Prediction and control for silicon content in pig iron of blast furnace by integrating artificial neural network with genetic algorithm

Abstract

Silicon content in pig iron has long been used as one of the most important indices to represent the thermal state of a blast furnace. The control of silicon at a low level has been regarded as one of the most important operational techniques. In this paper, a mathematical program was developed to predict and control the silicon content in pig iron. The program includes three main models: the self-learning model, the prediction model and the control model. The first is to train the program by the intelligent method integrating artificial neural network with genetic algorithm according to the past processing variables. The second is to predict the next silicon content according to the current processing variables. The last is to control the subsequent behaviour of silicon based on the expert knowledge. From practical applications it was found that the program can provide both accurate predicted silicon content and corresponding operational guidance. It is very useful to stabilise silicon content and as low as possible.

Keywords:

• SILICON CONTENT
• BLAST FURNACE
• ARTIFICIAL NEURAL NETWORKS
• GENETIC ALGORITHM
• PREDICTION
• CONTROL