Analytical device, learning method, estimation method, and program

Abstract

To estimate the vulnerability of communication facilities to disasters.SOLUTION: An analysis device includes a data acquisition unit that acquires learning data including data about topography and a failure history during disasters for each section where a communication facility is disposed, and a learning unit that learns a model for estimating vulnerability of the communication facility using the learning data.SELECTED DRAWING: Figure 4

Description

Technical Field

[0006] , ,

[0005] , , , It is calculated from mesh data obtained by dividing the terrain into predetermined meshes. , ,

[0007] , , ,

[0001] The present invention relates to a technology for evaluating the vulnerability of communication equipment.

Background Art

[0002] Currently, optical cables (which may also be called optical fiber cables) are widely laid across the country and are used for communication services.

[0003] Regarding optical cables, it is considered that their vulnerability (the likelihood of failure) to disasters such as heavy rain and earthquakes varies depending on the terrain and surrounding environment of the location where they are laid.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] If the vulnerability of an optical cable to disasters can be known before a disaster occurs, for example, when there is a disaster forecast, countermeasures can be taken in advance. This also applies to communication equipment other than optical cables (e.g., metal cables, antennas, towers, utility poles, etc.).

[0006] The present invention has been made in view of the above points, and an object thereof is to provide a technology for estimating the vulnerability of communication equipment to disasters. [[ID=​​​​​​​​​​​Using the aforementioned training data, the communication equipment In response to the heavy rain disaster A learning unit that trains a model for estimating vulnerabilities and An analytical device equipped with the following features is provided. [Effects of the Invention]

[0008] The disclosed technology provides a method for estimating the vulnerability of communication equipment to disasters. [Brief explanation of the drawing]

[0009] [Figure 1] This diagram shows an example of the overall system configuration. [Figure 2] This diagram illustrates the concept of data analysis. [Figure 3] This diagram shows an example of an optical cable connection configuration. [Figure 4] This figure shows an example configuration of the analytical device 100. [Figure 5] This is a diagram illustrating the processing procedure. [Figure 6] This is a diagram illustrating the processing procedure. [Figure 7] This figure shows an example of model construction. [Figure 8] This figure shows an example of model construction. [Figure 9] This figure shows an example configuration of the analytical device 500. [Figure 10] This figure shows an example of the output of the analyzer 500. [Figure 11] This figure shows an example of the device's hardware configuration. [Modes for carrying out the invention]

[0010] Hereinafter, embodiments of the present invention (this embodiment) will be described with reference to the drawings. The embodiments described below are merely examples, and the embodiments to which the present invention is applied are not limited to the embodiments described below.

[0011] Hereinafter, an embodiment for estimating vulnerability will be described as the first embodiment, and an embodiment related to the measurement of an optical cable by OTDR will be described as the second embodiment. The first embodiment and the second embodiment may be implemented independently or in combination. As an example of the case of combining the first embodiment and the second embodiment, the loss situation (such as a deteriorated location) of the optical cable obtained in the second embodiment can be used for model learning and estimation using the model in the vulnerability estimation in the first embodiment.

[0012] In the first embodiment, an optical cable is cited as the communication facility for which vulnerability is estimated, but this is an example. The technology according to the present invention can be applied to communication facilities (e.g., metal cables, antennas, towers, utility poles, ducts, pipelines, manholes, etc.) not limited to optical cables.

[0013] [First Embodiment] (System Configuration Example) FIG. 1 shows a configuration diagram of the analysis system in the present embodiment. As shown in FIG. 1, there are an analysis device 100, a database 200, and a terminal 300, and these are connected to a network 400. The network 400 may be the Internet, a LAN, or any network.

[0014] The analysis device 100 is a device for estimating the vulnerability of an optical fiber to disasters. The database 200 is a database that holds data used for model learning by the analysis device 100 and for estimating vulnerability. As will be described later, the data used for model learning and for estimating vulnerability can be obtained from various information sources, and in FIG. 1, the data obtained from those various information sources is stored in the database 200.

