Open-type TBM tunnel surrounding rock stability test device and stability identification method

Abstract

The invention relates to the technical field of tunneling, in particular to an open-type TBM tunnel surrounding rock stability test device and a stability identification method, and the method comprises the steps: constructing an open-type TBM surrounding rock stability test model, including a TBM test device and a simulated surrounding rock, calculating the power transmission efficiency in the TBM test device, and setting the test parameters of the TBM test device; adjusting simulated surrounding rock parameters based on the similarity ratio, and constructing a surrounding rock foundation in the model test; compared with the prior art, the method has the beneficial effects that the technical problem that the surrounding rock condition cannot be mastered during tunnel excavation is solved by designing the open-type TBM surrounding rock stability test, establishing the surrounding rock stability judgment database, training the Transform model through the database and utilizing the model to judge the tunnel surrounding rock stability during open-type TBM construction.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel boring, and particularly relates to an open TBM tunnel surrounding rock stability test device and a stability identification method. Background Art

[0002] With the rapid progress of technology, the full-face tunnel boring machine (abbreviation: TBM) is increasingly used in the fields of transportation, water conservancy, mining, etc. It has very significant advantages especially in the construction of long tunnels. Compared with the traditional drill and blast method for tunnel construction, TBM construction has the advantages of high construction efficiency, good safety, and little environmental impact.

[0003] However, after the tunnel is excavated by the drill and blast method, the visibility of the surrounding rock is better, and the deformation and stress parameters of the surrounding rock are relatively easy to obtain. Therefore, the stability identification of the drill and blast method tunnel is relatively fast and simple. On the contrary, during the TBM construction process, due to the shielding of the cutter head and the shield on the surrounding rock, the stability of the surrounding rock cannot be observed in time. At the same time, in actual projects, the current technology cannot realize the monitoring of the displacement, stress, etc. of the surrounding rock in the face and the shield section, which increases the difficulty of identifying the stability of the surrounding rock. For the above reasons, it is very difficult to identify the stability of the surrounding rock in the face and the shield section of the current TBM tunnel, which poses a huge challenge to the formulation of the support plan after the surrounding rock exits the shield.

[0004] For the stability identification of the surrounding rock in the face and the shield section of the TBM tunnel, there are currently the following several methods: The first is to observe the stability of the surrounding rock after it exits the shield, and then indirectly infer the stability of the surrounding rock in the face and the shield section in combination with engineering experience; the second is to analytically solve the displacement of the tunnel surrounding rock based on the traditional elastoplastic mechanics theory to judge whether the surrounding rock is unstable; the third method is to use numerical simulation technology to establish a simulation model of the TBM tunnel surrounding rock, and analyze the displacement characteristics and stress changes of the surrounding rock in the shield section to predict the stability of the surrounding rock in the shield section.

[0005] The above methods have played a positive role in promoting the stability of the TBM tunnel surrounding rock. However, the existing methods for identifying the confining pressure stability of the TBM tunnel surrounding rock still have some problems and deficiencies, which are mainly manifested in the following aspects:

[0006] 1. The first method is extremely dependent on the experience of engineering personnel and belongs to post-judgment;

[0007] 2. The second method based on elastoplastic mechanics has been well verified for the applicability of the drill and blast method tunnel, but its applicability to the TBM construction tunnel has not been verified;

[0008] 3. The third method based on numerical simulation has a large amount of calculation when facing dynamic tunneling, and the calculation results have a lag in guiding construction.

[0009] 4. The tunneling parameters during the TBM tunneling process are not fully utilized.

[0010] Therefore, developing an open-type TBM tunnel surrounding rock stability test device and a stability identification method not only has urgent research value, but also has good economic benefits and industrial application potential, which is the driving force and foundation for the completion of this invention. Summary of the Invention

[0011] In order to overcome the defects of the existing technology pointed out above, the inventor of the present invention has conducted in-depth research and completed this invention after a large amount of creative work.

[0012] Specifically, the present invention constructs a TBM tunnel surrounding rock stability sample by using the method of model test in combination with tunneling parameters, and finally realizes the real-time identification of the stability of the TBM tunnel surrounding rock by using tunneling parameters, providing technical support for TBM construction.

