基于多目标进化深度学习的钢铁产品质量缺陷识别方法
By automatically generating and optimizing a steel product quality defect identification model using multi-objective evolutionary deep learning technology, the problem of balancing model complexity and identification accuracy in existing technologies is solved, and efficient identification of surface defects in steel products is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEASTERN UNIV CHINA
- Filing Date
- 2024-08-08
- Publication Date
- 2026-07-17
AI Technical Summary
Existing deep learning technologies rely on manually defined neural network structures for identifying surface defects in steel products, making it difficult to balance model complexity and recognition accuracy, resulting in poor recognition performance.
A multi-objective evolutionary deep learning method is adopted, which combines automatic neural network deepening technology and multi-objective differential evolution algorithm to automatically generate and optimize the steel product quality defect identification model. By randomly generating residual blocks and performing model training and pruning, the model complexity and recognition accuracy are balanced.
Successfully breaking free from reliance on human experience, a steel product quality defect identification model with high accuracy and low computational complexity was constructed, achieving an ideal balance in model performance.
Smart Images
Figure CN119068240B_ABST