A geological dynamic inversion method using a shield tunneling database

Abstract

The application discloses a geological dynamic inversion method using a shield tunneling database, and particularly relates to a large-diameter shield, a shield for constructing an underwater tunnel or a soft upper and hard lower stratum, a large-diameter shield excavation face stratum, an underwater tunnel geological survey accuracy difference, a soft upper and hard lower stratum soil rock interface determination difficulty, through project geological survey data, tunnel longitudinal section drawing and the like, the real geological survey situation cannot be accurately reflected, through collection and analysis of shield equipment data, muck information and a plurality of sensor information, data-driven stratum inversion is formed, new electronic maps are continuously generated in shield tunneling to guide shield construction and the like, and the intelligent construction of the tunnel is promoted.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of shield intelligent tunneling, and particularly relates to a geological dynamic inversion method using a shield tunneling database. BACKGROUND

[0002] The shield is widely applied in the construction of highways, railways and subways due to its advantages of high construction speed and good safety, but the inaccuracy of geological survey data brings great troubles to shield construction, especially the complex stratum of large-diameter shield excavation face, the difficulty of underwater tunnel survey, and the undulating interface of soft upper and hard lower stratum. The inaccuracy of geological survey data cannot accurately reflect the real geological survey situation through the geological survey data and tunnel longitudinal section diagram of the project, which directly restricts the improvement of shield construction efficiency, quality and safety.

[0003] How to use the data in shield tunneling to implement dynamic inversion of geology through the fusion of multi-element heterogeneous data on the basis of the prior art, and further realize the dynamic inversion method of shield tunneling and updating of tunnel longitudinal section geological information, is of great significance to the improvement of shield construction quality and efficiency, and is a problem to be solved by those skilled in the art. SUMMARY

[0004] To solve the above technical problems, the purpose of the present application is to provide a geological dynamic inversion method using a shield tunneling database, which solves the problem that the inaccuracy of existing shield construction geological survey data restricts the improvement of construction efficiency and safety.

[0005] To solve the above technical problems, the present application adopts the following technical scheme:

[0006] The geological dynamic inversion method using a shield tunneling database comprises the following specific steps:

[0007] S1, load sensors and data acquisition and transmission systems are configured in the shield cutter and soft soil cutter to measure, transmit and store the load of the shield cutter; at the shield slag outlet, video monitoring equipment is configured to collect slag soil images, and a convolutional neural network deep learning algorithm is used to identify and classify the slag soil;

[0008] S2, the geological survey data of the project and the geological longitudinal section information are used to form data and generate an electronic map, the electronic map displays the geological longitudinal section diagram of the project line, contains the stratum information and the basic physical force parameters of the stratum corresponding to the mileage, and the electronic map can be updated in real time according to the data;

[0009] S3, shield tunneling is performed, and the load data of the shield cutter and soft soil cutter, the slag soil image and classification information, the thrust, torque, tunneling speed and cutter head speed data of the equipment in shield tunneling are synchronously collected;

[0010] S4, analyze the cutter and soft soil cutter load data, take the average value of the load to draw the load change curve of the cutter at different positions of the excavation surface, analyze the load size of different cutters at different positions of the excavation surface, and then through the soil layer information at the shield slagging position and combined with the original image information, the stratum information in the tunneling is identified again, and the electronic map is updated and corrected;

[0011] S5, collect machine data such as thrust, torque, tunneling speed and cutter load information, stratum type of the excavation surface, proportion of the corresponding stratum type on the excavation surface, and strength data information of the corresponding stratum type, form a database for model training;

[0012] S6, randomly divide the database into two parts, one part of the sample as the training set, and the other part of the sample as the validation set, use multiple machine learning algorithms to train the training set data sample, compare the effects of different algorithms, select the machine learning algorithm with high prediction accuracy and accuracy, and form a learning model, and establish the relationship between the tunneling data and the stratum;

[0013] S7, use the real-time data of shield tunneling, through the trained tunneling data and stratum matching learning model, output the stratum information of the stratum type of the excavation surface and the stratum proportion of the excavation surface, realize the inversion of the stratum in the shield tunneling;

[0014] S8, use the inverted stratum information to generate an updated electronic map to provide guidance for construction.

[0015] The technical scheme of the present application has the following advantages compared with the prior art:

[0016] In view of the complexity of the stratum of the large-diameter shield excavation surface, the difficulty of rock and soil investigation of underwater tunnel, and the difficulty of exploration of soft and hard stratum boundary, a geological dynamic inversion method based on shield tunneling database is proposed, which can update the stratum information of the tunnel longitudinal section in time, that is, based on machine learning, the real-time data of shield tunneling are used to drive the production of stratum information of the tunnel longitudinal section, and the shield construction is guided, which has positive significance for intelligent construction of tunnel and intelligent excavation of shield. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The flowchart of the geological dynamic inversion method based on the shield tunneling database of the present application. DETAILED DESCRIPTION

[0018] The technical scheme of the present application will be described in detail below in combination with the drawings and examples, so that those skilled in the art can more clearly understand the technical scheme. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application.

