JP7453654B1 - Computer program, structural abnormality determination device, and structural abnormality determination method - Google Patents

Abstract

The present invention provides a computer program, a structure abnormality determination device, and a structure abnormality determination method that can automatically determine the presence or absence of an abnormality in a structure from measurements without the intervention of an analyst. [Solution] A computer program acquires time-series data of XYZ components of acceleration measured by a sensor section provided at a predetermined location of a structure supported by columns, and adjusts the structure based on the acquired time-series data. Estimate one or more natural frequencies of A computer is caused to perform a process of calculating and determining whether or not there is an abnormality in the structure based on the calculated feature amount and natural frequency. [Selection diagram] Figure 1

Description

Application of Article 30, Paragraph 2 of the Patent Act Published on August 1, 2020, on the website of the 77th Annual Academic Lecture of the 2020 Japan Society of Civil Engineers National Conference https://confit. atlas. jp/guide/event/jsce2022/subject/CS9-21/advanced Published on September 15, 2020 at the 77th Annual Academic Lecture of the 2020 Japan Society of Civil Engineers National Conference

The present invention relates to a computer program, a structural abnormality determination device, and a structural abnormality determination method.

Patent Document 1 discloses that vibration is applied to a structure to obtain an elastic wave, the obtained elastic wave is Fourier transformed to obtain a frequency spectrum of the elastic wave, and a reference frequency spectrum of a normal elastic wave and the obtained frequency spectrum are A method for diagnosing a structure by comparing the two is disclosed.

JP 2021-81212 Publication

When determining abnormalities in a structure based on a frequency spectrum as in Patent Document 1, the natural frequency of the structure is visually confirmed and estimated from the acquired frequency spectrum, and the neighboring frequencies including the estimated natural frequency are estimated. Another method is to perform inverse Fourier transform on the frequency spectrum to generate a time-series waveform of acceleration, obtain particle motion based on the time-series waveform of acceleration, and identify the vibration mode. However, all such methods are performed manually. Therefore, for example, the search frequency range when estimating the natural frequency depends on the discretion of the analyst. Furthermore, when the natural vibration frequencies are close to each other, it is difficult to set the search frequency range.

The present invention has been made in view of such circumstances, and provides a computer program, a structure abnormality determination device, and a structure abnormality determination method that can automatically determine the presence or absence of abnormality in a structure from measurement without the intervention of an analyst. The purpose is to provide

The present application includes a plurality of means for solving the above-mentioned problems, but to give one example, a computer program calculates the acquire time series data, estimate one or more natural frequencies of the structure based on the acquired time series data, and calculate the frequency spectrum of the XYZ components of acceleration by Fourier transforming the acquired time series data. and calculating feature amounts of XYZ components based on the amplitude of the frequency spectrum at the natural frequency, and determining whether or not there is an abnormality in the structure based on the calculated feature amount and the natural frequency. Let the computer run it.

According to the present invention, the presence or absence of abnormality in a structure can be automatically determined from measurements without the intervention of an analyst.

1 is a diagram illustrating an example of the configuration of a structural abnormality determination system according to the present embodiment. It is a diagram showing an example of a coordinate system of a road information board (including pillars). It is a figure showing an example of time series data of acceleration. It is a figure showing an example of an analysis result by ERA method. It is a figure showing an example of a frequency spectrum. It is a figure which shows an example of a natural frequency and a feature amount. FIG. 3 is a diagram showing an example of a change in natural frequency over time. FIG. 3 is a diagram showing an example of the results of an abnormality reproduction test. It is a figure which shows the example which provided the sensor part in two places of a road information board. FIG. 3 is a diagram showing an example of natural frequencies and feature amounts when two acceleration sensors are used. It is a figure which shows an example of the processing procedure of abnormality determination of a structure when the ERA method is applied by a condition monitoring device. FIG. 3 is a diagram illustrating an example of a processing procedure for determining abnormality of a structure when a method other than the ERA method is applied by the condition monitoring device.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below based on drawings showing embodiments. FIG. 1 is a diagram showing an example of the configuration of a structural abnormality determination system according to the present embodiment. The structural abnormality determination system of this embodiment includes a condition monitoring device 50 as a structural abnormality determination device. The condition monitoring device 50 determines whether or not there is an abnormality in the facility structure 10 (in particular, the support supporting the facility structure 10), which is a determination target. Facility structures are also referred to as structures. The presence or absence of an abnormality is used to determine whether the structure is normal or abnormal in terms of its health. For example, the presence or absence of an event such as a crack or damage to a support that supports the facility structure 10 (including the support that supports the facility structure 10) is determined. represents. The facility structure 10 is, for example, a road information board installed on a road and supported by pillars, but the facility structure 10 is not limited to a road information board, and may include lighting devices, monitoring devices, ITS (Intelligent Transport systems) spots, road signs, and other appendages supported by columns. In this specification, a road information board 10 (including a support supporting the road information board 10) will be described as the facility structure 10.

