Outotec (St. Petersburg) (Russia):

Mashevskiy G. N., Doctor of Engineering Sciences, Chief Process Adviser, gennady.mashevsky@outotec.com
Romanenko S. A., Leading Process Engineer, Sergey.Romanenko@outotec.com

The article considers a new approach to flotation process modeling. It is proved, that modeling on the basis of multiple regression equations, practiced for several decades of the last century, is unpromising. It is shown, that modeling and control of flotation process on the basis of intelligent systems also does not bring positive results. Neural network modeling was put to test in cleaning flotation cycle treating copper-pyrite ores from the Named after the 50th October anniversary deposit. Specialties of flotation process are considered as those of an automation object, permitting to simplify the task to be solved through introduction of electrochemical control in slurry preparation, providing for process optimization regardless of treated ore type. Such special parameters include maintaining of optimal load in the cycle under consideration, which is a priori determined by design of flotation cells. By the example of copper-pyrite ores cleaning flotation cycle, multi-factor features
and nonlinearity of the automation object are shown, and, in principle, the automation object may be described only by neural networks. The mathematical model adequacy to the described object in accordance with the three years observations data package is estimated with respect to metal losses in tailings loop as R = 0.905, and with respect to concentrate loop as R = 0.82. A neural network model was developed for the automation object in question. A comparative analysis of the process control results is performed for intuitive and subjective manual actions of operator, as well as with regard to recommended actions in accord with the neural network model. Copper recovery with manual control is estimated to be 91.6 %, while with neural network model application it may be increased to 95.3 %.

1. Reznikov A. P. Obrabotka nakoplennoy informatsii v zatrudnennykh usloviyakh (Treatment of accumulated information in difficult conditions). Moscow, Nauka, 1976, 236 p.
2. Yeliseeva I. I., Yuzbashev M. M. Obshchaya teoriya statistiki. Uchebnik (General theory of statistics. Textbook). Ed. by I. I. Yeliseeva, 4th ed., rev. and enl., Moscow, Finansy i statistika, 2002, 480 p.
3. Obshchaya teoriya statistiki. Uchebnik (General theory of statistics. Textbook). Ed. by R. A. Shmoylova, 3d ed., rev., Moscow, Finansy i statistika, 2002, 560 p.
4. Kohonen T. Assotsiativnaya pamyat’ (Associative memory). Moscow, Mir, 1980, 239 p.
5. Kohonen T. Samoorganizuyushchiyesya karty: adaptivnyye i intellektualnyye sistemy (Self-organizing maps: adaptive and intelligent systems). Moscow, BINOM. Laboratoriya znaniy, 2010, 655 p.
6. Heikkinen S., Mashevsky G. N. Algorithmic base for flotation process control. Obogashchenie Rud, 2005, No. 6, pp. 32–37.
7. Mashevskiy G. N., Petrov A. V., Romanenko S. A., Sufyanov F. S., Balmanova A. Zh. A new approach to regulating the process of sulfide minerals selective flotation separation from pyrite in lime medium. Obogashchenie Rud, 2012, No. 1, pp. 12–16.
8. Petrov A. V., Romanenko S. A., Sufyanov F. S., Balmanova A. Zh. Regulation of copper sulfides separation from pyrite in lime medium in flotation process by pulp electrochemical potential. Obogashchenie Rud, 2012, No. 2, pp. 40–42.
9. Etelapaa M., Petrov A. V., Romanenko S. A., Sufianov F. S., Balmanova A. Zh. New approach to regulating the flotation process of selecting sulfide minerals from pyrite in a lime environment. XXVI Intern. Mineral Processing Congress IMPC-2012, New Delhi, India, September 24–28. Book of abstracts, 2012, Vol. 2, p. 680.
10. Romanenko S. A., Olennikov A. S. FrothMaster video system introduction experience practice at the «Green Mountain-2» plant. Obogashchenie Rud, 2014, No. 2, pp. 23–27.
11. Mashevskiy G. N., Petrov А. V., Lura M., Etelapaa M., Lamberg P. Chena^® multi-sensor systems for flotation process electrochemical control by Outotec. Obogashchenie Rud, 2009, No. 5, pp. 15–20.
12. Petrov А. V., Romanenko S. A., Olennikov A. S., Balmanova A. Zh. A new approach to regulating the process of sulfide minerals flotation separation from pyrite in lime medium by electrochemical. Gornyy Zhurnal Kazakhstana = Kazakhstan Mining Journal, 2013, No. 12, pp. 34–37.
13. Romanenko S. A. Principles and tools to optimize the flotation process Outotec. Globus. Geologiya i biznes = Globe. Geology and business, 2012, No. 5, pp. 32–35.
14. Mashevskiy G. N., Petrov A. V., Moilainen Ya., Timpere Yu., Saloheimo K., Kemppinen H. Principles of flotation process computer control on the basis of FrothMaster video system. Obogashchenie Rud, 2010, No. 3, pp. 39–43.
15. Mashevskiy G. N., Petrov A. V., Romanenko S. A., Sufyanov F. S. Development of ore types processing classification principles on the basis of flotation process parameters control and neural network modeling. Obogashchenie Rud, 2012, No. 4, pp. 36–42.
16. Romanenko S. A., Olennikov A. S. Application of modern mathematical analysis in flotation. Gorny Informatsionno-Analiticheskiy Byulleten (nauchno-tekhnicheskiy zhurnal) = Mining Informational and Analytical Bulletin (scientific and technical journal), Individual articles (preprints), 2013, 18 p.