Monitoring devices, systems, methods, and programs

Abstract

To estimate the natural frequency of a structure and the order of peak frequencies to estimate a physical value without executing settings to the structure.SOLUTION: A monitoring device of the present invention comprises: storage means that stores vibration frequency spectrum data based on a vibration waveform measured by detection means provided on an object to be monitored; peak data analysis means that derives peak data including a peak frequency and a peak height for every peak appearing in the vibration frequency spectrum data; natural frequency group derivation means that estimates the natural frequency of the object to be monitored and a coefficient related to the appearance of the natural frequency on the basis of the peak data for every peak derived by the peak data analysis means; and order changing means that checks the value of the coefficient that is related to the appearance of the natural frequency and estimated by the natural frequency group derivation means, and changes the value of the coefficient related to the appearance of the natural frequency.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a monitoring device, system, method, and program, and is applicable to a device that automatically estimates a natural frequency from a measured vibration frequency and monitors the soundness or detects an abnormality of a structure, such as a bridge.

Background Art

[0002] For example, in a structure using cables such as a cable-stayed bridge, it is necessary to periodically check whether the tension assumed at the time of design is acting on the cables.

[0003] Conventionally, a method of directly measuring the tension acting on the cables supporting a structure using a jack or the like has been used, but it has required a lot of time and labor. Therefore, in recent years, a measurement method using vibration has attracted attention.

[0004] For example, since a cable is constantly vibrating due to wind or the like, by attaching a highly accurate acceleration sensor to the cable, a vibration waveform can be measured without exciting the cable.

[0005] By performing a Fourier transform on the obtained vibration waveform, a vibration frequency can be obtained. Generally, in the vibration of a cable or the like, it is known that in addition to the natural frequency, peaks are also obtained at frequencies that are integer multiples of the natural frequency. Note that the peak derived from the natural frequency is not limited to being obtained by multiplying the natural frequency by an integer, and there is also a method of adding other elements or converting the natural frequency by some function to obtain it. Thus, a group including the natural frequency and frequencies that are integer multiples of the natural frequency is called a natural frequency group. Also, the value of the integer multiple is called the order.

[0006] It is known that there is a certain relationship between cable tension and natural frequency, and tension can be estimated from the natural frequency. Generally, natural frequencies are low frequencies of a few Hz or less, making accurate measurement with sensors difficult. Furthermore, there is always an error in frequency measurement. Therefore, using a natural frequency group allows for more accurate measurement.

[0007] Regarding infrastructure structures, the vibration source is often a passing vehicle (vehicles in motion) or wind, making it impossible to obtain a consistent vibration environment. In such cases, situations may occur where there is no vibration and therefore no peak can be obtained, or where the peak changes significantly due to anomalies on the vibration source side. Due to these problems, it is difficult to determine the natural frequency group from the spectrum obtained from a single measurement, and it is necessary to use the spectra obtained from multiple measurements superimposed. Simply extracting the natural frequency group is not a simple matter.

[0008] Furthermore, it is known that in bridge cables and similar structures, peak frequencies do not appear at equal multiples of the natural frequencies, but rather the spacing between them gradually widens as the frequency increases. This characteristic makes it more difficult to estimate the natural frequencies and the order of each peak (a numerical value indicating how many times the natural frequency it is).

[0009] Furthermore, vibrations originating from equipment and devices attached to the object being measured may occur, resulting in peak frequencies unrelated to the natural frequency. Such peaks need to be removed by some means.

[0010] To address these issues, traditional methods often involved humans pre-setting parameters for normal oscillation, such as natural frequencies and thresholds. Efforts are now being made to automate these settings using machine learning and other methods. [Prior art documents] [Patent Documents]

[0011] [Patent Document 1] International Publication Number WO2017 / 064854 [Overview of the project] [Problems that the invention aims to solve]

[0012] However, as the number of structures to be monitored increases, it becomes impractical for humans to manually set the natural frequencies for each structure. Even if the setting process is automated using machine learning, it is still necessary to verify the settings, and ensuring that they are set correctly is difficult. Furthermore, the appropriate settings may change over time, and it would be extremely time-consuming for humans to perform parameter settings and periodically verify them. It is desirable to automate these tasks and minimize the human effort involved.

