ANN Controller Design for Lime Kiln Process

Abstract

The lime kiln is a very complex multivariable process with severe non linearities, high degree of coupling and frequent disturbances. In this paper a 2x2 lime kiln process with two manipulated variables namely the fuel gas flowrate, and the percent opening of the induced draft damper and two controlled variables namely front end temperature and back end temperature has been considered. After its decoupling, artificial neural network (ANN) controllers have been designed to control the front end temperature. The performance of ANN controllers have been compared with that of conventional controllers.