed. Firstly, the surface point cloud data of the physical surface of parts is obtained by laser scanning equipment. Then the point cloud data are preprocessed by denoising, adaptive curvature simplification and other operations to reduce the computational time complexity. The IGES theoretical design model is imported and divided into several facets and the point cloud data model is roughly registered with the theoretical model by three-point method, so as to improve the accuracy of accurate registration and prevent the occurrence of local best fitting. Then one or two surfaces with obvious features are selected for accurate ICP registration with point cloud data. Finally, combined with the minimum bounding box and normal calculation methods, the error of the point cloud model is calculated and displayed as a whole. Experimental results of casting parts show that the method can accurately calculate and analyze the error distribution of the model. |
| Key words: point cloud data simplification IGES facets division three-point registration method ICP assembly error analysis |
