Smectite content estimation system and smectite content estimation method

Abstract

The distribution of the smectite content can be easily obtained.SOLUTION: There are provided an acquisition unit that acquires spectral data which is a result captured by a camera that observes the inside of a borehole in a forward natural ground from a face of a mountain tunnel for each target wavelength, a storage unit that stores a learned model in which a relationship between spectral data and a smectite content is learned, and an estimation unit that obtains a smectite content corresponding to the spectral data by inputting the acquired spectral data to the learned model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a smectite content estimation system and a smectite content estimation method. [Background technology]

[0002] At mountain tunnel construction sites, when the ground contains a large amount of smectite, large displacements within the tunnel often occur during construction, and closed structures are often adopted. Furthermore, after the tunnel is put into service, deterioration of the surrounding ground may cause deformation such as ground swelling, making it necessary to carry out countermeasures. As a countermeasure for these issues, geological surveys are carried out as appropriate for railway tunnels, and the tunnel structure is determined based on the ground strength ratio, water-logging collapse potential, and smectite content. In many cases, tests are conducted on core samples collected during advance drilling during tunnel construction to determine the smectite content in the ground ahead. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-181204 Summary of the Invention [Problem to be solved by the invention]

[0004] However, it takes a relatively long time, about one month, to obtain test results for smectite content, and there are issues with this method, such as only being able to determine the smectite content at the "point" where the core sample was collected. While collecting core samples at multiple locations along the longitudinal direction of advanced drilling and shortening the intervals between these locations can provide a detailed understanding of the longitudinal distribution of smectite content, this increases the number of objects to be analyzed, and also increases the burden of conducting the tests.

[0005] The present invention has been made in view of the above circumstances, and its object is to provide a smectite content estimation system and a smectite content estimation method that can easily obtain the distribution status of the smectite content. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems, one aspect of the present invention is a smectite content estimation system having an acquisition unit that acquires spectral data obtained by photographing the inside of a borehole in the ground ahead from the face of a mountain tunnel for each target wavelength, a memory unit that stores a trained model that has learned the relationship between the spectral data and the smectite content, and an estimation unit that inputs the acquired spectral data into the trained model to obtain the smectite content corresponding to the spectral data.

[0007] Another aspect of the present invention is a smectite content estimation method executed by a computer, which includes acquiring spectral data resulting from images taken by a camera that observes the inside of a borehole in the ground ahead from the face of a mountain tunnel for each target wavelength, and inputting the acquired spectral data into a trained model that has learned the relationship between the spectral data and the smectite content, thereby obtaining the smectite content corresponding to the spectral data. [Effects of the Invention]

[0008] As described above, according to the present invention, the distribution of the smectite content can be easily obtained. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic perspective view showing a mountain tunnel construction site in which a smectite content estimation system S according to one embodiment of the present invention is used. [Figure 2] FIG. 2 is a schematic functional block diagram illustrating the functions of the estimation device 10. [Figure 3] 1 is a flowchart illustrating the operation of the smectite content estimation system. [Figure 4] This is an image of the inside of borehole B taken by camera C. [Figure 5] 1 is a graph showing spectral data obtained at a location within borehole B. [Figure 6] FIG. 10 is a diagram illustrating an example of an analysis image displayed on a display screen of a terminal device. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, a smectite content estimation system according to one embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic perspective view showing a mountain tunnel construction site in which a smectite content estimation system S according to one embodiment of the present invention is used. Borehole B is formed by advance drilling from the face K of tunnel T into the ground ahead.

