Method Of Wet Zinc Smelting Goethite Process Of Heavy Iron End Ph Value Forecast Method Research

The iron removal by goethite method is one of key process in zinc hydrometallurgy. The process is designed to reduce the concentration of iron ion in zinc leaching solution, to void the negative influence of the iron ion in subsequent industrial sections. As a key factor and control parameter, an appropriate end-point pH value has to be maintained and timely adjusted by neutralizer feeding. In the industrial process, even the detection of the end-point pH value concentration is existed, the neutralizer feeding position is located a long distance from the end-point of the process, which makes the direct control of the end-point pH value unlikely. The swings of the end-point pH in will affect the normal production, even result in disqualification of the output leaching solution. Therefore, a thorough study of the end-point pH value prediction have theoretical significance and application value.On the basis of the research in industrial mechanism and the analysis of the CSTR chemical process in iron removal by goethite method, the influence of the main factors of pH value is investigated, and a mechanism-based prediction model is established. However, due to the complexity of the mechanism in iron removal process, the precision of the prediction model cannot meet the industrial requirements. In this situation, by analyzing the historical data and studying the temporal relation of the factors, a data model based on LS-SVM (least square support vector machine) is established, in which the model parameters settings using the experimental method to improve the generalization capacity. Considering the conditions changes as time goes on industrial production the prediction model based on historical data presenting misalignment phenomenon, the on-line corrections of the prediction model was studied, which enhance the adaptability of the model.The simulation research was done for prediction model of the end-point pH value and the correction method of the model by large quantity of actual industrial data. The result shows that the prediction model based on LS-SVM of the end-point pH value has a high precision capability. And the prediction precision would not decrease along with the changes of working condition by introducing the correction method in the model. By establishing the end-point pH prediction model, guidance to procedure parameter optimization of the iron removal prosess was afforded.