Statistical and machine learning application approach to predict mass flowrate

Abstract

Abstract Flow measurement is one of the critical points in the industrial process measurement. Mass flow measurement, as an alternative flow measuring method, is usually chosen to prevent temperature effects to the density that would be impacted directly to the flow measurement. One of the most popular direct mass flow measurements is Coriolis Mass Flowmeter (CMF). CMF which has an accuracy of mass flow measurement up to 0.1% of reading, recently became widely used for valuable fluids mass flowrate measurement. The challenge of the flow measurement will not stop in the selection and the smart capabilities of the flow instrument itself; another issue that appears during process measurement is the absence of the installed CMF as a mandatory installed, local pressure measurement could be a physical parameter that could be used to predict the mass flowrate using statistical and machine learning algorithms. This virtual flow measurement method will be an approach to face the absence of physical flowmeter issues, especially in critical or remote operation processes.