Building diagnosing system and program

Abstract

To provide a building diagnosing system and a program capable of easily and automatically diagnosing the deterioration state of a building.SOLUTION: Operations of a building diagnosis system according to the present invention include: acquiring bridge information related to a bridge; referring to a diagnosis model learned by using learning data in which the bridge information is input data and deterioration information related to the deterioration state of the bridge is output data; outputting the deterioration information based on the acquired bridge information; acquiring question information including questions about the outputted deterioration information; and outputting answer information indicating answers to the questions included in the question information based on the acquired question information.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a building structure diagnosis system and program. [Background technology]

[0002] There has been a need for diagnosing the deterioration state of buildings. Diagnosis of the deterioration state of buildings is performed by a professional engineer based on the damage state of the building. In this case, there is a possibility that the diagnosis result will differ depending on the professional engineer's judgment. In addition, since the number of buildings that need to be diagnosed is increasing dramatically every year, there is a need for a structure deterioration diagnosis system that can automatically diagnose the damage state of buildings, as shown in Patent Document 1.

[0003] Patent document 1 discloses a structure deterioration diagnosis system that includes a plurality of sensors that are installed on a structure and measure the live load displacement of the structure due to the live load, and a diagnosis unit that calculates the maximum displacement amount for each of the live load displacements measured by each of the plurality of sensors within a predetermined time range and performs a deterioration diagnosis of the structure from the change over time in the mutual relationship between the maximum displacement amounts calculated for each of the plurality of sensors. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2022-102230 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technology disclosed in Patent Document 1 requires sensors to be installed in the structure in advance to obtain deterioration information. This poses the problem that deterioration diagnosis cannot be performed on structures that have already been constructed. In addition, a sensor is required to measure the live load displacement of the structure during diagnosis. This poses the problem of high diagnosis costs.

[0006] The present invention has been devised in consideration of the above-mentioned problems, and its object is to provide a structure diagnosis system and program that can easily and automatically diagnose the deterioration state of a structure. [Means for solving the problem]

[0007] The structure diagnosis system of the first invention is characterized in that it comprises an acquisition means for acquiring bridge information related to a bridge, and a search means for referring to a diagnostic model trained using learning data in which the bridge information is input data and deterioration information related to the deterioration status of the bridge is output data, and outputting deterioration information based on the bridge information acquired by the acquisition means.

[0008] The structure diagnosis system of the second invention is characterized in that, in the first invention, the acquisition means acquires the bridge information including text information on attributes of the bridge and / or image information including the bridge, and the search means refers to the diagnostic model trained using learning data which has text information and / or image information as input data and deterioration information as output data, and outputs deterioration information based on the text information and / or image information included in the bridge information acquired by the acquisition means.

[0009] The structure diagnosis system of the third invention is characterized in that, in the first invention, the acquisition means further acquires differential information based on past bridge information regarding the bridge and the bridge information, and the search means refers to the diagnosis model trained using learning data in which the differential information is input data and deterioration information is output data, and outputs deterioration information based on the differential information acquired by the acquisition means.

[0010] The structure diagnosis system of the fourth invention is characterized in that, in the first invention, the acquisition means further acquires river information regarding the river on which the bridge is located, and the search means refers to the diagnosis model trained using learning data which has bridge information and river information as input data and deterioration information as output data, and outputs deterioration information based on the bridge information and river information acquired by the acquisition means.

[0011] The structure diagnosis system of the fifth invention is characterized in that, in the first invention, the acquisition means further acquires similar bridge information regarding other bridges crossing a river to the bridge, and the search means refers to the diagnosis model trained using learning data in which the bridge information and the similar bridge information are input data and deterioration information is output data, and outputs deterioration information based on the bridge information acquired by the acquisition means and the similar bridge information.

[0012] The structure diagnosis system of the sixth invention is characterized in that it comprises an acquisition means for acquiring image information indicating a bridge, an extraction means for referring to an extraction model trained using learning data which takes image information as input data and has as output data semantic information indicating the meaning of features contained in the image information and location information indicating the location of the features, and extracts semantic information and location information based on the image information acquired by the acquisition means, and a search means for referring to a diagnostic model trained using learning data which takes semantic information and location information as input data and has as output data deterioration information indicating the deterioration status of the bridge, and outputting deterioration information based on the semantic information and location information extracted by the extraction means.

[0013] The structure diagnosis program of the seventh invention is characterized in that it has a computer execute an acquisition step of acquiring bridge information related to a bridge, and a search step of referring to a diagnostic model trained using learning data in which the bridge information is input data and deterioration information related to the deterioration status of the bridge is output data, and outputting deterioration information based on the bridge information acquired by the acquisition step. Effect of the Invention

[0014] According to the first to seventh aspects of the present invention, the structure diagnosis system and program of the present invention refer to a diagnosis model and output deterioration information based on bridge information. This makes it possible to output deterioration information from bridge information such as bridge images. This makes it possible to easily and automatically diagnose the deterioration state of a structure.