[0013] In particular, a challenge exists in that the estimation of natural frequencies and order can fail due to the characteristic that the peak spacing gradually widens as the frequency increases. Failure to estimate the order can affect physical values ​​used to judge the structural integrity, such as tension, thus impacting highly accurate and nuanced assessments of structural integrity.

[0014] Therefore, in view of the above-mentioned problems, the present invention aims to provide a monitoring device, system, method, and program that can estimate the natural frequency of a structure and the order of individual peak frequencies in order to estimate physical values ​​such as tension without making settings for each individual structure. [Means for solving the problem]

[0015] To solve these problems, the first monitoring device of the present invention is: In a monitoring device that monitors structures supported by cables and detects abnormalities in the monitored structures based on the tension of the cables estimated from the natural frequencies derived from the vibration waveforms of the cables, Target of surveillance cable A storage means for storing vibration frequency spectrum data based on vibration waveforms measured by a detection means provided therein, Multiple stored in a memory device Vibration frequency spectrum data Each The peak frequency and peak height for each peak that appears. valuePeak data analysis means for deriving peak data including, and for each peak derived by the peak data analysis means multiple Based on the peak data The peak data is sorted in descending order of peak height, and the peak frequencies of the peaks that appear at the same relative height are found to be the highest. Of the monitoring target Estimate Natural frequency Derived as follows Natural frequency group derivation means, and The peak frequency of the peak data is divided by the estimated natural frequency, and the resulting integer value is derived as the estimated order of the peak, which represents an integer multiple of the natural frequency. Order Derivation Means, and is characterized by The order derivation means changes the estimated order of the peak of the peak data if the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated eigenfrequency. This.

[0016] The second monitoring system of the present invention is In a monitoring system that monitors structures supported by cables and detects abnormalities in the monitored structures based on the cable tension estimated from the natural frequency derived from the vibration waveform of the cables, Monitoring target cable Detection means provided therein, and storage means for storing vibration frequency spectrum data based on the vibration waveform measured by the detection means, Multiple stored in a memory device Vibration frequency spectrum data Each For each peak appearing in the peak frequency and peak height value Peak data analysis means for deriving peak data including, and for each peak derived by the peak data analysis means multiple Based on the peak data The peak data is sorted in descending order of peak height, and the peak frequencies of the peaks that appear at the same relative height are found to be the highest. Of the monitoring target Estimate Natural frequency Derived as follows Natural frequency group derivation means, and The peak frequency of the peak data is divided by the estimated natural frequency, and the resulting integer value is derived as the estimated order of the peak, which represents an integer multiple of the natural frequency. Order Derivation Means, and is characterized by The order derivation means changes the estimated order of the peak of the peak data if the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency. This.

[0017] The third monitoring method of the present invention is In a monitoring method that monitors structures supported by cables and detects abnormalities in the monitored structures based on the cable tension estimated from the natural frequency derived from the vibration waveform of the cables, The storage means stores vibration frequency spectrum data based on the vibration waveform measured by the detection means provided in the monitoring target, and the peak data analysis means cable Derives peak data including the peak frequency and peak height for each peak appearing in the vibration frequency spectrum data, and the natural frequency group derivation means, for each peak derived by the peak data analysis means Multiple stored in a memory device Vibration frequency spectrum data Each For each peak appearing in the peak frequency and peak height value Peak data including, and the natural frequency group derivation means, based on the peak data for each peak derived by the peak data analysis means multiple Based on the peak data The peak data is sorted in descending order of peak height, and the peak frequencies of the peaks that appear at the same relative height are found to be the highest. Of the monitoring target Estimate Natural frequency Derived as Order Derivation Means The estimated order of the peak is derived as an integer value obtained by dividing the peak frequency of the peak data by the estimated natural frequency, and if the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency, the estimated order of the peak of the peak data is changed. Is characterized by

