一种基于低秩分解与共同趋势解耦的故障检测方法及系统
By employing low-rank decomposition and common trend decoupling, the problem of separating industrial process data under the coexistence of noise interference and non-stationarity was solved, achieving effective decoupling of non-stationary trends and stationary features, and improving the accuracy and robustness of fault detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CENT SOUTH UNIV
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to effectively separate non-stationary common trends, stationary characteristics, and noise information in industrial process data when noise interference and non-stationarity coexist, leading to a decrease in fault detection accuracy.
We employ a method of low-rank decomposition and common trend decoupling. By performing low-rank decomposition on historical process data, we decompose it into clean data and sparse noisy data, construct a common trend learning model, apply stationarity constraints, decouple non-stationary common trends from stationary features, and obtain a fault detection model through joint optimization.
It improves the accuracy of fault detection under noise interference and non-stationary environments, enhances the robustness and applicability of the model, and can more accurately extract the time-invariant dependencies and fault characteristics between variables.
Smart Images
Figure CN122087455B_ABST