In-use Measurement of Ultrasonic Flowmeter based on Machine Learning

Mengna Li, Chengze Lv, Wenli Li, Chunhui Li

Abstract

To guarantee the accuracy of ultrasonic flowmeter, an in-use measurement system for ultrasonic flowmeter incorporating digital signal processors and machine learning approaches was proposed. Experimental analysis has been carried out to determine the variables affecting the accuracy of ultrasonic flowmeter. Based on random forest algorithm, we evaluated the contribution of different variables on the accuracy performance of ultrasonic flowmeter, and establish a model including variables extraction and prediction of flow deviationfor in-use measurement of ultrasonic flowmeter. By obtaining data of the flowmeter signal index, flow rate characteristics, sound velocity and flow velocity etc., the flow deviation of ultrasonic flow meter is predicted using random forest algorithm, and the difference between predicated value and observed value is smaller than 0.76 %. Furthermore, the degree of influence of different variables on the accuracy of ultrasonic flowmeter was analysed. The uncertainty of the prediction result was evaluated, with an extended uncertainty U = 0.92 % ~ 0.22 % (k = 2).

Keywords

Ultrasonic flowmeter; In-use measurement; Machine learning

Download

IMEKO-TC9-2019-116.pdf

DOI

10.21014/tc9-2022.116

IMEKO TC

TC9 - Flow Measurement

Back to the proceedings