Tunnel advance geological prediction method and device and electronic equipment

Abstract

The present application relates to the technical field of tunnel advanced geological prediction, in particular to a tunnel advanced geological prediction method, device and electronic equipment; the method comprises the following steps: acquiring a plurality of real-time images in front of a working face of a rock mass to be predicted for prediction; splicing the plurality of real-time images to obtain a prediction image to be processed; extracting features of the prediction image to obtain prediction features for prediction; inputting the prediction features into a pre-trained prediction model to obtain prediction information; the method provided in the present embodiment can realize prediction of abnormal structures in the working face, and the prediction information obtained through training of an artificial intelligence network has high accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tunnel advance prediction, in particular to a tunnel advance geological prediction method, device and electronic equipment. BACKGROUND

[0002] In the process of tunnel construction, it is usually necessary to collect the information of the structural surface of the working face in advance, further grasp the geological state of the tunnel construction site, so as to ensure the safety of the construction personnel and the rationality of the tunnel construction. At present, the most commonly used methods are geological radar, TSP and structural surface manual measurement. Based on the collected data, the geological information in front of the working face in the previous construction is obtained according to the manual experience. However, this method has the technical defects of low prediction efficiency and inaccurate judgment results due to manual judgment and identification. SUMMARY

[0003] In order to solve the above technical problems, the present application provides a tunnel advance geological prediction method, device and electronic equipment.

[0004] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0005] In a first aspect, a tunnel advance geological prediction method is provided. The method comprises: obtaining a plurality of real-time images in front of the working face of the rock mass to be predicted; splicing the plurality of real-time images to obtain a prediction image to be processed; extracting features of the prediction image to obtain prediction features; and inputting the prediction features into a pre-trained prediction model to obtain prediction information.

[0006] In a first implementation manner of the first aspect, obtaining a plurality of real-time images in front of the working face of the rock mass to be predicted and splicing the plurality of real-time images to obtain a plurality of prediction images to be processed comprises: inserting an ultrasonic scanning imaging probe into a target area of the hole wall in front of the working face of the rock mass inside the borehole, rotating, sliding or laterally tilting the ultrasonic scanning imaging probe, obtaining ultrasonic scanning imaging graphs of all positions and angles in the target area, filtering high-frequency and low-frequency noise signals of the ultrasonic scanning imaging graphs of all positions and angles, obtaining rock mass structure ultrasonic imaging profile graphs of all positions and angles, and splicing the rock mass structure ultrasonic imaging profile graphs of all positions and angles to obtain the prediction image to be processed.

[0007] In a second possible implementation manner of the first aspect, in the second possible implementation manner, the extracting the features of the prediction image to obtain prediction features for prediction comprises: performing binaryzation processing on the rock mass structure ultrasonic imaging profile to obtain a target rock mass structure ultrasonic imaging profile; processing the target rock mass structure ultrasonic imaging profile according to a prediction state neural network meeting a network convergence requirement to obtain prediction features for prediction in the target rock mass structure ultrasonic imaging profile, the prediction features including prediction image features and prediction vector features.

[0008] In a third possible implementation manner of the second aspect, in the third possible implementation manner, the training method of the prediction model comprises: configuring a virtual training scene in a same environment as the to-be-predicted working face; performing regulation and control on the virtual training scene to obtain a plurality of feedback data; obtaining a state image in the virtual training scene; training the prediction model based on the state image; and the plurality of feedback data includes a plurality of abnormal data.

[0009] In a fourth possible implementation manner of the third aspect, in the fourth possible implementation manner, the prediction model comprises a convolutional neural network model and a behavior tree configured at an output end of the convolutional neural network model.

[0010] In a fifth possible implementation manner of the fourth aspect, in the fifth possible implementation manner, the convolutional neural network model comprises a first convolutional layer, a second convolutional layer, a first full connection layer, a second full connection layer, a splicing layer, and a third full connection layer.

[0011] In a sixth possible implementation manner of the fifth aspect, in the sixth possible implementation manner, the training method of the prediction model further comprises: extracting image features and vector features of the first state image, processing the image features through the first convolutional layer and the second convolutional layer to obtain first feature information, processing the vector features through the first full connection layer and the second full connection layer to obtain second feature information; processing the first feature information and the second feature information through the splicing layer to obtain third feature information; inputting the third feature information into the third full connection layer to obtain a prediction probability distribution, and inputting the prediction probability distribution into the behavior tree to obtain an abnormal behavior.