[0015] In particular, according to the second aspect of the present invention, the structure diagnosis system of the present invention references a diagnosis model and outputs deterioration information based on text information and / or image information. This makes it possible to calculate deterioration information from multiple angles from image information such as bridge photographs and damage diagrams, and text information such as the size of the bridge. This makes it possible to easily and automatically diagnose the deterioration state of a structure with high accuracy.

[0016] In particular, according to the third aspect of the present invention, the structure diagnosis system of the present invention references the diagnosis model and outputs deterioration information based on the difference information. This makes it possible to calculate deterioration information taking into account the difference between the past bridge information and the bridge information, such as cracks in the bridge, for example. This makes it possible to easily and automatically diagnose the deterioration state of the structure with a high degree of accuracy.

[0017] In particular, according to the fourth aspect of the present invention, the structure diagnosis system of the present invention references a diagnosis model and outputs deterioration information based on bridge information and river information. This makes it possible to calculate deterioration information taking into account, for example, records of river flooding. This makes it possible to easily and automatically diagnose the deterioration state of a structure with high accuracy.

[0018] In particular, according to the fifth aspect of the present invention, the structure diagnosis system of the present invention refers to a diagnosis model and outputs deterioration information based on bridge information and related bridge information. This makes it possible to calculate deterioration information taking into account the conditions of other bridges across the same river, for example. This makes it possible to easily and automatically diagnose the deterioration state of a structure with high accuracy.

[0019] In particular, according to the sixth aspect of the present invention, the structure diagnosis system of the present invention refers to the extraction model, extracts semantic information and positional information based on image information, and outputs deterioration information based on the extracted semantic information and positional information. This makes it possible to extract semantic information and positional information from image information such as photographs and damage diagrams, and therefore makes it possible to extract information from image information such as photographs and damage diagrams with higher accuracy. [Brief description of the drawings]

[0020] [Figure 1] FIG. 1 is a schematic diagram showing an example of the configuration of a structure diagnosis system according to the first embodiment. [Diagram 2] FIG. 2(a) is a schematic diagram showing an example of the configuration of a structure diagnosis device in the first embodiment, and FIG. 2(b) is a schematic diagram showing an example of the functions of the structure diagnosis device in the first embodiment. [Diagram 3] FIG. 3 is a diagram showing an example of a bridge. [Figure 4] FIG. 4 is a flowchart showing an example of the operation of the structure diagnosis system 100 in the first embodiment. [Diagram 5] FIG. 5 is a diagram showing a first example of a periodic inspection record including bridge information. [Figure 6] FIG. 6 is a diagram showing a second example of a periodic inspection record including bridge information. [Figure 7] FIG. 7 shows images of each component of a bridge. [Figure 8] FIG. 8 shows a diagram of the damage to the bridge. [Figure 9] FIG. 9 is a diagram showing the degradation information. [Figure 10] FIG. 10 is a diagram showing the degree of association between bridge information and deterioration information. [Figure 11] FIG. 11 is a diagram showing the degree of association between text information, image information, and degradation information. [Figure 12] FIG. 12 is a diagram showing the degree of association between difference information and degradation information. [Figure 13]FIG. 13 is a diagram showing the degree of association between bridge information, river information, and deterioration information. [Figure 14] FIG. 14 is a diagram showing the degree of association between bridge information, related bridge information, and deterioration information. [Figure 15] FIG. 15 is a flowchart showing an example of the operation of the structure diagnosis system 100 in the second embodiment. [Figure 16] FIG. 16 is a diagram showing the degree of association between bridge information and semantic information and location information. [Figure 17] FIG. 17 is a diagram showing the degree of association between semantic information, location information, and degradation information. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] First embodiment Hereinafter, an example of a building diagnosis system according to a first embodiment of the present invention will be described with reference to the drawings.

[0022] Fig. 1 is a schematic diagram showing an example of the configuration of a structure diagnosis system 100 in the first embodiment. For example, as shown in Fig. 1, the structure diagnosis system 100 includes a structure diagnosis device 1, a server 3, and a user terminal 2 having an imaging device 6, which are connected via a public communication network 4. Alternatively, the structure diagnosis system 100 may include only the structure diagnosis device 1 and the imaging device 6, which are connected without going through the public communication network 4.

[0023] The server 3 is a storage medium that stores various data such as images transmitted from the structure diagnosis device 1 and the user terminal 2. Furthermore, the server 3 transmits the various stored data to the structure diagnosis device 1 and the user terminal 2 as necessary. The server 3 may, for example, have at least some of the functions of the structure diagnosis device 1, and may, for example, perform at least some of the processing in place of the structure diagnosis device 1.

