Flooding Prognostic in Packed Columns Based on Electrical Capacitance Tomography and Convolution Neural Network

The flooding of packed columns is accompanied by a steep increase in liquid hold-up and pressure drop, resulting lower mass transfer efficiency potential damage to equipment. This study aims investigate, for the first time, feasibility Electrical Capacitance Tomography Convolutional Neural Networks as an intensified alternative conventional prediction methods. allows variations predominant characteristics events be investigated greater detail than previous research. Combined with Networks, sensor enables high accuracy on calculation strong robustness against noise-contaminated measurements. In this work, detailed comparison made between results using more method based Maxwell equation. Both methods can accurately calculate at low gas flow rates. predicted according equation did not match measured values rates, showing discrepancies up 68%. contrast, much superior giving only 1% mean difference reference hold-up. supported shows great fidelity non-invasive monitoring local hold-up, allowing accurate, localized loading point columns.