[0018] The fourth monitoring program of the present invention is: A monitoring program that monitors structures supported by cables and detects abnormalities in the monitored structures based on the tension of the cables estimated from the natural frequencies based on the vibration waveforms of the cables, has a storage means for storing vibration frequency spectrum data based on vibration waveforms measured by detection means provided on the monitored cables. Computers, Multiple stored in a memory device Vibration frequency spectrum data Each The peak frequency and peak height for each peak that appears. value A peak data analysis means that derives peak data including the peak data derived by the peak data analysis means multiple Based on peak data, The peak data is sorted in descending order of peak height, and the peak frequencies of the peaks that appear at the same relative height are found to be the highest. Target of surveillance Estimate natural frequency Derived as follows Means for deriving natural frequency groups, The peak frequency of the peak data is divided by the estimated natural frequency, and the resulting integer value is derived as the estimated order of the peak, which represents an integer multiple of the natural frequency. order Derivation Use as a means to The order derivation means changes the estimated order of the peak of the peak data if the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated eigenfrequency. It is characterized by the following: [Effects of the Invention]

[0019] According to the present invention, in order to estimate physical values ​​such as tension, without making settings for individual structures, the natural frequencies of structures can be estimated and the order of individual peak frequencies can be estimated. [Brief explanation of the drawing]

[0020] [Figure 1] This is an internal configuration diagram showing the internal configuration of the monitoring device according to the embodiment. [Figure 2] This is an overall configuration diagram showing the overall configuration of the monitoring system according to the embodiment. [Figure 3] This is an internal configuration diagram showing the internal configuration of the sensor device according to the embodiment. [Figure 4] This is a flowchart showing the process for deriving the natural frequency in the monitoring device according to the embodiment. [Figure 5] This is an explanatory diagram illustrating the method for deriving prominence according to the embodiment. [Figure 6] This flowchart shows the procedure for deriving prominences according to the embodiment. [Figure 7] This spectral diagram shows an example of a frequency spectrum in an embodiment. [Figure 8]This is an explanatory diagram (part 1) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 9] This is an explanatory diagram (part 2) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 10] This is an explanatory diagram (part 3) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 11] This is an explanatory diagram (part 4) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 12] This is an explanatory diagram (part 5) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 13] This is an explanatory diagram (part 6) illustrating the method for deriving the natural frequency group of the embodiment. [Figure 14] This is an explanatory diagram (part 1) illustrating the method for estimating the order and deriving the natural frequency group in the embodiment. [Figure 15] This is an explanatory diagram (part 2) illustrating the method for estimating the order and deriving the natural frequency group in the embodiment. [Modes for carrying out the invention]

[0021] (A) Embodiment Hereinafter, embodiments of the monitoring device, system, method, and program according to the present invention will be described in detail with reference to the drawings.

[0022] This embodiment describes a monitoring device and calculation method that have the function of recognizing the natural frequency of a structure from spectral data and automatically determining anomalies in the structure, without requiring settings for each individual structure.

[0023] (A-1) Configuration of the embodiment (A-1-1) Overall composition Figure 2 is an overall configuration diagram showing the overall configuration of the monitoring system according to the embodiment.

[0024] In Figure 2, the monitoring system 1 comprises a master unit 10, multiple sensor devices 20 (20-1 to 20-n; n is a positive integer) as slave units, and a monitoring device 30.

[0025] The "structures" to be monitored can include so-called infrastructure structures, such as bridges supported by cables (e.g., cable-stayed bridges, extradosed bridges, etc.). However, the structures to be monitored are not limited to bridges; they can be broadly applied to any structure whose soundness can be determined based on vibration waveforms.

[0026] The sensor device 20 is installed on the structure to be monitored and measures vibrations of, for example, cables, pillars, bridge girders, etc., of the structure using sensors such as acceleration sensors. The sensor device 20 also transmits signals including vibration waveforms measured by the sensors to the master unit 10. The sensor device 20 mainly consists of sensors such as acceleration sensors, a computing unit for processing data, a communication device, a clock, a timer to wake the entire device from a dormant state, and a battery.

[0027] The master unit 10 receives signals from each of the multiple sensor devices 20, including vibration waveforms measured by the sensors. The master unit 10 also transmits the signals received from each sensor device 20 to the monitoring device 30.

[0028] The master unit 10 may have the same configuration as the sensor device 20. For example, the master unit 10 may have a sensor that measures vibrations of cables, pillars, etc., of a structure, and the master unit 10 may directly transmit a signal including the vibration waveform of that sensor to the monitoring device 30.

[0029] In another modification, the master unit 10 may be equipped with all or part of the functions of the monitoring device 30, which will be described later. In this modification, the master unit 10 may function as a receiver for the monitoring device 30 and collect vibration waveforms from multiple sensor devices 20. In this case as well, the master unit 10 (in other words, the monitoring device 30) may be equipped with sensors such as acceleration sensors.

