An Intelligent Control System for Continual Carbonation Decomposition Process of Sodium Aluminate Solutions Based on Expert and Prediction Strategy

In alumina production using the sintering process, it is one of the key processes to produce Al(OH)₃ using the method of continual carbonation decomposition of sodium aluminate solutions, and the resolution ratio and the last resolution ratio directly affect the output and quality of product. It is such a long time-delay and complex industrial process, which exits much uncertainty and is too complicated to describe with mathematical models, that it can not be controlled properly by traditional methods. In this paper, aimed to control optimal resolution ratio and the last decomposition ratio, the intelligent control system for continual carbonation decomposition process of sodium aluminate solutions is exploited which combined expert control with predictive control strategy. The principle knowledge and experts´ experience of continuous carbonation decomposition process of sodium aluminate solutions is analysed and applied to design a expert control model. And a neural network predicting model is set up to forecast the next output of system which compensated the output of expert control model. Thus, the influence of long time-delay was conquered effectively and the process of continual carbonation decomposition was optimal controlled. The practical results show that eligible ratio of decomposition ratio increases by 4%, and average value of decomposition ratio increases by 0.95%. The system is always running well