Based On Data-driven Biological Oxidation Metallurgical Process Control And Optimization

Bio-oxidation technology is an emerging gold pretreatment technology, thetechnology has the advantages of less investment, low cost, simple operation, smallenvironmental pollution, especially for refractory, low-grade gold mine, known asthe true sense of the "green metallurgy." The pretreatment effect a direct bearing onthe level of final gold extraction rate and gold grade is good or bad. With theresources, energy crisis and increasing competition in the market, under theconditions of existing process equipment, the use of optimal control techniques toimprove the extraction of gold and grade security to reduce energy consumption,improve the enterprise’s production and even improve the level of the entire field ofmetallurgy in the international the competitiveness of great significance.Biological pretreatment process, however, is a long process, and the processparameters to detect the presence of large delay strong nonlinear couplingrelationship between the parameters; interference and the process is affected by theuncertainty of the ambient temperature, etc. Therefore, the biological pretreatmentprocess is still manual control, presence of low gold extraction, different batches ofthe proposed rate fluctuations, high energy consumption, grade fluctuations andother issues. In response to these problems, a gold mine in Xinjiang for the studyproposed data-driven approach to study the factors that affect the biological goldextraction, and ultimately improve the control accuracy of the field parameters. Inthis paper, completed:Analysis of complex industrial process optimization research status at home andabroad, introduced bio-metallurgical process control situation, the international hottechnology; analysis of the mechanism of biological oxidation metallurgy, explainsthe process of microbial treatment of sulfide-containing gold minerals, with a focuson Analysis of several factors that affect the biological oxidation, and gives thecontrol target, the final selection of the data driver control As used herein theoptimization algorithm.Data-driven development process, a brief description of commonly useddata-driven techniques, and highlighted the forms of expression tight formats andpartial format model-free adaptive control and control strategies, and ultimatelyselect the model-free adaptive control method as this topic The final study method. Prove the convergence of the model-free adaptive controller and robust, andcarried out a detailed convergence analysis, the eventual adoption of computersimulation technology that change control method can be used for the complexstructure, large time delay nonlinear systems control; model-free adaptive controllerparameters can not be set automatically, parameters corresponding to the particleswarm optimization, the method will be based on valid information infeasiblesolution is not feasible calculation function augmented objective function, theformation of the fitness function of the approximate optimization problem, show thatthe simulation tested the effectiveness of the optimization algorithm; computersimulation results show that: the model-free adaptive controller in large delay by themodel-free adaptive PID control site, non-linear perturbation large complex systemscontrol effect than the PID controller, which proved the validity of the method of thistopic.Finally, in Matlab the oxidation tank temperature control model and model-freeadaptive control oxidation tank temperature control, to simulate the conditions offour different oxidation tank temperature control process, the simulation analysisshows: Fourth condition Model-Free Adaptive to oxidation tank temperature controlthan on-site control in the appropriate range, high accuracy, proved the effectivenessof the model-free adaptive control, and also shows the model-free adaptive controlcan be extended to The entire control system of the bioleaching.