[0011] Camera C is inserted into borehole B to observe the inner surface of the borehole. Camera C photographs the inner surface of borehole B by dispersing light for each wavelength of the observation target, and generates spectral data from the photographed results for each wavelength. The camera C may be any camera capable of obtaining spectral data. For example, the camera C may be a hyperspectral camera (hereinafter also referred to as HSI) or a multispectral camera. Here, the number of bands of a hyperspectral camera is generally greater than the number of bands of a multispectral camera, but either a hyperspectral camera or a multispectral camera may be used as long as spectral data of wavelengths that can be used to estimate the smectite content can be obtained. The camera C may also be a borehole camera having a hyperspectral camera or a multispectral camera and an illumination device, which may use, for example, a halogen light. A borehole camera is a camera that can take images along the longitudinal direction of a cylindrical borehole, thereby capturing images of the inner circumference (e.g., the entire circumference) of the borehole along the longitudinal direction of the cylindrical hole.

[0012] The estimation device 10 is connected to the camera C via a communication cable. The estimation device 10 may be, for example, a computer, a tablet terminal, or the like.

[0013] FIG. 2 is a schematic functional block diagram illustrating the functions of the estimation device 10. The estimation device 10 includes a communication unit 101, a memory unit 102, an acquisition unit 103, an estimation unit 104, an image generation unit 105, a control unit 106, an input unit 107, and an output unit 108. The estimation device 10 estimates the smectite content based on the image captured by the camera C, and outputs the estimation result.

[0014] The communication unit 101 is communicably connected to the camera C and receives various data output from the camera C. The storage unit 102 stores various types of data. For example, the storage unit 102 stores a trained model that has learned the relationship between spectral data and smectite content. Minerals have unique spectra. Samples are collected from drilling cores previously collected during advanced drilling, and the collected samples are photographed with a spectroscopic camera to obtain spectral data. The samples are then analyzed to determine the smectite content. In this way, a combination of spectral data and smectite content obtained from a single sample is generated as training data. Here, samples are collected from different positions in the drilling core, and multiple combinations of spectral data and smectite content obtained from the collected samples are used as training data and are used for learning by a learning device. The learning algorithm used by the learning device may be machine learning or deep learning. In this way, the learning device learns from data on combinations of spectral data and smectite content obtained from a large number of samples as training data, and generates a trained model that can output a smectite content corresponding to input spectral data when the spectral data is input. The trained model generated by the learning device in this way is input to the estimation device 10, and the memory unit 102 stores the trained model. Here, it is preferable to generate such a trained model in advance before actually photographing the borehole of the tunnel to be investigated for smectite content with camera C. When spectral data is obtained by photographing the inside of borehole B with camera C, the trained model will already be generated and stored in memory unit 102, allowing the smectite content to be determined quickly.

[0015] The acquisition unit 103 acquires spectral data as forward natural ground information from the camera C via the communication unit 101. The forward natural ground information is data including position information and spectral data at the position. The spectral data is data representing reflectance for each wavelength. The acquisition unit 103 acquires the spectral data (forward natural ground information) together with position information indicating the position from the face within the borehole B where the spectral data was obtained. The position information is data indicating a position on the inner surface of the borehole B, expressed as a two-dimensional position based on the circumferential and longitudinal directions of the borehole B. The camera C may use data indicating a correspondence between a position based on the circumferential and longitudinal directions of the borehole B and a position in the captured image as position data, and generate spectrum data for each of this position data. This makes it possible to determine the position within the borehole B from the position of the image (e.g., the position of a pixel) in the captured image by the camera C.

[0016] The estimation unit 104 inputs the acquired spectral data into the trained model, obtains the smectite content according to the spectral data, and estimates the smectite content. The estimation unit 104 estimates the smectite content according to the spectrum data for each position indicated by the position information.

[0017] The image generating unit 105 generates an image showing the relationship between the smectite content estimated by the estimating unit 104 and the position indicated by the position information.

[0018] The control unit 106 controls each unit of the estimation device 10 . The input unit 107 acquires input data from an input device such as a keyboard or a mouse in response to an operation input to the input device. The output unit 108 outputs various data to an external device. For example, the output unit outputs the output data to a display device, thereby displaying the various data on a display screen of the display device. The display device may be, for example, a liquid crystal display device.

