Evaluation System and Evaluation Method

Abstract

To enable highly accurate evaluation of risks at a face.SOLUTION: An evaluation system 1 comprises a face evaluation index acquisition unit 11 that acquires at least two face evaluation indexes regarding a tunnel face among a crack evaluation index that indicates the condition of cracks, a weathering degree evaluation index that indicates the degree of weathering, a rock strength evaluation index that indicates the strength of rock mass, and a spring water amount evaluation index that expresses spring water amount, a shotcrete index acquisition unit 12 that acquires a shotcrete index indicating the thickness of concrete sprayed on the face, a risk event information acquisition unit 13 that acquires risk event information representing the risk occurrence probability, which is the probability of occurrence of an event related to risk at the face, and an evaluation model generation unit 14 that calculates an evaluation model expressing the relationship between the face evaluation index, the shotcrete index, and the risk event information using a predetermined statistical analysis method.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an evaluation system and an evaluation method.

Background Art

[0002] Techniques for evaluating the face condition of a tunnel are known. For example, Patent Document 1 discloses a technique for dividing a face image obtained by imaging the face into a mesh shape, extracting regions where cracks intersect from each of the divided regions, and evaluating the face condition based on the extracted regions.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Conventionally, in order to prevent the occurrence of events that pose risks in the face, such as spalling, the thickness of the sprayed concrete sprayed on the face has been determined based on the evaluation result of the face condition. However, conventionally, since the thickness of the sprayed concrete has been determined only according to the face condition, the probability of occurrence of risks corresponding to the thickness of the sprayed concrete has not been considered, and the accuracy of the evaluation has not been sufficient.

[0005] Therefore, an object of the present invention is to enable highly accurate evaluation of risks in the face.

Means for Solving the Problems

[0006] An evaluation system according to an aspect of the present invention is an evaluation system for evaluating risks at the face of a tunnel, comprising: a face evaluation index acquisition unit that acquires at least two face evaluation indexes among a crack evaluation index representing the state of cracks, a weathering degree evaluation index representing the degree of weathering, a rock strength evaluation index representing the strength of the rock mass, and a water inflow evaluation index representing the amount of water inflow, which are face evaluation indexes related to the face of the tunnel; a sprayed concrete index acquisition unit that acquires a sprayed concrete index representing the thickness of the concrete sprayed on the face; a risk event information acquisition unit that acquires risk event information representing the risk occurrence probability, which is the probability of occurrence of an event related to the risk at the face; and an evaluation model generation unit that calculates an evaluation model representing the relationship among the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method.

[0007] An evaluation method according to an aspect of the present invention is an evaluation method in an evaluation system for evaluating risks at the face of a tunnel, comprising: a face evaluation index acquisition step of acquiring at least two face evaluation indexes among a crack evaluation index representing the state of cracks, a weathering degree evaluation index representing the degree of weathering, a rock strength evaluation index representing the strength of the rock mass, and a water inflow evaluation index representing the amount of water inflow, which are face evaluation indexes related to the face of the tunnel; a sprayed concrete index acquisition step of acquiring a sprayed concrete index representing the thickness of the concrete sprayed on the face; a risk event information acquisition step of acquiring risk event information representing the risk occurrence probability, which is the probability of occurrence of an event related to the risk at the face; and an evaluation model generation step of calculating an evaluation model representing the relationship among the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method.

[0008] According to the above-described embodiment, in addition to the face evaluation index indicating the state of the face and the risk event information representing the occurrence probability of the risk event, an evaluation model representing the relationship of the shotcrete index representing the thickness of the shotcrete is calculated. Based on the calculated evaluation model, knowledge regarding the relationship between the thickness of the shotcrete and the occurrence probability of the risk event can be obtained, so that it becomes possible to highly accurately evaluate the risk in the face according to the thickness of the shotcrete.

[0009] In an evaluation system according to another embodiment, the evaluation model generation unit calculates an evaluation model that takes the face evaluation index and the shotcrete index as inputs and outputs the risk event information, and the evaluation system further includes an evaluation unit that calculates the risk occurrence probability based on the face evaluation index and the shotcrete index using the evaluation model.

[0010] According to the above-described embodiment, it becomes possible to highly accurately calculate the occurrence probability of the risk in the face of the tunnel according to the thickness of the shotcrete.

