Study On Fuzzy-Neural Network Control System For The Nontoxic Extracting-gold Equipment

Nontoxic Process for Extraction Gold is also called "MLSSS (Modified Lime Slulfer Synthetic Sulution) Leaching Gold". It is a kind of "green" technology without cyanogens and polluter, which can bring remarkable social and economic efficience. Whereas it is difficult to acquire satisfied control effectiveness for this process using traditional PID control or modern control based on accurate mathematic model, because of its serious-nonlinear, time-variable, complexity and uncertainty characteristics. Therefore it is increasingly exigent of the study on innovative control algorithm for the Nontoxic Extracting-gold Equipment.Based on analyzing the characteristics of Nontoxic Process for Extraction Gold, a fuzzy-neural network control algorithm is prevailed. Firstly, the hybrid learning methods integrating genetic algorithm with BP algorithm to optimize fuzzy-neural network are employed. Secondly, the algorithm of system identification based on improved Elman dynamic recursive neural network is used. Finally, through MATLAB simulation demonstrates and verifies the control performance of the fuzzy-neural network control.The algorithm of system identification based on improved Elman network has quick convergent speed and strong approximation to actual object. And the proposed control algorithm with small overtop, quick dynamic response, high accuracy and strong robustness benefit,has good control of unknown-parameters and large-load-disturbance system, and has better performance in comparison with the traditionally designed fuzzy-neural network control. Therefore, it offers much significance for the problem of controlling Nontoxic Process for Extraction Gold.