The road information board 10 includes a sensor terminal 11 in addition to functional parts and devices such as an LED display board, a support supporting the LED display board, and a foundation (including anchor bolts) installed on the ground or a bridge. The sensor terminal 11 includes a data collection processing unit 12 and an acceleration sensor 16. The data collection processing unit 12 includes a data collection function 13, a data storage function 14, and a communication function 15. The shape of the support may be F-shaped, or may be gate-shaped or I-shaped. The power source for the road information board 10 may be an external commercial power source, or an internal battery or the like may be used in an installation location where there is no power source. The sensor terminal 11 is provided, for example, on the opposite side (predetermined location) of the support side of the road information board 10. Note that the installation location of the sensor terminal 11 is not limited to this. The acceleration sensor 16 includes, for example, one three-axis acceleration sensor. The acceleration sensor 16 measures the acceleration of the road information board 10 and transmits the measurement data to the data collection processing unit 12. The data collection processing unit 12 collects and processes the measurement data from the acceleration sensor 16 and transmits the measurement data to the condition monitoring device 50.

The acceleration sensor 16 measures vibrations of the road information board 10 (including the pillars), and detects vibrations not only during normal operation but also when the reinforcing ribs at the base of the pillar or the welded parts of the openings for wiring communication lines and power lines break. , Vibration in situations where axial force is lost, such as when an anchor bolt breaks or a nut falls off, is measured at all times or at any time. The acceleration sensor 16 acquires time series data of XYZ components of acceleration, and the acquired time series data is collected by the data collection function 13 of the data collection processing section 12 and once recorded in the data storage function 14. The road information board 10 (including the pillars) vibrates due to wind and passing vehicles, and the sensor terminal 11 measures the vibrations of the road information board 10 (including the pillars) and acquires it as time series data. For example, the acceleration time series data can measure vibrations for 3 minutes at predetermined intervals (eg, 10 minutes), but is not limited thereto. Further, the sampling frequency can be set to 200 Hz, but is not limited to this.

The data collection processing unit 12 of the road information board 10 is connected to the condition monitoring device 50 via the communication function 15. The acceleration time series data collected by the data collection processing unit 12 is stored in the condition monitoring device 50 together with the identifier of the road information board 10. Although only one road information board 10 is shown in FIG. 1 for convenience, in reality, time series data of acceleration is transmitted from a plurality (many) of road information boards 10 to the condition monitoring device 50 together with an identifier. The condition monitoring device 50 records the identifier of the road information board 10 and the time series data of acceleration acquired by the road information board 10 in association with each other.

The condition monitoring device 50 includes a processing section 51, a warning light 60, a display section 61, and an operation section 62. The processing unit 51 includes a structure attribute data management function 52, a data collection function 53, a data processing/detection function 54, a data accumulation function 55, a data output function 56, an abnormality detection function 57, a time correction function 58, and a sensor terminal time correction function. It has a function 59.

The processing unit 51 is connected to the sensor terminal 11 mounted on the road information board 10, receives measurement data, and performs analysis processing of the measurement data.

The warning light 60 alerts the user by blinking and sounding when an abnormality is detected in the structure or when the acceleration sensor 16 is abnormal.

The display section 61 includes a display panel, and can monitor the state of the road information board 10 through the processing section 51.

The operation section 62 includes a keyboard, a mouse, and the like, and can perform operations for monitoring the state of the road information board 10 through the processing section 51.

The structure attribute data management function 52 manages attribute data of structures. The attribute data of the structure includes, for example, the update date and time of the attribute data, identification information and IP address unique to the road information board 10, specification data of the road information board 10, information on the installation location, equipment specifications, construction method, and support specifications. , basic specifications, installation environment, etc.

The data collection function 53 automatically collects the measurement data from the road information board 10 by requesting the measurement data recorded in the data collection processing unit 12 of the road information board 10, for example, once every hour. Can receive. Note that data collection is not limited to automatic reception, and manual data collection is also possible.

The data processing/detection function 54 processes the collected measurement data, for example, (1) evaluation of the soundness of the support, (2) detection of loss of axial force of the anchor bolt, (3) detection of cracks in the support, (4) ) Detect structural abnormalities such as damage caused by flying objects, etc.

The data storage function 55 stores attribute data of the structure, measurement data for the required number of years (for example, about 3 years), and abnormality detection results.

The data output function 56 outputs various data accumulated in the data accumulation function 55 to a recording medium such as a USB memory.

The abnormality detection function 57 can call attention by blinking or sounding the warning light 60 when an abnormality is detected in the structure or when the acceleration sensor 16 is abnormal. Abnormalities in the acceleration sensor 16 include, for example, when the measured data exceeds the effective range, and when the measured data continues and shows no change. Note that the color of the warning light 60 may be changed depending on whether a structural abnormality is detected or when the acceleration sensor 16 is abnormal.

The time correction function 58 connects to an NTP (Network Time Protocol) server 30 connected to the status monitoring device 50 via a communication network, and automatically corrects the current time of the status monitoring device 50. The NTP server 30 is a server that provides time information to client devices on the network.

The sensor terminal time correction function 59 corrects the current time of the connected sensor terminal 11 based on the current time managed by the state monitoring device 50. Thereby, the current time can be synchronized between the state monitoring device 50 and the sensor terminal 11.

The processing unit 51 may include a required number of CPUs (Central Processing Units), MPUs (Micro-Processing Units), GPUs (Graphics Processing Units), and the like. Further, the processing unit 51 may be configured by combining DSPs (Digital Signal Processors), FPGAs (Field-Programmable Gate Arrays), and the like.

The data storage function 55 can be configured with, for example, a hard disk or a semiconductor memory, and may store a computer program (program product).

