Model construction device and disaster damage prediction device

Abstract

This model construction device is provided with a control unit for constructing a prediction model (11) that performs machine learning by using disaster damage information (50) indicating whether or not a first facility attached to a first bridge was damaged when an earthquake occurred, first facility information (51) indicating specifications related to the attachment of the first facility, first disaster information (52) indicating an earthquake motion occurred in a first area where the first bridge was provided when the earthquake occurred, and first environment information (53) indicating characteristics of the ground in the first area to predict the possibility that a second facility is damaged by a disaster when receiving an input of second facility information indicating specifications related to the attachment of a second facility attached to a second bridge, second disaster information indicating an earthquake motion occurred at another time point in a second area where the second bridge was provided, and second environment information indicating characteristics of the ground in the second area.

Description

Model construction device and disaster prediction device

[0001] The present disclosure relates to a model construction device and a disaster prediction device.

[0002] There is a known study that attempts to analyze the disaster of bridges affected by earthquakes and quantitatively evaluate their vulnerability. As disclosed in Non-Patent Document 1, for the bridge-mounted pipelines inspected during four past major earthquakes in Japan, data on various parameters and the presence or absence of disasters were analyzed by quantification type 1, and the magnitude of the contribution of various parameters was quantitatively evaluated. There is also a study of constructing a standard damage rate curve for ground motion parameters.

[0003] Patent Document 1 and Patent Document 2 disclose technologies for predicting the disasters of facilities due to natural disasters.

[0004] International Publication No. 2021 / 240650 International Publication No. 2022 / 013974

[0005] Mikihiro Terashima, et al., "Systematic Arrangement, Analysis, and Consideration of Past Earthquake Damage Data of Communication Bridge System Facilities", Journal of the Japan Society of Civil Engineers, Series A1 (Structural and Earthquake Engineering), Vol. 75, No. 4, I_170-I_188, at the end of 2019 Yujiro Tomifuchi, et al., "Estimation of Ground Amplification Degree by Integration of Topographic Classification and Boring Data and Estimation of Ground Motion Distribution in the 2004 Niigata Chuetsu Earthquake", Journal of the Japan Earthquake Engineering Society, Vol. 7, No. 3, pp. 1-12, 2007

[0006] Conventional studies cannot be directly applied to predicting the disasters of pipelines mounted on bridges. The technologies disclosed in Patent Document 1 and Patent Document 2 also mainly target pipelines buried underground, and it is difficult to directly apply them to predicting the disasters of pipelines mounted on bridges.

[0007] In view of such circumstances, an object of the present disclosure is to predict the disasters of facilities mounted on bridges due to earthquakes.

[0008] A model construction device according to one embodiment includes a control unit that acquires damage information indicating whether or not a first piece of equipment attached to a first bridge was damaged when an earthquake occurred, first piece of equipment information indicating the specifications for the attachment of the first piece of equipment, first disaster information indicating the seismic motion that occurred in the first area where the first bridge is installed when the earthquake occurred, and first environmental information indicating the characteristics of the ground in the first area. By performing machine learning using the acquired damage information, first piece of equipment information, first disaster information, and first environmental information, the control unit constructs a predictive model that predicts the possibility of the second piece of equipment being damaged when second piece of equipment information indicating the specifications for the attachment of a second piece of equipment attached to a second bridge, second disaster information indicating seismic motion that occurred in the second area where the second bridge is installed at a time different from when the earthquake occurred, and second environmental information indicating the characteristics of the ground in the second area are input.

[0009] A disaster prediction device according to one embodiment includes a control unit that acquires a prediction model constructed by machine learning using disaster information indicating whether or not a first piece of equipment attached to a first bridge was damaged when an earthquake occurred, first piece of equipment information indicating the specifications related to the attachment of the first piece of equipment, first disaster information indicating the seismic motion that occurred in the first area where the first bridge is installed when the earthquake occurred, and first environmental information indicating the characteristics of the ground in the first area; second piece of equipment information indicating the specifications related to the attachment of a second piece of equipment attached to a second bridge, second disaster information indicating seismic motion that occurred in the second area where the second bridge is installed at a time different from when the earthquake occurred, and second environmental information indicating the characteristics of the ground in the second area; and inputs the acquired second piece of equipment information, second disaster information and second environmental information into the acquired prediction model to predict the possibility that the second piece of equipment will be damaged using the prediction model.