[0024] The public communication network 4 is, for example, an Internet network to which the structure diagnosis device 1 is connected via a communication circuit. The public communication network 4 may be configured as a so-called optical fiber communication network. The public communication network 4 may be realized by a known communication technology such as a wireless communication network in addition to a wired communication network. The user terminal 2 is owned by, for example, a user of a service using the structure diagnosis system 100, and is connected to the structure diagnosis device 1 via the public communication network 4. The user terminal 2 may represent, for example, an electronic device that generates a database. The user terminal 2 may be, for example, an electronic device such as a personal computer or a tablet terminal. The user terminal 2 may have at least a part of the functions of the structure diagnosis device 1. The user terminal 2 may have a display or a speaker (not shown) that can present the diagnosis result to the user.

[0025] The imaging device 6 is a camera for capturing an image of the bridge 5. Any camera may be used as the imaging device 6. The imaging device 6 outputs the captured image to the user terminal 2. The imaging device 6 may also output the captured image to the structure diagnosis device 1 via the public communication network 4 without going through the user terminal 2.

[0026] The structure diagnosis device 1 outputs deterioration information indicating the deterioration state of the bridge 5 based on bridge information related to the bridge 5. The structure diagnosis device 1 may be an electronic device such as a personal computer (PC), or may be an electronic device such as a smartphone, a tablet terminal, a wearable terminal, an IoT (Internet of Things) device, or a single board computer such as Raspberry Pi (registered trademark), and may have an imaging device 6 built in.

[0027] Next, an example of the structure diagnosis device 1 in the first embodiment will be described with reference to Fig. 2. Fig. 2(a) is a schematic diagram showing an example of the configuration of the structure diagnosis device 1 in the first embodiment, and Fig. 2(b) is a schematic diagram showing an example of the function of the structure diagnosis device 1 in the first embodiment.

[0028] 2(a), the structure diagnosis device 1 includes a housing 10, a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a storage unit 104, and I / Fs 105 to 107. The CPU 101, the ROM 102, the RAM 103, the storage unit 104, and the I / Fs 105 to 107 are connected via an internal bus 110.

[0029] The CPU 101 controls the entire structure diagnosis device 1. The ROM 102 stores operation code for the CPU 101. The RAM 103 is a working area used when the CPU 101 is operating. The storage unit 104 stores various information such as bridge information, deterioration information, extraction models, and diagnosis models. The storage unit 104 may be, for example, a data storage device such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), an SD card, or a miniSD card. For example, the structure diagnosis device 1 may have a GPU (Graphics Processing Unit) (not shown).

[0030] The I / F 105 is an interface for transmitting and receiving various information via the public communication network 4. The I / F 106 is an interface for transmitting and receiving information to and from the input unit 108. For example, a keyboard is used as the input unit 108, and a user or the like who uses the structure diagnosis device 1 inputs various information or control commands for the structure diagnosis device 1 via the input unit 108. The I / F 107 is an interface for transmitting and receiving various information to and from the display unit 109. The display unit 109 outputs various information such as deterioration information stored in the storage unit 104, or the processing status of the structure diagnosis device 1, etc. A display is used as the display unit 109, and may be, for example, a touch panel type.

[0031] The structure diagnosis device 1 includes an acquisition unit 11, a processing unit 12, an output unit 14, and a storage unit 15. The acquisition unit 11, processing unit 12, output unit 14, and storage unit 15 shown in Fig. 2(b) are realized by the CPU 101 using the RAM 103 as a working area to execute a program stored in the storage unit 104 or the like, and may be controlled by, for example, artificial intelligence.

[0032] The acquisition unit 11 acquires various information such as bridge information. For example, the acquisition unit 11 acquires bridge information via the user terminal 2. In addition, the acquisition unit 11 may acquire various information from the server 3 or the like via the public communication network 4. Note that the frequency and period at which the acquisition unit 11 acquires various information are arbitrary.

[0033] The processing unit 12 performs various processes. The processing unit 12 outputs deterioration information based on various information such as bridge information acquired by the acquisition unit 11, for example.

[0034] The output unit 14 outputs various types of information and transmits the various types of information to the display unit 109 via the I / F 107.

[0035] The memory unit 15 retrieves, as necessary, various pieces of information stored in the storage unit 104. The memory unit 15 stores, in the storage unit 104, various pieces of information acquired or output by the acquisition unit 11, the processing unit 12, and the output unit 14.

