Multiphase flow metering using capacitance transducer and multivariate calibration

Abstract

The method of multivariate calibration is experimentally investigated to establish estimators of the required pertinent flow parameters in multiphase pipe flow. The unfiltered primary signals, provided by a capacitance sensor, are analysed as discrete time series and the signal characteristics are extracted. The multivariate model that is generated estimates the flow composition based on the extracted information existing in the broad-band capacitance signal. The data analysis and test results are presented.

Keywords

• Process measurements
• signal processing
• feature extraction
• chemometrics