[0015] The terminal 300 is, for example, a terminal held by a maintenance person who performs equipment maintenance. For example, the estimation result of the vulnerability of an optical fiber at a certain location is transmitted from the analysis device 100 to the terminal 300 and displayed on the terminal 300.

[0016] A maintenance worker with terminal 300 can, based on the vulnerability estimation results displayed on terminal 300, for example, learn that the probability of fiber optic cable failure at that location increases during heavy rain. In the event of a heavy rain forecast, they can take necessary measures in advance.

[0017] (Processing overview of analyzer 100) The processing overview will be explained with reference to Figure 2. First, data is collected for each section of the optical cable. In the example in Figure 2, data such as the construction date, whether the section is near a river, the elevation of the section, and the failure history during disasters (whether or not a failure occurred during a disaster) are collected. In addition to these, the presence or absence of faults in the section, rainfall (high, low), etc., may also be collected. Furthermore, in this embodiment, the range of equipment to be estimated is referred to as a "section," but this is just an example, and "points" may be used instead of "sections," or in addition to "sections." A "point" here refers to a location that appears as a point on a map, such as a manhole. If the range of equipment to be estimated is referred to as a "point," then equipment on and around that "point" will be the target of vulnerability estimation.

[0018] This data can be collected from sources such as diagrams describing the installation status of optical cables, on-site equipment maintenance records, equipment systems, and open data. The collected data is stored in, for example, database 200.

[0019] The analysis device 100 acquires data from the database 200 and uses that data to train a model. Then, the analysis device 100 inputs the data of the section for which vulnerability is to be estimated into the trained model, and outputs vulnerability information for that section. In this embodiment, the analysis device 100 outputs the probability that a failure will occur in the optical cable of that section in the event of a disaster as vulnerability information.

[0020] (Regarding the section) Let's explain the "section" mentioned above. Figure 3 shows an example of how optical cables are laid. Generally, optical cables are laid underground between communication buildings. In addition, optical cables may be laid above ground using utility poles, etc. Optical cables extend from the communication buildings to the user buildings.

[0021] Furthermore, there are locations (connection points) where optical cables are connected (fusion spliced, etc.) by optical closures. These connection points usually have manholes (MHs), allowing people to access the optical closures. In addition, there may be manholes along the optical cable route in places where there are no optical closures. While there are other facilities such as handholes (HHs) and utility poles, in this specification, when "manhole" is mentioned, it is assumed that the meaning also includes facilities such as "handholes" and "utility poles."

[0022] In this embodiment, the "section" of the optical cable can be anything that indicates a distance between two points. For example, the "section" could be the distance between two manholes, the distance between two communication buildings, the distance between a communication building and a user building, or the distance between two cities. Alternatively, in this embodiment, the "section" of the optical cable can be considered as the equipment through which the optical cable passes, or the equipment at the optical cable connection point.

[0023] (Regarding data collection methods) Data for training the model in the analysis device 100 and for vulnerability estimation using the model can be collected, for example, by the methods shown in (A) to (C) below. Hereafter, the person who collects the data used by the analysis device 100 will be referred to as the "user" (meaning the user of the analysis device 100).

[0024] (A) On-site equipment maintenance information / drawing information Typically, maintenance personnel performing on-site equipment maintenance visually inspect the locations where optical cables are laid (e.g., inside manholes) or their surroundings, or use recording devices such as cameras and measuring instruments, and record this information on paper as on-site equipment maintenance information. Additionally, diagrams showing the optical cable laying routes, for example, may exist. The user stores the on-site equipment maintenance information and diagram information collected visually or through recording devices in database 200.

[0025] (B) Equipment information Telecommunications carriers generally manage their communication equipment using equipment systems. Users collect equipment information from these systems and store it in database 200. Examples of equipment information obtained from the equipment system include the type and construction date (construction period) of communication equipment such as optical cables, the route of optical cables (location and distance of optical cables), and fault history.

[0026] (C) Open data (shape files, precipitation data, etc.) Shape files can be used as open data publicly available on the internet. A shape file is a collection of map data files containing both geometric information and attribute information.

