Graph matching-based MBD model retrieval method

Abstract

The invention discloses a graph matching-based MBD model retrieval method. The method comprises the steps of firstly extracting graph information of each historical MBD model and storing the graph information in an MBD model retrieval database; then building a retrieval input model and a graph matching objective function of each historical MBD model in the database; solving an optimal matrix of each graph matching objective function; and finally calculating the similarity between the retrieval input model and each historical MBD model in the database according to the optimal matrix of each graphic matching objective function, thereby finishing MBD model retrieval. According to the method, semantic information in the MBD models can be fully identified, so that the integrity and accuracy of retrieval are improved.

Description

technical field [0001] The invention relates to an MBD model retrieval method based on graph matching, and obtains existing similar MBD model products through similarity calculation, belonging to the field of computer-aided design. Background technique [0002] With the continuous advancement of manufacturing and processing technology and the widespread adoption of digital modeling technology, the large-scale equipment manufacturing industry represented by aerospace has gradually adopted the full 3D digital design and manufacturing method based on model definition MBD (Model Based Definition). For example, Boeing applied MBD technology in its 737-NX project, which shortened its development cycle by 50%, reduced design changes by half, reduced overall costs by 25%, and products better met user needs; in the 787 project, Boeing MBD technology is fully applied in the constructed CGE platform, gradually transitioning to the MBD manufacturing technology system. [0003] my count...

AI Technical Summary

Problems solved by technology

Retrieval based on text keywords only pays attention to the model name information, cannot reflect the semantic information contained in the model, cannot be applied to the MBD model that contains rich semantic information, and the retrieval results cannot meet the user's retrieval intention

Content-based model retrieval methods usually only consider the information of a single dimension of the model, and do not fully consider the semantic information in the model, resulting in low integrity and accuracy of model retrieval results

The semantic-based model retrieval method needs to build a complex logical hierarchy, and this method has no recognized standard model, and can only adapt to some special application environments, and has poor universality in MBD model retrieval

Images