Study On The Tunnel Fore Geological Forecast And Its Computer Assistant System

Various geological problems and hazards are the uppermost factors which often restricted tunnel construction; therefore geological forecast in tunnels are very important for the safety of tunnel construction. This paper researches the forecast of geological problems, the prediction of geological hazards during construction and synthetically forecast of geological condition ahead the workface with the disciplines of engineering geology, rock mass mechanics, geophysical exploration, digital image identification, nonlinear science theory and modern information technology. Finally, a computer assistant prediction software system for tunnel geological forecast has been obtained . The main results are as follows:(1) Base on the principles of TSP(Tunnel Seismic Prediction), GPR(Ground Penetrating Radar), TEM(Transient Electromagnetic Methods) and BEAM(Bore Tunneling Electrical Ahead Monitoring), typical examples on geological problems forecast are collected and analyzed. Then the seismic wave characters for empty cave, fault, hydrous crack by TSP, the electromagnetic wave characters for intact rock mass, rock with cracks, empty cave, hydrous cave by GPR, the response characters for hydrous fault, dry fault, hydrous cavity, dry cavity, hydrous crack,and low electric resistance by TEM, the response characters of hydrous cavity, hydrous crack, hydrous fault,dry fault by BEAM are analyzed and summarized. These results are helpful for creating elucidative criterion of tunnel geological problems forecast and improving prediction accuracy.(2) The application of digital image mode identify technology for tunnel geological problems forecast has been discussed. The intelligent identification methods of GPR image for cave has been explored by pre-deposal of the image and the artificial nervous network identify technology. A nervus network model which can intelligently identify hyperbola which means hydrous cave in GPR image has been gained. This method offers a new way for identifying other unfavorable geological problems in GPR image.(3) The relationships of rock burst and chamber physical dimension has been discussed. By analyzing the grade of rock burst with the condition of same dimension but different shape as well as the same shape but different dimension. Several conclusions have been gotten as fellows. Rock burst has the chamber shape effect. Ellipse is the best section shape of tunnel for reducing rock burst on condition of same chamber dimension. Rock burst has chamber dimensional effect. The grade of rock burst will enhance with the chamber dimension enlargement and finally stay at a constant on condition of same chamber shape. The methods were used to predict rock burst of a certain hydroelectric power station's and traffic tunnels. The result indicated that the grade of rock burst in the middle section of hydropower tunnels will increase 0.5 grade than traffic tunnel.(4) On the field of large deformation prediction, a method of predicting collapse time of by sidewall displacement-time monitoring data was expounded. The inflection point of displacement-time curve is regarded as the collapse time . the model of multistage nonlinear regression analysis has been created. The mathematic representation of surrounding rock collapse time has been solved. The parameters of the model were gained by fitting the data with least-squares procedure. The method is suit for critical prediction of surrounding rock collapse. The degree of reliability depends on the continuity of displacement-time monitoring data.(5) Advancing a new method for prediction of water gushing risk in front of the workface. A empirical equation about tunnel water gushing risk coefficient has been constructed. That isα=Log 10 ((A1×A2)/(B1×B2)), a is water gushing risk coefficient with no unit, A1 is thickness of wall rock, A2 is intensity of wall rock, B1 is degree of containing water, B2 is condition of water pressure. It judges degree of water contain from two factors: one is geologic factors such as rock character and geologic structure, the other is geophysical exploration parameters such as TSP, GPR and TEM. It is proved that the coefficient a can really show the water gushing risk degree in front of workface by some tunnel examples.(6) A principle of comprehensive prediction that"Take the geological analysis as the core, combine geological analysis and geophysical exploration, combine the condition inside and outside of the tunnel, combine long distance prediction methods with short distance prediction methods and fetch up mutually geophysical parameter"has been concluded. The system of geological comprehensive prediction in tunnel has been established. Then, the comprehensive prediction methods and workflows of common unfavorable geological condition in tunnel have been created. Through extracting geological and geophysical parameters, a comprehensive prediction model of geological problems has been constructed with fuzzy neural network. The comprehensive prediction system was used in TongLuoShan tunnel which leads to pretty good effect and ensure the tunnel's safety.(7) According to the theory of engineering geology, rock mass mechanics, geotechnical engineering monitoring, geophysical exploration and digital image identification, using nonlinear intelligent science technology such as artificial nervus network, fuzzy arrangement synthesis evaluate, fuzzy neural network, utilizing data-base and software engineering technique, a software about tunnel geological prediction computer assistant system (which named TGP-CAS) has been developed. This software has four main functions. The first, it can manage and predict the basic geology conditions of tunnel. The second, it can forecast tunnel's geological hazards such as rock burst ,l arge deformation, rock slip and gushing water. The third, it can manage the four geophysical explorations such as TSP, GPR, TEM and BEAM. Finally, it can synthetically predict tunnel geological problems and synthetically query all kinds of information. The software was applied in some over-length and deep tunnels, the results indicate that it has many advantages such as easy management, high prediction precision.