[0013] To achieve the above object, the present invention provides the following technical solutions:

[0014] An open-type TBM tunnel surrounding rock stability identification method, comprising:

[0015] Constructing an open-type TBM surrounding rock stability test model, including a TBM test device and a simulated surrounding rock, calculating the power transmission efficiency in the TBM test device, and setting the test parameters of the TBM test device;

[0016] Adjusting the simulated surrounding rock parameters based on the similarity ratio to construct the surrounding rock foundation in the model test;

[0017] Determining the influencing factors of the tunnel surrounding rock stability during TBM construction, conducting simulation tests, and obtaining the influence of TBM power data on the surrounding rock stability;

[0018] Constructing a TBM surrounding rock stability discrimination database based on the test data and the surrounding rock stability discrimination conditions;

[0019] Training a Transformer model using the database to establish an open-type TBM tunnel surrounding rock stability identification model.

[0020] In the present invention, as an improvement, the simulated surrounding rock includes a simulated rock mass and a surrounding rock loading system, and the surrounding rock loading system simulates the in-situ stress of the surrounding rock;

[0021] Calculating the transmission efficiency in the TBM test device includes:

[0022] Calculating the output power after loss in the TBM test device;

[0023] Establishing the relationship between the output power and the cutter head rotation speed and the propulsion speed, and setting the test parameters based on this relationship.

[0024] In the present invention, as an improvement, the output power after loss is calculated as follows:

[0025] The total power transmission efficiency of the transmission component is the product of the efficiencies of the transmission components. Among them, the transmission efficiency of a single component is η n , where n is the number of single components, and the total power transmission efficiency of the transmission component is:

[0026] η d = η 1 n · η 2 n · η 3 n ·……· η m n

[0027] Among them, η m n is the transmission efficiency of n m components;

[0028] The total output power of the motor is:

[0029]

[0030] Among them, η d is the total power transmission efficiency of the transmission component, T is the output torque of the motor, and L is the lead of the lead screw;

[0031] The output torque T of the motor is calculated as follows:

[0032]

[0033] Among them, η m is the motor efficiency, P is the actual output power of the motor, and n is the motor speed;

[0034] From the above formula, the total output power of the motor after loss can be obtained as:

[0035]

[0036] In the present invention, as an improvement, establishing the relationship between the output power and the cutter head speed and the propulsion speed includes:

[0037] Set the rated output power of the motor to a% of the total output power, start the TBM device, and calibrate the corresponding cutter head speed and propulsion speed as Rev a and v a , adjust the rated output power of the motor to b% of the total output power, and calibrate the corresponding cutter head speed and propulsion speed as Rev b and v b, adjust the rated output power of the motor to n% of the total output power, and calibrate the corresponding cutterhead rotation speed and propulsion speed to be Rev n and v n , establish the relationship between the calibrated cutterhead rotation speed and propulsion speed and the total output power, and obtain the relationship between the rated output power of the motor and the cutterhead rotation speed and propulsion speed.

[0038] In the present invention, as an improvement, the influencing factors of tunnel surrounding rock stability during TBM construction include the in-situ stress of the surrounding rock, elastic modulus, and cohesion.

[0039] In the present invention, as an improvement, the discrimination conditions for surrounding rock stability during TBM construction include:

[0040] The surrounding rock stability grades are divided into four grades: stable, face instability, instability inside the shield, and instability outside the shield. The specific grade discrimination conditions are as follows.

[0041] The discrimination condition for surrounding rock stability is that there is no rock fall or collapse of the surrounding rock.

[0042] The discrimination condition for face instability is that the displacement of the face surrounding rock exceeds 5 cm.

[0043] The discrimination condition for instability inside the shield is that the displacement of the surrounding rock inside the shield section is greater than 5 cm.

[0044] The discrimination condition for instability outside the shield is that the displacement of the surrounding rock inside the shield section is less than 5 cm, and the displacement of the surrounding rock after exiting the shield is greater than 5 cm.

[0045] In the present invention, as an improvement, constructing a TBM surrounding rock stability discrimination database includes the following steps:

[0046] 1) Obtain test data based on the discrimination conditions for surrounding rock stability, including shield pressure, cutterhead torque, total thrust, propulsion speed, parameters of influencing factors of surrounding rock stability, and surrounding rock deformation data;

[0047] 2) Convert the test data into actual engineering data according to the similarity ratio;

[0048] 3) Obtain the TBM tunnel stability grade under the test parameter combination according to the surrounding rock deformation data in the test;

[0049] 4) Record the test data, actual engineering data, and surrounding rock stability grade into the database;

[0050] 5) Adjust the parameters of the influencing factors of tunnel surrounding rock stability, repeat the above steps 1-4, and record the obtained test data, actual engineering data, and surrounding rock stability grade into the database to construct the surrounding rock stability discrimination database.