[0010] According to this disclosure, it is possible to predict the damage to equipment attached to bridges caused by earthquakes.

[0011] This is a block diagram showing the configuration of a system according to one embodiment. This is a schematic diagram showing an example of equipment attached to a bridge. This is a diagram showing the flow of constructing a prediction model. This is a table showing examples of values ​​for damage presence / absence, equipment characteristics, ground characteristics, and seismic motion. This is a table showing examples of categorization and standardization of the values ​​shown in Figure 4. This is a diagram showing an example of constructing a prediction model using a neural network. This is a diagram showing the flow of predicting the probability of damage and the presence / absence of damage. This is a table showing examples of damage prediction results. This is a diagram showing examples of damage prediction results.

[0012] One embodiment will be described below with reference to the figures.

[0013] In each figure, identical or corresponding parts are denoted by the same reference numerals. In the description of this embodiment, the description of identical or corresponding parts will be omitted or simplified as appropriate.

[0014] Referring to Figure 1, the configuration of the system 10 according to this embodiment will be described.

[0015] The system 10 according to this embodiment includes a model building device 20 and a disaster prediction device 30.

[0016] The model building device 20 is a device for building the prediction model 11. The model building device 20 may be a general-purpose computer such as a PC, a server computer such as a cloud server, or a dedicated computer. "PC" is an abbreviation for personal computer.

[0017] The model building device 20 may be able to communicate with the disaster prediction device 30 via a network. The network may include, for example, the internet, at least one WAN, at least one MAN, or any combination thereof. "WAN" is an abbreviation for wide area network. "MAN" is an abbreviation for metropolitan area network. The network may also include at least one wireless network, at least one optical network, or any combination thereof. The wireless network may be, for example, an ad hoc network, a cellular network, a wireless LAN, a satellite communication network, or a terrestrial microwave network. "LAN" is an abbreviation for local area network.

[0018] The disaster prediction device 30 is a device that uses a prediction model 11 constructed by the model construction device 20 to predict earthquake damage to equipment attached to the bridge 40, such as the pipeline 44 shown in Figure 2. The disaster prediction device 30 may be a general-purpose computer such as a PC, a server computer such as a cloud server, or a dedicated computer.

[0019] The bridge 40 is, for example, a road bridge built over a river. As shown in Figure 2, the bridge 40 comprises a deck slab 41, bridge girders 42, and support members 43 made of steel or the like. The pipeline 44 is attached to the bridge girders 42 below the deck slab 41 via the support members 43. The pipeline 44 is, for example, a structure for transporting water or gas, or a structure for laying communication cables or electric wires.

[0020] To predict damage to bridge 40, detailed information on the structural specifications of bridge 40 is necessary. However, it is difficult for the operator managing pipeline 44 to collect the same information for numerous facilities scattered over a wide area. Moreover, pipeline 44 differs from bridge 40 in terms of materials, structure, and dimensions, and pipeline 44 may be damaged even if bridge 40 is not. Therefore, in this embodiment, the model construction device 20 uses information on facilities owned by the operator managing pipeline 44 and information on seismic motion and ground conditions managed and published by government ministries or local authorities to construct a model for predicting damage to pipeline 44 due to earthquakes. The damage prediction device 30 uses this model to predict and output the probability of damage to pipeline 44 due to earthquakes or whether damage has occurred. Thus, according to this embodiment, it is possible to predict damage to pipeline 44 due to earthquakes.

[0021] Referring to Figure 1, the configuration of the model building apparatus 20 according to this embodiment will be described.

[0022] The model building device 20 comprises a control unit 21, a storage unit 22, a communication unit 23, and an input unit 24.

[0023] The control unit 21 includes at least one processor, at least one programmable circuit, at least one dedicated circuit, or any combination thereof. The processor is a general-purpose processor such as a CPU or GPU, or a dedicated processor specialized for a specific process. "CPU" is an abbreviation for central processing unit. "GPU" is an abbreviation for graphics processing unit. The programmable circuit is, for example, an FPGA. "FPGA" is an abbreviation for field-programmable gate array. The dedicated circuit is, for example, an ASIC. "ASIC" is an abbreviation for application specific integrated circuit. The control unit 21 controls each part of the model building device 20 and executes processes related to the operation of the model building device 20.