The computer program may be downloaded from an external device via the processing section 51 and stored in the processing section 51. Further, a computer program recorded on a recording medium (for example, an optically readable disk storage medium such as a CD-ROM) may be read by a recording medium reading section and stored in the processing section 51. A computer program can be deployed to run on a single computer or on multiple computers interconnected by a communications network.

The computer program can be loaded into the memory within the processing section 51, and the processing section 51 can execute the computer program. The processing unit 51 can execute processing determined by a computer program. That is, the processing by the processing unit 51 is also processing by a computer program.

Note that the display section 61 and the operation section 62 may be, for example, a monitoring terminal device configured with a personal computer, a tablet terminal, a smartphone, or the like. The monitoring terminal device is operated by a monitoring person to input processing commands to the condition monitoring device 50 or to display processing results by the condition monitoring device 50.

FIG. 2 is a diagram showing an example of the coordinate system of the road information board 10 (including the pillars). In this specification, as shown in FIG. 2, the XYZ coordinate system of the road information board 10 is as shown in FIG. That is, the X-axis direction is a direction perpendicular to the road (transverse direction), and is referred to as a "left-right" direction for convenience. The Y-axis direction is a direction parallel to the road (longitudinal direction), and is referred to as a "front-rear" direction for convenience. The Z-axis direction is the vertical direction. The acceleration sensor of the sensor unit 11 acquires time series data of acceleration in each of the XYZ directions.

FIG. 3 is a diagram showing an example of time series data of acceleration. When the road information board 10 vibrates due to wind or a vehicle running nearby, the acceleration sensor acquires an acceleration waveform as shown in FIG. 3 . In FIG. 3, the vertical axis represents acceleration (m/s ² ), and the horizontal axis represents time (seconds). In the example of FIG. 3, it can be seen that the acceleration of the X-axis component (left and right) is larger than the acceleration of the Y-axis and Z-axis components. Note that the acceleration waveform is just an example, and is not limited to the example shown in FIG. 3.

The processing unit 51 acquires time-series data of the XYZ components of acceleration measured by the sensor terminal 11 of the road information board 10, which is the target of determination for determining the presence or absence of an abnormality. The processing unit 51 estimates one or more natural frequencies of the road information board 10 (including the pillars) based on the acquired time series data. A known technique can be used as a method for estimating the natural frequency. In this embodiment, the ERA (Eigensystem Realization Algorithm) method is used, but other methods are also possible, such as (1) a method in which the frequency corresponding to the peak of the Fourier spectrum is set as the natural frequency, (2) a method in which a curve is fitted to the Fourier spectrum. (3) A method to calculate the correlation function and estimate the natural frequency from the peak of the Fourier spectrum, (4) an AR (Auto-Regressive) method, (5) a subspace method, ( 6) Ibrahim's method etc. may be used. The ERA method converts microtremor observation data into a cross-correlation function using NExT (National Excitation Technique), treating it as a free decay waveform, determining a characteristic matrix from the singular value decomposition of the Hankel matrix, and then determining the state transition matrix. The modal solution (natural frequency, damping ratio, modal amplitude) is estimated by performing eigenvalue analysis.

FIG. 4 is a diagram showing an example of an analysis result using the ERA method. In FIG. 4, the vertical axis shows the frequency, and the horizontal axis shows the number of times of analysis. In FIG. 4, four natural frequencies are estimated from the smallest frequency: 1.863 Hz, 1.932 Hz, 5.257 Hz, and 8.102 Hz. In addition, in order to avoid incorrectly selecting the natural frequency of another vibration mode when each natural frequency is close to each other or when false modes are mixed in the estimated value, the mode order is identified from the features of the estimated natural frequency. do. By identifying the mode order, the natural frequency change of that mode can be tracked. The characteristic amount of the natural frequency will be described later. Note that the processing unit 51 can display the analysis results based on the ERA method on the display unit 61.

When using the ERA method, the processing unit 51 acquires time-series data of XYZ components of acceleration measured by a sensor unit provided at a predetermined location of a structure supported by pillars, and performs processing based on the acquired time-series data. estimate one or more mode solutions of the structure, identify the mode order using the estimated mode solution as a feature quantity, and determine whether there is an abnormality in the structure based on the change over time of the feature quantity in the identified mode order. be able to. This makes it possible to automatically determine whether there is an abnormality in the structure from measurements without the intervention of an analyst.

When using a method other than the ERA method (for example, the AR method), the processing unit 51 may perform Fourier transform on the acquired time series data to calculate the frequency spectrum of the XYZ components of acceleration.

FIG. 5 is a diagram showing an example of a frequency spectrum. In FIG. 5, the vertical axis shows amplitude and the horizontal axis shows frequency. FIG. 5A shows the frequency spectrum of the X-axis component (left and right) of acceleration, and FIG. 5B shows the frequency spectrum of the Y-axis component (front and back) of acceleration. Note that the frequency spectrum of the Z-axis component of acceleration is not shown because no particular frequency peak was obtained. As shown in FIG. 5, it can be seen from the frequency spectrum of acceleration that there are four vibration modes. In FIG. 5A, two frequency peaks are seen in the X-axis component, and in FIG. 5B, two frequency peaks are seen in the Y-axis component. Note that the processing unit 51 may display the frequency spectrum on the display unit 61.

When using a method other than the ERA method (for example, the AR method), the processing unit 51 may calculate the feature amount of the XYZ components based on the amplitude of the frequency spectrum at the estimated natural frequency. . Specifically, the processing unit 51 calculates the feature amount of the XYZ component based on the estimated natural frequency and the amplitude of the frequency spectrum of the XYZ component at a plurality of frequencies separated before and after the natural frequency. It may be calculated.