[0027] By using shapefiles, topographic information such as the route (location) of optical cables or the location of manholes can be extracted. Precipitation data can also be obtained from open data. Various types of open data can be obtained from various open data sources, such as the presence or absence of hazard areas, the presence or absence of faults, and precipitation in a given section, and this data can be used for training and estimation.

[0028] (Configuration and operation of the analyzer 100) Figure 4 shows an example of the configuration of the analysis device 100. As shown in Figure 4, the analysis device 100 has a data acquisition unit 110, a learning unit 120, an analysis unit 130, an output unit 140, and a storage unit 150.

[0029] Refer to the flowchart in Figure 5 to explain the processing flow by the analyzer 100. Also, refer to Figure 6 as appropriate.

[0030] In S101 of Figure 5, the data acquisition unit 110 acquires training data from the database 200. The acquired data is stored in the storage unit 150. Figure 6 shows that the database 200 stores the following as training data for each section: the construction date of the optical fiber in that section, whether that section is near a river, the elevation of that section, ..., and the failure history. The "elevation" may be the average elevation of the section or the elevation of the highest point in the section.

[0031] Failure history indicates whether or not a failure occurred in a particular section during a specific disaster (e.g., Typhoon xx, xyz heavy rain). For example, when training a model to estimate vulnerability during heavy rain, one might use failure history from heavy rain disasters, and when training a model to estimate vulnerability during earthquakes, one might use failure history from earthquake disasters.

[0032] In S102 of Figure 5, the learning unit 120 trains a model using the data acquired in S101. The trained model is stored in the storage unit 150. Specifically, the model stored in the storage unit 150 consists of the equation and regression coefficients for regression analysis models, and the function and weight parameters for neural network models.

[0033] When using a neural network model for training, the data acquired for each interval (e.g., construction date, proximity to a river, elevation, etc., failure history) is used as input to the model, and the model's parameters are trained so that the output from the model correctly reflects the failure history (for example, if a failure exists, it outputs a high probability).

[0034] In S103 of Figure 5, the analysis unit 130 uses a trained model to estimate the vulnerability of the section for which vulnerability is to be estimated. Basically, by inputting data other than the failure history from the data items used in training for the section for which vulnerability is to be estimated into the model, the failure probability for that section is output. This failure probability corresponds to the probability that a failure will occur in the optical fiber of that section if a disaster of the same magnitude as the disaster assumed in the model training (e.g., Typhoon xx) occurs.

[0035] Figure 6 shows that intervals 5 to 8 are the intervals for which vulnerability is to be estimated, and that the failure probabilities for these intervals have been obtained by the analysis unit 130 (model).

[0036] In S104 of Figure 5, the output unit 140 outputs the analysis results (vulnerability estimation results) from the analysis unit 130. The output method can be any method. For example, the analysis results may be output on the display of the analysis device 100, or the analysis results may be transmitted via the network to the maintenance worker's terminal 300.

[0037] In this embodiment, one analysis device 100 performs both model learning and estimation using the model, but this is merely an example. For example, analysis device 100 may perform model learning, transmit the learned model to another analysis device 100, and that other analysis device 100 may perform vulnerability estimation. In this case, the analysis device 100 that performs model learning may be called a learning device.

[0038] (Model construction using logistic regression analysis) Next, as an example of model building (model learning), we will explain, with reference to Figure 7, how to build a model that calculates failure probability using logistic regression analysis. Note that using logistic regression analysis is just one example of model building. Any machine learning method other than logistic regression analysis can be used in model building. For example, random forests, SVMs, neural networks, etc., can be used.

[0039] As shown in Figure 7, the training data used here is "overlap with rivers, maximum slope angle, and failure history" for each section. The "maximum slope angle" is the largest slope angle among the surface topography slope angles in that section. More specifically, the "maximum slope angle" is calculated from mesh data obtained by dividing the topography into 5th-order mesh (250m mesh) sections, and is the data for the maximum slope angle in the areas where the cable passes. Alternatively, the average slope angle in the section where there is a slope may be used instead of the "maximum slope angle".

[0040] Logistic regression analysis is a method that can predict the probability of a binary outcome (dependent variable) occurring from multiple factors (explanatory variables). Here, we substitute the data for "overlap with rivers" into x1 and the "maximum slope angle" into x2 in the equation shown in Figure 7, and determine the regression coefficients a1, a2, and constant term b so that the result approaches the correct answer.

