ANN controller for binary distillation column — A Marquardt-Levenberg approach

Abstract

Artificial neural networks can provide good empirical controllers for complex nonlinear processes, because they are nets of basis functions that are useful for many purposes including process control. It is shown here that how artificial neural networks can design the column controller and demonstrate that the network controller is as good as or better than a fuzzy rule based controller. This paper investigates the design of a neural network based controller to control the concentration of the overhead and bottom product in the model of a distillation column.