For example, let the natural frequency be f ₀ and let f _-1 and f ₊₁ be a plurality of frequencies separated by the natural frequency. When the amplitude of the frequency spectrum of the X component at frequency f is α _x (f), the effective amplitude value (feature amount) A _x of the X component is A _x =√[{α _x (f _-1 ) ² + α _x (f ₀ ) ² + α _x (f ₊₁ ) ² }/3]. The effective amplitude value (feature quantity) A _y of the Y component and the effective amplitude value (feature quantity) A _z of the Z component can also be calculated in the same manner.

More specifically, the control unit 51 calculates the effective value of the amplitude of the frequency spectrum of the XYZ component at the estimated natural frequency and a plurality of frequencies separated before and after the natural frequency, and calculates the effective value of the amplitude of the frequency spectrum of the XYZ component. The feature values of the XYZ components may be calculated by normalizing the effective values. _{In this case ,} calculate _the ^{effective amplitude} _value ^of _the The effective amplitude values A _x , A _y , A _z of the components are expressed as vectors (A _x , A _y , A _z ), normalized (the vector size is set to 1), and the XYZ components of the vector are used as feature quantities. do.

When using a method other than the ERA method (e.g., the AR method), the processing unit 51 may determine that the estimated natural frequency has been correctly identified if the calculated feature quantity matches the features of the vibration mode.

FIG. 6 is a diagram showing an example of natural frequencies and feature amounts. In the frequency spectrum described above, frequency peaks can be seen around 1.93 Hz and 8.10 Hz in the frequency spectrum of the X-axis component, and around 1.86 Hz and 5.26 Hz in the frequency spectrum of the Y-axis component. Ta. Similar natural frequencies are identified in Figure 6 as well, and the magnitude relationship of the feature quantities X component A _{x ,} Y component A _{y , and} Z component A _z matches the characteristics of each order vibration mode. I understand.

In FIG. 6, mode indicates the eigenmode order. The primary mode is a mode in which the display part of the road information board 10 vibrates back and forth, and the upper end of the support and the display part vibrate in the same phase. Struts include bending and twisting. The secondary mode is a mode in which the display section of the road information board 10 vibrates from side to side, and the upper end of the support and the display section vibrate in the same phase. The strut includes bending. The tertiary mode is a mode in which the display part of the road information board 10 vibrates back and forth, but the torsion of the support is large, and the upper end of the support and the display part vibrate in opposite phases. The quaternary mode is a mode in which the display section of the road information board 10 vibrates left and right, but the vertical movement is larger than the left and right movement, and it is a vibration mode in which the driver shakes his head vertically as if pulling his chin. The strut includes bending.

When focusing on the XYZ component feature amounts, the Y component feature amounts are largest in the primary mode and the tertiary mode, and it is difficult to separate the primary mode and the tertiary mode using the feature amounts alone. Therefore, by focusing on the natural frequency, the vibration mode can be specified based on the feature amount and the natural frequency. In this case, in addition to or instead of the natural frequency, attention may be paid to the feature amounts of other components. That is, since there is a relatively large difference in the feature amount of the Z component between the primary mode and the tertiary mode, the vibration mode can be specified with higher accuracy by also considering the feature amount of the Z component. Alternatively, the mode order may be identified based on the trend similarity by focusing on all the feature amounts of the XYZ components.

That is, the processing unit 51 can identify one or more vibration modes of the road information board 10 (including the pillars) based on the natural frequency and the characteristic amounts of the XYZ components. Furthermore, the processing unit 51 may identify one or more vibration modes of the road information board 10 (including the pillars) based on the feature amounts of at least two or more components among the feature amounts of the XYZ components.

FIG. 7 is a diagram showing an example of a change in natural frequency over time. The processing unit 51 can obtain a change over time (time transition) in the natural frequency for each vibration mode specified based on the estimated natural frequency and the feature amount of the XYZ components. Fig. 7A shows the time course of the natural frequency in the first mode, Fig. 7B shows the time course of the natural frequency in the second mode, and Fig. 7C shows the time course of the natural frequency in the third mode. , FIG. 7D shows the time course of the natural frequency in the fourth mode.

The processing unit 51 determines whether or not there is an abnormality in the road information board 10 (including the pillars) based on the calculated feature amount and the estimated natural frequency. This makes it possible to automatically determine whether there is an abnormality in the structure from measurements without the intervention of an analyst.

Furthermore, the processing unit 51 can identify the eigenmode order using the calculated feature amount, and determine whether or not there is an abnormality in the structure based on the identified eigenmode order and eigenfrequency. Specifically, the processing unit 51 identifies a vibration mode based on the eigenmode order identified from the maximum value of the feature amount among the XYZ components and the eigenfrequency in the eigenmode order, and specifies the eigenvalue in the identified vibration mode. It is possible to determine whether there is an abnormality in the road information board 10 (including the pillars) based on the change in vibration frequency over time. For example, if the natural frequency changes (decreases) by more than a threshold value (threshold value for detecting changes over time in the natural frequency) compared to the normal natural frequency, it can be determined that there is an abnormality.