[0041] By inputting validation data (data for the interval in which vulnerability is to be estimated) into an equation using the learned regression coefficients a1, a2, and constant term b, the failure probability for that interval can be obtained. Figure 8 shows an example where specific values ​​of the validation data shown in Figure 7 are substituted.

[0042] For example, if a failure probability of "10%" is obtained here, it means that if there are 10 intervals with a failure probability of 10% under disaster conditions of a similar magnitude to those assumed in the training data, it can be estimated that one of those intervals will experience a failure.

[0043] (Regarding the data used for training and estimation) The data used for learning and estimation is not limited to construction date, proximity to a river (i.e., whether it runs along a river), elevation, rainfall, overlap with a river, maximum slope angle, average slope angle, presence of a hazard area, etc., but may also be used.

[0044] Furthermore, when using "construction date, proximity to a river (i.e., whether it is along a river), elevation, rainfall, overlap with a river, maximum slope angle, average slope angle, and whether there is a hazard area," you may use all of these, any one of them, or any multiple of them.

[0045] Furthermore, when estimating vulnerability to heavy rain disasters, data on "maximum slope angle" and "whether or not it is located along a river" may also be used. In other words, "maximum slope angle, whether or not it is located along a river, and failure history" may be used during training, and "maximum slope angle and whether or not it is located along a river" may be used during estimation.

[0046] "Maximum slope angle" and "whether or not it is along a river" are examples of data that can be used. Any data related to topography and rivers other than "maximum slope angle and whether or not it is along a river" may be used. Note that maximum slope angle, elevation, presence or absence of faults, and presence or absence of hazard areas are all examples of data related to topography. Furthermore, data related to topography may be used to estimate vulnerability to disasters other than heavy rain (e.g., vulnerability to earthquakes).

[0047] Furthermore, when using two sets of data—topographic data and river data—other data may be added in addition to these two sets. Alternatively, topographic data may be used without using river data.

[0048] [Second Embodiment] Optical Time Domain Reflectometers (OTDRs) are commonly used to measure the loss and relay status of optical cables (specifically, optical fiber cores). An OTDR is connected to one end of an optical cable and outputs an optical pulse signal. The OTDR measures the amount of light returning from the optical fiber and the time it takes for the light to return, thereby measuring the amount of loss, the location of the connection point, and other parameters.

[0049] Generally, the output (measurement result) of an OTDR is represented as waveform data (graph) with the vertical axis representing the optical loss level (attenuation) and the horizontal axis representing the distance from the OTDR (position along the length of the optical cable). In conventional technology, for example, a user would visually check the waveform data and determine that an abnormal loss was occurring at a certain point because the slope suddenly became steeper there.

[0050] However, the waveform data output by the OTDR displays the absolute value of attenuation, making it difficult to visually determine whether a change in the slope has occurred, etc. In other words, with conventional technology, it was sometimes difficult to determine the degree of attenuation in the optical cable or the section that was degraded from the waveform data obtained by the OTDR.

[0051] In this embodiment, the analysis device 500 described below makes it possible to easily determine the degree of attenuation in the optical cable and the deteriorated sections based on the waveform data obtained by the OTDR.

[0052] (Configuration and operation overview of the 500 analyzer) Figure 9 shows an example of the configuration of the analysis device 500. As shown in Figure 9, the analysis device 500 has a data acquisition unit 510, an analysis unit 520, an output unit 530, and a storage unit 550.

[0053] The data acquisition unit 510 acquires the measurement results of the OTDR (waveform data obtained using conventional technology). The acquired waveform data is stored in the storage unit 550. Alternatively, the waveform data may be stored in the database 200 shown in Figure 1, and the data acquisition unit 510 may acquire the waveform data from the database 200.

[0054] The analysis unit 520 reads waveform data from the storage unit 550 and performs processing on the waveform data, such as differentiation with respect to distance, as described later. The output unit 530 outputs the waveform data after processing by the analysis unit 520. Both the waveform data before processing and the waveform data after processing may be output.

