产品生命周期管理CAD模型数据格式与转换关系数据集
收藏资源简介：
产品生命周期管理CAD模型数据格式与转换关系数据集主要面向网络协同制造平台态势感知与安全管理过程中跨域跨网协同安全技术需求。为解决工业数据的长期保存问题，基于图卷积网络的工业流程敏感数据定位算法，提出一种基于数据挖掘的离散工业数据结构化方法；针对产品生命周期管理中的关键数据识别问题，提出一种基于复杂网络演化模型的大规模产品格式转换数据集生成方法。文件格式通常用于表示电子文档的媒体排列和布局。随着计算机和网络技术的快速发展，越来越多的文档以数字形式产生和保存。特别是，计算机辅助设计(CAD)文件格式的数量仍然不受控制。许多文件格式，如IGES、STEP、STL和DXF被认为是事实上的标准，而新的文件格式，如AMF和3MF，也被创建来取代它们。为了减少数据收集和制定过程中的繁重工作，自动构建格式交换网络是确保产品数据迁移的必要步骤。 网络爬虫用于有效地收集格式和交换数据。在线文件格式和软件目录，如AlternativeTo 和File-Extensions，提供了数千种文件类型及其相关应用的丰富信息。我们使用AlternativeTo收集CAD软件工具。选择cad和3d建模标签来查找相关软件。为了限制交换网络的规模并保持迁移的数据保真度，只包括交换矢量数据的交换工具。一般来说，对软件的关注较少，因此有时缺乏足够的产品数据迁移信息，因此只有具有10种以上导出或导入格式的软件才是交换格式的候选。File-Extensions网站是全球最大的计算机文件扩展名库，有每种文件类型的详细说明和相关软件程序的链接；这使得它适合于收集产品数据交换信息。 在收集了所有的转换关系之后，它们被转换成GraphML文件，用于构建图形和计算交换网络的度量。在清理了文件格式的扩展名之后，大约1673个节点和10123条边的数据最终可供使用。该数据集以工业产品为研究对象，重点选取了一些用于二维图形设计的常用软件，还包含部分三维图形设计、工程仿真及数据文档处理的知名软件。然后以这些格式为网络的节点，建立软件数据格式之间的转换网络。该网络是一个有向网络。网络的平均度约为6.15，网络直径为3，聚类系数约为0.15。
This dataset of CAD model data formats and conversion relationships for Product Lifecycle Management (PLM) targets the technical requirements of cross-domain and cross-network collaborative security in the situation awareness and security management process of networked collaborative manufacturing platforms. To address the long-term preservation of industrial data and draw on the industrial process sensitive data localization algorithm based on Graph Convolutional Networks (GCN), this work proposes a structured method for discrete industrial data based on data mining; aiming at the key data recognition problem in product lifecycle management, a generation method for large-scale product format conversion datasets based on complex network evolution models is put forward. File formats are typically used to represent the media arrangement and layout of electronic documents. With the rapid development of computer and network technologies, an increasing number of documents are generated and stored in digital form. Specifically, the number of Computer-Aided Design (CAD) file formats remains uncontrolled. Many file formats such as IGES, STEP, STL and DXF are regarded as de facto standards, while new formats like AMF and 3MF have also been created to replace them. To reduce the heavy workload in data collection and formulation, automatically constructing a format exchange network is a necessary step to ensure product data migration. Web crawlers are employed to efficiently collect format and exchange data. Online file format and software directories such as AlternativeTo and File-Extensions provide rich information on thousands of file types and their associated applications. We collected CAD software tools using AlternativeTo, by selecting the "cad" and "3d modeling" tags to find relevant software. To limit the scale of the exchange network and maintain the fidelity of migrated data, only exchange tools that handle vector data are included. Generally, less attention has been paid to software, so sufficient product data migration information is sometimes lacking. Therefore, only software with more than 10 export or import formats are considered candidates for exchange formats. The File-Extensions website is the world's largest computer file extension library, with detailed descriptions of each file type and links to related software programs, which makes it suitable for collecting product data exchange information. After collecting all conversion relationships, they are converted into GraphML files for graph construction and metric calculation of the exchange network. After cleaning the file format extensions, data with approximately 1673 nodes and 10123 edges is finally available for use. This dataset takes industrial products as the research object, focusing on several common software for 2D graphics design, and also includes well-known software for 3D graphics design, engineering simulation and data document processing. Then, taking these formats as the nodes of the network, a conversion network between software data formats is established. This network is a directed network, with an average degree of approximately 6.15, a network diameter of 3, and a clustering coefficient of approximately 0.15.
