Structural Optimization Profile Recognition for Editable CAD Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing structural optimization techniques produce discrete representations of mechanical parts as mesh structures that cannot be edited or directly input into manufacturing systems, requiring manual intervention and leading to overfitting issues when converting to 3D CAD models.
Innovation Solution
A machine-learning method using a graph neural network to convert 2D polyline profiles from structural optimization into primitive parametric curve classes, allowing automatic conversion and reducing overfitting by providing flexible output classes with fewer parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If structural optimization produces discrete mesh structures, then optimal mechanical properties are achieved, but the results cannot be edited or directly input into manufacturing systems
Solution Approach 1:
The patent introduces a mesh simplification algorithm and a parametric model generation module as intermediaries between the discrete mesh structure and the manufacturing system. The simplification algorithm reduces the mesh to a simplified representation, which is then converted into a parametric CAD model that maintains mechanical properties while being editable and manufacturable.
Solution Approach 2:
The patent transforms the discrete mesh representation into a parametric model representation, changing the data structure from fixed geometric elements to parameter-based definitions. This allows the model to be edited by modifying parameters rather than individual geometric elements, while still representing the optimal structure.
2Manufacturing precision
If manual intervention is used to convert optimization results to 3D CAD models, then model accuracy can be maintained, but the conversion process becomes tedious and time-consuming
Solution Approach 1:
The patent implements an automated conversion system where the mesh simplification algorithm and parametric model generation module automatically process the optimization results without requiring manual intervention. The system self-adjusts parameters and generates the final CAD model through algorithmic processes, maintaining accuracy while dramatically increasing conversion speed.
Solution Approach 2:
The patent replaces the manual mechanical process of converting mesh to CAD models with an automated computational system. The mesh simplification algorithm and parametric generation module substitute for manual modeling operations, using computational methods to achieve both accuracy and efficiency.
3Adaptability or versatility
If existing conversion solutions are used, then 3D CAD models can be obtained, but overfitting issues occur with too many parameters
Solution Approach 1:
The patent extracts only the essential parameters needed to define the parametric model from the detailed mesh structure. By selectively taking out critical geometric and topological parameters while discarding redundant details, the system achieves model flexibility with a reduced parameter set that avoids overfitting.
Solution Approach 2:
The patent segments the parametric model into hierarchical levels of abstraction, separating essential design parameters from detailed geometric parameters. This segmentation allows the model to maintain adaptability through essential parameters while reducing overall complexity by organizing parameters into manageable groups with different levels of detail.
Data Source
AI summary
A computer-implemented method for machine-learning a function. The method includes obtaining a dataset including 2D polyline profiles each representing respectively a portion of a manufacturing contour in a result of a structural optimization that represents a mechanical part, each 2D polyline profile being associated in the dataset with a respective primitive parametric curve class among a predetermined set of primitive parametric curve classes. The method further comprises training the function based on the dataset. The function is configured to take an input 2D polyline profile and to provide an output primitive parametric curve class. Such a method forms an improved solution for processing a result of a structural optimization that represents a mechanical part.