[0036] FIG. 3 is a diagram showing an example of a bridge 5. The bridge 5 is a bridge to be diagnosed by the structure diagnosis system 100. The bridge 5 is, for example, a bridge spanning a river 53. The bridge 5 may be a bridge using any type, configuration, and material. The bridge 5 includes, for example, a footing 58 installed on a shore (not shown), a footing 54 having a bearing 51 installed on the footing 58 at its end, a girder 55 supported by the footing 54, an abutment 56 installed on the upper surface of the girder 55, a balustrade 57 installed on the abutment 56, and a drainage pipe 52 installed on the side of the girder 55. The bridge 5 may also include, for example, a sign, a guardrail, a lighting lamp, an inspection road, an expansion joint, and the like (not shown). Each component of the bridge 5 may be made of any material, such as steel, concrete, iron, or aluminum.

[0037] Next, a description will be given of an example of the operation of the structure diagnosis system 100 in the first embodiment. Fig. 4 is a flowchart showing an example of the operation of the structure diagnosis system 100 in the first embodiment.

[0038] First, in step S1, the acquisition unit 11 acquires various information. The acquisition unit 11 acquires, for example, bridge information. In this case, the acquisition unit 11 may acquire, for example, bridge information including an image captured by the imaging device 6, via the public communication network 4. The acquisition unit 11 may also acquire, for example, bridge information stored in the server 3, via the public communication network 4. The acquisition unit 11 stores the bridge information in the storage unit 104, for example, via the memory unit 15.

[0039] Bridge information is information related to a bridge 5. FIG. 5 is a diagram showing a first example of a periodic inspection report including bridge information. FIG. 6 is a diagram showing a second example of a periodic inspection report including bridge information. Bridge information may include text information related to attributes of bridge 5, such as bridge name, route name, location, starting point, bridge ID, manager name, date of periodic inspection, road conditions, presence or absence of alternative road, motor vehicle road or general road, emergency transportation road, occupied property, periodic inspector, year of erection, horizontal length, width, bridge type, diagnosis of component unit, emergency treatment content, judgment category after emergency treatment, date of emergency treatment and judgment, and comments at the time of inspection, as shown in FIG. 5. The bridge information may also include text information such as the bridge number, number of spans, office name, bridge type, superstructure type, substructure type, deck material, foundation type, repair history, live load, grade, applicable specifications, traffic conditions, large vehicle traffic volume, load limit, traffic volume, width of each structure, distance from the coast, designation of priority route, coarse aggregate, painting specifications, intersection conditions, soundness assessment, third damage prevention, etc., as shown in Figure 6. Comments at the time of inspection may be comments such as the inspector's speculations, such as "There are two-way cracks and depressions on the underside of the deck, and it is assumed that fatigue due to vehicle load is the cause," "The depressions and the fallen parts of the road surface match," "It is assumed that the cover concrete spalled due to the progress of corrosion of the steel frame under the ground cover," "The impact on the structure is small. It can be fixed with maintenance and repair," and "The vertical cracks in the substructure are drying shrinkage cracks immediately after construction, and there is a suspicion of Alkali Silica Reaction (ASR), so a detailed investigation is considered necessary."

[0040] The bridge information may be, for example, information about the material, condition, etc. of each component of the bridge 5. The bridge information may include, for example, information about signs, guard rails, lighting, inspection paths, etc. (not shown) provided on the bridge 5. The bridge information may also be information about the amount of antifreeze sprayed, the distance from the coast, aggregate information, the year of construction, etc. The aggregate information may also be information about whether or not a material such as sea sand is used as the aggregate, whether or not a material that may cause an alkali-aggregate reaction is used, whether or not a material with a large amount of shrinkage is used, etc.

[0041] The bridge information may also include image information including the bridge 5. FIG. 7 is a diagram showing images of each component of the bridge 5. FIG. 8 is a diagram showing a damage diagram of the bridge 5. The bridge information may include a plurality of image information including each component such as the substructures of the left and right banks, for example, as shown in FIG. 7. The bridge information may also include image information including a damage diagram, for example, as shown in FIG. 8. The damage diagram is a diagram showing damage for each component of the bridge 5, and may be a diagram including symbols written in a two-dimensional space, for example, as shown in FIG. 8.

[0042] The image information may be a three-dimensional image. The image information may be a three-dimensional image including the entire bridge 5. The image information may be a three-dimensional image for each component of the bridge 5. In such a case, the image information may be a three-dimensional image generated from a plurality of two-dimensional images using a known technique such as photogrammetry.

[0043] Also, in step S1, the acquisition unit 11 may acquire river information regarding the river 53 over which the bridge 5 is built. The river information may include information such as the river's width, depth, flood records, and salinity concentration. Also, in step S1, the acquisition unit 11 may acquire environmental information regarding the external environment. The environmental information may be information such as temperature, weather, and disasters.