[0024] The storage unit 22 includes at least one semiconductor memory, at least one magnetic memory, at least one optical memory, or any combination thereof. The semiconductor memory is, for example, RAM, ROM, or flash memory. "RAM" is an abbreviation for random access memory. "ROM" is an abbreviation for read-only memory. The RAM is, for example, SRAM or DRAM. "SRAM" is an abbreviation for static random access memory. "DRAM" is an abbreviation for dynamic random access memory. The ROM is, for example, EEPROM. "EEPROM" is an abbreviation for electrically erasable programmable read-only memory. The flash memory is, for example, SSD. "SSD" is an abbreviation for solid-state drive. The magnetic memory is, for example, HDD. "HDD" is an abbreviation for hard disk drive. The storage unit 22 functions, for example, as main memory, auxiliary memory, or cache memory. The memory unit 22 stores information used for the operation of the model building device 20 and information obtained through the operation of the model building device 20.

[0025] The communication unit 23 includes at least one communication module. The communication module is, for example, a module compatible with a wired LAN communication standard such as Ethernet®, a wireless LAN communication standard such as IEEE 802.11, or a mobile communication standard such as LTE, 4G, or 5G. "IEEE" is an abbreviation for Institute of Electrical and Electronics Engineers. "LTE" is an abbreviation for Long Term Evolution. "4G" is an abbreviation for 4th generation. "5G" is an abbreviation for 5th generation. The communication unit 23 may communicate with the disaster prediction device 30. The communication unit 23 receives information used in the operation of the model building device 20 and transmits information obtained by the operation of the model building device 20. As shown in Figure 1, the communication unit 23 may transmit the prediction model 11 to the disaster prediction device 30.

[0026] The input unit 24 is, for example, a physical key, a capacitive key, a pointing device, a touchscreen integrated with a display, a camera, or a microphone. The input unit 24 accepts operations to input information used for the operation of the model building device 20. Instead of being provided in the model building device 20, the input unit 24 may be connected to the model building device 20 as an external input device. As a connection interface, an interface compatible with standards such as USB, HDMI®, or Bluetooth® can be used. "USB" is an abbreviation for Universal Serial Bus. "HDMI®" is an abbreviation for High-Definition Multimedia Interface.

[0027] The functions of the model building device 20 are realized by executing the model building program according to this embodiment on the processor acting as the control unit 21. In other words, the functions of the model building device 20 are realized by software. The model building program causes the computer to perform the operations of the model building device 20, thereby causing the computer to function as the model building device 20. That is, the computer functions as the model building device 20 by performing the operations of the model building device 20 according to the model building program.

[0028] Some or all of the functions of the model building device 20 may be implemented by a programmable circuit or a dedicated circuit as a control unit 21. In other words, some or all of the functions of the model building device 20 may be implemented by hardware.

[0029] Referring to Figure 1, the configuration of the disaster prediction device 30 according to this embodiment will be described.

[0030] The disaster prediction device 30 comprises a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, and a display unit 35.

[0031] The control unit 31 includes at least one processor, at least one programmable circuit, at least one dedicated circuit, or any combination thereof. The processor is a general-purpose processor such as a CPU or GPU, or a dedicated processor specialized for a specific process. The programmable circuit is, for example, an FPGA. The dedicated circuit is, for example, an ASIC. The control unit 31 controls each part of the disaster prediction device 30 and executes processes related to the operation of the disaster prediction device 30.

[0032] The storage unit 32 includes at least one semiconductor memory, at least one magnetic memory, at least one optical memory, or any combination thereof. The semiconductor memory is, for example, RAM, ROM, or flash memory. The RAM is, for example, SRAM or DRAM. The ROM is, for example, EEPROM. The flash memory is, for example, SSD. The magnetic memory is, for example, HDD. The storage unit 32 functions, for example, as a main memory, auxiliary memory, or cache memory. The storage unit 32 stores information used for the operation of the disaster prediction device 30 and information obtained by the operation of the disaster prediction device 30.

[0033] The communication unit 33 includes at least one communication module. The communication module is, for example, a module compatible with a wired LAN communication standard such as Ethernet®, a wireless LAN communication standard such as IEEE 802.11, or a mobile communication standard such as LTE, 4G, or 5G. The communication unit 33 may communicate with the model building device 20. The communication unit 33 receives information used in the operation of the disaster prediction device 30 and transmits information obtained by the operation of the disaster prediction device 30. As shown in Figure 1, the communication unit 33 may receive the prediction model 11 from the model building device 20.

