基于多模态大模型的起重机械智能检验方法及系统
By combining a multimodal large model with a numerical simulation program for fatigue damage evolution in fracture mechanics, a closed-loop coupled architecture driven by both data and mechanism is constructed, which solves the problem of identifying latent fatigue damage under varying working conditions of lifting machinery and realizes accurate detection and real-time early warning of early latent damage.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ANHUI SPECIAL EQUIP INSPECTION INST
- Filing Date
- 2026-05-07
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot effectively identify early latent fatigue damage under varying operating conditions of lifting machinery, resulting in missed detection of latent defects and delayed inspection results. This is mainly because multimodal deep learning models do not incorporate fracture mechanics damage evolution mechanisms and cannot distinguish between real fatigue damage and pseudo-features caused by operating condition disturbances.
By employing a multimodal large model combined with a numerical simulation program for fracture mechanics fatigue damage evolution, and through edge preprocessing and spatiotemporal decoupling and causal reasoning in the cloud, a closed-loop coupled architecture driven by both data and mechanism is constructed. False features are eliminated and damage-sensitive feature vectors are extracted. Feature extraction weight parameters are optimized to generate early latent damage features.
It effectively identifies early latent fatigue damage in lifting machinery, reduces the delay in inspection results, improves the detection accuracy of latent defects, reduces interference from false features, and realizes real-time monitoring and early warning of structural status.
Smart Images
Figure CN122133529B_ABST