[0011] In an evaluation system according to another embodiment, the shotcrete index acquisition unit may acquire a shotcrete index having, as an index value, the thickness of the concrete calculated based on the distance from a predetermined reference position to the face surface before the concrete spraying and the distance from the reference position to the face surface after the concrete spraying.

[0012] According to the above-described embodiment, since the distance from the reference position to the face surface is measured using, for example, the reflection of laser light, it is possible to easily acquire the thickness of the shotcrete with high accuracy.

[0013] In an evaluation system according to another embodiment, the evaluation model generation unit calculates an evaluation model that takes the face evaluation index and the risk event information as inputs and outputs the shotcrete index, and the evaluation system further includes an evaluation unit that calculates the shotcrete index based on the face evaluation index and the risk occurrence probability using the evaluation model.

[0014] According to the above-described embodiment, it is possible to calculate an appropriate thickness of the sprayed concrete in consideration of the occurrence probability of an event related to risk.

[0015] In an evaluation system according to another embodiment, the statistical analysis method is logistic regression analysis, and the evaluation model generation unit may generate an evaluation model for calculating risk event information based on the face evaluation index and the sprayed concrete index based on the logistic regression algorithm.

[0016] According to the above-described embodiment, it is possible to easily obtain an evaluation model in which the relationship between the face evaluation index, the sprayed concrete index, and the risk event information is accurately represented.

Advantages of the Invention

[0017] According to one aspect of the present invention, it is possible to highly accurately evaluate the risk at the face.

Brief Description of the Drawings

[0018]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Embodiments for Carrying Out the Invention

[0019] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same or equivalent elements are denoted by the same reference numerals, and duplicate descriptions are omitted.

[0020] Fig. 1 is a block diagram showing the functional configuration of the evaluation system and the evaluation device according to this embodiment. The evaluation system 1 is a system for evaluating risks at the face of a tunnel. The evaluation system 1 shown in Fig. 1 is configured to include an evaluation device 10.

[0021] The evaluation device 10 is composed of, for example, a computer such as a server. Functionally, the evaluation device 10 of this embodiment includes a face evaluation index acquisition unit 11, a sprayed concrete index acquisition unit 12, a risk event information acquisition unit 13, an evaluation model generation unit 14, an evaluation unit 15, and an output unit 16. These functional units will be described in detail later.

[0022] In addition, each functional unit of the evaluation device 10 is configured to be able to access storage means such as a face evaluation index storage unit 20, a sprayed concrete index storage unit 30, a risk event information storage unit 40, and an evaluation model storage unit 50. These storage units may be provided in the evaluation device 10, or may be configured as external storage means provided so as to be accessible from the evaluation device 10 as shown in FIG. 1.

[0023] FIG. 2 is a hardware configuration diagram of the evaluation device 10. Physically, the evaluation device 10 is configured as a computer system including a main storage device 102 composed of memories such as a processor 101, a RAM, and a ROM, an auxiliary storage device 103 composed of a hard disk, etc., and a communication control device 104 as shown in FIG. 2. The evaluation device 10 may further include an input device 105 such as a keyboard, a touch panel, a mouse, etc. which are input devices, and an output device 106 such as a display.

[0024] The processor 101 is an arithmetic device that executes an operating system and an application program. Examples of the processor include a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), but the type of the processor 101 is not limited to these. For example, the processor 101 may be a combination of a sensor and a dedicated circuit. The dedicated circuit may be a programmable circuit such as an FPGA (Field-Programmable Gate Array), or other types of circuits.

[0025] The main memory device 102 is a device that stores a program for realizing the evaluation device 10, the calculation results output from the processor 101, and the like. The main memory device 102 is constituted by at least one of, for example, a ROM (Read Only Memory) and a RAM (Random Access Memory).

[0026] The auxiliary storage device 103 is a device that can generally store a larger amount of data than the main memory device 102. The auxiliary storage device 103 is constituted by a non-volatile storage medium such as a hard disk and a flash memory, for example. The auxiliary storage device 103 stores the program P1 for making the computer function as the evaluation device 10 and various data. Also, when each of the storage units 20, 30, 40, 50 is included in the evaluation device 10, each of the storage units 20, 30, 40, 50 may be constituted by any of the main memory device 102, the auxiliary storage device 103, and other storage elements.