[0055] The output unit 530 outputs, for example, the processed waveform data to a display. Alternatively, the processed waveform data may be output to the database 200 and used as training data or estimation data for vulnerability estimation as described in the first embodiment. Furthermore, the degraded intervals (degraded locations) identified by a person viewing the processed waveform data may be output to the database 200 and the information of these degraded intervals (degraded locations) may be used as training data or estimation data for vulnerability estimation as described in the first embodiment.

[0056] (Detailed processing by analysis unit 520) The following describes in more detail the processing steps of the analysis unit 520.

[0057] <Example 1> In Example 1, the analysis unit 520 differentiates the original waveform data (attenuated waveform) with respect to distance. The output unit 530 may output the differentiated waveform data. Differentiation makes any changes in the slope of the original waveform data more prominent, making it easier to identify attenuated areas.

[0058] The analysis unit 520 may perform only differentiation, but it may also perform the following operations.

[0059] <Example 2> As explained in Example 1, differentiation can sometimes amplify noise, making it difficult to distinguish from loss. Therefore, in Example 2, the analysis unit 520 performs a moving average on the waveform data before differentiation and then differentiates the waveform data with reduced noise. The width of the moving average may be determined according to the pulse width, for example, as will be described later, or an appropriate width may be determined experimentally.

[0060] <Example 3> In Example 3, noise is further reduced by applying smoothing to the differentiated waveform data obtained in Example 1, or to the differentiated waveform data obtained in Example 2.

[0061] Smoothing itself is an existing technique, and any method can be used. For example, as a smoothing method, techniques such as removing high-frequency components using a digital filter based on polynomial approximation or Fourier transform can be used.

[0062] Figure 10 shows the waveform before differentiation (OTDR optical attenuation waveform) and the waveform obtained by applying "moving average + differentiation + polynomial approximation (smoothing)" to that optical attenuation waveform. As shown in Figure 10, the processing in Example 3 makes it easy to identify loss points that are not clearly visible in the waveform before differentiation.

[0063] <Example 4> The parameters for moving averages and smoothing described in Examples 2 and 3 may be adjusted to match the pulse width of the OTDR's test optical pulse. The resolution of the optical attenuation graph obtained from an OTDR changes depending on the optical pulse width. Therefore, the moving average or smoothing parameters may be adjusted to remove components finer than the resolution and extract components smaller than that.

[0064] <Example 5> Specifically, for example, the resolution of the optical pulse width is 100m for a pulse width of 1μs, and the waveform of the differentiated pulse will be the shape of the upper part of a 100m sine wave or a trapezoid (that is, a sine wave or trapezoid with a period of 200m. Whether it is a sine wave or a trapezoid depends on the pulse width and the parameter value of the moving average). In Example 5, the analysis unit 520 is set to smoothing (e.g., number of dimensions) and moving average (e.g., window width) that can recognize this, and the processing is executed.

[0065] <Example 6> In any of the processes in Examples 1 to 5 (or Examples 2 to 5 depending on the waveform), the location of the loss can be identified more precisely. For example, with a pulse width of 1 μs, it can be determined that the peak of the sine wave at the point where attenuation occurs, or the upper end point on the side closer to the trapezoidal test device, is 50 m behind the point where the loss occurs.

[0066] Furthermore, even when attenuation (loss) occurs in multiple locations and the attenuation waveforms overlap, the loss locations can be separated and recognized based on the length of the attenuation and the degree of sine wave overlap.

[0067] <Example 7> Furthermore, the analysis unit 520 may apply deep learning machine learning to the waveform data acquired in any of Examples 1 to 5, in the same manner as with time-series data. For example, the waveform data may be input to a neural network model, and the model may be trained to output degradation points (e.g., distance from the light pulse output point).

[0068] Subsequently, the waveform data obtained in one of the examples 1 to 5 is input into the trained model regarding the measurement results of the optical cable where the degradation is to be identified, and the model outputs the degradation locations.