[0030] The monitoring device 30 has, for example, a function to save vibration waveform data, a function to determine the natural frequency group from the sensor's vibration waveform, a function to estimate the order of the natural frequencies, a function to calculate (estimate) tension from the values ​​of the natural frequency group, a function to notify the results, and a function to issue a warning as needed.

[0031] The monitoring device 30 collects vibration waveforms measured by the sensors from each sensor device 20 and stores the collected data. The monitoring device 30 then estimates the natural frequency group using the previously accumulated spectral data of vibration frequencies and the latest spectral data of vibration frequencies, and uses the natural frequency group to determine the tension of the cable.

[0032] Furthermore, the monitoring device 30 may use the natural frequency group to perform processes such as monitoring the condition of the structure, detecting deterioration over time, detecting anomalies, and predicting deterioration. For example, the monitoring device 30 may issue a warning if the estimated tension value deviates from a predetermined range, or if it deviates significantly from the range of values ​​obtained from past measurements by exceeding a threshold. When the monitoring device 30 issues a warning, it may perform measurements again, estimate the tension by estimating the natural frequency group and the order of the natural frequencies, and then issue a warning after confirming the occurrence of an anomaly.

[0033] The monitoring device 30 collects and stores (stores) vibration waveform values ​​(data) measured by the acceleration sensor 203 from each sensor device 20 via the master unit 10. The monitoring device 30 then performs a Fourier transform on the vibration waveform of the acceleration sensor 203 to derive the vibration frequency spectrum. In other words, the monitoring device 30 derives the vibration frequency using a Fourier transform.

[0034] The monitoring device 30 analyzes the shape of the frequency spectrum, extracts specific frequency ranges where peaks appear under normal conditions of the structure, and stores peak data for each specific frequency range. For example, if the acceleration sensor 203 takes measurements once a day, the monitoring device 30 measures and stores peak data for each specific frequency range extracted from the frequency spectrum once a day.

[0035] In Figure 2, the network including the master unit 10, sensor equipment 20, and monitoring device 30 may use either a wireless or wired connection. For example, the sensor equipment 20 and the master unit 10 may constitute a sensor network SN. The communication method of the sensor network SN can be a low-speed wireless communication method, such as a specific low-power wireless communication method. For example, it may be a wireless network standard such as IEEE 802.11a / b / g / n, or a wireless communication method such as IEEE 802.15.4 or Bluetooth®. The sensor equipment 20 and the master unit 10 are each assigned a unique address (e.g., MAC address, short address, IP address, etc.) on the sensor network SN. Also, for example, the network between the master unit 10 and the monitoring device 30 may be a backbone network NT. The backbone network NT is, for example, the Internet or Ethernet®. The backbone network NT may be a wireless or wired connection.

[0036] (A-1-2) Internal configuration of the monitoring device Figure 1 is an internal configuration diagram showing the internal configuration of the monitoring device 30 according to the embodiment.

[0037] In Figure 1, the monitoring device 30 of the embodiment includes a communication unit 301, a control unit 302, a peak data analysis unit 303, a natural frequency group calculation unit 308, a determination unit 305, and a data storage unit 306. The natural frequency group calculation unit 308 also includes a natural frequency group derivation unit 304 and an order derivation unit 307.

[0038] The communication unit 301 is a communication interface for network communication connected to the master unit 10.

[0039] The control unit 302 is a device or processing unit that manages various functions of the monitoring device 30. The control unit 302 includes, for example, a CPU, ROM, RAM, EEPROM, input / output interface, etc. Various processes are realized when the CPU executes a processing program (for example, a monitoring program) stored in the ROM.

[0040] The data storage unit 306 stores information for determining the structural integrity of the structure. For example, the data storage unit 306 stores values ​​indicating the vibration waveform measured by the acceleration sensor 203, frequency spectrum data obtained by Fourier transforming the vibration waveform of the acceleration sensor 203, and peak data in a specific frequency range (for example, data including frequency, values ​​indicating peak height, date and time information, etc.) for each acceleration sensor 203.

