CN116341370A - TBM tunneling rock mass quality rapid determination method - Google Patents
TBM tunneling rock mass quality rapid determination method Download PDFInfo
- Publication number
- CN116341370A CN116341370A CN202310174462.4A CN202310174462A CN116341370A CN 116341370 A CN116341370 A CN 116341370A CN 202310174462 A CN202310174462 A CN 202310174462A CN 116341370 A CN116341370 A CN 116341370A
- Authority
- CN
- China
- Prior art keywords
- tbm
- rock mass
- parameters
- tunneling
- mass
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000011435 rock Substances 0.000 title claims abstract description 97
- 230000005641 tunneling Effects 0.000 title claims abstract description 72
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000013135 deep learning Methods 0.000 claims abstract description 31
- 238000013528 artificial neural network Methods 0.000 claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 26
- 239000002893 slag Substances 0.000 claims abstract description 17
- 230000035515 penetration Effects 0.000 claims abstract description 11
- 230000006870 function Effects 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000004075 alteration Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 4
- 239000004927 clay Substances 0.000 claims description 4
- 239000003673 groundwater Substances 0.000 claims description 4
- 239000000463 material Substances 0.000 claims description 4
- 210000002569 neuron Anatomy 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000003064 k means clustering Methods 0.000 claims description 2
- 238000005259 measurement Methods 0.000 abstract description 3
- 239000000945 filler Substances 0.000 abstract 1
- 238000005070 sampling Methods 0.000 abstract 1
- 238000010276 construction Methods 0.000 description 9
- 230000015572 biosynthetic process Effects 0.000 description 3
- 230000000903 blocking effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 206010057175 Mass conditions Diseases 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000009022 nonlinear effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21D—SHAFTS; TUNNELS; GALLERIES; LARGE UNDERGROUND CHAMBERS
- E21D9/00—Tunnels or galleries, with or without linings; Methods or apparatus for making thereof; Layout of tunnels or galleries
- E21D9/06—Making by using a driving shield, i.e. advanced by pushing means bearing against the already placed lining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mining & Mineral Resources (AREA)
- Evolutionary Biology (AREA)
- Geology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Environmental & Geological Engineering (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geochemistry & Mineralogy (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Excavating Of Shafts Or Tunnels (AREA)
Abstract
The invention provides a method for quickly determining the mass of TBM tunneling rock mass, and relates to the technical field of surrounding rock mass measurement. The invention discloses a method for rapidly determining the quality of surrounding rock mass in front of a face by utilizing an artificial neural network model based on real-time working parameters of TBM. In an excavated TBM tunnel, collecting TBM tunneling parameters including parameters such as cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed, cutter head penetration and the like, collecting parameters such as filler roughness in rock slag, sampling and collecting ground stress information for performing ground stress inversion, photographing bare rock mass between the TBM cutter head and a segment installing machine by using a circular camera for collecting surrounding rock mass parameter information, manually classifying and marking the above various data as training samples, and performing deep learning training by using RBF (radial basis) neural network based on the training samples and the Barton Q system classification, so that the surrounding rock mass quality rapid determination method based on the TBM tunneling parameters can be finally obtained.
Description
Technical Field
The invention relates to the technical field of surrounding rock quality measurement, in particular to a method for quickly determining the quality of TBM tunneling rock mass.
Background
The deep arming to the earth is a strategic and technological problem which we need to solve, so that the China is continuously developed to the deep in the aspects of traffic, water conservancy and hydropower, mine and the like. At present, china becomes the first major country of the world tunnel, and plentiful engineering experience is accumulated in the field of tunnel design, and a large number of deep tunnels are selected to be excavated by TBM tunneling. The TBM method has become the mainstream method in the construction of deep tunnel, and the method has the advantages of high tunneling speed, low construction cost, small construction disturbance, high safety and the like. But TBM construction is extremely sensitive to rock mass conditions, and the choice of tunneling parameters in the tunneling process is not reasonable, so that the tunneling efficiency of the TBM is greatly affected. When the TBM passes through a construction section with higher surrounding rock strength, tunneling control parameters such as total thrust, cutter torque, cutter penetration and the like need to be changed in time, otherwise abrasion of the cutter is accelerated, so that construction cost is increased, rock breaking difficulty is improved, and tunneling efficiency is reduced; when the tunnel boring machine passes through a construction section with low surrounding rock strength and poor stability, phenomena such as shield of TBM or machine blocking of a cutter head easily occur due to untimely adjustment of tunneling parameters. The essence of the phenomenon is that the cutter head rotates to cut the rock mass, so that the rock mass in front of the face can not be directly measured, and the quality of surrounding rock mass in front of the face to be excavated by the TBM is unclear, so that the TBM is excavated under the condition of unknown surrounding rock mass, and the TBM is provided with a certain blindness and subjectivity.
