CAD Design Support Device for Noise Reduction in Automated Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CAD-based design automatic check techniques suffer from noise, where non-conforming spots are over-detected, leading to time-consuming verification processes, and lack effective methods to reduce such noise.
Innovation Solution
A design support device comprising a decision rule definition section, a geometric shape recognition section, and a feature value recognition section, which defines and applies decision rules to CAD models, using geometric and feature value recognition to accurately identify non-conforming spots and reduce noise in the check results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automatic check is performed using geometric shape recognition according to decision rules, then check efficiency is improved, but noise increases causing over-detection of non-conforming spots
Solution Approach 1:
The automatic check process is divided into two distinct stages: first, geometric shape recognition performs initial screening to identify potential non-conforming spots efficiently; second, feature value recognition refines the results by analyzing additional characteristics to eliminate false positives. This segmentation allows each stage to specialize, improving both speed and accuracy.
Solution Approach 2:
Feature value recognition acts as an intermediary between the initial geometric shape recognition and the final determination of non-conforming spots. It processes the intermediate results, adding a layer of verification that filters out noise while preserving true non-conformities.
2Measurement precision
If multiple recognition stages are added to reduce noise, then detection accuracy is improved, but device complexity increases
Solution Approach 1:
The system is segmented into modular functional components: a geometric shape recognition module for initial detection and a feature value recognition module for refinement. This modular structure allows independent optimization of each module while maintaining overall system manageability.
Solution Approach 2:
The geometric shape recognition stage performs preliminary filtering to identify candidate non-conforming spots before the more complex feature value analysis is applied. This preliminary action reduces the workload for subsequent stages, making the overall complex system more efficient.
Data Source
AI summary
A design assistance device (10) comprises: a determination rule definition unit (102) that defines a determination rule for determining whether a CAD model (103) violates a design guideline (101); a geometric shape recognition unit (105) that extracts, from the CAD model (103), a violation location violating the design guidelines (101), by using geometric shape recognition based on the determination rule defined by the determination rule definition unit (102); and a feature amount recognition unit (108) that identifies, from among violation locations, a violation location violating the design guidelines (101), by using feature amount recognition based on a prescribed feature amount of the violation location extracted by the geometric shape recognition unit (105).