[0027] The communication control device 104 is a device that executes data communication with other computers via a communication network. The communication control device 104 is constituted by, for example, a network card or a wireless communication module.

[0028] Each function shown in FIG. 1 is realized by causing the processor 101 shown in FIG. 2 to read the program P1 onto hardware such as the main memory device 102 and causing the processor 101 to execute the program P1. The program P1 includes codes for realizing each functional element of the evaluation device 10. The processor 101 operates the communication control device 104 and the like according to the program P1, and executes reading and writing of data in the main memory device 102 and the auxiliary storage device 103. The data and databases necessary for the processing are stored in the main memory device 102 and the auxiliary storage device 103. In the present embodiment, although each of the functional units 11 to 16 is configured as the evaluation device 10, it may be configured to be distributed among a plurality of computers.

[0029] The program P1 may be provided after being fixedly recorded on a tangible recording medium such as a CD-ROM, DVD-ROM, or semiconductor memory. Alternatively, the program P1 may be provided via a communication network as a data signal superimposed on a carrier wave.

[0030] Here, general tunnel construction will be briefly described. In the construction methods mainly used for tunnel construction, a so-called blasting process, a squeezing-out process, a primary concrete spraying process, a support installation process, a secondary concrete spraying process, and a rock bolt installation process are repeated at predetermined intervals to construct the tunnel in the axial direction.

[0031] In operations such as the support installation process after the squeezing-out process and the primary concrete spraying process, workers may enter and work near the face of the tunnel. Since the face may peel off, in the primary concrete spraying process after the squeezing-out process, the face is stabilized by spraying a concrete material onto the face in order to improve the safety of the work near the face. At this time, the state of the face (the risk of face peeling) is evaluated, and based on the evaluation result, the spraying thickness of the concrete material, etc. is determined. Also, the face may peel off even after the concrete has been sprayed. The evaluation system 1 of the present embodiment is for appropriately evaluating the relationship between an evaluation index representing the state of the face, a sprayed concrete index indicating the thickness of the sprayed concrete, and the probability of occurrence of a risk event such as peeling.

[0032] Referring to FIG. 1 again, the functional units of the evaluation device 10 will be described. The face evaluation index acquisition unit 11 acquires a face evaluation index related to the state of the face of the tunnel. In the present embodiment, the face evaluation index acquisition unit 11 acquires the face evaluation index from the face evaluation index storage unit 20. The face evaluation index storage unit 20 is a storage means for storing the face evaluation index.

[0033] The shotcrete index acquisition unit 12 acquires a shotcrete index representing the thickness of the shotcrete sprayed on the face. In the present embodiment, the shotcrete index acquisition unit 12 acquires the shotcrete index from the shotcrete index storage unit 30. The shotcrete index storage unit 30 is a storage means for storing the shotcrete index.

[0034] The risk event information acquisition unit 13 acquires risk event information representing the risk occurrence probability, which is the probability of the occurrence of an event related to the risk at the face. In the present embodiment, the risk event information acquisition unit 13 acquires the risk event information from the risk event information storage unit 40. The risk event information storage unit 40 is a storage means for storing the risk event information.

[0035] FIG. 3 is a diagram showing a mesh set on the face of the tunnel. As shown in FIG. 3, in the evaluation system 1 of the present embodiment, a mesh group M is set on the face F of the tunnel. Hereinafter, each mesh will be denoted as mesh m.

[0036] The face evaluation index storage unit 20, the shotcrete index storage unit 30, and the risk event information storage unit 40 store the face evaluation index, the shotcrete index, and the risk event information associated with each mesh m, respectively. The face evaluation index acquisition unit 11, the shotcrete index acquisition unit 12, and the risk event information acquisition unit 13 acquire the face evaluation index, the shotcrete index, and the risk event information for each mesh m, respectively.

[0037] Note that in the present embodiment, three storage units 20, 30, and 40 that store the face evaluation index, the shotcrete index, and the risk event information, respectively, are exemplified, but these indexes and information may be stored in one storage unit. Also, in the present embodiment, three functional units 11, 12, and 13 for acquiring the face evaluation index, the shotcrete index, and the risk event information, respectively, are exemplified, but these three functional units may be configured as one functional unit.

