Study On Algorithm And Simulation Of Intelligente Control In Milling Industrial Processes

Grinding is an important industrial process of some industries such as metallurgy and non-ferrous materials, which is the most crucial part in beneficiation process. Grinding mechanism is complicated, it has lots of influencing factors, for instants ore property, grinding machine structure, grinding operation, etc. Also it carries traits of multivariate, non-linearity and indetermination. How to make good model, analysis and control for its industrial process becomes the recent focus, since these technical researches have much to do with economic benefit, resource utilization rate and environment protecting production, which is quite significant.Because the traditional control ways of grinding can't get appropriate values in 3 index, including ore quantity, grinding concentration, overflow concentration, and is hard put into use, the dissertation brings forward 3 ideas based on modern intelligent control theory: â‘  Apply controlling means of dynamic matrix to stable control of feeding quantity, improve every performance index, solve the problem that the traditional PID has poor result in stable control of feeding quantity;â‘¡Apply means of neural network and fuzzy control, check the grinding concentration by the corresponding relation between sound intensity and grinding concentration, get the grinding machine water controlling quantity through fuzzy control system, realize stable control of grinding concentration, sole the problem that grinding concentration is hard to measured;â‘¢Apply fuzzy control theory to set up overflow concentration fuzzy control system, avoid setting up cyclone model, meet the demand of overflow concentration control, and solve the problem that traditional ways can not control overflow concentration well.The dissertation also makes emulation of above control strategies with MATLAB, and compares with emulation results by traditional control means. The result shows that applying intelligent control means to grinding can not only meets demands of industrialprocess in every index, but also exceeds the traditional control means completely.