[0034] The input unit 34 is, for example, a physical key, a capacitive key, a pointing device, a touchscreen integrated with a display, a camera, or a microphone. The input unit 34 accepts operations to input information used for the operation of the disaster prediction device 30. Instead of being provided in the disaster prediction device 30, the input unit 34 may be connected to the disaster prediction device 30 as an external input device. As a connection interface, an interface compatible with standards such as USB, HDMI®, or Bluetooth® can be used.

[0035] The display unit 35 is, for example, an LCD or an organic EL display. "LCD" is an abbreviation for liquid crystal display. "EL" is an abbreviation for electroluminescent. The display unit 35 displays information obtained by the operation of the disaster prediction device 30. Instead of being provided in the disaster prediction device 30, the display unit 35 may be connected to the disaster prediction device 30 as an external display device. As a connection interface, an interface compatible with standards such as USB, HDMI (registered trademark), or Bluetooth (registered trademark) can be used.

[0036] The functions of the disaster prediction device 30 are realized by executing the disaster prediction program according to this embodiment on the processor acting as the control unit 31. In other words, the functions of the disaster prediction device 30 are realized by software. The disaster prediction program causes the computer to perform the operations of the disaster prediction device 30, thereby causing the computer to function as the disaster prediction device 30. That is, the computer functions as the disaster prediction device 30 by performing the operations of the disaster prediction device 30 according to the disaster prediction program.

[0037] Some or all of the functions of the disaster prediction device 30 may be implemented by a programmable circuit or a dedicated circuit as the control unit 31. In other words, some or all of the functions of the disaster prediction device 30 may be implemented by hardware.

[0038] Programs such as model building programs or disaster prediction programs can be stored on non-temporary computer-readable media. Examples of non-temporary computer-readable media include flash memory, magnetic recording devices, optical discs, magneto-optical recording media, or ROM. Program distribution is carried out, for example, by selling, transferring, or lending portable media such as SD cards, DVDs, or CD-ROMs containing the programs. "SD" is an abbreviation for Secure Digital. "DVD" is an abbreviation for digital versatile disc. "CD-ROM" is an abbreviation for compact disc read only memory. Programs may also be distributed by storing them in server storage and transferring them from the server to other computers. Programs may also be provided as program products.

[0039] A computer, for example, stores a program stored on a portable medium or a program transferred from a server in its main memory. Then, the computer reads the program stored in the main memory with its processor and executes the processing according to the read program. The computer may also read a program directly from the portable medium and execute the processing according to the program. The computer may also execute the processing according to the received program sequentially each time a program is transferred to it from a server. Processing may also be performed by a so-called ASP type service, which does not transfer programs from the server to the computer, but realizes its function only through execution instructions and result retrieval. "ASP" is an abbreviation for application service provider. A program includes information used for processing by an electronic computer that is equivalent to a program. For example, data that is not a direct instruction to the computer but has the nature of defining the computer's processing falls under "equivalent to a program".

[0040] Referring to Figure 3, the operation of the model building apparatus 20 according to this embodiment will be explained. The operation described below corresponds to the model building method according to this embodiment.

[0041] The control unit 21 of the model construction device 20 acquires disaster information 50, first facility information 51, first disaster information 52, and first environment information 53. The disaster information 50 is information indicating whether the first facility installed on the first bridge was damaged during an earthquake. The first facility information 51 is information indicating the specifications regarding the installation of the first facility. In the present embodiment, the first facility information 51 also includes information indicating other characteristics of the first facility. The first disaster information 52 is information indicating the ground motion that occurred during an earthquake in the first area where the first bridge is installed. In the present embodiment, the first disaster information 52 also includes information indicating the ground motion that occurred during an earthquake in an area different from the first area. That is, the first disaster information 52 is information indicating the distribution of the ground motion that occurred during an earthquake in the entire plurality of areas including the first area. The first environment information 53 is information indicating the characteristics of the ground in the first area. In the present embodiment, the first environment information 53 also includes information indicating the characteristics of the ground in an area different from the first area. That is, the first environment information 53 is information indicating the distribution of the characteristics of the ground in the entire plurality of areas including the first area.

