Intelligent based modelling, control and fault detection of chemical process

Abstract

Neutralizing pH value of sugar cane juice is the important craft in the control process in the clarifying process of sugar cane juice, which is the important factor to influence output and the quality of white sugar. On the one hand, it is an important content to control the neutralized pH value within a required range, which has the vital significance for acquiring high quality purified juice, reducing energy consumption and raising sucrose recovery. On the other hand, it is a complicated physical-chemistry process, which has the characteristics of strong non linearity, time-varying, large time-delay, and multi- input. Therefore, there has not a very good solution to control the neutralized pH value. Firstly, in this paper, neural network model for the clarifying process of sugar juice is established based on gathering 1200 groups of real-time sample data in a sugar factory.