[0044] Also, in step S1, the acquisition unit 11 may acquire past bridge information regarding a past bridge 5. The past bridge information may be, for example, bridge information at the time of past diagnosis or inspection. Also, the past bridge information may be, for example, deterioration information diagnosed for bridge information at the time of past diagnosis or inspection. Also, in step S1, the acquisition unit 11 may acquire difference information based on the bridge information and the past bridge information. The difference information may be, for example, information indicating the difference between bridge information at the time of past diagnosis or inspection and bridge information at the present time. The difference information may be, for example, the amount of change in cracks from the time of the previous inspection to the present time. Also, the difference information may be information such as a table or graph indicating changes from past bridge information at multiple times, as shown in Table 1. [Table 1]

[0045] Furthermore, in step S1, the acquisition unit 11 may acquire related bridge information regarding another bridge 5 crossing the river 53 of the bridge 5. The related bridge information may be bridge information of the other bridge 5 crossing the river 53 of the bridge 5. The related bridge information may also be deterioration information of the other bridge 5 crossing the river 53 of the bridge 5. The related bridge information may also be bridge information or deterioration information of the other bridge 5 on the road of the bridge 5.

[0046] The deterioration information is information indicating the deterioration state of the bridge 5. The deterioration information may be information indicating the deterioration state of each component of the bridge as a numerical value, for example, as in the judgment classification shown in FIG. 9. The deterioration information may also be information indicating the cause of deterioration, such as corrosion or water leakage. The deterioration information may also be information indicating the deterioration state of cracks, etc., included in the image information included in the bridge information. The deterioration information may also be information indicating the basis of the numerical value of the deterioration state. The information indicating the basis of the numerical value of the deterioration state may be, for example, specifications such as specifications for road bridges and concrete standard specifications, and guidelines for regular inspection of road bridges (guidelines issued by the Ministry of Land, Infrastructure, Transport and Tourism or the Road Bureau), bridge longevity repair plans of each prefecture, descriptions in papers, patent documents, etc.

[0047] Next, in step S2, the processing unit 12 outputs deterioration information based on the bridge information acquired in step S1 by the acquisition unit 11. In this case, the processing unit 12 refers to a diagnostic model trained using learning data in which, for example, the bridge information is input data and the deterioration information is output data, and outputs the deterioration information based on the bridge information acquired in step S1.

[0048] As a method for generating a diagnostic model, for example, machine learning based on a neural network model may be used to generate the diagnostic model. The diagnostic model may be trained using machine learning based on a neural network model such as a CNN (Convolution Neural Network), or any other model may be used. In addition, as a method for generating a diagnostic model, for example, linear discrimination, support vector machine, k-nearest neighbor method, random forest, deep learning, etc. may be used to generate the diagnostic model.

[0049] In such a case, the diagnostic model stores correlations having the correlation degree between bridge information, which is input data, and deterioration information, which is output data, as shown in FIG. 10. The correlation degree indicates the degree of correlation between the input data and the output data, and it can be determined that the higher the correlation degree, the stronger the connection between the data. The correlation degree may be indicated by three or more values ​​or three or more stages, such as a percentage, or may be indicated by two values ​​or two stages. The bridge information and deterioration information used to learn the correlation degree are, for example, bridge information and deterioration information for use in previously acquired learning data, but are not limited to this, and information acquired at any timing may be used.

[0050] For example, the association is constructed by the degree of association between a plurality of input data, pairs, and a plurality of output data. The association is appropriately updated in the process of machine learning, and indicates a classifier using a function optimized, for example, based on a plurality of input data and a plurality of output data. Note that the association may have, for example, a plurality of association degrees indicating the degree of association between each data. For example, when the database is constructed by a neural network, the association degree can be made to correspond to a weight variable. For example, as shown in FIG. 10, the association may indicate the degree of association between a plurality of input data and a plurality of output data. In this case, by using the association, the degree of relationship with a plurality of output data of "deterioration information A" to "deterioration information C" can be linked and stored for each input data of "bridge information A" to "bridge information C" in FIG. 10. Therefore, for example, a plurality of input data can be linked to one output data via the association. This makes it possible to realize a multifaceted selection of output data for the input data. In addition, the input data and the output data are not limited to this, and any type of information may be further used. For example, bridge information including multiple types of information such as the size of cracks in multiple components of the bridge 5 and the type of the bridge 5 may be used as the bridge information of the input data. Furthermore, the input data may be, for example, text information and / or image information included in the bridge information. Furthermore, deterioration information including multiple types of information such as a classification of the deterioration state and information indicating the basis for the numerical value of the deterioration state may be used as the deterioration information of the output data. Furthermore, text information written in a regular inspection report may be used as the input data.

