一种基于机器学习的反应釜用温度监测方法及系统
By combining anisotropic weighted gradient and image entropy with an LSTM model, the accuracy and adaptability issues of reactor temperature monitoring were solved, achieving efficient and accurate monitoring of reactor wall temperature.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 长青(湖北)生物科技有限公司
- Filing Date
- 2025-10-11
- Publication Date
- 2026-07-17
AI Technical Summary
Existing reactor temperature monitoring technologies suffer from several drawbacks: contact temperature sensing elements cannot reflect the overall temperature distribution of the reactor wall, have delayed response and are prone to corrosion; infrared thermal imaging data is large and susceptible to noise interference, resulting in low efficiency and frequent misjudgments during manual interpretation; and image processing and machine learning methods have poor adaptability and unsatisfactory recognition results.
Anisotropic weighted gradient calculation is used, combined with image entropy value and historical data analysis, to screen key pixels, and the temperature status is determined by fusing feature vectors through an LSTM model.
It significantly improves the accuracy and reliability of reactor temperature monitoring, reduces noise interference, adapts to different operating conditions, and reduces misjudgments and missed judgments.
Smart Images
Figure CN121526965B_ABST
Abstract
Citation Information
Patent Citations
Local gray level-entropy difference leak detection locating method based on infrared image
CN103217256A
Low-quality LDR image enhancement method based on pseudo HDR image generation
CN116245760A