[0019] Next, the operation of the smectite content estimation system will be described with reference to Fig. 3, which is a flowchart illustrating the operation of the smectite content estimation system. The camera C takes an image of the inside of the borehole B (step S101), and generates position information and spectrum data at that position (step S102). Here, the camera C can also obtain a photographed image.

[0020] Figure 4 shows an image of the inside of borehole B taken by camera C. The vertical axis represents the circumferential position of the borehole, and the horizontal axis represents the longitudinal position of the borehole. On the vertical axis, "D" represents the downward direction relative to the central axis of the borehole along its longitudinal direction, and similarly, "L" represents the left direction, "R" represents the right direction, and "U" represents the upward direction. From such an image, the relationship between the position on the inner surface of the borehole and the image results at that position can be understood based on the circumferential position and the longitudinal position of the borehole. Figure 5 is a graph showing spectral data obtained at a certain position in borehole B. The horizontal axis represents wavelength, and the vertical axis represents reflectance. This graph shows the reflectance for each wavelength at the position where the image was taken in borehole B.

[0021] The acquiring unit 103 acquires the position information and the spectrum data generated by the camera C as forward natural ground information from the camera C via the communication unit 101 (step S103). The acquiring unit 103 can also acquire the captured image generated by the camera C.

[0022] When the forward natural ground information is obtained, the estimation unit 104 reads out the trained model stored in the memory unit 102 and inputs the spectral data into the trained model to obtain the smectite content (step S104). Here, the estimation unit 104 inputs the spectral data for each position indicated by the position information into the trained model, thereby obtaining the smectite content for each position indicated by the position information.

[0023] Based on the smectite content at each position within borehole B obtained by the estimation unit 104, the image generation unit 105 generates an analysis image in which an image with a color tone corresponding to the smectite content is superimposed on a photographed image representing the inside of borehole B (step S105). The output unit 108 transmits the analysis image generated by the image generation unit 105 to an external terminal device such as a tablet or PC (step S106). This allows the analysis image to be displayed on the display screen of the terminal device such as a tablet or PC.

[0024] FIG. 6 is a diagram showing an example of an analysis image displayed on the display screen of a terminal device. In FIG. 6, the smectite content at each position is displayed in a different display mode depending on the smectite content, superimposed on the photographed image shown in FIG. 4. For example, the smectite content is displayed in a different display mode depending on the range of smectite content. The display mode may be any of colors, patterns, etc. In this example, different colors are displayed over the photographed image, allowing transparency, for different ranges of smectite content, such as "18% or less," "18-25%," and "25% or more." This allows the relationship between the position on the inner circumferential surface of the borehole and the smectite content to be easily understood.

[0025] According to the embodiment described above, a trained model is generated that has previously learned the relationship between spectral data and smectite content, and spectral data obtained by photographing the inner surface of a borehole formed by advanced drilling is input into the trained model, thereby enabling the smectite content to be determined early and immediately. For example, in the past, it took about 30 days to conduct advanced drilling, collect samples from the drill core, and obtain measurement results for the smectite content. However, according to the above-described embodiment, it is possible to shorten this time to about one-tenth or less, and by utilizing this shortened time, it is possible to plan support designs, countermeasures, etc. that are tailored to the construction situation.

[0026] Furthermore, since the time required to obtain measurement results for smectite content can be shortened, delays in tunnel construction work involving smectite-containing ground can be reduced, and the risk of increased construction costs due to delays can be avoided.

[0027] Furthermore, since spectral data can be acquired continuously in both the circumferential and longitudinal directions of two-dimensional positions based on the circumferential and longitudinal directions of the borehole, the smectite content of the foreground ground can be determined continuously and quickly for each linear position in the longitudinal direction. Furthermore, conventionally, the smectite content could only be investigated at positions sampled from drilling cores obtained by advanced drilling. However, according to the above-described embodiment, spectral data can be acquired at positions arranged in a plane based on the circumferential and longitudinal directions of the borehole, and the trained model can be used to determine the smectite content at each position in the plane.