Furthermore, the processing unit 51 can set a threshold value for detecting a change in the natural frequency over time. Specifically, the input threshold value is set by the monitor inputting the threshold value into the condition monitoring device 50 using the operation unit 62. The threshold values to be set are determined based on the installation location and seasonal fluctuations (season, temperature, weather, etc.) by referring to the normal/abnormal judgment history (abnormality occurrence history) of multiple road information boards 10 installed in various locations. An appropriate threshold value can be selected depending on the situation. Further, an appropriate threshold value may be selected depending on the type, shape, etc. of the structure. Note that instead of having the observer input the threshold value for detecting the change over time in the natural frequency, the threshold value may be automatically set depending on the temperature, time of day, amount of solar radiation, season, etc.

The rigidity of structures such as the road information board 10 (including pillars) changes depending on the temperature. As the temperature rises, the rigidity of the structure becomes weaker and the natural frequency tends to change (lower). For example, the natural frequency of a structure installed outdoors is affected by temperature and solar radiation, and exhibits diurnal fluctuations, with the natural frequency being low during the day and rising late at night. Therefore, in order to correct the influence of temperature, for example, the standard deviation of the natural frequency in short-term diurnal fluctuations (day and night) can be calculated, and the threshold value can be set in consideration of the calculated standard deviation. If the standard deviation of the natural frequency is, for example, about 0.008 Hz, the threshold value may be set larger than 0.008 Hz (the minimum threshold value is set to 0.008 Hz).

The condition monitoring device 50 of this embodiment can determine whether there is an abnormality in the structure without using an external vibration device for the structure. In the case of structures installed on roads, vibrations due to traffic vibration and wind are almost always occurring, so artificial excitation is not necessary.

FIG. 8 is a diagram showing an example of the results of the abnormality reproduction test. In the abnormality reproduction test, a road information board 10 (including pillars) as shown in FIG. This is a confirmation test to see if it can be detected. Abnormality item A is a breakage of an anchor bolt of a support pillar supporting the road information board 10 or a falling off of a nut. In a healthy state (fastened with specified torque), the natural frequencies of each mode from the first mode to the fourth mode are 1.86 Hz, 1.90 Hz, 5.45 Hz, and 8.01 Hz, in order. In the abnormal state in which one bolt at the tension center is removed, the natural frequencies of each mode are 1.86 Hz, 1.86 Hz, 5.44 Hz, and 7.93 Hz, in order. The natural frequency in the secondary mode is lowered by 0.04 Hz, and it can be seen that an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz. Further, the natural frequency in the fourth mode is lowered by 0.08 Hz, which indicates that an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz.

Abnormality item B is a crack in the base rib attachment weld of the pillar supporting the road information board 10. In a healthy state (welding length 310 mm), the natural frequencies of each mode from the first mode to the fourth mode are 1.86 Hz, 1.90 Hz, 5.45 Hz, and 8.01 Hz, in order. In the abnormal state where the welding length is 230 mm, the natural frequencies of each mode are 1.84 Hz, 1.88 Hz, 5.44 Hz, and 7.98 Hz in this order. It can be seen that the natural frequencies in the first and second modes are lowered by 0.02 Hz, and an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz.

Abnormality item C is a crack in the support at the upper end of the base rib of the support that supports the road information board 10. In a healthy state, the natural frequencies of the first, third, and fourth modes are 1.85 Hz, 5.46 Hz, and 8.02 Hz, respectively. In the abnormal state where the crack length is 100 mm, the natural frequencies of each mode are 1.86 Hz, 5.46 Hz, and 7.98 Hz, in order. The natural frequency in the fourth-order mode is lowered by 0.04 Hz, and it can be seen that an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz.

Abnormal item D is a crack in the opening corner of the support pillar that supports the road information board 10. In a healthy state, the natural frequencies of each mode from the first mode to the fourth mode are 1.87 Hz, 1.92 Hz, 5.47 Hz, and 8.04 Hz, in order. In the abnormal state where the crack length is 200 mm, the natural frequencies of each mode are 1.83 Hz, 1.91 Hz, 5.41 Hz, and 8.03 Hz, in order. It can be seen that the natural frequency in the first mode is lowered by 0.04 Hz, and an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz. Further, the natural frequency in the third mode is lowered by 0.06 Hz, which indicates that an abnormality can be detected even when considering the variation of the minimum threshold value of 0.008 Hz.

As described above, according to this embodiment, there is no need to visually estimate the natural frequency from the frequency spectrum, and in order to confirm whether the natural mode has been estimated correctly, the frequency spectrum is estimated by the natural frequency. There is no need to perform inverse Fourier transform to determine particle motion at an identified frequency, and it is possible to automatically determine the presence or absence of an abnormality in a structure without the intervention of an analyst.

Next, the number of sensor units used in determining abnormality of a structure by the condition monitoring device 50 will be explained. In order to specify (identify) the vibration mode of a structure, it is generally necessary to install sensor units at multiple points (for example, two locations) and acquire time-series data of acceleration at multiple points.

FIG. 9 is a diagram showing an example in which sensor sections are provided at two locations on the road information board 10. The road information board shown in FIG. 9 is a road information board equivalent to the road information board 10 shown in FIG. 1, but it does not have to be the same. The acceleration sensor A is installed on the opposite side of the road information board 10 from the pillar side. Acceleration sensor B is installed at the top of the column.

