Ural State Mining University, Ekaterinburg, Russia:
B. D. Polovov, Professor, Doctor of Engineering Sciences
N. G. Valiev, Professor, Doctor of Engineering Sciences
K. V. Kokarev, Associate Professor, Candidate of Engineering Sciences, konstantin.kokarev@m.ursmu.ru

The modern geomechanics is based on substantial experimental and analytical research findings, generalized practical knowledge and on experience of mine management. Thereupon, a series of procedures is developed to perform geomechanical analyses with the specified methods to acquire and process information on state and properties of enclosing rock mass, and to design and control variouspurpose mining objects. The governing members in these procedures are the deterministic methods operating constant input parameters that are set based on statistical means with corrections. The deterministic methods, in view of the random nature of source information on enclosing rocks and their properties, offer insecure decisions on operating safety and efficiency. In this context, indubitably promising seems to be simulation modeling with Monte Carlo method. With a view to materializing capabilities of the Monte Carlo method, procedures are developed to improve quality of processing of source data on enclosing rocks for prompt simulation modeling, to refine standard simulation procedures and error reduction measures and to ensure reliability of modeling. As a result, a new approach to geomechanical analyses of mining objects is shaped, including combination of deterministic and probabilistic calculations, considerable improvement of quality of source data on enclosing rocks based on small samplings and methods of distribution-free statistics and differential construction of simulation models of geomechanical analysis: creation, debugging and demonstration in master “simulator” program with transition to “run-time versions”, with multi-stage screening of statistical and physical nature errors, with total number of generations up to 105–106. Such performance on common computers with small capacity of short-term memory is ensured by series implementation from 500 to 800 generations per cycle with the accumulation of data after each cycle completion. The time of simulation within 105 generations with plotting a bar chart of hash output and with quantitative estimation of levels of risks and reliability is no more than 25 min. The cloud simulation generator — a complete analog of simulation generator — is nearly unlimited in the number of generations. The high accuracy and calculating speed of the cloud computing is ensured by high short-term memory capacity, which enables not less than 1 million cycles within 50–200 s. On the whole, the put forward approach enhances reliability of a geomechanical analysis of mining objects in terms of estimation of risks of injuries, accidents and reduction in cost due to surplus reserves introduced in the conventional deterministic and probabilistic models. Application of the obtained results is ensured by a software system intended for mass use.

1. State Standard GOST 20522–2012. Soils. Methods of statistical treatment of test results. Moscow : Standartinform, 2013. 19 p. (in Russian)
2. Emblemsvåg J., Kjølstag L. E. Qualitative risk analysis: some problems and remedies. Management Decision. 2006. Vol. 44, No. 3. pp. 395–408.
3. Mun J. Modeling Risk: applying Monte Carlo simulation, real options analysis, forecasting, and optimization techniques. New Jersey : John Wiley & Sons, 2006. 623 p.
4. Yoe C. Principles of risk analysis: decision making under uncertainty. USA : CRC Press, 2011. 576 p.
5. Rejda G. E., McNamara M. Principles of risk management and insurance. Boston : Pearson, 2011. 745 p.
6. Berman A. F., Nikolaychuk O. A., Yurin A. Yu., Pavlov A. I. A methodology for the investigation of the reliability and safety of unique technical systems. Journal of Risk and Reliability. 2014. Vol. 228. pp. 29—38.
7. Polovov B. D., Valiev N. G., Kokarev K. V. Geomechanical analysis of mining units on basis of nonparametric statistics methods. Izvestiya vuzov. Gornyy zhurnal. 2016. No. 1. pp. 66–74.
8. Polovov B. D., Volkov M. N., Mikhaylov D. V., Romanov D. A., Tokarev P. V., Kharchenko D. I. Geomechanical analysis of mining facilities: points of growth. Izvestiya vuzov. Gornyy zhurnal. 2013. No. 7. pp. 40–50.
9. Kloke J., Mckean J. Nonparametric statistical methods using R. USA : CRC Press, 2014. 287 p.
10. Efron В. Bootstrap methods. Another look at the jackknife. The Annals of Statistics. 1979. Vol. 7, No. l. pp. 1–26.
11. State Standard GOST R 54500.3.1–2011. Uncertainty of measurement. Part 3. Guide to the expression of uncertainty in measurement. Supplement 1. Propagation of distributions using a Monte Carlo method. Moscow : Standartinform, 2012. Introduced: 01.10.2012. 82 p. (in Russian)
12. State Standard GOST R ISO 28640–2012. Statistical methods. Random variate generation. Introduced: 01.12.2013. Moscow : Standartinform, 2014. 39 p.(in Russian)