[0051] A method for training a Transformer model includes:

[0052] Divide the data in the database into a training set, a validation set, and a test set, and use stratified random sampling to ensure data representativeness;

[0053] Select and adjust the structural parameters of the deep neural network to find the optimal parameters;

[0054] Use the database to train the deep neural network, optimize the deep learning algorithm, and update the weights and biases of the deep neural network;

[0055] Use the database to monitor and evaluate the deep neural network with metrics to verify the performance of the deep neural network;

[0056] According to the performance monitoring results, perform model parameter fine-tuning and model integration to train an open TBM surrounding rock stability identification model.

[0057] An open TBM tunnel surrounding rock stability test device includes a micro open TBM test device, a simulated rock mass, and a surrounding rock loading system. The micro open TBM test device includes:

[0058] A cutter head, detachably installed on a power transmission bearing, and rotated by a power assembly through the transmission bearing;

[0059] A propulsion and rotation power system, including a transmission component and a power assembly. The power generated by the power assembly is transmitted to the cutter head through the transmission component to push the cutter head to rotate forward;

[0060] A test platform, which is a steel structure frame and is connected to a ground anchor at the bottom;

[0061] A control and monitoring system. The monitoring system monitors the test data through instruments and transmits it to the control system, and the control system adjusts the operation of each test component;

[0062] The surrounding rock loading system is arranged outside the simulated rock mass to provide the rock mass boundary conditions and simulate the rock stress.

[0063] The surrounding rock loading system includes a loading test box, a reaction frame, and a jack. The simulated surrounding rock is placed in the loading test box, the reaction frame is placed outside the loading test box, and the jack is borne between the reaction frame and the test box.

[0064] Compared with the prior art, the beneficial effects of the present invention are:

[0065] (1) By designing an open TBM surrounding rock stability test, establishing a surrounding rock stability discrimination database, and training a Transformer model through the database, using the model to discriminate the surrounding rock stability in open TBM construction, the technical problem of being unable to master the surrounding rock conditions during tunnel excavation is solved.

[0066] (2) Based on the discriminant conditions of surrounding rock stability in the database and combined with test and actual engineering data, after deep learning of the model, the stability state of the surrounding rock during tunnel excavation can be judged quickly, accurately and in real time, providing a reference for determining the support measures after the surrounding rock exits the shield. Brief Description of the Drawings

[0067] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. In all the drawings, similar elements or parts are generally denoted by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0068] Figure 1 It is a schematic structural diagram of the micro TBM test device of the present invention;

[0069] Figure 2 It is a schematic structural diagram of the surrounding rock simulation system of the present invention;

[0070] Figure 3 It is a schematic structural diagram of the shield fixing device of the present invention;

[0071] Figure 4 It is a schematic flow structural diagram of the method for judging the stability of the surrounding rock of the present invention;

[0072] Figure 5 It is a schematic diagram of the data line type of the model test of the present invention;

[0073] Figure 6 It is a schematic diagram of the data line type of the actual engineering after the similarity ratio conversion of the present invention;

[0074] Figure 7 It is a schematic diagram of the data line type after the data noise reduction processing of the present invention;

[0075] In the figure, 1. Cutter head, 2. Shield, 3. Shield fixing device, 4. Test platform, 5. Lead screw, 6. Slide rail, 7. Propulsion motor, 8. Control panel, 10. Jack, 11. Reaction frame, 12. Model test box, 13. Simulated rock mass. Detailed Embodiments

[0076] The following will describe in detail the embodiments of the technical solutions of the present invention in conjunction with the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0077] An open TBM tunnel surrounding rock stability test device, comprising a micro TBM test device and a surrounding rock simulation system. The micro open TBM test device includes a shield 2, a cutter head 1, a power system and a control and monitoring system. The power system is divided into a rotary power system and a propulsion power system. The rotary power system controls the rotation of the cutter head 1, and the propulsion power system provides the equipment propulsion force. The whole test device is installed on a test platform 4.

[0078] The shield 2 is made of high-strength aluminum oxynitride alloy, and the transparent material is used to monitor the deformation of the surrounding rock inside the shield. The diameter of the shield 2 is smaller than that of the cutter head 1 to simulate the gap between the shield 2 and the surrounding rock. A stainless-steel baffle is installed at the front end of the shield 2, with a central hole in the baffle, and a central bearing for power transmission is installed in the hole. In the rotary power system, the rotary drive motor is connected to the cutter head 1 through this central bearing to drive the cutter head 1 to rotate. The cutter head 1 is composed of stainless steel and is of a detachable design. The cutter head 1 is connected to the central bearing by bolts.