[0038] Referring again to FIG. 1, the face evaluation model generation unit 14 calculates an evaluation model representing the relationship between the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method. Specifically, the face evaluation model generation unit 14 generates an evaluation model using the learning data including the face evaluation index, the sprayed concrete index, and the risk event information obtained for each mesh m based on the face at each stage of the tunnel excavation work.

[0039] FIG. 4 is a diagram showing an example of learning data used for generating an evaluation model, including risk event information, a sprayed concrete index, and a face evaluation index obtained for each mesh of the face.

[0040] The learning data includes the face evaluation index, the sprayed concrete index, and the risk event information obtained by the face evaluation index acquisition unit 11, the sprayed concrete index acquisition unit 12, and the risk event information acquisition unit 13. Specifically, the learning data consists of the face evaluation index, the sprayed concrete index, and the risk event information for each mesh m specified by the location ID and the mesh ID.

[0041] The face evaluation model generation unit 14 may construct the learning data exemplified as one table in FIG. 4 based on the face evaluation index, the sprayed concrete index, and the risk event information respectively obtained by the face evaluation index acquisition unit 11, the sprayed concrete index acquisition unit 12, and the risk event information acquisition unit 13. Further, when the face evaluation index storage unit 20, the sprayed concrete index storage unit 30, and the risk event information storage unit 40 are configured as one storage unit, the face evaluation model generation unit 14 may obtain the learning data exemplified as one table in FIG. 4 via the face evaluation index acquisition unit 11, the sprayed concrete index acquisition unit 12, and the risk event information acquisition unit 13.

[0042] The location ID in the learning data shown in FIG. 4 is, for example, identification information for specifying the face at any stage in the process of tunneling. The location ID may be information indicating the distance from a reference position (for example, the tunnel entrance). The mesh ID is identification information for specifying one mesh m described with reference to FIG. 3.

[0043] The presence or absence of spalling is information indicating the presence or absence of spalling, for example, in mesh m of the face, and constitutes an example of risk event information representing the risk occurrence probability, which is the probability of occurrence of spalling, a risk event at the face. Since the presence or absence of spalling in the learning data indicates the actual results, it is 1 (100%) when spalling has occurred in the mesh m, and 0 (0%) when spalling has not occurred. The presence or absence of spalling is determined, for example, by face observation. Also, the presence or absence of spalling may be automatically determined based on a captured image or video of the face. The risk event at the face is not limited to only the presence or absence of spalling, and the risk event information may be a probability (0% to 100%) determined based on the size of spalling, the size of rockfall, the deformation and damage of installed supports, etc.

[0044] The spraying thickness is information indicating the thickness of the sprayed concrete, for example, in mesh m of the face, and constitutes an example of a sprayed concrete index. The spraying thickness may be calculated based on the distance from a predetermined reference position to the face before spraying of the concrete and the distance from the reference position to the face after spraying of the concrete. The distance from the reference position to the face can be measured by a well-known distance sensor, and may be measured, for example, by a laser distance meter that measures the distance to an object by reflection of laser light. Also, since the spraying thickness can be calculated based on the known area of the face and the amount of concrete to be sprayed, a planned value of the spraying thickness may be used. Note that the sprayed concrete index is not limited to the thickness of the sprayed concrete, and may be specifications such as the strength or composition of the sprayed concrete.

[0045] The face evaluation index is information indicating the properties of the face, and includes at least two of the crack evaluation index, weathering degree evaluation index, rock strength evaluation index, and water inflow evaluation index, in which various information for each mesh m of the face is appropriately aggregated. In the example shown in FIG. 4, the face evaluation index includes the crack evaluation index, the weathering degree evaluation index, and the rock strength evaluation index.

[0046] The crack evaluation index is information quantitatively indicating the properties of the cracks generated on the face surface, and may be, for example, at least one of the crack interval and the crack intersection density. In the example shown in FIG. 4, the crack evaluation index includes the crack interval and the crack intersection density.

