Control of neutralization process using neuro and fuzzy controller

Abstract

Due to the extreme nonlinearity in the pH characteristics, control of pH in any chemical or biochemical process is a difficult task. Different approaches for pH control are proposed in various literatures. In the present study, control of pH neutralization process using neural and fuzzy controller is proposed. Initially a conventional linear controller is tried to control the pH at different linear regions. Based on the pH process characteristics, the nonlinear operating region is divided into three linear regions like pH low, pH middle, and pH high. Within each region, a local linear model is used to represent the process and a controller is designed. Control of pH by conventional PI controller based on the local linear model fails to provide satisfactory performance over the entire region, because of the extreme nonlinearity in the pH dynamics. It is required to tune the controller gain for different operating regions. Hence to overcome this drawback a neuro controller and a fuzzy controller are used. In this paper a novel fuzzy controller is used. Most fuzzy controllers use control error (e) and change in the control error (Deltae) as controller inputs and hence not able to differentiate the region in which the process operates, which is important information, necessary to control the nonlinear process. This controller uses set point as third input to select the region in which the process is operating.