Research On Prediction Model Of Mineral Flotation Process Based On Multi-source Data

At present,with the continuous exploitation of mineral resources,the grade of mineral resources is gradually reduced.In the future,the rational use of lower grade mineral resources will become even more important.Therefore,many non-ferrous metal smelting enterprises in China have added flotation agents to the slurry under certain process conditions,and raised the mineral content through flotation foam with minerals attached.Finally,a flotation production line was established to meet the requirements for smelting minerals.Foam flotation is a method of mineral separation and beneficiation based on the difference of physical and chemical properties of ore surface.Different minerals are separated by different physical and chemical properties of different minerals.This property separates different minerals.Foam flotation is widely used in the beneficiation industry..Concentrate grade is an important quality indicator for the froth flotation process.At the industrial site,process parameters,such as feed volume and ventilation,are usually adjusted based on the current concentrate grade to provide timely control of the current concentrate grade for optimal control.Determination of the determination of concentrate grades Off-line testing is used in most flotation equipment to obtain the current concentrate grade.The flotation process is a complex multiphase,polymorphic,multiple input and output,and coupling correlation system.There are many factors affecting the final product quality,including raw material grade,grinding method,grinding fineness,pulp concentration,pulp size,feed rate,flotation level,temperature,PH value,reagent type,reagent amount,reaction time,air volume,air pressure,etc.Mineral flotation process has a long process,including rough selection,sweeping and selection.Each process includes multiple flotation cells,and the variation characteristics of bubbles in each flotation cell vary greatly.In the traditional flotation operation,the field personnel rely on experience to judge the flotation production status according to the shape,color and dynamic characteristics of the flotation froth,and make appropriate adjustments according to experience.This mode of operation is more dependent on the experience of the field personnel,easy to produce large product quality fluctuations,and due to the long time lag of the system itself,when quality problems occur,the adjustment system needs a long time to adjust normal.At present,the basic reason for manual adjustment is that there is not enough knowledge about flotation process.Because of the complexity of the system itself,the flotation process can not be effectively modeled,and the precise control can not be discussed.In recent years,people are gradually inclined to use artificial intelligence to simulate the judgment and operation of field personnel,and use expert system to optimize the judgment after data accumulation.This is a relatively effective way to deal with.This research project involves research contents such as mineral processing,applied mathematics,computer science and so on.This project will start with the analysis of the relationship between the process flow and the characteristic parameters in the flotation process,study the relationship between the characteristic parameters and the output grade,and carry on the data-based integrated modeling.1.In view of the past data modeling is confined to the bubble image characteristics,the project will model the industrial site production data and foam image feature data in the form of information entropy in order to improve the accuracy and accuracy of the prediction model.2.In view of the model errors caused by the long flotation process,it is difficult to accurately describe the flotation process of the long flotation process by the established prediction model.Therefore,the prediction model is optimized and analyzed,and the accuracy of the model is improved by the principal component analysis algorithm and clustering analysis algorithm.3.Establish the software system of prediction model based on SQL Server and MATLAB,use SQL Server to store data and interact with MATLAB graphical interface through ODBC,complete data analysis and prediction,and output the prediction results.