Development of Self-Organized Group Method of Data Handling (GMDH) Algorithm to Increase Permeate Flux (%) of Helical-Shaped Membrane

Abstract

The chapter focuses on enhancing the permeate flux of helical shaped membrane using group method of data handling (GMDH) algorithm. The variables such as operating pressure, pore size, and feed velocity were selected as input parameters, and permeate flux as model output. The uncertainty analysis evaluates the acceptability of the model, and it was found that values of Nash-Sutcliffe efficiency (NSE), the ratio of the root mean squared error to the standard deviation (RSR), percent bias (PBIAS) were close to the best value which shows the model acceptability. The effect of input parameters on model output is calibrated using sensitivity analysis. It shows that pore size is the most sensitive parameter followed by feed velocity. The optimum values of pore size, operating pressure, and feed velocity were calibrated and found to be 2.21µm, 1.31×10-03KPa, and 0.37m/sec, respectively. The errors in GMDH model were compared with multi linear regression (MLR) model. It shows that GMDH predicts results with minimum error. The predicted variable follows the actual variables with good accuracy.