Therefore, the engineering rock mass is classified in time by acquiring accurate indexes such as surrounding rock mass quality in real time, the tunneling control parameters and the construction scheme are adjusted in a direct relation, and the method has important significance in preventing TBM from being blocked by a machine and cutter head from being worn too fast and improving tunneling efficiency. There are a number of engineering rock classifications in the world today: classification by uniaxial compressive strength of rock, stability of tunnel rock, integrity of rock mass, etc., but the above classifications are all based on direct measurements on site and are not comprehensive enough. However, the following problems remain in the class of the barton Q system currently in widespread international use:
1. the differences in formation production of the excavated cavity, i.e. the inclination and strike of the formation, are not taken into account.
The RQD value in the 2.Q system does not consider the relation between the geologic body structural plane occurrence and the tunnel axis direction, the Q system does not consider the unfavorable combination condition of rock strength and structural plane in consideration of factors affecting surrounding rock stability, the RQD value is obviously affected by the structural plane direction, the directional relation between the tunnel boring direction and the dominant joint group is considered, and the unfavorable combination relation is particularly emphasized.
The main direction of the ground stress is not considered in the 3.Q system design method, the main directions of the ground stress are different, the stress concentration and the destruction position of the surrounding rock of the tunnel are necessarily different, and in the normal condition, tangential stress on the cross section after tunnel excavation is increased, radial stress is reduced, and axial stress is almost unchanged.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a rapid determination method for the mass of TBM tunneling rock mass. And collecting tunneling data, rock mass data and ground stress inversion on site, and combining deep learning of an artificial neural network to quickly and accurately determine the rock mass parameters of the TBM tunneling surrounding rock.
A TBM tunneling rock mass quality rapid determination method specifically comprises the following steps:
step 1: TBM tunneling parameters during TBM tunneling are collected, wherein the TBM tunneling parameters comprise cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration;
step 2: collecting rock slag generated in tunneling, so as to obtain rock slag information of the rock mass, wherein the rock slag information comprises filling materials, roughness and whether clay is contained in the rock mass;
step 3: determining the structural stress field direction of the engineering area according to the actually measured ground stress points of the engineering area, and performing three-dimensional ground stress field inversion analysis on the engineering area according to the actually measured ground stress inversion data to determine the structural stress field direction of the area;
step 4: a circular automatic camera is arranged between a TBM cutter disc and a pipe piece mounting machine to record shooting of the rock stratum, the record shooting is stored in computer software, and rock mass quality index RQD of the rock stratum and group number coefficient J of joint are manually carried out n Roughness coefficient J of joint r Coefficient of alteration of joint J a Coefficient of influence J of groundwater ω And the ground stress influence coefficient SRF, manually classifying, calculating the TBM tunneling parameters, the rock slag information and the ground stress inversion data in the steps 1,2 and 3, and marking the data as training samples;
the calculation formula is as follows:
wherein k is 1 Is the correlation coefficient of the included angle between the structural plane shape and the tunnel axis, k 2 A correlation coefficient that is the principal direction of ground stress;
step 5: establishing a deep learning artificial network identification model, and correlating six parameters of cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration with geomechanical characteristics and ground stress of the rock mass;
the RBF neural network is adopted to complete the establishment of the deep learning artificial network identification model; the RBF neural network is a three-layer neural network and comprises an input layer, an hidden layer and an output layer; the transformation from the input layer to the hidden layer is nonlinear, while the transformation from the hidden layer to the output layer is linear;
the RBF neural network activation function is expressed as:
wherein,,
for the p-th input sample; />
The central vector of the radial basis function of the node of the ith hidden layer; p=1, 2, …, P being the total number of samples; II x P -c i II is European norms; sigma is the variance of the gaussian function;
normalizing the TBM tunneling parameters in the step 1, namely uniformly converting the parameters into [0,1 ]]In the self-organizing selection center learning method, a supervised learning process is carried out, weights between hidden layers and output layers are solved, and firstly, h numbers are selectedThe center is a k-means cluster center, and for the radial basis of the Gaussian kernel function, the variance is solved by the formula:
c max for the maximum distance between the selected center points, the connection weight of the neurons between the hidden layer and the output layer is directly calculated by using a least square method, namely, the partial derivative of omega is solved for the loss function, so that the partial derivative is equal to 0, and a calculation formula is obtained:
the RBF neural network selects tunneling parameters of TBM and inversion ground stress data as input layer sample parameters, and the H value representing rock mass quality as output layer sample parameters;
step 6: performing deep learning training on the deep learning artificial network identification model in the step 5 by utilizing TBM tunneling parameters, rock slag information, ground stress inversion data and training samples;
training the deep learning artificial network identification model until the grids are converged, and obtaining the trained deep learning artificial network identification model;
step 7: testing the trained deep learning artificial network identification model;
and (3) inputting TBM tunneling parameters and ground stress data of a section of the excavated tunnel, and comparing whether the output H value and the actual surrounding rock mass difference value of the tunnel are within a set error threshold value range. If the identification requirement is met, performing step 8, if the identification requirement is not met, adding training sample data, adjusting RBF neural network training parameters, and repeating step 5-6 until a deep learning artificial network identification model meeting the actual engineering requirement is obtained;
step 8: uploading real-time TBM tunneling parameters and ground stress to a computer, analyzing and processing in a deep learning artificial network identification model, and obtaining an H value according to a calculation formula to finish rock mass quality prediction.
