Babets D. Mathematical modeling of geomechanical processes in a technogenic disturbed environment with stochastic physical and mechanical properties

Dissertation for the degree of Doctor of Technical Sciences in the specialty 05.15.09 – "Geotechnics and Geomechanics". – National Technical University "Dnipro Polytechnic", Dnipro, 2021. The dissertation is devoted to solving an important scientific problem of improving the forecast of geomechanical processes in structurally and technogenic disturbed rock mass based on the development of probabilistic and statistical approaches and establishing pat-terns of changes in the state of the rocks at different stages of mining, taking into account the stochastic nature of physical and mechanical properties. Algorithms for taking into ac-count such patterns of rock mass disturbance as joint spacing, crack inclination angle, joint conditions, and rock water saturation has been improved. Based on three-dimensional mod-eling the dependences of the displacement of the contour of long underground excavation and the height of the zone of inelastic deformations in its vicinity on the depth of mining, rock strength, joint spacing, joint condition index, and distance to the face at which the movements are stabilized. A method of multifactor structural identification based on the group method of data handling (GMDH) was improved, which allowed to determine the most dominant factors and build a prognostic model of the stability of the preparatory exca-vation conjugated with longwall. The geomechanical substantiation of excavations reuse under conditions of the jointed mass is executed. The problem of nonlinear optimization is solved by the method of generalized reduced gradient and the optimal combination of factors is established, which ensures the preservation of the maximum cross-sectional area of pre-paratory excavation. Key words: excavation stability, probabilistic-statistical model, mathematical model-ing, rock mass strength, excavation reuse, structural factor, finite element method, predictive models