[0051] The correlation has, for example, multiple correlations linking each input data with each output data. The correlation is shown in three or more levels, for example, a percentage, a 10-level scale, or a 5-level scale, and is shown, for example, by the characteristics of the line (for example, thickness, etc.). For example, "bridge information A" included in the input data shows a correlation AA of "73%" with "deterioration information A" included in the output data, and a correlation AB of "12%" with "deterioration information B" included in the output data. In other words, the "correlation" indicates the degree of connection between each data, and for example, the higher the correlation, the stronger the connection between each data.

[0052] Three or more levels of correlation as shown in Fig. 10 are acquired in advance. In other words, when determining the actual solution, past data sets are accumulated to show which of the input data and output data was adopted and evaluated, and these are analyzed to create the correlation shown in Fig. 10.

[0053] For example, suppose that in the past, "Deterioration Information B" was judged to be the most suitable input data for "Bridge Information B" and was evaluated. By collecting and analyzing such data sets, the degree of correlation between the input data and the output data becomes stronger.

[0054] This analysis may be performed by artificial intelligence. In such a case, for example, if there are many cases in which "deterioration information B" is estimated for input data of "bridge information B," the degree of correlation between this "bridge information B" and "deterioration information B" is set higher.

[0055] This correlation may be formed by a node of a neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of the neural network corresponds to the correlation. In addition, it is not limited to a neural network, and may be formed by any decision-making factor that constitutes artificial intelligence.

[0056] Furthermore, the diagnostic model may be machine-learned by providing at least one or more hidden layers between the input data and the output data. The above-mentioned correlation is set in either the input data or the hidden layer data, or in both, and this becomes the weighting of each data, and the output is selected based on this. Then, when this correlation exceeds a certain threshold, the output may be selected.

[0057] Such correlations are what is called learning data in artificial intelligence. Such learning data is learned in advance, and in step S2, the processing unit 12 actually outputs deterioration information based on new bridge information. When outputting, for example, the correlations shown in FIG. 10 obtained in advance are referenced. For example, if the newly acquired bridge information is the same as or similar to "bridge information A", it is associated with "deterioration information A" through the correlations at correlations AA "73%" and with "deterioration information B" at correlations AB "12%". In this case, "bridge information A" with the highest correlation is selected as the optimal solution. However, it is not essential to select the one with the highest correlation as the optimal solution, and "deterioration information B" with a low correlation but a correlation itself that is recognized may be selected as the optimal solution. In addition, it is of course possible to select an output solution that is not connected by an arrow other than the above, and any other priority order may be selected as long as it is based on the correlations.

[0058] By referring to such correlations, it is possible to quantitatively select output data suitable for input data not only when the bridge information is identical or similar to the input data, but also when the bridge information is dissimilar.

[0059] Also, in step S2, the processing unit 12 may refer to a diagnostic model trained using learning data in which text information and image information are input data and deterioration information is output data, and output deterioration information based on the text information and image information included in the bridge information acquired in step S1. In such a case, the diagnostic model records the correlation trained using learning data in which text information and image information as shown in FIG. 11 are input data and deterioration information is output data. Also, in such a case, various information included in the bridge information may be arbitrarily selected and used as input data. This makes it possible to output deterioration information from multifaceted information such as the state of cracks shown in the image information and the materials used for the bridge 5 included in the text information, and therefore it is possible to output deterioration information with higher accuracy.

[0060] Also, in step S2, the processing unit 12 may output deterioration information based on the bridge information and the past bridge information by referring to a diagnostic model that uses the bridge information and the past bridge information as input data and the deterioration information as output data. In such a case, the diagnostic model uses the bridge information and the past bridge information as input data, and records the correlation learned by learning data that uses the deterioration information as output data. In such a case, the input data may be any information included in the bridge information or the past bridge information. Also, the input information may be difference information. In such a case, a diagnostic model that uses the difference information as input data and the deterioration information as output data as shown in FIG. 12 may be referred to, and deterioration information may be output based on the difference information. This makes it possible to output deterioration information from multifaceted information such as changes from the past state of the bridge 5, for example, and to output deterioration information with higher accuracy.

[0061] Furthermore, in step S2, the processing unit 12 may output deterioration information based on the bridge information and river information by referring to a diagnostic model trained using learning data in which the bridge information and river information are input data and the deterioration information is output data. In such a case, the diagnostic model records the correlation trained using learning data in which the bridge information and river information as shown in FIG. 13 are input data and the deterioration information is output data. In such a case, various information included in the bridge information or river information may be arbitrarily selected and used as the input data. This makes it possible to output deterioration information taking into account the state of the river 53, such as flooding of the river 53 and salt damage, for example, and makes it possible to output deterioration information with higher accuracy.