The beneficial effects of adopting above-mentioned technical scheme to produce lie in:
the invention provides a method for quickly determining the mass of TBM tunneling rock mass. Compared with the existing method, the method improves the classification of the Barton Q system, and considers the influence factors such as different formation shapes, the relation between the structure surface shapes and the tunnel axis, the main direction of the ground stress and the like; the method carries out deep learning through the RBF (radial basis function) neural network, has higher timeliness than other neural networks, accords with the characteristic of quick confirmation parameters, only needs artificial auxiliary operation when carrying out artificial neural network learning, and can be highly artificial and intelligent after the deep learning. The rock mass quality parameters in front of the tunnel face can not be measured directly in the TBM tunneling process, the rock mass between the TBM cutter head and the segment erector can be used as a training sample, the tunnel face tunneling parameters are used as input parameters (the cutter head rotating speed, the cutter head torque, the total propelling force, the propelling cylinder pressure, the tunneling speed and the cutter head penetration degree), the rock mass in front of the tunnel face is predicted, a basis is provided for rapid tunneling, and the problems of excessively rapid cutter wear or machine blocking and the like are avoided.
Drawings
FIG. 1 is a flow chart of a method for quickly determining the mass of a TBM tunneling rock mass in an embodiment of the invention;
FIG. 2 is a schematic view of the ground stress and tunnel axis in an embodiment of the present invention;
FIG. 3 is a schematic flow diagram of an artificial neural network in an embodiment of the invention;
fig. 4 is a conceptual diagram of TBM construction in an embodiment of the present invention.
Detailed Description
The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.
A TBM tunneling rock mass quality rapid determination method is shown in fig. 1, and specifically comprises the following steps:
step 1: TBM tunneling parameters during TBM tunneling are collected, wherein the TBM tunneling parameters comprise cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration;
in the embodiment, the cutter rotating speed, the propelling cylinder pressure, the tunneling speed, the cutter penetration, the cutter torque and the total propelling force can be obtained through monitoring through a TBM console.
Step 2: collecting rock slag generated in tunneling, so as to obtain rock slag information of the rock mass, wherein the rock slag information comprises filling materials, roughness and whether clay is contained in the rock mass;
in the embodiment, the residue collecting device is used for picking up the rock crushed by the cutter head, the infrared ranging mode is adopted for three-dimensional imaging of the rock residue to be detected, and the information of the filling material of the rock mass, the roughness of the joints, whether the rock mass contains clay and the like is identified.