[0041] The peak data analysis unit 303 refers to the data storage unit 306 and performs a Fourier transform on the vibration waveform measured by the acceleration sensor 203 to obtain the frequency spectrum. The peak data analysis unit 303 also uses the concept of peak height to determine the frequency at which the peak is the maximum value (hereinafter referred to as the "peak frequency") from among the many peaks appearing in the frequency spectrum. At this time, there are also maximum values ​​that are unsuitable as peaks, so the peak data analysis unit 303 removes the unsuitable peaks.

[0042] The peak data analysis unit 303 stores data including the peak frequency and the peak value of the spectrum of that peak frequency (hereinafter also referred to as "peak height") as peak data in the data storage unit 306.

[0043] The natural frequency group derivation unit 304 uses the peak data and past peak data stored in the data storage unit 306 to derive the natural frequency group of the structure to be monitored. The natural frequency group derivation unit 304 also determines the natural frequencies from the natural frequency group. A detailed explanation of the natural frequency group derivation unit 304 will be provided in the operation section.

[0044] The order derivation unit 307 estimates the order of the peak frequency (i.e., a coefficient related to the natural frequency). Using the characteristic that the interval between peaks widens as the frequency increases, it determines whether the estimated order of the peak frequency is correct or not. If it is determined to be incorrect, it re-estimates the order of the peak frequency. This calculation of estimating the order of the peak frequency is repeated multiple times until an estimated value of the order that is considered correct is derived.

[0045] The determination unit 305 uses the natural frequencies obtained by the natural frequency group derivation unit 304 to determine the tension (for example, the tension acting on a cable supporting a structure). The relationship between the tension on the cable and the natural frequency can be determined by applying an existing relational formula, and the tension can be determined by substituting the natural frequency into the relational formula.

[0046] (A-1-3) Internal configuration of sensor equipment Figure 3 is an internal configuration diagram showing the internal configuration of the sensor device 20 according to the embodiment.

[0047] In Figure 3, the sensor device 20 includes a communication unit 201, a control unit 202, an acceleration sensor 203, a schedule determination unit 204, a timer unit 205, a clock unit 206, and a data storage unit 207. The sensor device 20 also has a power source such as a battery.

[0048] The sensor device 20 may be configured using hardware, or some of its components may be configured using software.

[0049] The communication unit 201 is a communication interface that communicates with the network connected to the master unit 10.

[0050] The acceleration sensor 203 measures vibrations of structures such as bridges. The acceleration sensor 203 measures instantaneous acceleration and provides a measurement signal (vibration waveform) to the control unit 202. A single sensor device 20 may be equipped with multiple acceleration sensors 203.

[0051] Furthermore, the sensor device 20 may be equipped with various types of sensors in addition to, or instead of, the acceleration sensor 203, such as a temperature sensor, humidity sensor, vibration sensor, infrared image sensor, etc.

[0052] The control unit 202 is a device and processing unit that manages various functions in the sensor device 20. The control unit 202 includes, for example, a CPU, ROM, RAM, EEPROM, input / output interface, etc. Processing is realized when the CPU executes a processing program (for example, a measurement program) stored in the ROM.

[0053] The control unit 202 receives information regarding measurement operations from the monitoring device 30 via the communication unit 201 and provides measurement operation schedule information for operating the acceleration sensor 203 to the schedule determination unit 204.

[0054] The schedule determination unit 204 determines the periodic measurement timing for operating the acceleration sensor 203 based on the measurement operation schedule information received from the monitoring device 30. The schedule determination unit 204 sets the periodic measurement timing of the acceleration sensor 203 in the timer unit 205.

[0055] The timer unit 205 manages the measurement timing set by the schedule determination unit 204. When the time on the clock unit 206 reaches the measurement time of the measurement timing, the timer unit 205 activates the control unit 202. Upon activation of the control unit 202, the acceleration sensor 203 also activates and begins sensing.

[0056] Furthermore, to conserve power, the sensor device 20 enters a dormant state when not measuring. In other words, all components except the timer unit 205 and the clock unit 206 are shut down.

[0057] The clock unit 206 is a clock that displays the current time.

[0058] (A-2) Operation of the embodiment Figure 4 is a flowchart showing the processing in the monitoring device 30 according to the embodiment.

[0059] For example, an acceleration sensor 203 is installed on an infrastructure structure where the vibration source is not constant. The acceleration sensor 203 operates for a certain period of time and measures vibrations. Then, the sensor device 20 transmits a signal containing the vibration waveform values ​​measured by the acceleration sensor 203 to the monitoring device 30.