[0012] In a seventh possible implementation manner of the sixth aspect, in the seventh possible implementation manner, the training method of the prediction model further comprises: comparing the abnormal behavior with the feedback data to obtain a difference value, comparing the difference value with a preset difference value threshold, performing multiple times of training based on a comparison result until the difference value is less than the preset difference value threshold, and adjusting key parameters in the neural network model.

[0013] In a second aspect, the embodiments of the present application further provide a tunnel face disaster prediction device, comprising: an acquisition module configured to acquire a plurality of real-time images of a rock mass to be predicted for prediction; a splicing module configured to splice the plurality of real-time images to obtain a prediction image to be processed; a feature extraction module configured to extract features of the prediction image to obtain prediction features for prediction; and an identification module configured to identify the prediction features by using a pre-trained prediction model to obtain prediction information.

[0014] In a third aspect, the embodiments of the present application provide an electronic device, comprising: a memory configured to store executable instructions; and a processor configured to execute the executable instructions stored in the memory to implement the tunnel advanced geological prediction method.

[0015] In the technical scheme provided by the embodiments of the present application, the prediction model of the face is constructed to identify the acquired real-time images of the face to obtain prediction information, and the acquisition of the real-time images of the face is based on the ultrasonic probe arranged in the face and the image splicing method. The method provided in the embodiments can predict the abnormal structure in the face, and the prediction information obtained by training the artificial intelligence network has high accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] The methods, systems and / or programs in the accompanying drawings will be further described according to exemplary embodiments. These exemplary embodiments will be described in detail with reference to the drawings. These exemplary embodiments are non-limiting exemplary embodiments, in which the example numbers represent similar mechanisms in each view of the drawings.

[0018] Figure 1 is a block schematic diagram of the device provided by the embodiments of the present application.

[0019] Figure 2 is a flowchart of a tunnel advanced geological prediction method according to some embodiments of the present application. DETAILED DESCRIPTION

[0020] For better understanding of the above technical solutions, the technical solutions of the present application are described in detail below through the drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, and not limitations of the technical solutions of the present application. In the case of no conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0021] In the following detailed description, many specific details are set forth in order to provide a thorough understanding of the relevant teachings. However, it will be apparent to one skilled in the art that the present application can be practiced without these details. In other instances, well-known methods, procedures, systems, components, and / or circuits have been described at a relatively high level, without detail, in order to avoid unnecessarily obscuring aspects of the present application.

[0022] The flowcharts in the present application illustrate the execution processes performed by the system according to the embodiments of the present application. It should be explicitly understood that the execution processes of the flowcharts can not be executed in sequence. Instead, these execution processes can be executed in reverse order or simultaneously. In addition, at least one other execution process can be added to the flowchart. One or more execution processes can be deleted from the flowchart.

[0023] The technical solutions provided by the embodiments of the present application are mainly based on the identification of prediction information in ultrasonic images, and the abnormal information in the abnormal structure in the working face is obtained through computer processing technology.

[0024] The terminal device provided by the embodiments of the present application comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the tunnel advance geological prediction method based on ultrasonic images to identify and judge the prediction information in the abnormal structure in the working face.

[0025] In the present embodiment, the terminal can be a server, and the physical structure of the server comprises a memory, a processor and a communication unit. The memory, the processor and the communication unit are directly or indirectly electrically connected to each other to realize the transmission or interaction of data. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines. The memory is used to store specific information and programs, and the communication unit is used to send the processed information to the corresponding user end.

[0026] The embodiment divides the storage module into two storage areas, one of which is a program storage unit and the other is a data storage unit. The program storage unit is equivalent to a firmware area, and the read-write permission of the area is set to a read-only mode, and the data stored in the area cannot be erased and changed. The data in the data storage unit can be erased or read and written, and when the capacity of the data storage area is full, the newly written data will overwrite the earliest historical data.

[0027] The memory can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc.