[0042] Specifically, if part or all of the disaster information 50, first facility information 51, first disaster information 52, and first environment information 53 are stored in the storage unit 22 in advance, the control unit 21 reads that information from the storage unit 22. Alternatively, the control unit 21 may receive part or all of the disaster information 50, first facility information 51, first disaster information 52, and first environment information 53 from another device via the communication unit 23. Alternatively, the control unit 21 may accept input of part or all of the disaster information 50, first facility information 51, first disaster information 52, and first environment information 53 from the user via the input unit 24.

[0043] In the present embodiment, the first facility is a pipeline 44 as shown in FIG. 2.

[0044] The disaster information 50 is classified into two states, "with damage" and "without damage", based on the investigation results of the pipeline 44 in past major earthquakes.

[0045] The definition of "disaster-affected" is set by the person who conducts disaster prediction according to the required performance of pipeline 44. When there is a surface damage that does not penetrate inside the pipeline 44, it is necessary to take countermeasures for pipes transporting combustibles or pipes under internal pressure, but for pipes accommodating cables, it may be subject to progress observation. In the present embodiment, the disaster of the first facility means a state in which the first facility cannot be continuously used without taking repair or reinforcement measures for the first facility due to seismic motion. As forms of disaster to pipeline 44 caused by an earthquake, for example, there are breakage of the pipeline body or detachment of joints. When the bridge girder 42 of the bridge 40 main body is damaged, or when the bridge pier is inclined or toppled, pipeline 44 is also affected by the disaster. However, since it is difficult to predict the disaster of the bridge 40 main body without detailed information on the structural specifications of the bridge 40 main body, the disaster of pipeline 44 when the bridge 40 main body is affected by the disaster is excluded from the prediction target.

[0046] In the present embodiment, the disaster information 50 includes, together with the presence or absence of disaster, the facility name and number for uniquely identifying pipeline 44.

[0047] The first facility information 51 includes pipe type, mounting position, or section length.

[0048] The pipe type is the type of pipeline 44. The pipe type is related to, for example, the strength of the material of pipeline 44. When there are multiple combinations of the outer diameter and inner diameter of pipeline 44, the pipe type is determined for each combination of the material, outer diameter, and inner diameter.

[0049] The mounting position is the position where pipeline 44 is installed with respect to the bridge girder 42. The mounting position is, for example, the side surface on the upstream or downstream side of the river where the bridge 40 is spanned, or the back surface between two main girders.

[0050] The section length is the length of the section where pipeline 44 is mounted. The section length is related to vibration characteristics. The section length is the bridge length when pipeline 44 is mounted over the entire length of bridge 40. The bridge length can be regarded as the width of the river where the bridge 40 is spanned.

[0051] The first equipment information 51 may further include the type of joint of the pipeline 44. The type of joint relates to the limit values ​​of the axial load, displacement, bending moment, and rotation angle at the joint location. The first equipment information 51 may further include the mounting method of the pipeline 44 or the spacing of the support members 43. Examples of mounting methods include end-supported beams, cantilever beams, or suspension structures. If the first equipment includes multiple pipelines 44, the first equipment information 51 may further include the total number of pipelines 44, or the total number of pipelines 44 arranged in the horizontal and vertical directions, respectively. The first equipment information 51 may further include items related to the strength deterioration of the pipeline 44, such as the year the pipeline 44 was mounted on the bridge 40, or the number of years elapsed since the year the pipeline 44 was mounted on the bridge 40, the distance from the coast to the bridge 40, the amount of de-icing agent spread on the pipeline 44, or the amount of sunshine at the location where the bridge 40 is located.

[0052] The attachment position, section length, attachment method, and spacing of the support members 43 are examples of specifications related to attachment. The pipe type, joint type, total number of pipes, years, elapsed years, distance from the coast, amount of de-icing agent, and sunshine hours are examples of other characteristics.

[0053] In this embodiment, the first equipment information 51 includes the pipe type, attachment location, and section length, as well as the equipment name and number, and equipment coordinates for identifying the location of the pipeline 44.

[0054] Disaster Information 52, the first item, includes the estimated distribution of seismic motion.

