PARAMETRIC AND NONPARAMETRIC MODELS FOR PREDICTING OFF-NORMAL SITUATIONS IN UNDERGROUND MINES

Under discussion is an approach to the decision-making on risk management in anthropogenic systems. The parametric and nonparametric methods to predict multi-dimensional processes are compared. The perspective of the nonparametric prediction methods under conditions when the variables to be predicted exceed the number of initial data sampling points is demonstrated. The simulation prediction models of off-normal situations (rock falls and collapse in roadways, flooding, shutdown of mine hoist and main mine fan) are constructed. The prediction models were the selection algorithms of objective and system analysis by the consistency criterion and the pattern-analysis based on the orthogonal Karhunen–Loève transform by searching analog in historical data.

It has been found that out of thirty off-normal situations, the satisfactory prediction is only obtained for seven ONS while the prediction is fuzzy in other cases. Each of the unpredicted off-normal situations is described by a number of models in the form of difference equations depending on the conditions of hazardous factors. The conclusion is made that underground accidents are fuzzy objects with varied dynamics equations and non-optimality is possible in their detailed description.

Keywords

Variable, history, prediction, underground accidents, roadway, data sampling, off-normal situation, nonparametric method, selection algorithm, fuzzy prediction, pattern-analysis.

Issue number: 3
Year: 2018
ISBN:
UDK: 622.831:681.513
DOI: 10.25018/0236-1493-2018-3-0-200-207
Authors: Kupriyanov V. V., Matskevich O. A., Bondarenko I. S.

About authors: Kupriyanov V.V., Doctor of Technical Sciences, Professor, e-mail: msmu_asu@mail.ru, Matskevich O.A., Master’s Degree Student, e-mail: olya_9414@mail.ru, Bondarenko I.S., Candidate of Technical Sciences, Assistant Professor, National University of Science and Technology «MISiS», 119049, Moscow, Russia.

REFERENCES:

1. Bakhvalov L. A., Mogirev A. M. Gornyy informatsionno-analiticheskiy byulleten'. 2011, no 6, pp. 94—101.

2. Bendat Dzh., Pirsol A. Izmerenie i analiz sluchaynykh protsessov (Random data: analysis and measurement procedures), Moscow, Mir, 1998, 464 p.

3. Ivakhnenko A. G. Avtomatika. 1989, no 4, pp. 82—93.

4. Ivakhnenko A. G., Timchenko I. K., Ivakhnenko D. A. Avtomatika. 1990, no 1, pp. 20—31.

5. Kupriyanov V. V., Solov'ev A. E. Gornyy informatsionno-analiticheskiy byulleten'. 2010, no 5, pp. 143—149.

6. Kupriyanov V. V. Prikladnaya matematika. Uchebnoe posobie, ch. 1 (Applied mathematics. Educational aid, part 1), Moscow, Izdatel'skiy Dom MISiS, 2016, 176 p.

7. Nalimov V. V. Chelovek i biosfera. 1983, no 2, pp. 115—140.

8. Puchkov L. A., Ayurov V. D. Sinergetika gorno-tekhnologicheskikh protsessov (Synergetics of mi-

ning — technological processes), Moscow, MGGU, 1997, 264 p.

9. Seber Dzh. Lineynyy regressionnyy analiz (Linear regression analysis), Moscow, Mir, 1980, 410 p.

10. Solov'ev A. E., Kupriyanov V. V. Gornyy informatsionno-analiticheskiy byulleten'. 2010, no 5, pp. 297—301.

11. Khan G. Shapiro S. Statisticheskie modeli v inzhenernykh zadachakh (Statistical models in engineering problems), Moscow, Mir, 1980, 395 p.

12. Khimmel'blau D. Analiz protsessov statisticheskimi metodami (Process analysis by statistical methods), Moscow, Mir, 1983, 957 p.

13. Shek V. M., Pasechnik I. A. Gornyy informatsionno-analiticheskiy byulleten'. 2010, no 5, pp. 363—368.

14. Ezekiel M., Fox K. A. Methods of correlation and regression analysis. New York: Wiley, 1966, 430 p.

15. Hannan E. J. Time series analysis. New York: Wiley, 1977, 400 p.

16. Leeming J. J. Statistical methods for engineers. London, 1968. 390 p.

17. Lorence E. N. Atmospheric predictability is revealed by naturaly occurring analogues. Atmo-

spherics Science. 1969. Vol. 4, no 4. pp. 636—646.

Mining World Russia
Subscribe for our dispatch