Machine learning-enhanced mass flow measurements using a Coriolis mass flow sensor

Romas Zubavicius; Dennis Alveringh; Mannes Poel; Jarno Groenesteijn; Remco G.P. Sanders; Remco J. Wiegerink; Joost Conrad Lötters

Machine learning-enhanced mass flow measurements using a Coriolis mass flow sensor

Romas Zubavicius, Dennis Alveringh, Mannes Poel, Jarno Groenesteijn, Remco G.P. Sanders, Remco J. Wiegerink, Joost Conrad Lötters

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

55 Downloads (Pure)

Abstract

We present machine learning-enhanced mass flow measurements based on microfabricated Coriolis mass flow sensors with integrated pressure and temperature sensors. Two machine learning techniques have been applied: linear regression (LR) and support vector regres- sion (SVR) on four features extracted from the raw sensor data to improve mass flow estimation. In contrast to conventional mass flow detection, LR and SVR use information from all integrated sensors to estimate the mass flow, which results in a full-scale accuracy improvement of a factor 4 for trained fluids.

Original language	English
Title of host publication	The 5th Conference On MicroFluidic Handling Systems (MFHS 2024)
Pages	65-68
Publication status	Published - 21 Feb 2024
Event	5th Conference on MicroFluidic Handling Systems, MFHS 2024 - Munich, Germany Duration: 21 Feb 2024 → 23 Feb 2024 Conference number: 5

Conference

Conference	5th Conference on MicroFluidic Handling Systems, MFHS 2024
Abbreviated title	MFHS 2024
Country/Territory	Germany
City	Munich
Period	21/02/24 → 23/02/24

Keywords

Coriolis
Mass flow
Machine learning
Linear regression
Support vector regression
Sensors
Microfluidics

Access to Document

mfhs2024_ml_coriolisFinal published version, 1.51 MB

Cite this

@inproceedings{1cf18cf9a33543f085a21a192d1b335a,

title = "Machine learning-enhanced mass flow measurements using a Coriolis mass flow sensor",

abstract = "We present machine learning-enhanced mass flow measurements based on microfabricated Coriolis mass flow sensors with integrated pressure and temperature sensors. Two machine learning techniques have been applied: linear regression (LR) and support vector regres- sion (SVR) on four features extracted from the raw sensor data to improve mass flow estimation. In contrast to conventional mass flow detection, LR and SVR use information from all integrated sensors to estimate the mass flow, which results in a full-scale accuracy improvement of a factor 4 for trained fluids.",

keywords = "Coriolis, Mass flow, Machine learning, Linear regression, Support vector regression, Sensors, Microfluidics",

author = "Romas Zubavicius and Dennis Alveringh and Mannes Poel and Jarno Groenesteijn and Sanders, {Remco G.P.} and Wiegerink, {Remco J.} and L{\"o}tters, {Joost Conrad}",

year = "2024",

month = feb,

day = "21",

language = "English",

pages = "65--68",

booktitle = "The 5th Conference On MicroFluidic Handling Systems (MFHS 2024)",

note = "5th Conference on MicroFluidic Handling Systems, MFHS 2024, MFHS 2024 ; Conference date: 21-02-2024 Through 23-02-2024",

}

TY - GEN

T1 - Machine learning-enhanced mass flow measurements using a Coriolis mass flow sensor

AU - Zubavicius, Romas

AU - Alveringh, Dennis

AU - Poel, Mannes

AU - Groenesteijn, Jarno

AU - Sanders, Remco G.P.

AU - Wiegerink, Remco J.

AU - Lötters, Joost Conrad

N1 - Conference code: 5

PY - 2024/2/21

Y1 - 2024/2/21

N2 - We present machine learning-enhanced mass flow measurements based on microfabricated Coriolis mass flow sensors with integrated pressure and temperature sensors. Two machine learning techniques have been applied: linear regression (LR) and support vector regres- sion (SVR) on four features extracted from the raw sensor data to improve mass flow estimation. In contrast to conventional mass flow detection, LR and SVR use information from all integrated sensors to estimate the mass flow, which results in a full-scale accuracy improvement of a factor 4 for trained fluids.

AB - We present machine learning-enhanced mass flow measurements based on microfabricated Coriolis mass flow sensors with integrated pressure and temperature sensors. Two machine learning techniques have been applied: linear regression (LR) and support vector regres- sion (SVR) on four features extracted from the raw sensor data to improve mass flow estimation. In contrast to conventional mass flow detection, LR and SVR use information from all integrated sensors to estimate the mass flow, which results in a full-scale accuracy improvement of a factor 4 for trained fluids.

KW - Coriolis

KW - Mass flow

KW - Machine learning

KW - Linear regression

KW - Support vector regression

KW - Sensors

KW - Microfluidics

M3 - Conference contribution

SP - 65

EP - 68

BT - The 5th Conference On MicroFluidic Handling Systems (MFHS 2024)

T2 - 5th Conference on MicroFluidic Handling Systems, MFHS 2024

Y2 - 21 February 2024 through 23 February 2024

ER -

Machine learning-enhanced mass flow measurements using a Coriolis mass flow sensor

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this