[0069] [Example of device hardware configuration] (Example hardware configuration) Both the analytical device 100 and the analytical device 500 can be implemented, for example, by having a computer run a program. This computer may be a physical computer or a virtual machine on the cloud. Hereinafter, the analytical device 100 and the analytical device 500 will be collectively referred to as "the device".

[0070] In other words, the device can be realized by using hardware resources such as the CPU and memory built into the computer to execute a program corresponding to the processing performed by the determination device 100. The program can be recorded on a computer-readable recording medium (such as portable memory), saved, and distributed. It can also be provided via a network such as the Internet or email.

[0071] Figure 11 shows an example of the hardware configuration of the computer described above. The computer in Figure 11 has a drive device 1000, an auxiliary storage device 1002, a memory device 1003, a CPU 1004, an interface device 1005, a display device 1006, an input device 1007, an output device 1008, etc., all of which are interconnected by a bus BS.

[0072] The program that enables processing on the computer is provided, for example, on a recording medium 1001 such as a CD-ROM or memory card. When the recording medium 1001 containing the program is set in the drive device 1000, the program is installed from the recording medium 1001 to the auxiliary storage device 1002 via the drive device 1000. However, the program does not necessarily have to be installed from the recording medium 1001; it may also be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program as well as necessary files and data.

[0073] When a program startup command is received, the memory device 1003 reads the program from the auxiliary storage device 1002 and stores it. The CPU 1004 then implements the functions related to the memory device 1003 according to the program stored in the memory device 1003.

[0074] Interface device 1005 is used as an interface for connecting to a network, etc. Display device 1006 displays a GUI (Graphical User Interface) or the like using a program. Input device 1007 consists of a keyboard and mouse, buttons, or a touch panel, and is used to input various operation instructions. Output device 1008 outputs the calculation results.

[0075] (Effects of the embodiment) According to the technology of the first embodiment, it is possible to estimate the vulnerability of communication equipment to disasters. According to the technology of the second embodiment, it is possible to easily understand the condition of optical cables based on waveform data obtained by OTDR.

[0076] (Note 1) This specification discloses at least the following analytical devices, learning methods, estimation methods, and programs. (Additional note 1) A data acquisition unit acquires data related to topography and learning data including failure history during disasters for each section where communication equipment is installed. A learning unit that learns a model for estimating the vulnerability of the communication equipment using the aforementioned training data. An analytical device equipped with the following features. (Additional note 2) An analysis unit estimates the vulnerability of the communication equipment in a given section by inputting estimation data, including data on the topography of a given section, into the model learned by the learning unit. The analytical apparatus described in Appendix 1, further comprising: (Additional note 3) The aforementioned training data includes the data relating to the topography and the data relating to the rivers. The analytical apparatus described in Appendix 1 or 2. (Additional note 4) The aforementioned communication equipment is an optical cable. The analytical apparatus described in any one of the appendices 1 to 3. (Additional note 5) A data acquisition unit that acquires estimation data including data on the terrain in a target section where communication equipment is installed, For each section in which the communication equipment is located, an analysis unit estimates the vulnerability of the communication equipment in the target section by inputting the estimation data into a model trained using topographic data and training data including failure history during disasters. An analytical device equipped with the following features. (Additional note 6) A learning method performed by an analytical instrument, A data acquisition step involves acquiring data on terrain and learning data including disaster failure history for each section where communication equipment is deployed. A learning step in which a model for estimating the vulnerability of the communication equipment is trained using the aforementioned training data. A learning method that includes [the following features]. (Additional note 7) An estimation method performed by an analytical instrument, A data acquisition step to acquire estimation data including data on the terrain in a target section where communication equipment is installed, For each section in which the communication equipment is located, an analysis step is performed to estimate the vulnerability of the communication equipment in the target section by inputting the estimation data into a model trained using topographic data and training data including failure history during disasters. An estimation method comprising the following: (Additional note 8) A program for causing a computer to function as one of the components of an analytical apparatus described in any one of the appendices 1 through 5.