FIG. 10 is a diagram showing an example of the natural frequency and feature amount when two acceleration sensors A and B are used. In FIG. 10, the natural frequency (frequency) and the feature amount are estimated or calculated using the same method as described above. The feature quantities A _x , A _y , and A _z use time series data of acceleration measured by acceleration sensor A, and the feature quantities B _x , B _y , and B _z use time series data of acceleration measured by acceleration sensor B. It uses data.

The processing unit 51 is provided at a predetermined location (a location similar to the example in FIG. 1) of a structure (for example, the road information board in FIG. 9) equivalent to a structure (for example, the road information board 10 in FIG. 1). The first time of the XYZ components of acceleration obtained from acceleration sensor A (first sensor section) and acceleration sensor B (second sensor section) provided at a location distant from a predetermined location (in the example of FIG. 9, the top of the column) Feature quantities A _x , A _y , A _z (first feature quantities) and feature quantities B _x , B _y , B _z (second feature quantities) of XYZ components calculated based on series data and second time series data , as well as related information that associates the equivalent natural frequency of the equivalent structure with the vibration mode of the equivalent structure. The related information is information as shown in FIG. 10, and can be obtained or collected in advance. Note that the related information may be statistical values (eg, average value, median value, mode value, etc.) of related information regarding a plurality of equivalent structures.

On the other hand, as described above, one sensor terminal 11 is installed on the road information board 10 (cantilever structure including pillars) that is subject to abnormality determination. The processing unit 51 calculates the feature amount and natural frequency of the XYZ components calculated based on the time series data of the XYZ component of the acceleration measured by the sensor terminal 11, and the feature amount A _x in the related information shown in FIG. By associating A _y , A _z (first feature quantity) and the equivalent natural frequency, simply installing one acceleration sensor can achieve the same result as installing multiple (for example, two) acceleration sensors. The vibration modes of a structure can be identified with precision. For example, among the feature quantities of XYZ components, a combination in which the combination of the feature quantity component with the largest value and the natural frequency matches (the numerical values do not have to match) is selected from the related information, and the selected combination The desired vibration mode may be the vibration mode corresponding to . Specifically, among the XYZ component feature quantities, if the feature quantity component with the largest value is A _y and the natural frequency is 1.863 Hz or its vicinity, the vibration mode of the first mode (back-and-forth/torsion) ), and if the natural frequency is 5.261 Hz or around 5.261 Hz, it is understood that the vibration mode is the third-order mode (back-and-forth/torsion). Thereby, even when one sensor is used, the vibration mode can be specified. Further, in the example of FIG. 10, the related information using two acceleration sensors is shown, but the number of acceleration sensors may be three or more.

As described above, the condition monitoring device 50 of this embodiment can determine whether there is an abnormality in the structure by simply using one sensor terminal 11.

FIG. 11 is a diagram showing an example of a processing procedure for determining abnormality of a structure when the ERA method is applied by the condition monitoring device 50. The processing unit 51 acquires time series data of the XYZ components of acceleration measured by the sensor terminal 11 installed at a predetermined location of a structure (for example, the road information board 10) (S11), and applies the acquired time series data to the time series data. The modal solution (natural frequency, damping ratio, modal amplitude) of the structure is estimated by applying the ERA method (S12).

The processing unit 51 identifies the mode order using the estimated mode solution as a feature quantity (S13), and determines whether or not there is an abnormality in the structure based on the temporal change of the feature quantity in the identified mode order (S14). The processing unit 51 determines whether or not to end the process (S15), and if the process is not to end (NO in S15), it continues the process from step S11 onwards. If the process is to be ended (YES in S15), the processing unit 51 ends the process.

As described above, the processing unit 51 acquires time series data of the XYZ components of acceleration measured by the sensor unit provided at a predetermined location of the structure supported by the pillar, and based on the acquired time series data. Estimating one or more mode solutions of the structure, identifying the mode order using the estimated mode solution as a feature quantity, and determining whether there is an abnormality in the structure based on the change over time of the feature quantity in the identified mode order. Can be done. This makes it possible to automatically determine whether there is an abnormality in the structure from measurements without the intervention of an analyst.

FIG. 12 is a diagram illustrating an example of a processing procedure for determining abnormality of a structure when a method other than the ERA method is applied by the condition monitoring device 50. The processing unit 51 acquires time series data of XYZ components of acceleration measured by the sensor terminal 11 installed at a predetermined location of a structure (for example, the road information board 10) (S21), and applies the time series data to the acquired time series data. The natural frequency of the structure is estimated by applying the AR method (a method other than the ERA method) (S22).

The processing unit 51 calculates the frequency spectrum of the XYZ components of acceleration by applying Fourier transform to the acquired time series data (S23), and calculates the effective amplitude value of the frequency spectrum at the estimated natural frequency (S24). . The processing unit 51 converts the calculated effective amplitude value into a vector representation, normalizes it, and calculates the feature quantity of the XYZ components (S25).

The processing unit 51 identifies the vibration mode based on the calculated feature amount and the estimated natural frequency (S26), and identifies the abnormality of the structure based on the calculated feature amount and the estimated natural frequency for each vibration mode. The presence or absence is determined (S27). The processing unit 51 determines whether or not to end the process (S28), and if the process is not to end (NO in S28), it continues the process from step S21 onwards. If the process is to be ended (YES in S28), the processing unit 51 ends the process.