[0060] The monitoring device 30 stores the vibration waveform value for each acceleration sensor 203. The monitoring device 30 performs a Fourier transform on the vibration waveform to obtain the vibration frequency spectrum.

[0061] Furthermore, the monitoring device 30 estimates the natural frequency group of the structure using past vibration waveform data, as described below. Then, the monitoring device 30 estimates the tension using the natural frequencies of the structure.

[0062] The natural frequency group is determined from the frequencies (peak frequencies) at which the frequency spectrum takes its maximum value. However, even if a frequency takes its maximum value, there are some maximum values ​​that are unsuitable to use as peaks. For example, peaks caused by vibrations resulting from other factors are considered such. Such unsuitable maximum values ​​may be deleted.

[0063] Since the peak frequencies generated from a specific vibration source are equally spaced, unwanted peaks can be removed. Depending on the characteristics of the vibrating material, the peak spacing may gradually increase as the frequency increases, and it is necessary to address these situations. In addition, the acceleration sensor 203 has an effective frequency range. For frequencies outside this specific range, the reliability of the obtained data may not be ensured.

[0064] The following explanation will be given with reference to the diagrams.

[0065] In the monitoring device 30, the peak data analysis unit 303 derives the peak height using the vibration waveform values ​​measured by the acceleration sensor 203, which are stored in the data storage unit 306 (S101).

[0066] The frequency spectrum contains numerous peaks. Each peak has a height. The taller peaks are mainly those obtained from the vibration source and are considered to be of high value.

[0067] Here, we will use the concept of "prominence" as an example of deriving peak height. Figures 5 and 6 will be used to explain how to derive prominence.

[0068] Figure 5(A) shows an example of a frequency spectrum. For ease of explanation, the leftmost point of the frequency spectrum is denoted as "a" and its rightmost point as "g". The five maximum values ​​from "a" to "g" are designated as peak numbers "1" to "5", and the five minimum values ​​are designated as "b" to "f".

[0069] The peak data analysis unit 303 uses prominences to indicate the height of each peak, according to the flowchart in Figure 6.

[0070] First, the peak data analysis unit 303 selects a peak (S11), and then the peak data analysis unit 303 moves to the left and right of the peak until the traced value reaches the next state, thereby deriving the endpoint of the horizontal line (S12). • Higher peaks and crossover values • The value reaches the leftmost or rightmost limit.

[0071] Next, the peak data analysis unit 303 finds the minimum point (minimum value) in the left and right intervals of the peak derived in S12 (S13), and sets the higher of the two minimum values ​​derived in S13 as the reference level (highest local minimum) (S14). The peak data analysis unit 303 sets the difference between the reference level and the peak as the peak height (prominence) (S15).

[0072] The peak data analysis unit 303 determines whether or not all peaks have been selected (S16). If not all peaks have been selected (S16 / No), the peak data analysis unit 303 proceeds to S11 and repeats the process. On the other hand, if all peaks have been selected (S16 / Yes), the peak data analysis unit 303 terminates the process.

[0073] For example, we will specifically explain an example of how to derive the prominence of the peak labeled "Peak Number: 1" in Figure 5(A).

[0074] First, select the peak with "Peak Number: 1" (S11). Moving the value to the left from this peak, the value reaches "Left End a," so "Left End a" becomes the endpoint of the horizontal line for the left section of the peak. On the other hand, moving the value to the right from the peak, the value intersects with the peak with "Peak Number: 2," which is higher than the peak with "Peak Number: 1," so the point where the value intersects with Peak 2 becomes the endpoint of the horizontal line for the right section of the peak (S12).

[0075] Next, since the minimum point in the left section of the peak is "leftmost point a" and the minimum point in the right section of the peak is "b" (S13), the higher of "leftmost point a" and "b" is taken as the reference level (S14). Then, the height "P1" from the reference level "b" to "peak number: 1" is taken as the height (prominence) of the peak "peak number: 1" (S15).

[0076] The peak data analysis unit 303 performs the above-described processing on all peaks, and the results are shown in Figure 5(B).

[0077] Next, the peak data analysis unit 303 compares the height of each peak in the frequency spectrum with the heights of the peaks appearing before and after each peak, and, taking into account the difference in frequency, deletes the smaller peak if the ratio is below a certain level (S102).