Step 3: determining the structural stress field direction of the engineering area according to the actually measured ground stress points of the engineering area, and performing three-dimensional ground stress field inversion analysis on the engineering area according to the actually measured ground stress inversion data to determine the structural stress field direction of the area;
in the embodiment, each 30-50 m apart, the tunnel rock mass is sampled to obtain the ground stress data of the section, and the ground stress inversion is carried out on the data by utilizing the particle swarm-differential PSO-DE basic principle and FLAC3D modeling, as shown in figure 2;
step 4: a circular automatic camera is arranged between a TBM cutter disc and a pipe piece mounting machine to record shooting of the rock stratum, the record shooting is stored in computer software, and rock mass quality index RQD of the rock stratum and group number coefficient J of joint are manually carried out n Roughness coefficient J of joint r Coefficient of alteration of joint J a Coefficient of influence J of groundwater ω Manually classifying the ground stress influence coefficient SRF, calculating the data, TBM tunneling parameters, rock slag information and ground stress inversion data in the steps 1,2 and 3, and marking the data as training samples;
in the embodiment, the annular automatic photographing machine is installed to photograph the record of the excavated and unsupported bare rock stratum between the tunneling cutter head and the segment installing machine, the record is stored in the computer in real time, and the data in the period are manually recorded according to the rock mass quality index RQD and the jointCoefficient of group number J n Roughness coefficient J of joint r Coefficient of alteration of joint J a Coefficient of influence J of groundwater ω And performing value marking and classification combination on the six indexes of the ground stress influence coefficient SRF, and matching with TBM tunneling parameters, rock slag information and ground stress inversion data at the moment to form a training sample for use in training an artificial neural network.
The calculation formula is as follows:
wherein k is 1 Is the correlation coefficient of the included angle between the structural plane shape and the tunnel axis, k 2 Is the correlation coefficient of the principal direction of the ground stress. Details are shown in Table 1.
Table 1k1 and k2 coefficients
The values of the coefficients are as shown in tables 2 to 7.
TABLE 2 rock quality index RQD description and values
TABLE 3 number of joint groups (J) n ) Description and value
TABLE 4 Joint roughness coefficient (J) r ) Description and value
TABLE 5 Joint Change coefficient (J) a ) Description and value
TABLE 6 Water reduction coefficient of joints (J) ω ) Description and value
TABLE 7 Stress Reduction Factor (SRF) description and values
Step 5: establishing a deep learning artificial network identification model, and correlating six parameters including cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration with geomechanical characteristics and ground stress of a rock mass as shown in fig. 3;
in order to ensure quick learning and meet the real-time requirement while drilling to the greatest extent, an RBF neural network is adopted to complete the establishment of a deep learning artificial network identification model; the RBF neural network is a three-layer neural network and comprises an input layer, an hidden layer and an output layer; the transformation from the input layer to the hidden layer is nonlinear, while the transformation from the hidden layer to the output layer is linear;
the RBF neural network activation function is expressed as:
wherein,,
for the p-th input sample; />
The central vector of the radial basis function of the node of the ith hidden layer; p=1, 2, …, P being the total number of samples; II x P -c i II is European norms; σ is the gaussian variance.
In order to effectively utilize the characteristics of the S-shaped function to ensure the nonlinear effect of the network neurons and reduce the errors of the neural network, the TBM tunneling parameters in the step 1 are normalized, i.e. the parameters are uniformly converted into [0,1 ]]In the self-organizing selection center learning method, a supervised learning process is carried out, the weight between a hidden layer and an output layer is solved, firstly, h centers are selected as k-means clustering centers, and for the radial basis of a Gaussian kernel function, the variance is solved by a formula:
c max for the maximum distance between the selected center points, the connection weight of the neurons between the hidden layer and the output layer is directly calculated by using a least square method, namely, the partial derivative of omega is solved for the loss function, so that the partial derivative is equal to 0, and a calculation formula is obtained:
the RBF neural network selects tunneling parameters of TBM and inversion ground stress data as input layer sample parameters, and the characteristic rock mass H value is used as output layer sample parameters.
Step 6: and (3) performing deep learning training on the deep learning artificial network identification model in the step (5) by using TBM tunneling parameters, rock slag information, ground stress inversion data and training samples.
Training the deep learning artificial network identification model until the grids are converged, and obtaining the trained deep learning artificial network identification model; the sample data adopted during training are all parameters collected in the steps 1,2, 3 and 4;
step 7: testing the trained deep learning artificial network identification model;
and (3) inputting TBM tunneling parameters and ground stress data of a section of the excavated tunnel, and comparing the output H value with the actual surrounding rock mass difference value of the tunnel within a set error threshold value range as shown in fig. 4. If the identification requirement is met, performing step 8, if the identification requirement is not met, adding training sample data, adjusting RBF neural network training parameters, and repeating step 5-6 until a deep learning artificial network identification model meeting the actual engineering requirement is obtained;
step 8: uploading real-time TBM tunneling parameters and ground stress to a computer, analyzing and processing in a deep learning artificial network identification model, and obtaining an H value according to a calculation formula to finish rock mass quality prediction.