[0047] FIG. 5 is a diagram showing an example of obtaining the crack interval, which is an example of the crack evaluation index. The crack interval is extracted based on an image obtained by imaging the face surface. Since there is a difference in the reflection of illumination between the concave and convex portions of the cracks in the face, the pixels of the portion where the luminance changes greatly (by a predetermined degree or more) in the image are extracted as the pixels of the crack portion. As shown in FIG. 5, in the image of the mesh m, the crack portions c1 to c6 are extracted. Then, the number (4) of the crack portions intersecting the scanning line sc in the mesh m is measured. In the example shown in FIG. 5, assuming that the length of one side of the mesh m is 100 cm, the crack interval is calculated as 25 cm (= 100 cm / 4). By performing such calculation of the crack interval for all the scanning lines in the mesh, statistical values such as the maximum value (the maximum value in the mesh, hereinafter referred to as "mesh maximum value"), the minimum value (mesh minimum value), and the average value (mesh average value) of the crack interval for each mesh m can be calculated. One or more of the statistical values such as the mesh maximum value, the mesh minimum value, and the mesh average value may be used for the training data.

[0048] FIG. 6 is a diagram showing an example of obtaining the crack intersection density which is an example of a crack evaluation index. The obtaining of the crack intersection density, similar to the crack interval, is performed by extracting pixels with a luminance change of a predetermined degree or more as pixels of the crack portion based on an image obtained by imaging the cross-section. As shown in FIG. 6, in the image of the mesh m, the crack portion cr is extracted. Then, for example, the top two components of the main dominant directions of the crack portion cr are detected as the dominant directions md1, md2 (md). And an index indicating the degree of crack intersection called the crack intersection density is calculated by comparing the dominant directions md between adjacent meshes m. The unit of the crack intersection density is, for example, percent (%), and the crack intersection density increases as the dominant direction md of the crack between adjacent meshes m is different.

[0049] The weathered area ratio constitutes an example of a weathering degree evaluation index which is an index representing the degree of weathering of the cross-section. The weathered area ratio is, for example, in each mesh m of the cross-section, the ratio of the area of the region where weathering has occurred to the area of the entire mesh m. When weathering occurs in the cross-section, the color tone changes due to the generation of clay minerals, the elution of metal ions, the generation of oxides, etc. Therefore, by referring to the association between a given degree of weathering and the pixel value, the presence and degree of weathering of each pixel portion of the mesh m are obtained. In the example shown in FIG. 4, the weathered area ratio indicating the presence or absence of weathering is used as the weathering degree evaluation index, but other indexes reflecting the degree of weathering may be used as the weathering degree index.

[0050] The fracture energy coefficient constitutes an example of a rock strength evaluation index representing the strength of the rock mass at the face. The fracture energy coefficient is an index value indicating the hardness or softness of the ground at each mesh m on the face, for example. The fracture energy coefficient may be obtained by a computer jumbo used for tunnel excavation. A computer jumbo is a large machine for drilling blast holes at accurate positions, angles, and depths according to a pre-planned drilling pattern. The computer jumbo can obtain the fracture energy coefficient based on various data acquired during the drilling operation. In the example shown in FIG. 4, the fracture energy coefficient is used as a rock strength evaluation index, but other index values reflecting the strength of the face for each mesh m may be used as the rock strength evaluation index. Also, as the rock strength evaluation index, the uniaxial compressive strength and triaxial compressive strength of face samples may be used, or the coefficient of restitution of the face impact test, etc. may be used.

[0051] Also, the water inflow evaluation index, which is an example of a face evaluation index not illustrated in FIG. 4, is an index value indicating the water inflow for each mesh m, and has units such as (l / m 2 ), for example.

[0052] The evaluation model generation unit 14 calculates an evaluation model by a predetermined statistical analysis method using learning data as exemplified in FIG. 4. Specifically, the evaluation model generation unit 14 uses logistic regression analysis as a predetermined statistical analysis method and generates an evaluation model for calculating risk event information based on the face evaluation index and the sprayed concrete index according to the algorithm of logistic regression.

[0053] The evaluation model generation unit 14 calculates the evaluation model represented by the following equations (1) and (2) by machine learning. p = 1 / (1 + exp(-z)) ···(1) z = w1x1 + w2x2 + ··· + w n x n + b ···(2)

[0054] In Equation (1), p is the probability of spalling occurrence as risk event information, which becomes 0 or 1 during learning as described with reference to FIG. 4. In Equations (1) and (2), z is the ground condition in the face mesh m. In Equation (2), x1, x2, ···, x n are respectively the spraying thickness and various face evaluation indexes, and the respective values illustrated in FIG. 4 may be input, or they may be values normalized as necessary. In Equation (2), w1, w2, ···, w n are coefficients, and b is a constant. That is, the evaluation model generation unit 14 generates an evaluation model by calculating the coefficients w1, w2, ···, w n and the constant b by machine learning. Then, the evaluation model generation unit 14 outputs and stores the generated evaluation model in the evaluation model storage unit 50.