[0055] The estimated seismic motion distribution is represented by one or more index values ​​that indicate the magnitude of the seismic motion generated when an earthquake occurs in the first area, such as PGA or PGV. "PGA" is an abbreviation for peak ground acceleration. "PGV" is an abbreviation for peak ground velocity. In this embodiment, the estimated seismic motion distribution uses multiple types of index values. These index values ​​are, for example, seismic motion data published by the Japan Meteorological Agency, observation records published by the National Research Institute for Earth Science and Disaster Resilience, or data on assumed earthquakes published by local governments. For example, since seismic motion data and observation records are point information, when estimating a spatial distribution using these, it is conceivable to use the method proposed in Non-Patent Document 2.

[0056] In this embodiment, the first disaster information 52 includes a mesh code along with index values ​​such as PGA and PGV. The mesh code is a code for identifying a mesh. A mesh is an example of an area, such as the first area.

[0057] The first environmental information item 53 includes one or more items related to the hardness of the ground, such as microtopographic classification and AVS of the surface ground, which affect the ease of transmission or amplification of seismic motion. "AVS" is an abbreviation for average shear-wave velocity of the surface ground.

[0058] Microtopographic classification and AVS are examples of ground characteristics, respectively.

[0059] In this embodiment, the first environmental information 53 includes a mesh code along with the values ​​of each item, similar to the first disaster information 52.

[0060] The control unit 21 of the model building device 20 constructs a prediction model 11 by performing machine learning using the acquired damage information 50, first equipment information 51, first disaster information 52, and first environmental information 53. The prediction model 11 is a model that predicts the possibility of damage to the second equipment when second equipment information 61, second disaster information 62, and second environmental information 63, as shown in Figure 7, are input. The second equipment information 61 is information indicating the specifications related to the attachment of the second equipment attached to the second bridge. In this embodiment, the second equipment information 61 also includes information indicating other characteristics of the second equipment. The second disaster information 62 is information indicating seismic motion that occurred in the second area where the second bridge is installed, at a time different from the time of the earthquake related to the first disaster information 52. In this embodiment, the second disaster information 62 also includes information indicating seismic motion that occurred in an area different from the second area at that time. In other words, the second disaster information 62 is information showing the distribution of seismic motion that occurred at that other time point in the entirety of multiple areas, including the second area. The second environmental information 63 is information showing the characteristics of the ground in the second area. In this embodiment, the second environmental information 63 also includes information showing the characteristics of the ground in areas other than the second area. In other words, the second environmental information 63 is information showing the distribution of ground characteristics in the entirety of multiple areas, including the second area.

[0061] In this embodiment, the second equipment is a pipeline 44 as shown in Figure 2, similar to the first equipment. Damage to the second equipment refers to a state in which, similar to damage to the first equipment, the second equipment cannot be used continuously unless repair or reinforcement measures are taken due to seismic activity.

[0062] Details of the second equipment information 61, the second disaster information 62, and the second environmental information 63 are the same as those of the first equipment information 51, the first disaster information 52, and the first environmental information 53, respectively, so their explanation will be omitted.

[0063] In this embodiment, the control unit 21 of the model building device 20 constructs a prediction model 11 using the various information described above in order to calculate the predicted probability of damage Re for any attached pipeline. The damage information 50 includes whether or not attached pipelines were damaged during past major earthquakes.

[0064] An example of the values ​​to be used is shown in Figure 4. The dependent variable, whether or not an accident occurred, is set to 1 if an accident occurred and 0 if no accident occurred. Of the values ​​used as explanatory variables, pipe type and attachment location are qualitative data, while section length, AVS, PGA, and PGV are quantitative data. As shown in Figure 5, the qualitative data of pipe type and attachment location are divided into several categories, and each piece of equipment is set to a dummy variable of 1 if it falls into a category and 0 if it does not. The quantitative data of section length, AVS, PGA, and PGV are different physical quantities and have different ranges of values, so they are standardized to have a mean of 0 and a variance of 1. In other words, when performing machine learning, the control unit 21 categorizes the qualitative data values ​​such as pipe type or attachment location included in the first equipment information 51. When performing machine learning, the control unit 21 standardizes the quantitative data values ​​such as section length included in the first equipment information 51. When performing machine learning, the control unit 21 standardizes the index values ​​included in the first disaster information 52. When performing machine learning, the control unit 21 standardizes the values ​​of the items included in the first environmental information 53.

[0065] The prediction model 11 can have any structure as long as it can predict the probability of damage Re or whether or not any attached pipeline is damaged, for example, it can be a decision tree or a neural network. A neural network is one of the machine learning methods that attempts to artificially realize information processing in the brain. An example of constructing a prediction model 11 using a neural network is shown in Figure 6.

