A method for leak monitoring and locating based on flow signals
By calculating the upstream and downstream flow difference of the pipeline flow signal and extracting features, combined with bipolarization processing and support vector data description model, the problems of high false alarm rate and inaccurate positioning in the existing technology are solved, and real-time detection and accurate positioning of pipeline leaks are realized.
CN118189071BActive Publication Date: 2026-07-21BEIJING UNIV OF CHEM TECH
View PDF 2 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING UNIV OF CHEM TECH
- Filing Date
- 2024-04-22
- Publication Date
- 2026-07-21
AI Technical Summary
Technical Problem
Existing flow balance-based leak detection technologies have a high false alarm rate in pipeline leak detection and cannot accurately locate the leak point.
Method used
By calculating the average flow rate and the difference in flow rate between the upstream and downstream of the pipeline, trend extraction and curve fitting elimination are performed. Combined with bipolarization processing of the flow signal, abnormal flow sub-signals are detected, and a support vector data description model is used for leak diagnosis and location.
Benefits of technology
It enables real-time detection of both slowly changing micro-leaks and sudden leaks, reducing the false alarm rate and accurately locating the leak point.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN118189071B_ABST
Abstract
The application discloses a kind of leakage monitoring and positioning method based on flow signal, specifically relates to pipeline leakage monitoring technical field, and the application is obtained by spline interpolation to the difference of delivery to upper and lower envelope line, and then the mean value of delivery difference fitting curve is obtained;According to the average of non-exceptional signal part calculated by alarm time selection, the leakage amount and leakage rate are calculated when the signal is abnormal, and the average is calculated again after confirming the end of abnormal signal;Then the leakage diagnosis model is established by support vector data description, the model is a hypersphere, after the flow signal is de-meaned, the leakage signal is enhanced using digital high-pass filter, the characteristic value is extracted from the enhanced signal using interval sub-signal amplitude multi-scale representation method, and the characteristic value obtained is calculated, and if the hypersphere radius is greater than the hypersphere radius of diagnostic model, it is an abnormal signal.
Need to check novelty before this filing date? Find Prior Art