[0055] With reference to FIG. 7, an example of the generated evaluation model will be described. The graphs lc shown in FIGS. 7(a) and 7(b) show examples of evaluation models generated using a learning data group consisting of the spalling occurrence probability (presence or absence of spalling), spraying thickness, and various face evaluation indexes for each location and mesh. The graph lc shows the relationship (see Equation (1)) of the spalling occurrence probability p with respect to the ground condition z (see Equation (2)) calculated based on the spraying thickness and various face evaluation indexes.

[0056] The point pl1 shown in FIG. 7(a) is a plot of the spalling occurrence probability p obtained by inputting the ground condition z (face evaluation index) and spraying thickness of the learning data without spalling (spalling occurrence probability p = 0) in the learning data group into the generated Equation (1). As shown in FIG. 7(a), the points pl1 indicating the spalling occurrence probability calculated based on the ground condition z where spalling did not occur are mostly distributed in the region of p < 0.5 on the graph lc. That is, such a distribution state of the points pl1 indicates that the validity of the generated evaluation model is high.

[0057] The point pl2 shown in Fig. 7(b) is a plot of the face stability condition z (face evaluation index) and the spraying thickness of the training data with spalling (spalling occurrence probability p = 1) among the training data groups, obtained by inputting them into the generated formula (1), with the spalling occurrence probability p plotted. As shown in Fig. 7(b), the points pl2 indicating the spalling occurrence probability calculated based on the face stability condition z where spalling has occurred are mostly distributed in the region where p > 0.5 on the graph lc. That is, such a distribution situation of the points pl2 indicates that the validity of the generated evaluation model is high.

[0058] Note that the evaluation model generation unit 14 may calculate an evaluation model that takes the face evaluation index and the spraying concrete index (spraying thickness) as inputs and outputs risk event information (spalling occurrence probability), as shown in formula (1).

[0059] Also, the evaluation model generation unit 14 may calculate an evaluation model that takes the face evaluation index and the risk event information as inputs and outputs the spraying concrete index (spraying thickness) by transforming formula (1) and formula (2).

[0060] With reference to Fig. 8, an example of the use of the generated evaluation model will be described. The graph fp1 shown in Fig. 8(a) is a graph showing the relationship between the spalling occurrence probability p and the spraying thickness x1 obtained by substituting the face evaluation index of a certain point p1 where spalling did not occur in the training data group into the evaluation model generated by machine learning using the training data group consisting of the spalling occurrence probability p, the spraying thickness x1, and various face evaluation indexes (x2,..., x n ). At the point p1, it is assumed that as a result of setting the spraying thickness to 100 mm, spalling did not occur.

[0061] According to the reference of graph fp1, at point p1, the spraying thickness was 100 mm, but in that case, the spalling occurrence probability was 0.1 (10%), and it can be judged that the spraying thickness was excessive. And if the appropriate construction condition is that the spalling occurrence probability p is 30%, then in the ground condition at point p1, as shown by point pt1, it can be judged that it is appropriate to set the spraying thickness to 50 mm.

[0062] The graph fp2 shown in Fig. 8(b) is a graph showing the relationship between the spalling occurrence probability p and the spraying thickness x1 obtained by substituting the face evaluation index (x2,..., x n ) of a certain point p2 where spalling occurred in the learning data group used for machine learning consisting of ) into the evaluation model generated by machine learning. Note that at point p2, it is assumed that spalling occurred as a result of the spraying thickness being 31 mm.

[0063] According to the reference of graph fp2, at point p2, the spraying thickness was 31 mm, but in that case, the spalling occurrence probability exceeded 0.7 (70%), and it can be judged that the spraying thickness was too small. And if the appropriate construction condition is that the spalling occurrence probability p is 30%, then in the ground condition at point p2, as shown by point pt2, it can be judged that it is appropriate to set the spraying thickness to 100 mm.