[0062] Also, in step S2, the processing unit 12 may output deterioration information based on the bridge information and the homologous bridge information by referring to a diagnostic model in which the bridge information and the homologous bridge information are input data and the deterioration information is output data. In such a case, the diagnostic model records the correlation degree learned by learning data in which the bridge information and the homologous bridge information as shown in FIG. 14 are input data and the deterioration information is output data. Also, in such a case, the input data may be arbitrarily selected and used from various information included in the bridge information or the homologous bridge information. As a result, for example, when a tendency of salt damage is observed in other bridges 5 of the same river 53, it becomes possible to output deterioration information taking into account the state of the other bridges 5, and it is possible to output deterioration information with higher accuracy. It is also possible to output deterioration information taking into account the traffic volume of other bridges 5 on the same road.

[0063] Next, in step S3, the acquisition unit 11 acquires question information including a question for the degradation information output in step S2. The question information is information including a question for the degradation information, and may be, for example, text-based information, but is not limited thereto. The acquisition unit 11 may acquire question information input by a user, for example, via the input unit 108. In this case, for example, a voice input by a microphone or the like (not shown) may be converted into text-based question information using a known voice recognition technology.

[0064] Next, in step S4, the processing unit 12 outputs answer information indicating an answer to the question included in the question information based on the question information acquired in step S3. In this case, the processing unit 12 may output the answer information based on the question information acquired in step S3 by referring to an answer model trained using training data in which the input data is the question information and the output data is the answer information. In addition, the processing unit 12 may perform a morphological analysis on the question information, for example, and output the answer information based on the analysis result. In this case, the processing unit 12 may use any known morphological analysis technique.

[0065] Also, in step S4, the processing unit 12 may refer to an answer model trained using learning data in which the input data is question information and deterioration information and the output data is answer information, and output answer information based on the deterioration information output in step S2 and the question information acquired in step S3. As a result, for example, when the user inputs question information such as "Please tell me the specification on which the diagnosis is based" in step S3 in response to the deterioration information output in step S2, the processing unit 12 can output answer information such as "It is based on the description of XX in XX item of XX section of the Standard Specifications for Concrete" in step S4 based on the deterioration information and the question information.

[0066] In addition, in step S4, the processing unit 12 may further add deterioration information previously diagnosed for the bridge 5 as input data, for example. In this case, the processing unit 12 may refer to an answer model learned using learning data in which the input data is question information, deterioration information, and past deterioration information, and the output data is answer information, and output answer information based on the deterioration information output in step S2, the question information acquired in step S3, and the past deterioration information previously acquired. As a result, for example, when a user inputs question information in step S3 for the deterioration information output in step S2, such as "Last time, the judgment category was 1, but why is it now judgment category 2?", in step S4, the processing unit 12 can output answer information such as "The previous judgment was 1, but the judgment category was set to 2 because it corresponds to the example described in the photo example on page 53 of Appendix 3 Judgment Guide of the Road Bridge Periodic Inspection Guidelines. When checking the previous record, deterioration that would lead to the above judgment was observed, but it is thought that the judgment was incorrect." based on the deterioration information, question information, and past deterioration information.

[0067] Furthermore, the deterioration information or the answer information may be output by the output unit 14 at any timing. The output unit 14 outputs the deterioration information or the answer information to the display unit 109 or the like. Furthermore, the output unit 14 may output the deterioration information or the answer information to the user terminal 2 or the server 3 via the public communication network 4.

[0068] This completes the operation of the structure diagnosis system 100 in the first embodiment. This makes it possible to output deterioration information from bridge information such as bridge images. This makes it possible to easily and automatically diagnose the deterioration state of a structure.

[0069] Second Embodiment A second embodiment of the present invention will be described below with reference to the drawings. Fig. 15 is a flowchart showing the operation of a building diagnosis system 100 in the second embodiment of the present invention. The building diagnosis system 100 in the second embodiment differs from the first embodiment in that it refers to an extraction model trained using learning data in which image information is used as input data and semantic information indicating features contained in the image information and positional information indicating the position of the feature are used as output data, and extracts semantic information and positional information based on the acquired image information. Below, a description of the same parts as in the first embodiment will be omitted.

[0070] First, in step S11, the acquisition unit 11 acquires various pieces of information. In this case, the acquisition unit 11 acquires various pieces of information including image information in the same manner as in step S1.

[0071] Next, in step S12, the processing unit 12 extracts semantic information and positional information based on the image information acquired in step S11. The semantic information is information indicating the semantic content of the feature included in the image information. The semantic information may be, for example, text information indicating a feature such as a crack or rust included in the image information. The semantic information may also be text information indicating the meaning of a feature such as a symbol indicated by a damage diagram included in the image information. The positional information is information indicating the position of the feature. The positional information may be, for example, data indicating the coordinates of the feature on the image information. The positional information may be, for example, data indicating the coordinates of the feature in a two-dimensional or three-dimensional space indicated by the image information. The positional information may also be, for example, information such as the distance or direction between multiple features.

