RSM and ANN modeling of hold up, slip, and characteristic velocities in standard systems using pulsed disc-and-doughnut contactor column
Alireza Hemmati, Ahad Ghaemi, Mehdi Asadollahzadeh
Abstract
The liquid-liquid extraction process is numerically studied based on two modeling approaches of artificial neural network (ANN) and response surface methodology (RSM). Four influential parameters are considered to study their impacts on three responses of dispersed phase holdup, slip, and characteristic velocities. Three correlations for the responses are proposed based on the quadratic model via the RSM-central composite design approach. The performance of two algorithms of radial basis function and multi-layer perceptron are compared in ANN modeling. Both models of ANN and RSM indicated an acceptable prediction of experimental data with a max R2 value of 0.996 and 0.987, respectively.
Topics & Concepts
Response surface methodologyContactorArtificial neural networkCentral composite designQuadratic equationSlip (aerodynamics)ChemistryMultilayer perceptronQuadratic functionPerceptronBiological systemChromatographyArtificial intelligenceComputer scienceMathematicsThermodynamicsGeometryPhysicsPower (physics)BiologyFault Detection and Control SystemsFluid Dynamics and MixingAdvanced Control Systems Optimization