| Tunnel boring machine(TBM)has become the main equipment in tunnel and underground engineering construction because of its advantages of high excavation efficiency,small disturbance to surrounding rock,high quality of tunnel completion and strong safety.However,the selection and adjustment of TBM operating parameters and tunnelling direction in the past are lack of the science basis and mainly rely on human experience.It is difficult to adjust and match with the change of geological conditions in time.The main parameters related to the tunnelling and direction of TBM often do not match and adapt to the rock conditions,which lead to slow excavated speed,sharp increase in construction cost,abnormal wear of cutter head and cutter,even jamming,machine damage,engineering accidents and other serious consequences.Therefore,the research on optimization method of TBM operating parameters is of great significance for ensuring TBM safe,efficient and low-cost excavation,and has become a research hotspot in the field of international TBM construction.This paper puts forward the idea and implementation method of TBM operating parameters optimization and tunnelling direction based on field-tested geological information and intelligent method.The principle is to use high-precision imaging equipment,combined with detection information and artificial intelligence algorithm,the parameters such as the uniaxial compressive strength,the joint frequency and the rock muck size of the surrounding rock along the tunnel.Based on the above rock mass information,on the one hand,we carry out full-scale linear cutting tests and big data mining analysis,reveal the rock-machine interfeed law and relationship based on various intelligent algorithms,optimize the TBM tunnelling parameters under different geological conditions in steps with the optimization objectives of optimal specific energy and efficient tunnelling,and then establish a progressive step-by-step decision-making method for the main control parameters based on the rock-machine interfeed law.On the other hand,combined with the mechanical principle of TBM tunnelling direction control and the depth neural network method,the bidrived intelligent control model of TBM tunnelling direction is constructed.The intelligent decision-making of TBM tunneling parameters and the intelligent control method of tunneling direction have been successfully verified in practical projects.The main research work and achievements of this paper are as follows:(1)Fast perception method of rock mass parameters in the front of the tunnel face.Aiming at the rock uniaxial compressive strength,integrity and rock muck size,the in-situ fast perception method is studied.In terms of rock unixial compressive strength,taking the measured seismic wave velocity and TBM tunneling data as input,the BP neural network algorithm is used to mine the correlation of parameters,and the calculation model of rock mass compressive strength is established based on this.In terms of rock mass integrity,the joint frequency is used to represent the integrity of rock mass.Based on the multi vision imaging and image recognition algorithm,the distribution of joints on the surface of surrounding rock is captured,and the joint frequency of surrounding rock is calculated.In terms of rock muck information,taking the average size of rock muck as the key perception target,based on the linear array laser scanning method,the 3D point cloud image of belt conveyor with mucks is quickly captured,and then the average size of rock muck is calculated in real time,so as to establish a rapid recognition method of rock debris information based on linear array laser scanning.The rock mass data perceived by the above methods will serve as an important data basis.(2)Intelligent decision-making method for TBM operating parameters.The selection of main control parameters for TBM tunneling depends on human experience and is difficult to match the geological conditions of rock mass.In this paper,we propose an intelligent decisionmaking method for rock-machine feedbacks and main control parameters in TBM tunnelling.Based on the rock parameters and the main control parameters of TBM tunnelling obtained from the real-time sensing of the method in the first part of the study,the paper explores the correlation between these two types of parameters based on various machine learning algorithms,partitions the rock parameters according to the performance of multiple models in different rock conditions,and differentiates and weights each model to establish a rock-machine feeder relationship model,and step-by-step decision making of the main control parameters with the objective of optimal specific energy and boring efficiency,finally forming an intelligent decision making method of TBM boring main control parameters based on the law of rock-machine interfeed in boring,which has good accuracy and geological adaptability as verified by field actual measurement data.(3)In terms of the TBM tunnelling direction intelligent control method,aiming at the problem that TBM tunneling direction control depends on human experience and lacks scientific basis,this paper firstly establishes a theoretical model between the strokes of the left and right main gripper cylinders and the left and right reaction torque cylinders and the existing offset of the TBM under known rock mass conditions.Based on this model and combined with deep learning,the TBM attitude control scheme under different rock mass conditions and existing tunneling attitudes is formed,and finally the intelligent control method of TBM tunneling attitude is formed.Specifically,this method divied the TBM tunnelling direction control into two sub problems:horizontal and vertical attitude control.Taking the depth neural network as the method,taking the measured rock geological conditions and the existing horizontal and vertical angle and distance offset of TBM as the data basis and key input,and taking the stroke of the main thrust and counter torque cylinders as the control objectives,an intelligent TBM heading attitude control method is constructed.This method can give the stroke of main gripper and counter torque oil cylinder according to the geological conditions of rock mass and the existing tunnelling direction.By controlling the stroke of the gripper and counter torque cylinders,the tunnelling direction can be effectively adjusted.(4)The engineering verification of TBM operating parameter optimization method.The TBM operating parameter optimization model is verified by two methods of field test data deduction and field driving test.First of all,based on the Pearl River Delta water resources allocation project,in terms of data deduction,through the decision-making calculation of the intelligent decision-making of TBM tunneling parameters and the intelligent control method of tunneling direction under various designed working conditions,the changing rules of the decision-making results of operating parameters and direction control parameters with the geological information of rock mass is obtained,and it is theoretically analyzed to judge the rationality of the model decision-making.In terms of field test,relying on the the Pearl River Delta water resources allocation project,this paper selects test section at the mileage of SL 7+655～SL 7+396,and carries out the tunneling test with a total length of 259 meters.In the test section,based on the rock mass parameters obtained from the actual measurement,the intelligent decision method of TBM tunneling parameters is used to calculate the decision results of parameters such as penetration and cutter head speed ring by ring,and the recommended range is given for the reference of the main driver.At the same time,in the test section,the built TBM tunnelling direcion intelligent control method is also used to calculate the left and right support shoe cylinder stroke difference and the left and right counter torque cylinder stroke sum ring by ring.The results show that the experience tunnelling results of the main driver and the intelligent decision-making results of the model have high consistency and coincidence,which proves that the intelligent decision-making model of TBM tunneling parameters and the intelligent control method of tunneling direction established in this paper have certain scientific and reference value. |
PDF Full Text Request