| TBM construction is greatly affected by geological conditions.When encountered with bad geology,TBM was stuck and trapped occasionally because the tunneling parameters and tunneling strategies couldn’t be adjusted in time.If the change trend of surrounding rock conditions can be known and the adverse geological conditions can be warned during TBM excavation,it will greatly reduce the risk of construction.Therefore,based on the Zhuxi Reservoir diversion tunnel project and so on,this paper studies the intelligent identification and early warning method of surrounding rock in TBM tunneling based on rock muck and tunneling parameter information.As a direct product of TBM cutting rock mass,rock muck has abundant geological information.In this paper,based on the Zhuxi Reservoir diversion tunnel project and so on,some on-site images of rock muck were collected and the characteristics of rock muck were analyzed.Firstly,the correspondence between rock muck and rock mass integrity was summarized.Secondly,according to the characteristics of rock muck images,a new segmentation algorithm based on the improved marked watershed transform was proposed to extract and classify the muck image.Finally,the accuracy of the method was verified using 160 pieces of onsite rock muck images under different degree of rock mass integrity.The results show that the algorithm of rock mack image segmentation has good performance and ideal segmentation effect.And the average accuracies of A,B and C image classification are 96.3%,94% and 86.7%,respectively,indicating that the method can effectively realize automatic analysis and recognition of rock muck.The automatic recognition technology of rock muck is of great significance to the real-time judgment of surrounding rock conditions of the tunnel face and the early warning of poor geological conditions.In addition,the tunneling parameters generated by TBM construction are also the main bases to judge the surrounding rock conditions.Based on the statistical analysis and fitting of field data collected at several tunnels with different geological conditions and different types of TBM,a general relationship model between intact rock mass strength(namely rock strength)and FPI was established.At the same time,the concept of rock mass equivalent strength was put forward.Further,FPI was divided according to the equivalent strength estimated by the model.Combined with the analysis of engineering examples,the corresponding relationship between FPI and rock mass conditions was summarized,and the early warning limit values of FPI about extremely hard rock and weak or broken surrounding rock were given.Therefore,the real-time judgment and early warning can be made on the surrounding rock conditions of the tunnel face according to the FPI.Finally,the rock muck analysis was combined with the value of FPI to make a comprehensive judgment on the rock mass conditions.The method of surrounding rock identification,classification and early warning based on the information fusion of rock muck and tunneling parameter was proposed.According to the change of rock muck and FPI’s value,this method can quickly analyze the degree of boreability and integrity of the rock mass,judge geological conditions,and give early warning to the bad geology.At the same time,a real-time recognition and warning software of surrounding rock was designed.The software can display the value of FPI and its changing trend,the recognition results of rock muck image,the analysis results of surrounding rock and the adverse geological warning results on the display screen at TBM operation cabin in real-time.And the research results can be better applied to engineering practice.The real-time identification and early warning method of surrounding rock in TBM construction proposed in this paper is reliable and practical.From the viewpoint of site construction requirements,the analysis and early warning of surrounding rock in tunneling by TBM are carried out according to the rock muck and FPI,which provides technical support for TBM’s safe and efficient tunneling. |
PDF Full Text Request