[0077] (Note 2) This specification discloses at least the analytical apparatus, analytical methods, and programs described in the following sections. (Additional note 1) A data acquisition unit that acquires the original waveform data for the optical cable obtained by OTDR, Regarding the original waveform data, an analysis unit performs differentiation in the distance direction, An output unit that outputs waveform data obtained by the analysis unit. An analytical device equipped with the following features. (Additional note 2) The analysis unit processes the original waveform data or the differentiated waveform data according to the pulse width of the optical pulse signal used to measure the original waveform data. The analytical apparatus described in Appendix 1. (Additional note 3) The analysis unit differentiates the processed waveform data obtained by taking a moving average with respect to the original waveform data. The analytical apparatus described in Appendix 1 or 2. (Additional note 4) The analysis unit applies a smoothing process to the processed waveform data. The analytical apparatus described in Appendix 3. (Additional note 5) The analysis unit uses a trained model to estimate the degradation locations from the waveform data obtained by the analysis unit. The analytical apparatus described in any one of the appendices 1 through 4. (Additional note 6) An analytical method performed by an analytical device, A data acquisition step to obtain the original waveform data for the optical cable obtained by OTDR, The analysis step involves performing differentiation in the distance direction with respect to the original waveform data, An output step that outputs waveform data obtained by the analysis step, An analytical method that includes the following features. (Additional note 7) A program for causing a computer to function as one of the components of an analytical apparatus described in any one of the appendices 1 through 5.

[0078] Although this embodiment has been described above, the present invention is not limited to this specific embodiment, and various modifications and changes are possible within the scope of the gist of the invention as described in the claims. [Explanation of Symbols]

[0079] 100, 500 analyzer 200 databases 300 devices 400 Networks 110 Data Acquisition Unit 120 Learning Department 130 Analysis Department 140 Output section 150 Storage section 510 Data Acquisition Unit 520 Analysis Department 530 Output section 550 Storage section 1000 drive unit 1001 Recording media 1002 Auxiliary storage device 1003 Memory device 1004 CPU 1005 Interface device 1006 Display device 1007 Input device 1008 Output device

Claims

1. A data acquisition unit acquires learning data for each section where communication equipment is installed, including the maximum slope angle calculated from mesh data obtained by dividing the terrain into predetermined meshes, whether or not it is along a river, and failure history during heavy rain disasters. A learning unit that uses the aforementioned training data to train a model for estimating the vulnerability of the communication equipment to heavy rain disasters. An analytical device equipped with the following features.

2. An analysis unit estimates the vulnerability of the communication equipment to heavy rain damage in a given section by inputting estimation data, including data on the topography of a given section, into the model learned by the learning unit. The analytical apparatus according to claim 1, further comprising:

3. The aforementioned communication equipment is an optical cable. The analytical apparatus according to claim 1.

4. A data acquisition unit that acquires estimation data including the maximum slope angle and whether or not it is along a river in a target section where communication equipment is installed, For each section in which the communication equipment is located, an analysis unit estimates the vulnerability of the communication equipment to heavy rain disasters in the target section by inputting the estimation data into a model trained using training data that includes the maximum slope angle calculated from mesh data obtained by dividing the terrain into predetermined meshes, whether or not it is along a river, and failure history during heavy rain disasters. An analytical device equipped with the following features.

5. A learning method performed by an analytical instrument, A data acquisition step involves acquiring training data for each section where communication equipment is installed, including the maximum slope angle calculated from mesh data obtained by dividing the terrain into predetermined meshes, whether or not it is along a river, and failure history during heavy rain disasters. A learning step in which a model is trained to estimate the vulnerability of the communication equipment to heavy rain disasters using the aforementioned training data. A learning method that includes [the following features].

6. An estimation method performed by an analytical instrument, A data acquisition step to obtain estimation data including the maximum slope angle and whether or not it is along a river in a target section where communication equipment is installed, For each section in which the communication equipment is located, an analysis step is performed to estimate the vulnerability of the communication equipment to heavy rain disasters in the target section by inputting the estimation data into a model trained using training data that includes the maximum slope angle calculated from mesh data obtained by dividing the terrain into predetermined meshes, whether or not it is along a river, and failure history during heavy rain disasters. An estimation method comprising the following:

7. A program for causing a computer to function as a component of the analytical apparatus described in any one of claims 1 to 4.

Images