[0079] The test fuselage is also provided with a shield fixing device 3, which is designed with a full circle plus a semi-circle. The shield 2 is fixed to the full-circle part of the shield fixing device 3 by bolts. The inner diameter of the full-circle part of the shield fixing device 3 is the outer diameter of the shield to ensure their tight combination. The whole shield fixing device 3 is welded by a support plate and a bottom plate. A hoop and a threaded hoop are fixed under the bottom plate. The hoop is connected to the sliding guide rail of the test platform 4 to ensure the propulsion direction of tunnel excavation; the threaded hoop is connected to the threaded steel shaft of the test platform 4, and the micro TBM tunnel boring machine can be advanced step by step by rotating the threaded steel shaft.

[0080] The propulsion power system consists of a sliding guide rail 6, a lead screw 5, a propulsion motor 7 and a transmission system. The fuselage of the micro TBM boring machine is sleeved on the sliding guide rail 6 through a hoop to control the boring direction. The lead screw 5 is installed on the test platform 4 and is connected to the upper fuselage of the boring machine through a threaded hoop. The propulsion motor 7 is welded on the test platform 4, and the power of the propulsion motor 7 is also transmitted to the lead screw 5 through gears and a synchronous belt. By changing the gear ratio, the rotation speed of the lead screw 5 can be controlled, and then the propulsion speed of the micro boring machine can be controlled.

[0081] The test platform 4 is made of steel structure, and the bottom of the test platform 4 is connected to a ground anchor to ensure the stability of the platform.

[0082] A main control system is provided on the test platform 4, including a control system and a monitoring system. The control system has a control panel 8. Signals are sent by the control system to control the operation and stop of the two motors and set the output power of the two motors. The forward and reverse rotation of the drive motor can be realized through the control system to simulate the forward movement of the micro TBM and the backward movement after the test is completed.

[0083] The monitoring system consists of a pressure cell and a monitoring camera. The pressure cell is used to monitor the shield pressure after the surrounding rock contacts the shield, and the monitoring camera monitors the displacement of the surrounding rock inside the shield section.

[0084] The surrounding rock loading system includes a jack 10, a reaction frame 11 and a model test box 12. The simulated rock mass 13 is placed in the model test box 12 to control the boundary of the rock mass, and the reaction frame 11 and the jack 10 are used to apply forces to the rock mass to simulate the in-situ stress of the tunnel rock mass.

[0085] An open TBM tunnel surrounding rock stability identification method, the method comprising:

[0086] Based on the above test device, an open TBM tunnel surrounding rock stability test model is constructed, the power transmission efficiency in the TBM test device is calculated, and the test parameters are obtained. The required power transmission efficiency includes the loss of output power and the relationship between the output power, the cutterhead rotation speed and the propulsion speed;

[0087] Determine the influencing factors of tunnel surrounding rock stability during TBM construction. The influencing factors are the stability index and mechanical parameter index of the surrounding rock itself, including in-situ stress, elastic modulus and cohesion, etc. Based on the similarity ratio principle, a simulated rock mass is constructed with the surrounding rock indexes in the actual project, and a surrounding rock loading system is used to provide the mechanical conditions of the surrounding rock to construct a surrounding rock foundation that meets the test requirements;

[0088] By conducting tests, obtain the test data of shield pressure, cutterhead torque, total thrust, propulsion speed and surrounding rock deformation, and obtain the influence of TBM power data on surrounding rock stability;

[0089] Based on the test data and the surrounding rock stability discrimination conditions, construct a TBM surrounding rock stability discrimination database;

[0090] Use the database to train the Transformer model to establish an open TBM tunnel surrounding rock stability identification model.

[0091] The output power loss is jointly determined by the motor output power and the transmission efficiency of the transmission components. The total transmission efficiency of the transmission components is:

[0092] η d = η 1 n · η 2 n · η 3 n ·……· η m n

[0093] where n is the number of single components, and η m n is the transmission efficiency of n m components.

[0094] According to the mechanical transmission principle, the total power output by the motor is:

[0095]

[0096] In the formula, T is the output torque of the motor, with the unit of N·m, and L is the lead of the lead screw, with the unit of m.