[0028] The results of the smectite content determination can be displayed on a tablet display screen or a display device connected to a computer. Using the positional information of the spectral data, it is possible to determine which position on the borehole's inner surface the smectite content corresponds to. This allows the relationship between each position on the borehole's inner surface and the smectite content to be easily determined.

[0029] In addition, it becomes possible to adopt a support structure that corresponds to the smectite content at each position that is continuously arranged in the longitudinal direction, which can contribute to improving the quality of the tunnel structure.

[0030] The Sustainable Development Goals (SDGs) are 17 international goals adopted at the United Nations Summit in September 2015. The smectite content estimation system and smectite content estimation method according to this embodiment can contribute to achieving one of the 17 SDGs, for example, goal 9, "Build resilient infrastructure, promote inclusive and sustainable industrialization, and foster innovation."

[0031] In the above-described embodiment, the estimation device 10 is described as a terminal device such as a single computer, but at least one of the functions of the storage unit 102, the estimation unit 104, the image generation unit 105, etc. may be provided in a server device connected to the terminal device via a communication network. In this case, the server device may be a physical server or a cloud server provided by a cloud computing service.

[0032] The functions of the estimation device 10 in the above-described embodiment may be implemented by a computer. In this case, a program for implementing the functions may be recorded on a computer-readable recording medium, and the program may be loaded into a computer system and executed. Note that the term "computer system" as used herein includes hardware such as an OS and peripheral devices. Furthermore, the term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into a computer system. Furthermore, the term "computer-readable recording medium" may also include devices that dynamically store programs for a short period of time, such as communication lines used when transmitting programs via networks such as the Internet or telephone lines, or devices that store programs for a fixed period of time, such as volatile memory within a computer system serving as a server or client. The program may be for implementing only a portion of the above-described functions, or may be capable of implementing the above-described functions in combination with a program already stored in the computer system, or may be implemented using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0033] The program running on the estimation device 10 according to one aspect of the present invention may be a program that controls one or more processors, such as a central processing unit (CPU), to implement the functions described in the above-described embodiments and modifications of the present invention (a program that causes a computer to function). The term "computer" as used herein also includes quantum computers. Information handled by each of these devices may be temporarily stored in random access memory (RAM) during processing, and then stored in various storage devices such as flash memory and hard disk drives (HDDs), and may be read, modified, or written by the CPU or the like as needed. [Explanation of symbols]

[0034] 10 Estimation device 101 Communications Department 102 Storage section 103 Acquisition Department 104 Estimation part 105 Image generation unit 106 Control Unit 107 Input section 108 Output section Borehole B C Camera K Cutting edge S Smectite Content Estimation System T-Tunnel

Claims

1. an acquisition unit that acquires spectral data obtained by photographing the inside of a borehole in the natural ground ahead of the mountain tunnel for each target wavelength using a camera; a memory unit that stores a trained model that has learned the relationship between the spectral data and the smectite content; an estimation unit that inputs the acquired spectral data into the trained model to obtain a smectite content according to the spectral data; A smectite content estimation system having:

2. The acquisition unit Acquiring the spectral data together with position information indicating a position from a face within the borehole where the spectral data was obtained; The estimation unit A smectite content corresponding to the spectral data is estimated for each position indicated by the position information. The smectite content estimation system according to claim 1 .

3. an image generating unit that generates an image showing the relationship between the smectite content estimated by the estimation unit and the position indicated by the position information; The smectite content estimation system according to claim 2, comprising:

4. 1. A computer-implemented method for estimating smectite content, comprising: The spectral data is obtained from the borehole in the front ground of the mountain tunnel, which is photographed by a camera observing each wavelength of the target. The acquired spectral data is input to a trained model that has learned the relationship between spectral data and smectite content, thereby obtaining the smectite content corresponding to the spectral data. A method for estimating smectite content, comprising:

Images