[0028] The processor can be an integrated circuit chip having a processing capability of a signal. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; or can be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0029] Referring to Figure 2 In the embodiment, the tunnel running advanced geological prediction method comprises the following specific methods:

[0030] Step S210. Obtain a plurality of real-time images of the rock mass to be predicted for prediction.

[0031] In the embodiment, the obtaining of the image information of the tunnel face mainly based on the ultrasonic scanning mode, specifically, the ultrasonic scanning imaging probe is put into the target region of the hole wall of the rock mass inside the drill hole, the ultrasonic scanning imaging probe is rotated, slid or laterally inclined, the ultrasonic scanning imaging graphs of all positions and angles in the target region are obtained, the ultrasonic scanning imaging graphs of all positions and angles are filtered to remove high-frequency and low-frequency noise signals, and the rock mass structure ultrasonic imaging profile graphs of all positions and angles are obtained.

[0032] Step S220. The plurality of real-time images are spliced to obtain a prediction image to be processed.

[0033] In the embodiment, because the real-time images obtained in step S210 are rock mass structure ultrasonic imaging images of multiple positions and multiple angles, if the overall tunnel face anomaly prediction processing is to be performed, the image of the overall tunnel face needs to be constructed, and therefore the plurality of rock mass structure ultrasonic imaging images need to be spliced to obtain a global tunnel face image.

[0034] In the embodiment, the splicing method can be to label the plurality of rock mass structure ultrasonic profile graphs, configure the position information of the target region corresponding to the corresponding image in the label, and splice the images based on the position information. For example, the position information of the first obtained image is coordinate information, the second image also has coordinate information, the adjacent images are obtained based on the coordinate information, and the adjacent images are spliced to obtain the final global image.

[0035] Step S230. Features of the prediction image are extracted to obtain prediction features for prediction.

[0036] The feature extraction process further includes a binarization process of the final global image to obtain a gray-scale image, wherein the gray-scale image represents a target rock mass structure ultrasonic imaging profile graph.

[0037] The target rock mass structure ultrasonic imaging profile graph is processed according to a prediction state neural network that meets the network convergence requirement to obtain prediction features for prediction in the target rock mass structure ultrasonic imaging profile graph, and the prediction features include prediction image features and prediction vector features.

[0038] Step S240. The prediction features are input into a pre-trained prediction model to obtain prediction information.

[0039] In the embodiment, the process is a recognition process, and the recognition mode is based on the prediction model. The prediction model needs to be configured in advance, and the configuration of the prediction model is achieved based on an artificial intelligence training mode. The process includes the following methods.

[0040] Step S241. A virtual training scene in the same environment as the tunnel face to be predicted is configured.

[0041] In the embodiment, because the prediction model is obtained by training, and the training is based on a sample data set, the sample data set of the tunnel face is difficult to obtain, and because each tunnel face cannot collect data from other tunnel face structures due to environmental reasons and rock mass composition reasons, a method for constructing a sample data set, i.e., a method for obtaining a sample data set, is needed. In the embodiment, the sample data set of the tunnel face is based on field rock mass data collection, and then the rock mass stability is analyzed by simulation, a digital model of the same rock type in the same environment as the tunnel face to be identified is configured in the simulation model, a plurality of tunnel face stability evaluation data and tunnel face state images are obtained based on the simulation of specific disaster scenes in digital analysis, the above information is used to construct a sample data set for training, and the prediction model is trained based on the above sample data set and stability analysis, thereby obtaining a prediction model with high accuracy.

[0042] Step S242. A state image in the virtual training scene is obtained.

[0043] Step S243. The prediction model is trained based on the state image.

[0044] In the embodiment, the prediction model includes a convolutional neural network model and a behavior tree configured at the output end of the convolutional neural network model. The convolutional neural network model includes a first convolutional layer, a second convolutional layer, a first fully connected layer, a second fully connected layer, a splicing layer, and a third fully connected layer.

[0045] The training method of the prediction model further includes extracting image features and vector features of the first state image, processing the image features through the first convolutional layer and the second convolutional layer to obtain first feature information, processing the vector features through the first fully connected layer and the second fully connected layer to obtain second feature information, inputting the first feature information and the second feature information into the splicing layer to obtain third feature information, inputting the third feature information into the third fully connected layer to obtain a prediction probability distribution, and inputting the prediction probability distribution into the behavior tree to obtain an abnormal behavior.