[0078] For example, taking Figure 5(A) as an example, the height of the peak with "Peak Number: 3" is smaller than the height of the adjacent peak with "Peak Number: 4," so the peak with "Peak Number: 3" is to be deleted.

[0079] More specifically, the peak data analysis unit 303 calculates the ratio of the peak height P3 of "Peak Number: 3" to the peak height P4 of "Peak Number: 4". If this ratio is below a threshold, it determines that the peak of "Peak Number: 3" has a small height and is to be deleted.

[0080] Alternatively, the peak data analysis unit 303 may take the difference between the peak height P3 of "peak number: 3" and the peak height P4 of "peak number: 4," and if this difference is less than or equal to a threshold, it may determine that the peak of "peak number: 3" has a small height. The threshold used when determining which peaks to adopt based on peak height may be changed as appropriate. Furthermore, as another method, the peak data analysis unit 303 may sort the peak data in order of peak height and adopt only the data with a fixed ratio from the top as peaks.

[0081] Figure 7 shows an example of a frequency spectrum. The horizontal axis represents frequency (Hz), and the vertical axis represents the amplitude spectrum (gal × s). For example, high peaks appear around 8 Hz and 11 Hz. In contrast, peaks also appear around 7 Hz and 10 Hz, but the height of the peak around 7 Hz is lower than that of the peak around 8 Hz, and similarly, the height of the peak around 10 Hz is lower than that of the peak around 11 Hz. Therefore, the peaks around 7 Hz and 11 Hz are not adopted as peaks in this embodiment. In such cases, the peak data analysis unit 303 deletes the peaks around 7 Hz and 11 Hz.

[0082] Let's return to Figure 4 for explanation. The natural frequency group derivation unit 304 retrieves the peak data stored in the data storage unit 306 (see Figure 8(A)) (S103), and rearranges the peak data in descending order of peak height, as shown in Figure 8(B) (S104).

[0083] The natural frequency group derivation unit 304 extracts peak data above a certain height, as shown in Figure 8(C), and rearranges them in order of frequency, as shown in Figure 8(D) (S105).

[0084] For example, in S105, peaks with high heights are extracted. Various methods can be applied to this. For example, all peak data could be used to extract those with relatively high peak heights based on variance, standard deviation, etc. Alternatively, a threshold for peak height could be set in advance, and peak data with a height above that threshold could be extracted. In the example in Figure 8(C), peaks with relatively high heights were extracted. In Figure 8(C), the peak data selected is limited to those with high peak heights up to "Frequency;Height = 13.67;17.66".

[0085] Next, the natural frequency group derivation unit 304 uses the peak data sorted in frequency order in S105 to calculate the difference in frequencies between two peaks (S106).

[0086] Figure 9 shows the difference between the frequencies of two peaks (difference values). Here, when calculating the difference between the frequencies of peaks X positions apart, it is denoted as "difference X," and the corresponding difference values ​​of the peak frequencies are shown in those columns.

[0087] For example, when calculating the difference in frequency between adjacent peaks, X=1, so it is written as "Difference 1". More specifically, when calculating the difference in frequency between adjacent peaks such as "3.77" and "7.61", X=1 is set, and "Difference 1" is the difference value "3.84". Also, for example, when calculating the difference in frequency between two peaks that are far apart, such as "3.71" and "9.56", X=2 is set, and "Difference 2" is the difference value "5.79".

[0088] The natural frequency group derivation unit 304 sorts the differences in the frequencies of the two peaks in ascending order. Then, the natural frequency group derivation unit 304 takes the difference between the two consecutive values ​​of the "frequency difference" sorted in ascending order (S107).

[0089] Here, the value of the "frequency difference" is also called the "first difference value." The difference between the two values ​​immediately preceding and succeeding the "frequency difference" is denoted as the "difference of the difference." Furthermore, the value of the "difference of the difference" is also called the "second difference value." The "difference" value and the "difference of the difference" value are treated as a pair of data.

[0090] The natural frequency group derivation unit 304 sorts the "difference of differences" values ​​of a pair of data in ascending order and assigns a marker to the range in which the "difference of differences" values ​​are large (S108).