[0097] The output torque T of the motor is calculated as follows:

[0098]

[0099] Among them, η m is the motor efficiency, P is the actual output power of the motor, and n is the motor speed;

[0100] From the above formula, the total output power of the motor after loss can be obtained as:

[0101]

[0102] The relationship between the output power and the cutter head speed and the propulsion speed can be calibrated in the following way:

[0103] Different output powers correspond to different cutter head speeds and propulsion speeds. Adjust the rated output power of the motor to a% of the total power, and the corresponding calibrated cutter head speeds and propulsion speeds are Rev a and v a , adjust the rated output power of the motor to b% of the total power, and the corresponding calibrated cutter head speeds and propulsion speeds are Rev b and v b , adjust the rated output power of the motor to n% of the total output power, and the calibrated corresponding cutter head speeds and propulsion speeds are Rev n and v n . After multiple calibrations, the relationship between the rated output power of the motor and the cutter head speed and the propulsion speed is obtained, and a relational expression is established. Adjust the output power according to the required propulsion speed and cutter head speed in the test.

[0104] Before conducting the TBM tunnel surrounding rock stability test, it is also necessary to clarify the self-influence factors of the surrounding rock that affect the tunnel surrounding rock stability, including the in-situ stress on the surrounding rock, the elastic modulus of the surrounding rock, the cohesion, etc., which are the indexes affecting the surrounding rock stability. Based on the indexes, different index values are selected, and the orthogonal test method is used to design the test.

[0105] Based on the surrounding rock stability discrimination conditions, test data under different surrounding rock stability grades are obtained, including shield pressure, cutter head torque, total thrust, propulsion speed, and surrounding rock deformation data;

[0106] According to the similarity ratio, the test data is converted into actual engineering data;

[0107] According to the surrounding rock deformation data in the test, the TBM tunnel stability grade under the test parameter combination is obtained;

[0108] Incorporate the test data, actual engineering data, and the surrounding rock stability grades into the database to construct a discriminant database for surrounding rock stability.

[0109] Among them, the surrounding rock stability grades are defined as follows:

[0110] Based on past engineering cases, obtain the surrounding rock deformation data, analyze the surrounding rock deformation data to obtain the surrounding rock stability conditions, and divide the surrounding rock stability into 4 grades according to different surrounding rock stability conditions, namely stable, face instability, instability inside the shield, and instability outside the shield. The specific instability conditions are shown in Table 1:

[0111] Table 1 Judgment of surrounding rock stability grades

[0112]

[0113] Also incorporate the surrounding rock stability judgment conditions into the surrounding rock stability discriminant database.

[0114] Use the database to train the Transformer model. Through the backpropagation algorithm, update the weights and biases of the network to minimize the error between the output value of the network and the true value of the sample, and establish an open TBM construction tunnel surrounding rock stability identification model. The establishment of the TBM tunnel surrounding rock stability identification model includes the following steps:

[0115] Design the deep neural network structure and use the surrounding rock stability discriminant database to train the Transformer model;

[0116] Divide the dataset into a training set, a validation set, and a test set, and use stratified random sampling to ensure data representativeness;

[0117] Select and adjust the deep neural network structure parameters, including the loss function, optimizer, initial learning rate, number of nodes, and Patience, to find the optimal parameters. Among them, the loss function uses categorical cross-entropy, the optimizer uses Adam, and SGD is used to optimize the global search at specific stages;

[0118] Use the surrounding rock stability discriminant database to monitor and evaluate the performance of the deep neural network. The indicators include accuracy, precision, recall, and F1 score to verify the performance of the deep neural network;

[0119] According to the performance monitoring results, perform model parameter fine-tuning and model integration. The parameter fine-tuning adjusts the number of attention heads and the hidden layer size according to the performance feedback of the validation set. The model integration uses bagging or boosting techniques to combine multiple models to improve the prediction stability and accuracy;

[0120] Based on the surrounding rock stability discriminant database and the deep neural network training, obtain the TBM tunnel surrounding rock stability identification model.

[0121] The monitoring metrics of the deep neural network are calculated as follows:

[0122] Calculation of accuracy rate:

[0123]

[0124] Calculation of precision rate:

[0125]

[0126] Calculation of recall rate:

[0127]

[0128] Calculation of F1 score:

[0129]

[0130] In the above formula, Numberofcorrectpredictions is the number of correct samples, Total number of prediction is the total number of samples, TP is the true positive, FP is the false positive, and FN is the false negative.

[0131] The stability of the tunnel surrounding rock during open TBM construction is discriminated by using the surrounding rock stability discrimination model. The real-time data of the tunnel construction site is collected through sensors, and the above data is input into the open TBM construction tunnel surrounding rock stability discrimination model. After the model analysis, the surrounding rock stability grade is output in real time, and corresponding measures are taken based on the instability modes corresponding to different surrounding rock positions.