[0066] The output, the probability of damage Re, will be a value between 0 and 1. When making a binary prediction of whether or not damage has occurred based on the probability of damage Re, the threshold T for the probability of damage Re should be set according to the purpose. For example, if you want to predict damage as comprehensively as possible, you can make the threshold T small, and if you want to extract only the equipment with a high probability of being damaged, you can make the threshold T large.

[0067] Figure 6 illustrates the concept of model construction using a neural network. The number of hidden or intermediate layers may be one or multiple. The input explanatory variables and the number of nodes in the intermediate layers are not limited to those shown in the figure.

[0068] Referring to Figure 7, the operation of the disaster prediction device 30 according to this embodiment will be explained. The operation described below corresponds to the disaster prediction method according to this embodiment.

[0069] The control unit 31 of the disaster prediction device 30 acquires the second equipment information 61, the second disaster information 62, and the second environmental information 63. Specifically, if some or all of the second equipment information 61, the second disaster information 62, and the second environmental information 63 are pre-stored in the storage unit 32, the control unit 31 reads that information from the storage unit 32. Alternatively, the control unit 31 may receive some or all of the second equipment information 61, the second disaster information 62, and the second environmental information 63 from another device via the communication unit 33. Alternatively, the control unit 31 may accept input of some or all of the second equipment information 61, the second disaster information 62, and the second environmental information 63 from a user via the input unit 34.

[0070] The second piece of equipment information 61 is, for example, prepared in advance and stored in a database. In the event of a major earthquake, data on seismic motion is obtained from the Japan Meteorological Agency, the National Research Institute for Earth Science and Disaster Resilience, or local governments. From the obtained data, the distribution of seismic motion may be estimated using the method proposed in Non-Patent Document 2. Alternatively, seismic motion distributions published by research institutions or meteorological companies may be used. When predictions are used in disaster prevention planning, such as identifying weak points during peacetime, the assumed seismic motion data used in the earthquake damage assessment or hazard map of the local government where the equipment is located is used.

[0071] The control unit 31 of the disaster prediction device 30 extracts information used to predict damage to the second equipment from the second disaster information 62 by associating the second equipment with the index values ​​included in the second disaster information 62, using the equipment coordinates included in the second equipment information 61 and the mesh code included in the second disaster information 62. Specifically, the control unit 31 obtains index values ​​such as PGA and PGV, which indicate the magnitude of the seismic motion that occurred in the second area. Similarly, the control unit 31 extracts information used to predict damage to the second equipment from the second environmental information 63 by associating the second equipment with the values ​​of items included in the second environmental information 63, using the equipment coordinates included in the second equipment information 61 and the mesh code included in the second environmental information 63. Specifically, the control unit 31 obtains values ​​such as AVS, which indicate the characteristics of the ground in the second area.

[0072] The control unit 31 of the disaster prediction device 30 acquires the prediction model 11 constructed by the model construction device 20. Specifically, the control unit 31 receives the prediction model 11 from the model construction device 20 via the communication unit 33. Alternatively, if the prediction model 11 is pre-stored on an external storage medium, the control unit 31 may read the prediction model 11 from that storage medium.

[0073] The control unit 31 of the disaster prediction device 30 inputs the acquired second equipment information 61, second disaster information 62, and second environmental information 63 into the acquired prediction model 11, thereby predicting the possibility of the second equipment being damaged using the prediction model 11. Specifically, the control unit 31 inputs variables into the prediction model 11 to predict the probability of damage Re for each attached pipeline. The control unit 31 compares the predicted probability of damage Re with a threshold T to predict whether each attached pipeline is damaged or not.

[0074] The control unit 31 of the damage prediction device 30 outputs the result of predicting the likelihood of the second facility being damaged. Specifically, the control unit 31 displays the prediction result on the display unit 35 as a tabular list as shown in Figure 8, or as information on a geographic information system map as shown in Figure 9. To make it easier to understand visually, the prediction result may be displayed in different colors according to the probability of damage Re or whether or not damage has occurred.