[0019] A geological dynamic inversion method using a shield tunneling database, comprising the following steps:

[0020] s1, configure sensors and data acquisition and transmission system: configure load sensors and data acquisition and transmission system in the shield cutter and soft soil cutter to measure, transmit and store the load of the shield cutter. At the shield residue outlet, configure video monitoring equipment to collect residue image, and use deep learning algorithms such as convolutional neural network to identify and classify the residue.

[0021] s2, input project geological exploration information into electronic map: use the geological exploration data and geological profile information of the project to form data and generate an electronic map, which displays the geological profile graph of the project line, including the stratum information corresponding to the mileage, the basic physical force parameters of the stratum, and the electronic map can be updated in real time according to the data;

[0022] s3, shield construction, collecting tunneling data: shield tunneling, synchronously collecting shield cutter and soft soil cutter load data, residue image and classification information, and data such as thrust, torque, tunneling speed and cutter head speed during shield tunneling.

[0023] s4, analyze data and update electronic map: analyze the cutter and soft soil cutter load data, take the average load value to draw the load change curve of the cutter at different positions of the excavation surface, analyze the load size of different cutters at different positions of the excavation surface at different mileages, and then update and correct the electronic map by combining the soil layer information at the shield residue position with the original image information;

[0024] s5, form a training sample database: collect machine data such as thrust, torque, tunneling speed and cutter head speed during tunneling, as well as cutter load information, stratum type of the excavation surface, proportion of the corresponding stratum type on the excavation surface, and strength data information of the corresponding stratum type, to form a database for model training;

[0025] s6, form a machine learning model and establish a matching relationship between tunneling data and stratum: randomly divide the database into two parts, one part as the training set and the other part as the validation set, use multiple machine learning algorithms to train the training set data samples, compare the effects of different algorithms, select the machine learning algorithm with high prediction accuracy and accuracy, and form a learning model to establish the matching relationship between the tunneling data and the stratum;

[0026] s7, use real-time tunneling data to invert stratum information: use real-time data of shield tunneling, through the trained tunneling data and stratum matching learning model, output the stratum information of the stratum type and the stratum proportion of the excavation surface, and realize the inversion of the stratum during shield tunneling;

[0027] s8, generating a dynamic electronic map: using the inverted stratum information to generate an updated electronic map to guide the construction.

[0028] The above-mentioned geological dynamic inversion method can update the tunnel longitudinal section stratum information in time, form data-driven stratum inversion, continuously generate new electronic maps to guide the development of shield construction, etc., and has a promoting effect on the intelligent construction of tunnels.

[0029] The above-mentioned specific embodiments further explain the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-mentioned is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A geological dynamic inversion method using a shield tunneling database, characterized by: Comprise the following specific steps: S1, configure load sensors and data acquisition and transmission systems in the shield cutter and soft soil cutters to measure, transmit and store the load of the shield cutters; at the shield residue outlet, configure video monitoring equipment to collect residue soil images, and use convolutional neural network deep learning algorithm to identify and classify the residue soil; S2, use the geological exploration data and geological profile information of the project to form data and generate an electronic map, which shows the geological profile of the project route, including the stratum information and basic physical force parameters of the corresponding mileage, and the electronic map can be updated in real time according to the data; S3, carry out shield tunneling, and synchronously collect shield cutter and soft soil cutter load data, residue soil images and classification information, and thrust, torque, tunneling speed, cutter head speed data of the equipment in shield tunneling; S4, analyze the cutter and soft soil cutter load data, take the average load value to draw the load change curve of the cutter at different positions of the excavation face, analyze the load size of different cutters at different positions of the excavation face at different mileages, and then update and correct the electronic map by using the soil layer information at the shield residue outlet and combining the original image information to identify the stratum information again and update the electronic map; S5, collect machine data of thrust, torque, tunneling speed, and cutter head speed, as well as cutter load information, stratum type of the excavation face, proportion of the corresponding stratum type in the excavation face, and strength data information of the corresponding stratum type, form a database for model training; S6, randomly divide the database into two parts, one part as the training set and the other part as the validation set, use multiple machine learning algorithms to train the model with the training set data samples, compare the effects of different algorithms, select the machine learning algorithm with high prediction accuracy and accuracy, and form a learning model to establish the relationship between the tunneling data and the stratum; S7, use the real-time data of shield tunneling, through the trained tunneling data and stratum matching learning model, output the stratum type of the excavation face and the stratum information of the stratum proportion of the excavation face, realize the inversion of the stratum in shield tunneling; S8, use the inverted stratum information to generate an updated electronic map to provide guidance for construction.

Images