Based On The RBF Neural ESN Chaotic Time Series Network Analysis For The Forecast Of Flotation Economic Indications

Mineral processing enterprises as a typical continuous flow type enterprise,the concentrate grade and recovery rate of the key production indicators usually refers to the,for flotation,concentrate grade is stable or not is play a decisive role.Traditional dressing enterprises usually refers to to achieve the economic target root on the basis of the beneficiation process mechanism and factory production accumulated experience to achieve the production index is decomposed into corresponding parameters,such as grinding ore particle size,slurry concentration and dosage etc.,the production operation personnel's job is to keep the process index within the scope of the provisions of,and production management personnel,according to the technical index is within the scope of the provisions to check the pros and cons of production operation.This study is mainly through the optimization of flotation process in order to achieve optimal better control of the flotation process key process index.In this paper,we select the flotation of economic indicators are: to taste,the level of feed,feed concentration,feed flow rate four data as input of RBF neural system,Concentrate Taste and homework recovery of two indexes as the output of neural network,using simulink toolbox in MATLAB software compiler,simulation.Contrast the error between actual curve and the expected curve,observe whether the actual curve is smooth,if the difference is too much,can change the system parameters,adjust the neural network combined with chaotic time series after the system precision,study system after adjustment parameters of fitting degree is up to par value.Can be concluded that using RBF neural network to forecast chaos system analysis,the algorithm is simple,fast response,save a lot of tedious steps,improve operation efficiency.At the same time through the Mackey-Glass and Lorenz chaos system simulation also shows that directly using the neural network modeling and analysis of the chaotic system can effectively improve the precision of the system.At the same time,for the modeling and simulation of flotation process are fully illustrates the RBF neural network can be effective for chaotic time series prediction,and can be applied to the production practice the effective method.Also the research laid the foundation for the future.