[0075] As described above, in this embodiment, the model building device 20 constructs a prediction model 11 by performing machine learning with equipment information related to bridge-attached pipelines, environmental information related to the ground, and disaster information related to seismic motion as explanatory variables, and the presence or absence of damage as the dependent variable, regarding damage to bridge-attached pipelines during past earthquakes. The damage prediction device 30 uses the prediction model 11 to predict the probability of damage due to earthquakes or whether or not damage has occurred, based on the equipment information, environmental information, and disaster information corresponding to the bridge-attached pipeline to be predicted.

[0076] According to this embodiment, it is possible to predict the probability of damage to bridge-attached conduits due to earthquakes. By applying this embodiment, it becomes possible to identify conduits that are highly likely to be damaged, enabling the following responses: (1) If critical lines are laid on bridge-attached conduits with a high probability of damage, service interruptions can be prevented by switching the cable route to bridge-attached conduits with a low probability of damage in advance, or by constructing a backup route in preparation for damage. (2) Early identification of damaged areas and shortening of the recovery period can be achieved by formulating and implementing an efficient emergency inspection plan that prioritizes inspections of equipment with a high probability of damage. (3) Early recovery can be achieved by preparing and deploying recovery materials and personnel estimated in the necessary quantities based on the damage prediction.

[0077] For equipment that is predicted to have a high probability of being damaged even under normal circumstances, it is possible to build a highly reliable network by implementing measures such as relocating it to a safe location, securing alternative routes, securing alternative means of transportation, or reinforcing it.

[0078] This disclosure is not limited to the embodiments described above. For example, two or more blocks shown in the block diagram may be combined, or one block may be divided. Instead of executing two or more steps shown in the flowchart in chronological order as described, they may be executed in parallel or in a different order, depending on the processing capacity of the device performing each step, or as necessary. Other modifications are possible without departing from the spirit of this disclosure.

[0079] 10 System 11 Prediction Model 20 Model Construction Device 21 Control Unit 22 Memory Unit 23 Communication Unit 24 Input Unit 30 Disaster Prediction Device 31 Control Unit 32 Memory Unit 33 Communication Unit 34 Input Unit 35 Display Unit 40 Bridge 41 Deck 42 Bridge Girder 43 Support Member 44 Pipeline 50 Disaster Information 51 First Equipment Information 52 First Disaster Information 53 First Environmental Information 61 Second Equipment Information 62 Second Disaster Information 63 Second Environmental Information

Claims

1. A model building device comprising a control unit that acquires damage information indicating whether or not the first equipment attached to the first bridge was damaged when an earthquake occurred, first equipment information indicating the specifications related to the attachment of the first equipment, first disaster information indicating the seismic motion that occurred in the first area where the first bridge is installed when the earthquake occurred, and first environmental information indicating the characteristics of the ground in the first area, and performs machine learning using the acquired damage information, first equipment information, first disaster information, and first environmental information to construct a predictive model that predicts the possibility of the second equipment being damaged when second equipment information indicating the specifications related to the attachment of the second equipment attached to the second bridge, second disaster information indicating seismic motion that occurred in the second area where the second bridge is installed at a time different from when the earthquake occurred, and second environmental information indicating the characteristics of the ground in the second area is input.

2. The model construction apparatus according to claim 1, wherein the first equipment and the second equipment are each pipelines, and the first equipment information and the second equipment information each include the location to which the pipeline is attached, the length of the section to which the pipeline is attached, or the method used for attaching the pipeline.

3. The model building apparatus according to claim 1 or claim 2, wherein the first disaster information includes a plurality of index values ​​indicating the magnitude of the seismic motion that occurred in the first area when the earthquake occurred, and the control unit standardizes the plurality of index values ​​included in the first disaster information when performing machine learning.

4. A disaster prediction device comprising: a control unit that obtains a prediction model constructed by machine learning using damage information indicating whether or not the first equipment attached to the first bridge was damaged when an earthquake occurred, first equipment information indicating the specifications related to the attachment of the first equipment, first disaster information indicating the seismic motion that occurred in the first area where the first bridge is installed when the earthquake occurred, and first environmental information indicating the characteristics of the ground in the first area; second equipment information indicating the specifications related to the attachment of the second equipment attached to the second bridge, second disaster information indicating the seismic motion that occurred in the second area where the second bridge is installed at a time different from when the earthquake occurred, and second environmental information indicating the characteristics of the ground in the second area; and inputs the obtained second equipment information, second disaster information and second environmental information into the obtained prediction model to predict the possibility that the second equipment will be damaged using the prediction model.

Images