In the embodiment described above, the sensor terminal 11 of the road information board 10 may be integrated into a compact unit by being mounted on a board. Furthermore, the sensor terminal 11 can be mounted on the road information board 10 regardless of whether it is newly installed equipment or existing equipment, and can be used to determine the presence or absence of an abnormality, monitor it, and the like. In addition, by providing an earthquake detector near the road information board 10, when an earthquake occurs, the data collection processing unit 12 of the road information board 10 directly sends an analysis request along with the recorded acceleration time series data. The information may be transmitted to the condition monitoring device 50 so that the condition monitoring device 50 immediately determines whether there is an abnormality in the structure.

In the above-described embodiment, the determination function of the condition monitoring device 50 may be built into each structure such as the road information board 10, and the determination result determined by the structure may be transmitted to the condition monitoring device 50. good. That is, a structure such as the road information board 10 may be provided with a configuration equivalent to the processing unit 51 of the condition monitoring device 50, so that the structure executes the processing shown in FIG. 11.

(Additional Note 1) A computer program acquires time series data of XYZ components of acceleration measured by a sensor section provided at a predetermined location of a structure supported by columns, and based on the acquired time series data, Estimate one or more natural frequencies of the object, perform Fourier transform on the acquired time series data to calculate the frequency spectrum of the XYZ components of acceleration, and calculate the XYZ components based on the amplitude of the frequency spectrum at the natural frequency. A computer is caused to execute a process of calculating a feature amount of and determining whether or not there is an abnormality in the structure based on the calculated feature amount and the natural frequency.

(Additional Note 2) The computer program performs the process in Appendix 1 of identifying the eigenmode order using the calculated feature amount and determining whether or not there is an abnormality in the structure based on the identified eigenmode order and the eigenfrequency. Let the computer run it.

(Additional note 3) In Appendix 1 or 2, the computer program determines the characteristics of the XYZ component based on the amplitude of the frequency spectrum of the XYZ component at the natural frequency and a plurality of frequencies separated by the natural frequency. Calculate the amount and have the computer perform the process.

(Appendix 4) In any one of Appendices 1 to 3, the computer program calculates the effective amplitude of the frequency spectrum of the XYZ components at the natural frequency and a plurality of frequencies spaced apart from the natural frequency. The computer is caused to perform processing of calculating the values, normalizing the calculated effective values, and calculating the feature amounts of the XYZ components.

(Additional Note 5) In any one of Appendixes 1 to 4, the computer program calculates the eigenmode order identified from the maximum value of the feature amount among the XYZ components, and the change over time of the eigenfrequency in the eigenmode order. A computer is caused to execute a process of determining whether or not there is an abnormality in the structure based on the above.

(Additional Note 6) The computer program causes the computer to execute a process of setting a threshold value for detecting a change in the natural frequency over time in any one of Addendums 1 to 5.

(Additional Note 7) In any one of Additional Notes 1 to 6, the computer program performs a process of identifying one or more vibration modes of the structure based on the characteristic quantities of the natural frequency and the XYZ components. Let the computer run it.

(Additional Note 8) In Addendum 7, the computer program is configured to detect accelerations obtained from a first sensor section and a second sensor section provided at a predetermined location of an equivalent structure equivalent to the structure and a location separated from the predetermined location. The first feature amount and the second feature amount of the XYZ component calculated based on the first time series data and the second time series data of the XYZ component, the equivalent natural frequency of the equivalent structure, and the vibration of the equivalent structure The relevant information associated with the mode is acquired, and the feature amount and natural frequency of the XYZ component calculated based on the time series data of the XYZ component of the acceleration measured by the sensor unit, the first feature amount and the equivalent natural frequency A computer is caused to execute a process of identifying the vibration mode of the structure by associating it with the vibration frequency.

(Additional note 9) The computer program acquires time-series data of XYZ components of acceleration measured by a sensor section provided at a predetermined location of a structure supported by columns, and based on the acquired time-series data, estimating one or more mode solutions of the object, identifying the mode order using the estimated mode solution as a feature quantity, and determining the presence or absence of an abnormality in the structure based on the change over time of the feature quantity in the identified mode order; Have a computer perform a process.

(Additional Note 10) The structural abnormality determination device includes a processing unit, and the processing unit collects time series data of XYZ components of acceleration measured by a sensor unit provided at a predetermined location of the structure supported by a support. One or more natural frequencies of the structure are estimated based on the acquired time series data, the acquired time series data is Fourier transformed to calculate the frequency spectrum of the XYZ components of acceleration, and the natural frequency is estimated based on the acquired time series data. A feature amount of the XYZ component is calculated based on the amplitude of the frequency spectrum in numerical terms, and presence or absence of an abnormality in the structure is determined based on the calculated feature amount and the natural frequency.

(Additional Note 11) The structural abnormality determination device includes a processing unit, and the processing unit collects time series data of XYZ components of acceleration measured by a sensor unit provided at a predetermined location of the structure supported by a support. Estimate one or more mode solutions of the structure based on the acquired time-series data, identify the mode order using the estimated mode solution as a feature, and calculate the change over time of the feature in the identified mode order. Based on this, it is determined whether or not there is an abnormality in the structure.

(Additional note 12) The structure abnormality determination method is based on the time series data of the XYZ components of acceleration measured by the sensor section installed at a predetermined location of the structure supported by the pillars. estimate one or more natural frequencies of the structure, calculate the frequency spectrum of the XYZ components of acceleration by Fourier transforming the acquired time series data, and calculate the frequency spectrum of the XYZ components of acceleration based on the amplitude of the frequency spectrum at the natural frequency. The feature amounts of the XYZ components are calculated, and the presence or absence of an abnormality in the structure is determined based on the calculated feature amounts and the natural frequency.