[0072] In step S12, the processing unit 12 refers to an extraction model trained using training data in which image information is used as input data and semantic information and position information are used as output data, and extracts semantic information and position information based on the acquired image information. The extraction model is a model trained using training data, similar to the diagnostic model. The extraction model is trained using training data in which image information is used as input data and semantic information and position information are used as output data. In this case, the extraction model stores associations having the degree of association between the image information, which is the input data, and the semantic information and position information, which are the output data, as shown in FIG. 16, for example. In step S12, the processing unit 12 refers to an extraction model trained in advance, and extracts semantic information and position information based on the image information acquired in step S11.

[0073] Furthermore, in step S12, when the image information is a three-dimensional image including the entire bridge 5, the processing unit 12 may acquire, based on the image information, two-dimensional images each showing an image including each component included in the bridge 5. The processing unit 12 may extract each of the semantic information and position information based on each of the acquired two-dimensional images.

[0074] Next, in step S13, the processing unit 12 outputs the deterioration information based on the semantic information and the position information extracted in step S12. In this case, the processing unit 12 may output the deterioration information based on the semantic information and the position information extracted in step S12, by referring to a diagnostic model trained using learning data in which the semantic information and the position information are input data and the deterioration information is output data. In this case, the diagnostic model records the correlation trained by the learning data in which the semantic information and the position information as shown in FIG. 17 are input data and the deterioration information is output data. In this case, the input data may further include any information such as text information included in the bridge information. This makes it possible to extract the semantic information and the position information from image information such as a photograph or a damage diagram, and therefore makes it possible to extract information from image information such as a photograph or a damage diagram with higher accuracy.

[0075] Although the first and second embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0076] 1. Building diagnostic equipment 2. User terminal 3 Server 4 Public communication network 5. Bridges 6. Imaging device 10. Chassis 11 Acquisition Department 12 Processing section 14 Output section 15 Storage section 51 Bearing 52 Drain pipe 53 Rivers 54 Legs 55 digits 56 Bridge Abutment 57 Parapet 58 Leg foundation 100 Building diagnostic system 101 CPU 102 ROM 103 RAM 104 Preservation Department 105 Interface 106 Interfaces 107 Interface 108 Input section 109 Display section 110 Internal Bus

Claims

1. An acquisition means for acquiring bridge information relating to a bridge; a search means for referring to a diagnostic model trained using learning data in which bridge information is input data and deterioration information on the deterioration state of the bridge is output data, and outputting deterioration information based on the bridge information acquired by the acquisition means. A building diagnostic system characterized by:

2. the acquiring means acquires the bridge information including text information on attributes of the bridge and / or image information including the bridge, The search means refers to the diagnostic model trained using training data having text information and / or image information as input data and deterioration information as output data, and outputs deterioration information based on the text information and / or image information included in the bridge information acquired by the acquisition means. The building diagnostic system according to claim 1 .

3. The acquisition means further acquires difference information based on past bridge information regarding the bridge in the past and the bridge information, The searching means refers to the diagnostic model trained using training data having difference information as input data and deterioration information as output data, and outputs deterioration information based on the difference information acquired by the acquiring means. The building diagnostic system according to claim 1 .

4. The acquiring means further acquires river information related to the river on which the bridge is located, The search means refers to the diagnostic model trained using learning data in which bridge information and river information are input data and deterioration information is output data, and outputs deterioration information based on the bridge information and river information acquired by the acquisition means. The building diagnostic system according to claim 1 .

5. The acquiring means further acquires related bridge information regarding other bridges crossing the river of the bridge, The search means refers to the diagnostic model trained using learning data in which bridge information and related bridge information are input data and deterioration information is output data, and outputs deterioration information based on the bridge information and related bridge information acquired by the acquisition means. The building diagnostic system according to claim 1 .

6. An acquisition means for acquiring image information including a bridge; an extraction means for extracting the semantic information and the positional information based on the image information acquired by the acquisition means, by referring to an extraction model trained using training data in which image information is used as input data and semantic information indicating the meaning of features included in the image information and positional information indicating the position of the features are used as output data; a search means for referring to a diagnostic model trained using training data in which semantic information and positional information are input data and deterioration information showing the deterioration state of the bridge is output data, and outputting deterioration information based on the semantic information and positional information extracted by the extraction means. A building diagnostic system characterized by:

7. An acquisition step of acquiring bridge information relating to a bridge; a search step of referring to a diagnostic model trained using training data in which bridge information is input data and deterioration information relating to the deterioration state of the bridge is output data, and outputting deterioration information based on the bridge information acquired in the acquisition step. A building diagnostic program that features:

Images