[0064] In this embodiment, the evaluation model generation unit 14 generates an evaluation model based on the algorithm of logistic regression with logistic regression analysis as a predetermined statistical analysis method. However, the type of the algorithm is not limited, and the evaluation model generation unit 14 may generate an evaluation model based on any one of algorithms such as neural networks, support vector machines, and decision trees.

[0065] Referring to FIG. 1 again, the evaluation unit 15 evaluates the risk at the face of the tunnel using the calculated evaluation model. As an example, the evaluation unit 15 calculates the risk occurrence probability based on the face evaluation index and the shotcrete index using the evaluation model.

[0066] Specifically, the evaluation unit 15 acquires the face evaluation index and the shotcrete index for each mesh m of the face to be evaluated, inputs the acquired face evaluation index and shotcrete index into the evaluation model stored in the evaluation model storage unit 50, and acquires the spalling occurrence probability as the risk occurrence probability output from the evaluation model. Then, the output unit 16 outputs the spalling occurrence probability acquired by the evaluation unit 15 in a predetermined manner. The output manner may be display on a predetermined display means, transmission to a predetermined device, storage in a predetermined storage means, etc.

[0067] By configuring the evaluation model in this way, it becomes possible to calculate the occurrence probability of the risk at the face of the tunnel according to the thickness of the shotcrete with high accuracy. Therefore, since the occurrence probability of the risk after installing (spraying) concrete on the face in the shotcrete process can be calculated, the subsequent work safety management can be appropriately performed. Specifically, by appropriately setting the allowable risk occurrence probability, the thickness of the shotcrete can be calculated based on the allowable occurrence probability, so that waste and unevenness of the concrete amount can be eliminated.

[0068] Also, as another example, the evaluation unit 15 may calculate the shotcrete index based on the face evaluation index and the risk occurrence probability using the evaluation model. Specifically, the evaluation unit 15 acquires the face evaluation index and the spalling occurrence probability for each mesh m of the face to be evaluated, inputs the acquired face evaluation index and spalling occurrence probability into the evaluation model stored in the evaluation model storage unit 50, and acquires the spraying thickness of the concrete as the shotcrete index output from the evaluation model. Then, the output unit 16 outputs the spraying thickness of the concrete acquired by the evaluation unit 15 in a predetermined manner.

[0069] By configuring the evaluation model in this way, it becomes possible to calculate the appropriate thickness of the sprayed concrete while taking into account the probability of occurrence of events related to risks.

[0070] Next, with reference to FIGS. 9 to 11, the operation of the evaluation system 1 of the present embodiment will be described. FIG. 9 is a flowchart showing the processing contents of an evaluation method related to the generation of an evaluation model implemented in the evaluation system 1.

[0071] In step S1, the face evaluation index acquisition unit 11 acquires a face evaluation index related to the state of the face of the tunnel.

[0072] In step S2, the sprayed concrete index acquisition unit 12 acquires a sprayed concrete index representing the thickness of the concrete sprayed on the face.

[0073] In step S3, the risk event information acquisition unit 13 acquires risk event information representing the risk occurrence probability, which is the probability of occurrence of an event related to the risk at the face. Note that the processes in steps S1 to S3 may be performed in any order or simultaneously.

[0074] In step S4, the evaluation model generation unit 14 calculates an evaluation model representing the relationship between the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method. Then, the evaluation model generation unit 14 outputs the calculated evaluation model in a manner of storing it in, for example, the evaluation model storage unit 50.

[0075] FIG. 10 is a flowchart showing the processing contents of an evaluation method related to the calculation of the risk occurrence probability implemented in the evaluation system 1.

[0076] In step S11, the face evaluation index acquisition unit 11 and the sprayed concrete index acquisition unit 12 each acquire the face evaluation index and the sprayed concrete index.

[0077] In step S12, the evaluation unit 15 inputs the face evaluation index and the sprayed concrete index into the evaluation model.

[0078] In step S13, the evaluation unit 15 obtains the risk occurrence probability output from the evaluation model. Then, the output unit 16 outputs the obtained risk occurrence probability in a predetermined manner.

[0079] FIG. 11 is a flowchart showing the processing content of the evaluation method regarding the calculation of the thickness of the sprayed concrete implemented in the evaluation system 1.

[0080] In step S21, the face evaluation index acquisition unit 11 and the risk event information acquisition unit 13 respectively acquire the face evaluation index and the risk event information.