[0046] The training method of the prediction model further includes comparing the abnormal behavior with the rock mass stability feedback data to obtain a difference, comparing the difference with a preset difference threshold, performing multiple training based on the comparison result until the difference is less than the preset difference threshold, and adjusting key parameters in the neural network model.

[0047] And, referring toFigure 1 In the embodiment, a device 100 is configured to perform the method described above, comprising: an acquisition module 110 configured to acquire a plurality of real-time images for prediction of the rock mass to be predicted; a splicing module 120 configured to splice the plurality of real-time images to obtain a prediction image to be processed; a feature extraction module 130 configured to extract features of the prediction image to obtain prediction features for prediction; and an identification module 140 configured to identify the prediction features by using a pre-trained prediction model to obtain prediction information.

[0048] It should be understood that, for the technical terms not explained above, those skilled in the art can determine their meanings without any doubt based on the above disclosure, for example, for some threshold values, coefficients, and the like, those skilled in the art can deduce and determine based on the logical relationship before and after, and the value range of these numerical values can be selected according to the actual situation, for example, 0.1-1, for example, 1-10, for example, 50-100, which are not limited herein.

[0049] Those skilled in the art can determine some preset, reference, predetermined, set, and preference label technical features / technical terms without any doubt based on the above disclosed content, such as threshold values, threshold intervals, threshold ranges, and the like. For some technical features and terms not explained, those skilled in the art can reasonably deduce based on the logical relationship before and after, so as to clearly and completely implement the above technical solutions. The prefix of the technical feature term not explained, such as "first", "second", "example", "target", and the like, can be deduced and determined without any doubt based on the context. The suffix of the technical feature term not explained, such as "set", "list", and the like, can also be deduced and determined without any doubt based on the context.

[0050] The above content disclosed by the embodiments of the present application is clear and complete for those skilled in the art. It should be understood that the process of deducing and analyzing the technical terms not explained by those skilled in the art based on the above disclosure is based on the content recorded in the present application, and therefore the above content is not a creative judgment of the overall scheme.

[0051] The above has described the basic concept, and it is obvious that the above detailed disclosure is only as an example for those skilled in the art, and does not constitute a limitation on the present application. Although it is not explicitly stated here, those skilled in the art can make various modifications, improvements, and corrections to the present application. Such modifications, improvements, and corrections are suggested in the present application, so such modifications, improvements, and corrections still belong to the spirit and scope of the exemplary embodiments of the present application.

[0052] Also, certain terminology can also be used in the description for the purposes of reference only, and thus are not necessarily limiting. For example, the terms "upper", "lower", "right", "left", "rear", "front", "rearward", "forward", "upward", "downward", "vertical", "horizontal", "up", "down", "top", "bottom", "lateral", "longitudinal", and "transverse" merely describe the relative position, direction, or orientation, in use or operation, of an item, individual parts or components, or portions of items, individual parts, or components, as the case can be, and are not otherwise intended to limit the items, individual parts or components, or portions of items, individual parts, or components, nor the positional, directional or orientational use or operation of the items, individual parts or components, or portions of items, individual parts, or components, to or as described only in the recitations of such terms.

[0053] Further, a person of ordinary skill in the art will recognize that the various aspects of the present application can be described in terms of certain terminology used to describe embodiments of the application. Such terminology includes, for example, terms such as "one embodiment", "an embodiment", and / or "some embodiments". It is also noted that such terminology is not necessarily describing the same embodiment(s) or aspect(s) of the application. Rather, such terminology is used by a person of ordinary skill in the art to describe different embodiments or aspects of the application. Moreover, it is further noted that certain features, structures, or characteristics described in the specification can be combined in any suitable manner in one or more embodiments or aspects of the application.

[0054] Computer-readable signal media can include a propagated data signal with computer- program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal can take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. Computer-readable signal media can be any computer-readable medium that is not a computer-readable storage medium and that can communicate, propagate, or transport program code. A computer-readable storage medium can be any computer-readable medium except for a propagated signal per se.

[0055] Computer program code for carrying out operations required by aspects of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, Javascript, or the like, conventional procedural programming languages, such as the "C" programming language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider), or cloud computing environment, or a service such as Software as a Service (SaaS).

