零件CAD模型与点云数据集

该数据集来源于工业制造过程中的典型小批量-定制化零件，大体上分为叶轮、起落架等航空类零件和机床定位块、刀架、刀夹等机床类零件。数据集包括零件CAD模型库和对应点云数据库，其中点云数据由CAD模型转换而来。数据库共包含7类零件，每类零件包括10-20个特征类型、尺寸和空间结构不同的零件。加工特征识别效率主要通过零件的多种待识别加工特征验证，由于不同类别零件存在圆孔、倒角、台阶面等通用基础加工特征，进一步将不同类别零件的同类特征合并为一类，最终7类零件共包括自由曲面、复合特征和异性特征等待识别加工特征。三维点云数据可以应用于零件模型分类和工艺信息检索，实现零件级的工艺重用、验证和反馈优化。

This dataset is sourced from typical small-batch customized parts in industrial manufacturing processes, and these parts are broadly categorized into two groups: aerospace components including impellers and landing gears, and machine tool components such as machine tool positioning blocks, tool holders and tool clamps. The dataset includes a part CAD model library and a corresponding point cloud database, where the point cloud data is converted from the CAD models. The database contains 7 categories of parts, with each category comprising 10 to 20 parts that vary in feature types, dimensions and spatial structures. The efficiency of machining feature recognition is primarily validated using various machining features to be recognized from the parts. As parts of different categories share common basic machining features such as circular holes, chamfers and stepped surfaces, similar features from different part categories are further merged into a single unified category. Ultimately, the 7 part categories cover a total of machining features to be recognized, including free-form surfaces, composite features and special-shaped features. 3D point cloud data can be applied to part model classification and process information retrieval, enabling part-level process reuse, verification and feedback optimization.

搜集汇总

背景与挑战

背景概述

该数据集聚焦工业制造中的小批量-定制化零件，包含航空类和机床类共7类零件的CAD模型及对应点云数据，每类零件有10-20个不同特征、尺寸和空间结构的变体。数据集旨在支持加工特征识别、零件模型分类和工艺信息检索，可用于智能制造领域的工艺重用、验证和反馈优化，数据格式包括pcd、obj和docx，总大小为152.21MB。