19-08-2014, 10:34 AM
The relationship between actual no of stages and reflux ratio is complex function in distillation column that is invariably used in Chemical Engg. Data required for the study is generated by simulating the distillation column. The column is simulated using general system irrespective of number of component to be separated. The coulmn is simulated to study the effect of relative volatality ratio , ratio of number of plates in stripping section to the number of plates in rectifying section, average relative volatality and ratio of latent heat of vaporization of number of liquids for the desired separation.Scaled conjugate gradient alogoritham is used as learning rule. Model prediction for testing points is quite accurate. Hence ANN can be used for better design of distillation column.
The relationship between actual no of stages and reflux ratio is complex function in distillation column that is invariably used in Chemical Engg. Data required for the study is generated by simulating the distillation column. The column is simulated using general system irrespective of number of component to be separated. The coulmn is simulated to study the effect of relative volatality ratio , ratio of number of plates in stripping section to the number of plates in rectifying section, average relative volatality and ratio of latent heat of vaporization of number of liquids for the desired separation.Scaled conjugate gradient alogoritham is used as learning rule. Model prediction for testing points is quite accurate. Hence ANN can be used for better design of distillation column