Research On Dynamic Controlling Of Ground Pressure For Room And Pillar Mining And Application Of Artificial Intelligence

To the earth, mining is a production activity that human being destroys the stratum. During the mining, with the exploitation of useful ores, the original equilibrium state of the rockmass around the mined-out area is boken, subsequently the wall rock will come to a new equilibrium state through deformation, failure (crack, collapse, and so on) and stress redistribution. The processes of stress redistribution, deformation and failure of wall rock are called as Ground Pressure Phenomenon (GPP). In room-and-pillar stope, the GPP is mainly shown as the release of stress of stope roof, settlement deformation, collapse of stope roof, stress concentration in pillars, failure of pillars, and so on. Serious GPP, stope roof collapsing with large area or failure of too many pillars, will effect badly on mining of orebody by causing heavy losing of ore resources and deteriorating the economic benefit of mine enterprises. Thus, the ground pressure must be efficiently controlled in order to ensure that the orebody is mined out successfully. But, orebody and their wall rock were formed by endless course of rock formation and frequent geological activities, the special distribution of their original state and properties are asymmetric. That is, the mining objects (orebody) and their media environments are metamorphic. To different parts of one orebody, the embedding situations are different, and the engineering environment of stopes is continuously changed with the mining progress. Again, having the property to remember the load history, rockmass will show different mechanical affects (such as stress distribution, deformation and failure result) under different excavating order and load time. In a word, mining is a dynamic changing process, the ground pressure in stope is dynamic. Based on the understanding of the dynamic property of ground pressure in stope, with the dynamic viewpoint, this thesis studied and discussed in-depth the methods and also their implement steps for controlling ground pressure in stopes, and proposed scientific guidance for control of ground pressure in stopes of NanPing Ta-Ni mine in FuJian province and other similar mines. It is useful to improve security of mining and increase the recovery ratio and economic profit of mines to carry out dynamic control of ground pressure in stopes. In order to perform dynamic control of ground pressure in stopes, the following work should be done. Failure modes of pillars and stope roofs should be judged rapidly and correctly; Parametric indices of ground pressure and physical mechanical properties of rockmass should be obtained in time and continuously corrected and refined during mining; Stress and deformation of wall rock in stopes should be quickly analyzed; Measurements should be carried out to know real-time ground pressure in stopes. Therefor, this thesis made detailed study and discuss on these aspects. In the lingt of the facts that the failure modes of stope roofs and pillars are influenced by geological, engineering and production factors, and a great lot of uncertain reasoning is needed to gain a judge, and certain computation or theoretical analysis methods seldom gave satisfied results, the author has established an expert knowledge database of the failure modes of stope roofs and pillars in room-and-pillar stopes, and has proposed an method of expert system reasoning. Practices has showed that the knowledge database is scientifically reasonable and is feasible to judge the failure modes of stope roofs and pillars with expert system. Mechanical parameters of rocks are the fundament for performing mechanical analysis for rock engineering. Based on the fact that there is some mapping relation between the mechanical parameters of rock and the deformation of wall rock of roadway, using the strong self-learning and highly nonlinear mapping abilities of Artificial Neural Network (ANN), the author proposed a new method using the ANN to recognize the mechanical parameters of rock by back-analysis of the measured deformation of roadway. Also the feasibility of parametric recognition of rock and the stability of recoginized results are discussed in detail. Practical results showed that it is feasible to recognize the mechanical parameters of rock from the measured deformation of roadway with ANN. The new method can be fulfilled easily and cheaply. It is convenient for use in mine and is worth popularizing. Numerical simulation is one of the most powerful tools for quickly analyzing the stress and deformation of wall rock in stope. Based on practical experiences in mine, the author gave a detailed discussion on how to select the exact numerical methods and constitute the computational model and analyze the computing results for mechanical analysis for rock engineering. This gives a scientific guidance for applying numerical simulation methods in mining engineering. To realize dynamic control of ground pressure, the GPP occurred during mining process must be observed systematically. Based on long-time practical experiences of scientific research, the author give some good advices on how to lay observation system and carry out observation. Finally, with the above-mentioned methods for control of ground pressure in stopes, this thesis programmed the mining scheme of the orebody above the level 475m of the NO. 31 vein of NanPing Ta-Ni mine in FuJian province, and proposed scientific direction for mining.