Residual life prediction method under mechanical multi-source monitoring data based on evidence theory
By integrating multi-source sensor data and using evidence theory to perform uncertainty modeling, the problem of low credibility of single sensor prediction is solved, and accurate equipment residual life prediction and uncertainty quantification is achieved, which is suitable for aerospace, nuclear energy and large-scale energy systems.
Patent Information
- Application Number
- CN202510411069.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-02
- Publication Date
- 2025-07-11
AI Technical Summary
The existing residual life prediction methods are mostly based on a single type of sensor, which cannot fully capture the health status of the device, and the prediction credibility is low when facing unknown distributed data, especially in industries with high security requirements.
Using a mechanical multi-source monitoring data prediction method based on evidence theory, a variety of sensor monitoring data are integrated and joint data and pseudo-data are constructed, uncertainty modeling is carried out through evidence coding and regression reasoning, and accurate life prediction and uncertainty quantification are achieved.
Accurate lifetime prediction under multi-source data, while quantifying accidents and cognitive uncertainties, improving the credibility and applicability of predictions, and is suitable for industries with high safety requirements such as aerospace, nuclear energy and large-scale energy systems.
Smart Images
Figure CN120296698A_ABST