[0056] Furthermore, the order of execution or performance of the operations of the aspects of the application illustrated and described herein is not essential, unless otherwise specified. That is, the operations can be performed in any order, unless otherwise specified, and the examples described herein will not limit or restrict the application in this regard. Additionally, some of the aspects of the application can be performed by hardware, software, firmware or combinations thereof, as can be readily understood by one of ordinary skill in the art. Also, it is noted that the examples can be described throughout this application in terms of functional or logical block components, but such description is not limited to literal hardware implementations. For example, a processing step, function, component or block can be implemented as hardware, software, firmware or combinations thereof.

[0057] It is also to be understood that the following description is only illustrative of the aspects of the present application and that no limitation of the scope of the application is intended by the description.

Claims

1. A method for advanced geological prediction of tunnels, characterized in that, The method includes: To acquire multiple real-time images of the rock mass in front of the tunnel face for prediction, specifically, the following steps are taken: inserting an ultrasonic scanning imaging probe into the target area of ​​the borehole wall in front of the tunnel face inside the borehole; by rotating, sliding, or tilting the ultrasonic scanning imaging probe laterally, acquiring ultrasonic scanning images of all positions and angles within the target area; filtering out high-frequency and low-frequency noise signals from the ultrasonic scanning images of all positions and angles to obtain ultrasonic imaging profiles of the rock mass structure at all positions and angles. The prediction image to be processed is obtained by stitching together multiple real-time images, specifically including: stitching together ultrasonic imaging profiles of the rock face structure at all positions and angles to obtain the prediction image to be processed; Extracting features from the forecast image to obtain forecast features for forecasting specifically includes: binarizing the ultrasonic imaging profile of the rock face structure to obtain the ultrasonic imaging profile of the target rock structure; processing the forecast state neural network according to the network convergence requirements to obtain forecast features for forecasting in the ultrasonic imaging profile of the target rock structure, wherein the forecast features include forecast image features and forecast vector features; The forecast features are input into a pre-trained forecast model for identification to obtain forecast information. The training method of the forecast model includes: configuring a virtual training scene under the same environment as the face to be forecasted; adjusting the virtual training scene to obtain multiple feedback data; and acquiring a state image under the virtual training scene; training the forecast model based on the state image; the multiple feedback data include multiple abnormal data.

2. The tunnel advanced geological prediction method according to claim 1, characterized in that, The prediction model includes a convolutional neural network model and a behavior tree configured at the output of the convolutional neural network model.

3. The tunnel advanced geological prediction method according to claim 2, characterized in that, The convolutional neural network model includes a first convolutional layer, a second convolutional layer, a first fully connected layer, a second fully connected layer, a splicing layer, and a third fully connected layer.

4. The tunnel advanced geological prediction method according to claim 3, characterized in that, The training method of the prediction model further includes: extracting image features and vector features of the first state image; processing the image features through the first convolutional layer and the second convolutional layer to obtain first feature information; processing the vector features through the first fully connected layer and the second fully connected layer to obtain second feature information; processing the first feature information and the second feature information through the concatenation layer to obtain third feature information; inputting the third feature information into the third fully connected layer to obtain a prediction probability distribution; and inputting the prediction probability distribution into a behavior tree to obtain abnormal behavior.

5. The tunnel advanced geological prediction method according to claim 4, characterized in that, The training method of the prediction model further includes: comparing the abnormal behavior with the feedback data to obtain a difference, comparing the difference with a preset difference threshold, performing multiple training based on the comparison results until the difference is less than the preset difference threshold, and adjusting the key parameters in the neural network model.

6. A tunnel advanced geological prediction device based on the tunnel advanced geological prediction method according to any one of claims 1-5, characterized in that, include: The acquisition module is used to acquire multiple real-time images of the rock mass to be predicted for prediction purposes. The stitching module is used to stitch together multiple real-time images to obtain the forecast image to be processed; The feature extraction module is used to extract features from the forecast image to obtain forecast features for forecasting. The identification module is used to identify the forecast features through a pre-trained forecast model to obtain forecast information.

7. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the tunnel advanced geological prediction method according to any one of claims 1 to 5.

Images