[0081] In step S22, the evaluation unit 15 inputs the face evaluation index and the risk event information into the evaluation model.

[0082] In step S23, the evaluation unit 15 obtains the sprayed concrete index output from the evaluation model. Then, the output unit 16 outputs the obtained sprayed concrete index in a predetermined manner.

[0083] According to the evaluation system 1, the evaluation method, and the program P1 of the present embodiment described above, in addition to the face evaluation index indicating the state of the face and the risk event information representing the occurrence probability of the risk event, an evaluation model representing the relationship of the sprayed concrete index representing the thickness of the sprayed concrete is calculated. Based on the calculated evaluation model, knowledge regarding the relationship between the thickness of the sprayed concrete and the occurrence probability of the risk event can be obtained, so that it becomes possible to highly accurately evaluate the risk at the face according to the thickness of the sprayed concrete.

[0084] As described above, the present invention has been described in detail based on its embodiments. However, the present invention is not limited to the above embodiments. The present invention can be variously modified without departing from the gist thereof.

Description of Symbols

[0085] 1…Evaluation system, 10…Evaluation device, 11…Face evaluation index acquisition unit, 12…Sprayed concrete index acquisition unit, 13…Risk event information acquisition unit, 14…Evaluation model generation unit, 15…Evaluation unit, 16…Output unit, 20…Face evaluation index storage unit, 30…Sprayed concrete index storage unit, 40…Risk event information storage unit, 50…Evaluation model storage unit, P1…Program.

Claims

1. An evaluation system for evaluating risks at the face of a tunnel, comprising: A face evaluation index acquisition unit that acquires at least two face evaluation indexes out of a crack evaluation index representing the state of cracks, a weathering degree evaluation index representing the degree of weathering, a rock strength evaluation index representing the strength of the rock mass, and a water inflow evaluation index representing the amount of water inflow, which are face evaluation indexes related to the face of the tunnel; A sprayed concrete index acquisition unit that acquires a sprayed concrete index representing the thickness of the concrete sprayed on the face; A risk event information acquisition unit that acquires risk event information representing the probability of occurrence of an event related to the risk at the face; An evaluation model generation unit that calculates an evaluation model representing the relationship between the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method; An evaluation system comprising the above.

2. The evaluation model generation unit calculates the evaluation model that takes the face evaluation index and the sprayed concrete index as inputs and outputs the risk event information, The evaluation system further comprises: An evaluation unit that calculates the risk occurrence probability based on the face evaluation index and the sprayed concrete index using the evaluation model. The evaluation system according to Claim 1.

3. The sprayed concrete index acquisition unit acquires the sprayed concrete index having, as an index value, the thickness of the concrete calculated based on the distance from a predetermined reference position to the face before spraying the concrete and the distance from the reference position to the face after spraying the concrete. , The evaluation system according to Claim 2.

4. The evaluation model generation unit calculates the evaluation model that takes the face evaluation index and the risk event information as inputs and outputs the sprayed concrete index. The evaluation system further includes an evaluation unit that calculates the sprayed concrete index based on the face evaluation index and the risk occurrence probability using the evaluation model. The evaluation system according to claim 1.

5. The statistical analysis method is logistic regression analysis, and the evaluation model generation unit generates the evaluation model that calculates the risk event information based on the face evaluation index and the sprayed concrete index based on the algorithm of the logistic regression. The evaluation system according to any one of claims 1 to 4.

6. An evaluation method in an evaluation system for evaluating risks at the face of a tunnel, comprising: a face evaluation index acquisition step of acquiring at least two face evaluation indexes from among a crack evaluation index representing the state of cracks, a weathering degree evaluation index representing the degree of weathering, a rock strength evaluation index representing the strength of the rock mass, and a water inflow evaluation index representing the water inflow, which are face evaluation indexes related to the face of the tunnel; a sprayed concrete index acquisition step of acquiring a sprayed concrete index representing the thickness of the concrete sprayed on the face; a risk event information acquisition step of acquiring risk event information representing the probability of occurrence of an event related to the risk at the face; an evaluation model generation step of calculating an evaluation model representing the relationship among the face evaluation index, the sprayed concrete index, and the risk event information by a predetermined statistical analysis method; and having the evaluation method.

Images