The Application Research Of Non-linear Prediction Algorithms On Heap Leaching Process Of Uranium

The bioleaching technology usually refers to Bacterial oxidation or biological oxida-tion of ore by Microorganisms exist in nature. That it is a technology to extract the valuab-le metals in the ore or solution, through use of microorganisms or its metabolites having anoxidation, reduction, dissolution or absorption of certain minerals orion effect. Its applicat-ion to uranium began in the1950s. Studies have shown that bacterial heap leaching techno-logy has good economic and environmental benefits. China’s uranium ore heap lea-chingindustrial production began in the early1980s and in2002successfully achieved industria-lization, and achieved good technical and economic benefits.In recent years, research on uranium heap leaching technology, in particular the heapleaching process modeling and analysis has become increasingly a cause for concern. Howto establish a reasonable mathematical model of heap leaching, to provide reliable quantit-ative research methods to constantly improve the theory and practice of uranium ore heapleaching system, is a hot and difficult problem in heap leaching of uranium. In the specificproduction process, many factors affect the heap leaching of uranium leaching rate. Andthe data generated on the process of operation crafts has the nonlinear and timing charac-teristics, this requires us to apply the nonlinear time series analysis method to study heapleaching phenomenon, to explore the variation.The vast majority of the nature of things in the real world are chaos. In this thesis,based on the sample data generated by a batch of uranium ore heap leaching, first, have asingle GM (1,n) model on the accumulated uranium leaching rate forecasts and fitting anal-ysis, and the use of BP neural network prediction research. Then have chaos identificationon the system of uranium ore heap leaching, and then by means of phase space reconstruc-tion theory to build a nearest neighbor prediction algorithm based on local phase space rec-onstruction, gray prediction algorithm based on the phase space reconstruction, calculatedcumulative leaching rate of uranium by using joint forecasting and fitting analysis.The accumulated uranium leaching rate prediction is as follows:(1)Based on singleGM (1,n) model predictions for on the average relative prediction error is0.0739, on prob-ability of the small error is P=0.6154, on ratio value of the Posterior error is C=1.0041.(2)On the classic BP neural network algorithm on the average prediction relative error is 0.0081.(3)Base on the reconstructed phase space points as nearest neighbors, use of localnearest neighbor forecasting techniques to do four times of predictions, Forecast results foron the maximum average re lative prediction error is R1=4.8664e-004, a minimum of R2=7.5822e-005.(4) Based on Phase Space Reconstruction and GM (1,n) model combin-ation prediction algorithm predicted results for on the average relative prediction error is0.0063, on probability of the small error is P=1, on ratio value of the Posterior error is C=0.1831.In this thesis, on the effect on the uranium on the cumulative leaching rate fitting is:(1) Based on single GM (1,n) model fitting result for on the average relative fitting error is0.0087, on probability of the small error is P=1, on ratio value of the Posterior error is C=0.1831.(2)Based on Phase Space Reconstruction and GM (1, n) model combination algo-rithm fitting results for on the average relative fitting error is0.0046,on probability of thesmall error is P=1, on ratio value of the Posterior error is C=0.1631.It is thus clear that, uranium heap leaching system has the nature on the Chaos, the va-rious prediction algorithm based on the phase space reconstruction predicted effects are be-tter than the single predicted models. Prediction algorithms based on the phase space reco-nstruction is a prediction algorithm of more scientific.In this thesis, computer simulation using MATLAB language, referenced data come f-rom a number of measured data in process of production in heap leaching, of heap leachingColumn of ZQ8in a mining area.