[0132] Example 1:

[0133] The specific steps in combination with this example are as follows:

[0134] (1) Construct the test similarity ratio

[0135] In the test, the geometric similarity ratio C l = 30 and the unit weight similarity ratio C γ = 1 are used as the basic similarity ratios. According to the similarity theory, the surrounding rock strain similarity ratio C ε , Poisson's ratio C v , friction angle similarity ratio stress similarity ratio C σ , elastic modulus similarity ratio C E , cohesion similarity ratio C c , displacement similarity ratio C δ are determined, as shown in Table 2:

[0136] Table 2 Model test similarity ratio

[0137]

[0138] Construct a simulated surrounding rock using the above parameters. For a certain on-site open TBM tunnel railway with a diameter of 10.3 m, the corresponding cutterhead diameter is 10.28 m, and the shield diameter is 10.3 m; the geometric similarity ratio is 30. Therefore, the cross-section of the model tunnel is 34.3 cm, the shield diameter of the micro open TBM test device is 34.3 cm, and the cutterhead diameter is 34.2 cm.

[0139] Select Class III surrounding rock as the prototype material with reference to the "Code for Design of Railway Tunnels".

[0140] Since it is a hard rock tunnel, four raw materials, namely cement, barite, and quartz sand, are used. Through indoor proportioning tests, the proportion of the surrounding rock similar material is cement : barite : quartz sand = 0.5 : 2 : 0.6, meeting the test requirements, as shown in Table 3:

[0141] Table 3 Physical and Mechanical Parameters of Surrounding Rock Materials

[0142]

[0143] The elastic modulus of the shield of the TBM is 210 GPa, and the Poisson's ratio is 0.3. According to the similarity ratio principle, the shield material of the TBM test device is made of aluminum oxynitride alloy, with an elastic modulus of 7 GPa and a Poisson's ratio of 0.29.

[0144] (2) Construct an experimental device for the stability of the surrounding rock of an open TBM tunnel

[0145] Carry out the test on the experimental device for the stability of the surrounding rock of the TBM tunnel. The experimental device for the stability of the surrounding rock of the open TBM tunnel mainly consists of a micro tunnel boring machine (abbreviated as micro TBM) and a surrounding rock loading system.

[0146] Among them, the traditional efficiency of the transmission system components of the simulation experimental device is shown in Table 4:

[0147] Table 4 Basic Parameters of the Transmission System

[0148] Component Quantity Transmission efficiency Synchronous belt 1 <![CDATA[η 1 > Sliding screw 1 <![CDATA[η 2 > Roller bearing 2 <![CDATA[η 3 > Sliding bearing 2 <![CDATA[η 4 >

[0149] The total efficiency of the transmission system is calculated by the following formula:

[0150]

[0151] Set the motor output power to 10% of the rated power, and the corresponding cutterhead rotation speed and propulsion speed are Rev1 and v 1 respectively. Set the motor output power to 20% of the rated power, and the corresponding cutterhead rotation speed and propulsion speed are Rev2 and v 2 respectively. Set the motor output power to 30% of the rated power, and the corresponding cutterhead rotation speed and propulsion speed are Rev3 and v3 , construct the relational expressions of cutterhead rotation speed, propulsion speed and motor output power based on the above relationships between rated power, cutterhead rotation speed and propulsion speed. The relational expressions are data summary expressions.

[0152] (3) Construct the TBM tunnel surrounding rock stability test database

[0153] Carry out the open-type TBM tunnel surrounding rock stability test and construct the test sample library, including the following sub-steps:

[0154] ① Carry out indoor model tests according to the parameters in Table 2, calculate the cutterhead torque, total thrust, cutterhead rotation speed and propulsion speed through the TBM tunnel surrounding rock stability model test and the above calculation formulas. The above calculation formulas are automatically embedded in the system, and the system automatically collects, displays and stores data during the test. The cutterhead torque, total thrust, propulsion speed, cutterhead rotation speed and shield pressure are respectively as Figure 5 shown in the model test data.

[0155] In the model test, the deformation value of the surrounding rock is collected by the micro-camera of the monitoring system for the images before and after tunneling, and the displacement values of the surrounding rock before and after tunneling are obtained through image processing technology. The final displacement value of the surrounding rock in this test is 0.1 cm.

[0156] ② According to the similarity ratio, convert the test data to the on-site tunnel construction level. The specific data is as Figure 6 shown.

[0157] Noise reduction is performed on the data. The data after noise reduction is as Figure 7 shown.

[0158] Remove the data in the early tunneling stage and during shutdown from the data after noise reduction, and then calculate the average values of the cutterhead torque, total thrust, propulsion speed, cutterhead rotation speed and shield pressure within this section of the test to obtain the data of this test, which are 5404 kN·m, 30190 kN, 27.1 mm / min, 4.9 rev / min and 37.3 bar respectively. The displacement value of the surrounding rock in the test is converted to the on-site surrounding rock displacement value of 3 cm according to the similarity ratio. At this time, the surrounding rock is stable.

