Valve Production Data Clustering for Faster Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing valve production detection methods rely on inspector experience and instrument precision, leading to subjective errors and long detection times, with limitations in detecting small surface defects and internal quality issues.
Innovation Solution
A method for intelligent management of valve production process data using signal collection instruments, denoising, grid classification, coverage density distance calculation, and clustering algorithms to improve detection accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional visual observation and measuring tools are used for quality detection, then the detection method is simple to operate, but the detection accuracy is limited by inspector experience and instrument precision
Solution Approach 1:
The patent replaces traditional mechanical measuring tools (calipers, altimeters) and human visual inspection with an optical detection system using cameras to capture valve images. This substitution eliminates the dependence on inspector experience and mechanical instrument precision, achieving more consistent and accurate measurements through digital image analysis.
Solution Approach 2:
The patent creates a digital copy of the valve by capturing its image through a camera system. This digital replica allows for precise measurement and analysis of valve features without physical contact, enabling repeated measurements with consistent accuracy and facilitating automated defect detection through image processing algorithms.
2Productivity
If traditional detection methods are used, then the detection process is straightforward, but the detection time is relatively long, affecting production efficiency
Solution Approach 1:
The patent implements continuous quality detection by integrating the camera system into the production line, allowing valves to be inspected automatically as they pass through the manufacturing process. This continuous detection approach eliminates the need for stopping production for manual inspection, thereby maintaining high productivity while reducing detection time through automated real-time monitoring.
3Measurement precision
If traditional measuring tools are used, then the equipment is simple, but small surface defects such as small cracks or burrs cannot be detected
Solution Approach 1:
The patent transitions from one-dimensional mechanical measurements to two-dimensional optical imaging, capturing the entire valve surface in high resolution. This dimensional change enables the detection of small surface defects like cracks and burrs that are invisible to traditional point-based measuring tools, as the camera captures spatial information across the entire valve surface simultaneously.
Data Source
AI summary
The present invention proposes a method for intelligent management of valve production process data, a medium, and an apparatus. Through results of grid classification of sampling points, the method constructs information entropy of grid distribution density levels as well as grid distribution density, which takes into account the circumstance that sampling points may not exist in the classification space, so as to avoid grids where sampling points do not exist from affecting the local density of the sampling points; secondly, based on the core relative distance of the sampling points in a seed grid, the method constructs a coverage density distance, which takes into account the magnitude of the local density of the sampling points in the seed grid as well as differences in the distance from sampling points that have a relatively large local density in the other seed grids.