After the test is finished, the real-time TBM tunneling parameters, namely the cutter head rotating speed, the cutter head torque, the total propelling force, the propelling cylinder pressure, the tunneling speed, the cutter head penetration and inversion ground stress data of the tunnel being excavated are transmitted into a computer, and the surrounding rock quality H value in front of the face of the tunnel face can be predicted after calculation and analysis are carried out through the artificial neural network.
The foregoing description is only of the preferred embodiments of the present disclosure and description of the principles of the technology being employed. It will be appreciated by those skilled in the art that the scope of the invention in the embodiments of the present disclosure is not limited to the specific combination of the above technical features, but encompasses other technical features formed by any combination of the above technical features or their equivalents without departing from the spirit of the invention. Such as the above-described features, are mutually substituted with (but not limited to) the features having similar functions disclosed in the embodiments of the present disclosure.
Claims (7)
1. The method for quickly determining the mass of the TBM tunneling rock mass is characterized by comprising the following steps of:
step 1: TBM tunneling parameters during TBM tunneling are collected, wherein the TBM tunneling parameters comprise cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration;
step 2: collecting rock slag generated in tunneling, so as to obtain rock slag information of the rock mass, wherein the rock slag information comprises filling materials, roughness and whether clay is contained in the rock mass;
step 3: determining the structural stress field direction of the engineering area according to the actually measured ground stress points of the engineering area, and performing three-dimensional ground stress field inversion analysis on the engineering area according to the actually measured ground stress inversion data to determine the structural stress field direction of the area;
step 4: a circular automatic camera is arranged between a TBM cutter disc and a pipe piece mounting machine to record shooting of the rock stratum, the record shooting is stored in computer software, and rock mass quality index RQD of the rock stratum and group number coefficient J of joint are manually carried out n Roughness coefficient J of joint r Coefficient of alteration of joint J a Coefficient of influence J of groundwater ω Manually classifying the ground stress influence coefficient SRF, calculating the data, TBM tunneling parameters, rock slag information and ground stress inversion data in the steps 1,2 and 3, and marking the data as training samples;
step 5: establishing a deep learning artificial network identification model, and correlating six parameters of cutter head rotating speed, cutter head torque, total propelling force, propelling cylinder pressure, tunneling speed and cutter head penetration with geomechanical characteristics and ground stress of the rock mass;
step 6: performing deep learning training on the deep learning artificial network identification model in the step 5 by utilizing TBM tunneling parameters, rock slag information, ground stress inversion data and training samples;
step 7: testing the trained deep learning artificial network identification model;
step 8: uploading real-time TBM tunneling parameters and ground stress to a computer, analyzing and processing in a deep learning artificial network identification model, and obtaining an H value according to a calculation formula to finish rock mass quality prediction.
2. The method for quickly determining the mass of a TBM tunneling rock mass according to claim 1, wherein the formula calculated in the step 4 is:
wherein k is 1 Is the correlation coefficient of the included angle between the structural plane shape and the tunnel axis, k 2 Is the correlation coefficient of the principal direction of the ground stress.
3. The method for quickly determining the mass of the TBM tunneling rock mass according to claim 1, wherein the step 5 is specifically: the RBF neural network is adopted to complete the establishment of the deep learning artificial network identification model; the RBF neural network is a three-layer neural network and comprises an input layer, an hidden layer and an output layer; the transformation from the input layer to the hidden layer is nonlinear, while the transformation from the hidden layer to the output layer is linear.
4. A method for rapidly determining mass of a TBM-tunnelling rock mass as claimed in claim 3 wherein the RBF neural network activation function is expressed as:
wherein,,
for the p-th input sample; />
The central vector of the radial basis function of the node of the ith hidden layer; p=1, 2, …, P being the total number of samples; II x P -c i II is European norms; sigma is the variance of the gaussian function;
normalizing the TBM tunneling parameters in the step 1, namely uniformly converting the parameters into [0,1 ]]The number between them, in self-organizing selection center learning method, supervised learning process, solving hidden layer to inputFirstly, selecting h centers as k-means clustering centers, and solving a radial basis of a Gaussian kernel function by a formula:
c max for the maximum distance between the selected center points, the connection weight of the neurons between the hidden layer and the output layer is directly calculated by using a least square method, namely, the partial derivative of omega is solved for the loss function, so that the partial derivative is equal to 0, and a calculation formula is obtained:
5. a method for rapidly determining mass of a TBM tunnelling rock mass as claimed in claim 3 wherein the RBF neural network selects tunnelling parameters of the TBM, inversion ground stress data as input layer sample parameters and a value indicative of mass H as output layer sample parameters.