[0159] The above are the data of a single test. Based on the data of multiple tests, establish a test sample library, which is used as a data set to provide data for subsequent training of the deep neural network.

[0160] (4) Based on the Transformer deep learning model, establish a TBM tunnel surrounding rock stability identification model, including the following sub-steps:

[0161] Design the deep neural network structure and use the TBM tunnel surrounding rock stability model test database to train the Transformer.

[0162] Both the decoder and the encoder have 6 layers, and the dimension of the hidden layer is 512. Multi-head attention mechanism: 8 heads per layer, 64 dimensions per head, and a total dimension of 512. Feed-forward network: dimension 2048, using the ReLU activation function. Positional encoding: Using a combination of sine and cosine functions to add the positional information of time series data.

[0163] Using the numerical sample library of the TBM tunnel surrounding rock stability model, the dataset is divided into a training set of 70%, a validation set of 20%, and a test set of 10%. Stratified random sampling is used to ensure data representativeness.

[0164] Select and adjust the structure parameters of the deep neural network, such as the loss function (categorical cross-entropy), optimizer (Adam), initial learning rate, number of nodes, Patience, etc. to find the optimal parameters.

[0165] Select and adjust the structure parameters of the deep neural network. The loss function uses categorical cross-entropy, and the optimizer uses Adam, combined with SGD to optimize the global search at specific stages.

[0166] The initial learning rate is set to 1×10 -4 , apply learning rate decay. If the validation set loss does not improve for 5 epochs, it is reduced by 10%. Implement the learning rate restart strategy to periodically reset the learning rate to promote jumping out of local minima.

[0167] Set Patience to 10 epochs, and the maximum number of training epochs is 100. It can be terminated early according to the early stopping strategy.

[0168] Using the experimental sample library of the TBM tunnel surrounding rock stability model, monitor and evaluate the performance of the deep neural network. The indicators include accuracy, precision, recall, and F1 score to verify the performance of the deep neural network.

[0169] Parameter fine-tuning, adjust according to the performance feedback of the validation set, such as the number of attention heads and the hidden layer size. Model integration uses bagging or boosting techniques to combine multiple models to improve the prediction stability and accuracy.

[0170] Based on the experimental sample library of the TBM tunnel surrounding rock stability model and deep neural network training, a TBM tunnel surrounding rock stability identification model is obtained.

[0171] (5) Using the surrounding rock stability identification model, realize the identification of the TBM tunnel surrounding rock stability, including the following sub-steps:

[0172] Collect real-time data of the TBM tunnel construction site through sensors, such as shield pressure, cutterhead torque, cutterhead thrust, propulsion speed, and geological parameters of rock mechanics, such as in-situ stress, elastic modulus, cohesion, etc.; input the above data into the TBM tunnel surrounding rock stability identification model, and output the stability level 1 or 2 or 3 or 4 in real time, corresponding to stable surrounding rock, face instability, instability inside the shield, and instability outside the shield respectively. Based on the stability level, on-site construction personnel can formulate corresponding engineering measures.

[0173] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the description of the present invention.

Claims

1. A method for identifying the stability of surrounding rock of an open TBM tunnel, characterized in that: include: Construct an open TBM surrounding rock stability test model, where the test model includes a TBM test device and simulated surrounding rock, calculate the power transmission efficiency in the TBM test device, and set the test parameters of the TBM test device; Adjust the simulated surrounding rock parameters based on the similarity ratio and construct the surrounding rock foundation in the model test; Establish the factors affecting the stability of tunnel surrounding rock during TBM construction, conduct simulation tests, and obtain the influence of TBM dynamic data on the stability of surrounding rock; Construct a TBM surrounding rock stability judgment database based on test data and surrounding rock stability judgment conditions; The Transformer model is trained using the database, and a surrounding rock stability identification model for open TBM tunnels is established.

2. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 1, characterized in that: Calculation of the transmission efficiency in the TBM test rig includes: Calculate the output power after losses in the TBM test device; The relationship between the output power and the cutter head rotation speed and propulsion speed was established, and the test parameters were set based on the relationship.

3. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 2, characterized in that: The output power after loss is calculated as follows: The total power transmission efficiency of the transmission component is the product of the transmission component efficiencies, where the transmission efficiency of a single component is η n , n is the number of single components, and the total power transmission efficiency of the transmission component is: or d =η1 n ·η2 n ·h3 n ·……·or m n Among them, η m n is the transfer efficiency of n m components; The total output power of the motor is: Among them, η d is the total power transmission efficiency of the transmission assembly, T is the motor output torque, and L is the lead of the screw; The motor output torque T is calculated as follows: Among them, η m is the motor efficiency, P is the actual output power of the motor, and n is the motor speed; From the above formula, it can be concluded that the total power output of the motor after loss is:

4. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 2, characterized in that: Establishing the relationship between output power and cutter head speed and propulsion speed includes: Set the rated output power of the motor to a% of the total output power, start the TBM device, and calibrate the corresponding cutter head speed and propulsion speed to Rev. a and v a , adjust the rated output power of the motor to b% of the total output power, and calibrate the corresponding cutter head speed and propulsion speed to Rev. b and v b , adjust the rated output power of the motor to n% of the total output power, and calibrate the corresponding cutter head speed and propulsion speed to Rev n and v n , establish the relationship between the calibrated cutter head speed and propulsion speed and the total output power, and obtain the relationship between the motor rated output power and the cutter head speed and propulsion speed.

5. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 1, characterized in that: The factors affecting the stability of tunnel surrounding rock during TBM construction include surrounding rock ground stress, elastic modulus and cohesion.

6. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 1, characterized in that: The conditions for judging the stability of surrounding rock in TBM construction include: The surrounding rock stability level is divided into four levels: stable, face instability, shield instability and shield outstability. The specific level judgment conditions are as follows: The condition for judging the stability of surrounding rock is that there is no block falling or collapse of surrounding rock; The criterion for determining the instability of the tunnel face is that the displacement of the surrounding rock of the tunnel face exceeds 5 cm; The judgment condition for instability in the shield is that the displacement of the surrounding rock in the shield section is greater than 5 cm; The criteria for judging instability after the shield is that the displacement of the surrounding rock inside the shield section is less than 5 cm, and the displacement of the surrounding rock after the shield is removed is greater than 5 cm.

7. The method for identifying surrounding rock stability of an open TBM tunnel according to claim 1, characterized in that: The construction of TBM surrounding rock stability judgment database includes the following steps: 1) Based on the surrounding rock stability judgment conditions, test data are obtained, including shield pressure, cutter head torque, total thrust, propulsion speed, parameters of factors affecting surrounding rock stability and surrounding rock deformation data; 2) Convert the test data into actual engineering data based on the similarity ratio; 3) Based on the surrounding rock deformation data during the test, the stability level of the TBM tunnel under the test parameter combination is obtained; 4) The test data, actual engineering data and surrounding rock stability grade are recorded in the database; 5) Adjust the parameters of factors affecting tunnel surrounding rock stability, repeat the above steps 1-4, enter the obtained test data, actual engineering data and surrounding rock stability grade into the database, and build a surrounding rock stability judgment database.

8. A Transformer model training method used in the identification method according to claim 1, characterized in that: include: The data in the database are divided into training set, validation set and test set, and stratified random sampling is used to ensure data representativeness; Select and adjust the structural parameters of deep neural networks to find the optimal parameters; Use the database to train the deep neural network, deep learning optimization algorithm, and update the weights and biases of the deep neural network; Use the database to monitor and evaluate deep neural networks with indicators and verify the performance of deep neural networks; Model parameters were fine-tuned and model integration was performed based on the performance monitoring results, and an open TBM surrounding rock stability identification model was obtained through training.

9. An open TBM tunnel surrounding rock stability test device, comprising a micro open TBM test device, a simulated rock mass and a surrounding rock loading system, characterized in that: The micro open TBM test device comprises: The cutter disc is detachably mounted on the power transmission bearing and is driven to rotate by the power assembly through the transmission bearing; The propulsion and rotation power system includes a transmission component and a power assembly. The power generated by the power assembly is transmitted to the cutter disc through the transmission component, driving the cutter disc to rotate and move forward; The test platform is a steel structure frame with the bottom connected to the ground anchor; Control and monitoring system: the monitoring system monitors the test data through instruments and transmits it to the control system, which adjusts the operation of each test component; The surrounding rock loading system is set outside the simulated rock mass to provide rock mass boundary conditions and simulate rock stress.

10. The open TBM tunnel surrounding rock stability test device according to claim 9, characterized in that: The surrounding rock loading system comprises a loading test box, a reaction frame and a jack, wherein the simulated surrounding rock is placed in the loading test box, the reaction frame is placed outside the loading test box, and the jack is supported between the reaction frame and the test box.

Images