Quantitative Optimization of Interoperability During Feature-Based Data Exchange

D. J. Zhang,F. Z. He,S. H. Han,X. X. Li
DOI: https://doi.org/10.3233/ica-150499
IF: 6.137
2015-01-01
Integrated Computer-Aided Engineering
Abstract:Sharing feature-based computer-aided design (CAD) modelsis a challenging problem that is frequently encountered among heterogeneousCAD systems. In this work, a new asymmetric strategy is presented to enrichthe theory of feature-based interoperability, particularly when addressing asingular feature or singular sketch. This paper analyzes the semanticasymmetry singular feature interoperability (SA-SFI) and parameter asymmetrysingular sketch interoperability (PA-SSI) in detail. We pay specialattention to the problem of PA-SSI, which is universally significant incollaborative product development (CPD). The objective of PA-SSI is todevelop an optimized model to exchange a singular sketch (spline) to ensurethat the exchanged model both maintains high geometric fidelity and can beeffectively edited in the target CAD system. The proposed method applies theestimation of distribution algorithm (EDA) to automatically solve thisproblem, and a Gaussian mixture model (GMM) is built according to thepromising solutions. Furthermore, Hausdorff distance is adopted to calculatethe fitness, and a local optimization operator is designed to enhance theglobal search capability of the population. Experimental results demonstratethat the proposed approach can maintain a sufficiently high geometricfidelity, and ensure that the exchanged model of the target CAD system canbe parametrically edited.
What problem does this paper attempt to address?