6. A method for rapidly determining mass of a TBM tunnelling rock mass as claimed in claim 1 wherein the training in step 6 is specifically: training the deep learning artificial network identification model until the grids are converged, and obtaining the trained deep learning artificial network identification model.
7. The method for quickly determining the mass of a TBM tunneling rock mass according to claim 1, wherein the step 7 is specifically: inputting TBM tunneling parameters and ground stress data of a section of the tunnel which is excavated, and comparing whether the output H value and the actual surrounding rock mass difference value of the tunnel are within a set error threshold value range or not; if the identification requirement is met, the step 8 is carried out, if the identification requirement is not met, training sample data are added, RBF neural network training parameters are adjusted, and the steps 5-6 are repeated until a deep learning artificial network identification model meeting the actual engineering requirement is obtained.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310174462.4A CN116341370A (en) | 2023-02-28 | 2023-02-28 | TBM tunneling rock mass quality rapid determination method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310174462.4A CN116341370A (en) | 2023-02-28 | 2023-02-28 | TBM tunneling rock mass quality rapid determination method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116341370A true CN116341370A (en) | 2023-06-27 |
Family
ID=86876680
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310174462.4A Pending CN116341370A (en) | 2023-02-28 | 2023-02-28 | TBM tunneling rock mass quality rapid determination method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116341370A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114692273A (en) * | 2022-03-29 | 2022-07-01 | 中铁工程装备集团有限公司 | TBM (tunnel boring machine) -oriented construction tunnel geological dictionary establishing method and system |
-
2023
- 2023-02-28 CN CN202310174462.4A patent/CN116341370A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114692273A (en) * | 2022-03-29 | 2022-07-01 | 中铁工程装备集团有限公司 | TBM (tunnel boring machine) -oriented construction tunnel geological dictionary establishing method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110109895B (en) | Surrounding rock grading combined prediction method suitable for TBM tunneling tunnel and application | |
| CN107605536B (en) | Coal and gas prominent real-time early warning device and method based on Multi-source Information Fusion | |
| CN108930539A (en) | A method of it is controlled based on BIM Tunnel Overbreak & Underbreak | |
| CN110889532B (en) | Intelligent selection and optimization method and system for tunnel excavation and support parameters | |
| CN106157181B (en) | Method for quickly grading soil surrounding rock in real time based on while-drilling parameters | |
| CN113685188B (en) | TBM tunneling optimization method based on physical characteristics of rock slag | |
| CN108868805A (en) | Shield method for correcting error based on statistical analysis in conjunction with XGboost | |
| CN108415079B (en) | Rock stratum interface delineation method based on rock drilling impact sound identification | |
| CN110805469A (en) | Stability grading method for construction tunnel face by mountain tunnel drilling and blasting method | |
| CN116341370A (en) | TBM tunneling rock mass quality rapid determination method | |
| CN116522692B (en) | Underground engineering surrounding rock structural feature in-situ detection and classification method | |
| CN113175302A (en) | Intelligent rock mass quality sensing small-sized drilling machine system and evaluation method | |
| CN115263432A (en) | Stability and safety monitoring and analyzing system for deeply-buried tunnel | |
| Agrawal et al. | Prediction of TBM disc cutter wear and penetration rate in tunneling through hard and abrasive rock using multi-layer shallow neural network and response surface methods | |
| CN114386290A (en) | Method for testing five-dimensional space effect caused by surface subsidence based on dual-mode shield construction | |
| CN115586086A (en) | Borehole wall instability analysis method based on big data | |
| CN111505733A (en) | Multi-source heterogeneous information fusion analysis and dynamic prediction method for occurrence of special geologic body of underground engineering fractured rock mass | |
| Guo et al. | Influence analysis and relationship evolution between construction parameters and ground settlements induced by shield tunneling under soil-rock mixed-face conditions | |
| CN109543268A (en) | The recognition methods of TBM propulsive force major influence factors based on kriging model | |
| CN116721223A (en) | Rock mass stability analysis method based on oblique photography and measurement while drilling technology | |
| CN112115537A (en) | TBM diggeability grading method | |
| CN110119522B (en) | Stability rating method for excavated rock slope damage risk analysis | |
| CN116378769A (en) | Mine ground pressure activity monitoring system | |
| LIANG et al. | Stratum identification based on multiple drilling parameters and probability classification | |
| CN115434696A (en) | Lithology identification database construction method and lithology identification method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |