A VAE-k-means-based ultrasonic flowmeter health state evaluation method
By combining variational autoencoders and K-means clustering algorithms with Mahalanobis distance, the health status assessment problem of ultrasonic flow meters was solved, enabling efficient online status monitoring and assessment, and improving the metering accuracy and industrial production stability.
CN122432800APending Publication Date: 2026-07-21CHINA JILIANG UNIV
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA JILIANG UNIV
- Filing Date
- 2026-04-29
- Publication Date
- 2026-07-21
Smart Images
Figure CN122432800A_ABST
Abstract
The present application relates to a kind of VAE-Kmeans-based ultrasonic flowmeter health state evaluation method, belongs to flowmeter equipment state monitoring technical field.The method includes offline stage and online stage: offline stage acquires ultrasonic flowmeter historical operation data, including flow, echo amplitude, receiving gain value and signal-to-noise ratio, after pretreatment, feature extraction and dimension reduction are carried out using variational autoencoder VAE, adopt K-means clustering algorithm to construct health state benchmark model and record health state characteristic data;Real-time operation data of sensor are collected in online stage, after pretreatment and dimension reduction, the Mahalanobis distance of real-time characteristic data and health state characteristic data is calculated, health state evaluation is completed and sensor health index HI is output.The flowmeter state is divided according to health index HI.The present application can realize sensor health state online quantitative evaluation, improve sub-health state recognition ability, real-time is good, with good engineering application value.
Need to check novelty before this filing date? Find Prior Art