Research On The Key Techniques In The Online And Offline Integrations Of CAD Models

With the development of economic globalization, Collaborative Product Development (CPD) has become the most competitive way on the development of products. On one hand, CPD makes the most companies in supply chain more specialized; On the other hand, it enhances the corporation between companies from different areas through dynamic alliance of enterprises. Because of differences in history and cultural background, different companies often adopt heterogeneous Computed-Aided Design (CAD) systems to develop products. Even within a company, designers would apply different CAD systems due to the differences among the different design stages. Therefore, the integration and interoperability among heterogeneous CAD systems has become inevitable technological problems. In this thesis, we study the problem of consistency maintenance of heterogeneous CAD systems, theory of feature-based interoperability among heterogeneous CAD systems,3D model similarity comparison, and the main contribution of the paper is highlighted in the following:A collaborative CAD concurrency control method based on the modeling history consistency is proposed. This paper analyzes the relations between modeling operations in detail and constructs a concurrency control model considering the modeling history. The model maintains causality relationship by state vector, solves partial concurrent relationship by reconstructing operation history, and classifies the modeling operations with concurrent relationship into commute, mask and mutex and handle them respectively. Finally, we design a collaborative CAD concurrency control method which supports modeling history consistency. The operation generated by local site is executed immediately, while remote site calls processing mechanism of concurrency control model to implement concurrency control. The experiment result shows this work can maintain modeling history consistency of collaborative sites as well as support less-constrained multi-user interaction.A singular feature interoperability of heterogeneous CAD model based on directed mutation particle swarm optimization is proposed. This paper explores the framework of feature-based interoperability and studies the method of singular feature interoperability based on feature-based data exchange. A spline element interoperability algorithm based on feature fitting is proposed and an optimization model for spline element fitting supporting feature interoperability is established. Particle Swarm Optimization (PSO) algorithm is introduced to solve the problem. To get better optimization results, an improved PSO algorithm named PSO-DM including time-varying acceleration coefficient mechanism and directed mutation mechanism was presented. Experimental examples are presented to demonstrate the effectiveness of our method. It not only maintenances geometric likeness of singular features, but also has the ability of interoperability.An asymmetric method for singular feature interoperability in feature-based data exchange is proposed. We study the method of interoperability of singular sketch and enrich the theory of feature-based interoperability based on the framework of feature-based interoperability. A semantic asymmetry singular feature interoperability and parameter asymmetry singular sketch interoperability is proposed. The optimization model for spline sketch fitting supporting feature interoperability is improved and Hausdorff distance estimation of distribution algorithm is introduced to solve the problem. Experimental results demonstrate that our approach can not only maintain the sufficiently high geometric fidelity, but also preserve the features in the process of singular feature. Our approach can simplify algorithm and improve precision in the process of singular sketch.An adaptive and efficient Hausdorff distance computing approach is proposed. According to whether the pair of point sets are separated or overlapped, we propose Non-Overlap Hausdorff Distance (NOHD) and Overlap Hausdorff Distance (OHD), respectively. Based on the algorithms of NOHD and OHD, we present an adaptive and efficient Hausdorff distance computing approach. Given the spatial relationship of the pair of point sets, we adaptively choose the NOHD or OHD algorithm to compute Hausdorff distance. Experiments show that, when processing two non-overlap point sets, the NOHD algorithm can effectively reduce the number of the nodes in the octree, while the OHD algorithm can reduce the number of the traversed points when the point sets are overlapped. We have performed lots of experiments on point clouds and CAD models, and compared our approach with state-of-the-art algorithms. The results demonstrated the effectiveness of our method.In summary, this thesis completes the collaborative CAD concurrency control method based on the modeling history consistency, enriches the theory of feature-based interoperability and improves the performance of computing the Hausdorff distance. It provides a new way to the heterogeneous CAD online and offline integrations on the theory and key technologies.