(Additional note 13) The structural abnormality determination method is based on the time series data of the XYZ components of acceleration measured by the sensor section installed at a predetermined location of the structure supported by the pillars. estimating one or more mode solutions of the structure, identifying the mode order using the estimated mode solution as a feature quantity, and determining whether or not there is an abnormality in the structure based on the change over time of the feature quantity in the identified mode order. judge.

Items described in each embodiment can be combined with each other. Moreover, the independent claims and dependent claims recited in the claims may be combined with each other in any and all combinations, regardless of the form in which they are cited. Furthermore, although the scope of claims uses a format in which claims refer to two or more other claims (multi-claim format), the invention is not limited to this format. It may be written using a multi-claim format that cites at least one multi-claim.

10 Road information board (facility structure)
11 Sensor terminal 12 Data collection processing unit 13 Data collection function 14 Data storage function 15 Communication function 16 Acceleration sensor 30 NTP server 50 Condition monitoring device 51 Processing unit 52 Structure attribute data management function 53 Data collection function 54 Data processing/detection function 55 Data accumulation function 56 Data output function 57 Abnormality detection function 58 Time correction function 59 Sensor terminal time correction function 60 Warning light 61 Display section 62 Operation section

Claims (12)

Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,

Estimating one or more natural frequencies of the structure based on the acquired time series data,
Fourier transform the acquired time series data to calculate the frequency spectrum of the XYZ components of acceleration,
Calculating feature amounts of XYZ components based on the amplitude of the frequency spectrum at the natural frequency,
Identify the eigenmode order using the calculated features,
determining whether there is an abnormality in the structure based on the identified eigenmode order and the eigenfrequency;
A computer program that causes a computer to perform a process. calculating a feature amount of the XYZ component based on the amplitude of the frequency spectrum of the XYZ component at the natural frequency and a plurality of frequencies separated by the natural frequency;
The computer program according to claim 1, which causes a computer to execute the process. Calculating the effective value of the amplitude of the frequency spectrum of the XYZ component at the natural frequency and a plurality of frequencies spaced apart before and after the natural frequency,
Normalize the calculated effective value and calculate the feature amount of the XYZ components,
The computer program according to claim 1, which causes a computer to execute the process. Determining the presence or absence of an abnormality in the structure based on the eigenmode order identified from the maximum value of the feature amount among the XYZ components and the change over time of the eigenfrequency in the eigenmode order;
The computer program according to any one of claims 1 to 3 , which causes a computer to execute the process. setting a threshold for detecting a change in the natural frequency over time;
The computer program according to any one of claims 1 to 3 , which causes a computer to execute the process. identifying one or more vibration modes of the structure based on the characteristic quantities of the natural frequency and the XYZ components;
The computer program according to any one of claims 1 to 3 , which causes a computer to execute the process. First time series data and second time series data of XYZ components of acceleration obtained from a first sensor section and a second sensor section provided at a predetermined location and a location separated from the predetermined location of an equivalent structure equivalent to the structure. obtaining related information that associates the first feature amount and the second feature amount of the XYZ components calculated based on the series data, and the equivalent natural frequency of the equivalent structure and the vibration mode of the equivalent structure;
The feature amount and natural frequency of the XYZ component calculated based on the time series data of the XYZ component of the acceleration measured by the sensor unit are associated with the first feature amount and the equivalent natural frequency, and the Identify vibration mode,
The computer program according to claim 6 , which causes a computer to execute the process. Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,
One or more mode solutions of the structure are estimated based on the acquired time series data, the mode order is identified using the estimated mode solution as a feature quantity, and the Determine whether there is an abnormality in the structure,
A computer program that causes a computer to perform a process. Equipped with a processing section,
The processing unit includes:
Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,
Estimating one or more natural frequencies of the structure based on the acquired time series data,
Fourier transform the acquired time series data to calculate the frequency spectrum of the XYZ components of acceleration,
Calculating feature amounts of XYZ components based on the amplitude of the frequency spectrum at the natural frequency,
Identify the eigenmode order using the calculated features,
determining whether there is an abnormality in the structure based on the identified eigenmode order and the eigenfrequency;
Structural abnormality determination device. Equipped with a processing section,
The processing unit includes:
Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,
One or more mode solutions of the structure are estimated based on the acquired time series data, the mode order is identified using the estimated mode solution as a feature quantity, and the Determine whether there is an abnormality in the structure,
Structural abnormality determination device. Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,
Estimating one or more natural frequencies of the structure based on the acquired time series data,
Fourier transform the acquired time series data to calculate the frequency spectrum of the XYZ components of acceleration,
Calculating feature amounts of XYZ components based on the amplitude of the frequency spectrum at the natural frequency,
Identify the eigenmode order using the calculated features,
determining whether there is an abnormality in the structure based on the identified eigenmode order and the eigenfrequency;
Structural abnormality determination method. Obtain time series data of the XYZ components of acceleration measured by a sensor section installed at a predetermined location of a structure supported by pillars,
One or more mode solutions of the structure are estimated based on the acquired time series data, the mode order is identified using the estimated mode solution as a feature quantity, and the Determine whether there is an abnormality in the structure,
Structural abnormality determination method.

Images