Modeling of Microfiltration Membrane for Oily Wastewater Treatment Using Hermia Model and Neural Network

In this paper, results of the oily wastewater treatment by ceramic membranes in microfiltration process are simulated using neural network and hermia model. The process simulation by neural networks is written using MATLAB software program. The network input consisted of oil concentration (Coil), cross-flow velocity (CFV), temperature (T), trans-membrane pressure (TMP), and filtration time (t) and the network output consisted of permeation flux. The operational data is divided in three categories: training, validation and testing. In this study, three modes (20-20-60 and 15-15-70 and 10-10-80) are investigated for training the network. Also formation of precipitate is modeled using hermia model includes four different behavior in obstruction. In all states are observed prediction of permeation flux using neural network is better than hermia model.