The fuzzy inference algorithm based on a measurement of 3D raw ECT data for two-phase gas-liquid flow nature identification

P. Fiderek, T. Jaworski, R. Wajman, J. Kucharski

Institute of Applied Computer Science, Lodz University of Technology, Poland

Abstract

The new approach to a two-phase gas-liquid flow recognition using Electrical Capacitance Tomography (ECT) has been described in this paper. The presented fuzzy logic based algorithm is capable of recognizing flow regimes in both horizontal and vertical pipeline sections of its three given diameters: 38mm, 60mm and 90mm. Raw measurement data acquired from the 3D electrical capacitance tomograph represent the capacitive relationship between electrodes in the sensor. The presented paper shows that observation of changes in electrode pair-wise capacitance provides valuable information about a specific flow regime. The authors claim that the two-phase flow recognition process can be performed, even before the image reconstruction process has been done.

Keywords: fuzzy logic, FCM, two-phase flows, AI