Study on the Application of Neural Networks in Flotation Process Modelling

"In this present paper, the authors introduce the artificial neural networks into the modeling of the industrial flotation process, integrating information technology, control theory and intelligent optimizing method to flotation technology. The new idea gives a new way to solve the difficulty in this field, and therefore brings the study on the automatic control of flotation process in China into a new start.The Radial Basis Function (RBF) network, featuring a quick learning speed, is one of the feedforward neural networks. It has attracted much attention from automatic control recently for its implementing the nonlinearity mapping through adjusting the parameters of the nonlinearity transfer function. One of the online adaptive learning algorithm—the nearest neighbor-clustering algorithm, which can automatically determine the number of hidden units, has been implemented in the MATLAB programming language in this paper, laying the foundation for the flotation models.The relation model between ore pulp characteristics and reagents consumption has been proposed using the RBF networks based on the industrial lead rougher flotation of lead & zinc ore, and the simulation has been conducted in the laboratory with the results matching the samples data very well. The Xilin Lead & Zinc Mine Flotation Process Intelligent Consultation System of Reagents Consumption, which suitable for industrial process, has been developed. The reagents consumption advised by this system would stabilize the flotation process, and increase the recoveries. It could be used for providing the technological consultation to flotation operators in site, and for training new operators."