Semi-supervised fatigue test condition monitoring method based on adaptive confidence active learning
Through a semi-supervised method of adaptive confidence active learning, long short-term memory networks and active learning mechanisms are used to screen low-confidence samples for manual labeling, which solves the problems of high labeling cost and poor adaptability of traditional monitoring methods and realizes efficient and low-cost equipment status monitoring.
Patent Information
- Application Number
- CN202510953471.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-11
- Publication Date
- 2025-09-16
- Estimated Expiration
- 2045-07-11
AI Technical Summary
Traditional equipment condition monitoring methods require a large amount of labeled data for training, which has high labeling costs and poor adaptability in different environments or working conditions. Semi-supervised methods are limited by the quality and representativeness of a small amount of labeled data.
A semi-supervised fatigue test state monitoring method based on adaptive confidence active learning is adopted. A state recognition model is constructed through a long short-term memory network. Combined with the active learning mechanism of adaptive confidence, low-confidence samples are screened for manual labeling and retraining, which reduces the labeling cost and improves the adaptability of the model.
It significantly reduces labeling costs, improves monitoring accuracy and the ability to adapt to complex working conditions, realizes adaptive evaluation of equipment operating status, and avoids economic losses and R&D interruptions caused by untimely monitoring.