[0091] Here, we use markers, but you may also add information that distinguishes it from other "differences of differences" values, such as flags. A value to which such distinguishable information as a marker has been added is called a "delimiter."

[0092] For example, the "Difference" column in Figure 10(A) represents the difference between the frequencies of the two peaks obtained in S106, and the "Difference of Difference" column represents the difference between two values ​​that are adjacent in the "Difference" column. The natural frequency group derivation unit 304 sorts the "Difference of Difference" in ascending order to obtain the table in Figure 10(B). The natural frequency group derivation unit 304 then extracts the ranges where the values ​​of other "Differences of Difference" are clearly large and marks these "Differences of Difference" values. Various methods can be applied to determine whether the values ​​of other "Differences of Difference" are large or not. For example, a method can be used that uses all the data to determine whether the values ​​of other "Differences of Difference" are large or not using variance or standard deviation.

[0093] Next, the natural frequency group derivation unit 304 sorts the data in ascending order by the "difference" value of the frequencies. Then, the natural frequency group derivation unit 304 groups the data according to the delimiters assigned in S108 (S109).

[0094] Then, the natural frequency group derivation unit 304 calculates the average value of the frequency "difference" for each group, and calculates the value obtained by dividing the average value of the frequency "difference" by the group number for each group (S110). If the values ​​obtained by dividing the frequency "difference" for each group by the group number are approximately the same, it means that the natural frequencies have been correctly determined.

[0095] The processes in S109 and S110 will be explained using Figure 11. For example, the natural frequency group derivation unit 304 sorts the pairs of data from S108, namely the "difference" value and the "difference of the difference" value, in ascending order by the "difference" value. As shown in Figure 11(A), the natural frequency group derivation unit 304 groups the data using the "difference of the difference" value to which a marker has been added as a "delimiter" for grouping.

[0096] Here, the data up to the "separator" is treated as one group. For example, in Figure 11(A), the "difference" values ​​from "1.95" to "2.13" are designated as "Group Number: 1," and the values ​​from "3.84" to "4.20" are designated as "Group Number: 2." In this example, the natural frequency group derivation unit 304 divides the data of the "difference" values ​​and the "difference of differences" values ​​into eight groups, from "Group Number: 1" to "Group Number: 8."

[0097] Furthermore, as shown in Figure 11(B), the natural frequency group derivation unit 304 calculates, for example, the average value of the "difference" for the six data belonging to "group number: 1", which is "2.030".

[0098] Furthermore, the natural frequency group derivation unit 304 divides the average value "2.030" by "1 (group number)" and obtains "2.030" as the quotient. Similarly, for the other groups, the average value of the frequency "difference" and the value obtained by dividing that average value by the group number are calculated.

[0099] In the example shown in Figure 11(B), the "average value / group number" calculated for each group is approximately "2.0," which means that in this example, the natural frequency is approximately 2 Hz.

[0100] The order derivation unit 307 divides the frequency of each peak data in Figure 8(A) by the natural frequency. In this example, the natural frequency is set to 2 Hz. If the obtained value (quotient) is close to an integer, the order derivation unit 307 determines that it is a valid peak, rounds the value, and estimates the order (S111).

[0101] Here, when the order derivation unit 307 estimates the order, if it overlaps with the order of the preceding or succeeding peaks, or if the obtained value (quotient) is far from an integer value, or if the height of the peak is small compared to the preceding or succeeding peaks, the order derivation unit 307 will refrain from estimating the order. In this example, all peaks other than the first order are detected, but it should be noted that there are many cases where the order is not determined.

[0102] A modified example of degree estimation will be explained using Figure 13.

[0103] The peak interval calculated in Figure 11(B) is the overall average value, and therefore may be inappropriate in the low-frequency and high-frequency ranges. Accordingly, the following configuration should also be considered.

[0104] The order derivation unit 307 uses the frequencies of each peak data in Figure 8(A). First, the order derivation unit 307 divides the smallest frequency (3.77 in this example) by the natural frequency (2.01 in this example). This value (1.876 in this example) is taken as the estimated order. The order derivation unit 307 rounds the estimated order value (1.876 in this example) to the nearest integer (2 in this example), and takes this integer value as the determined order.

[0105] Next, the natural frequency group derivation unit 304 divides the peak frequency by the determination order and recalculates the natural frequency. For example, the frequency "3.77" is divided by the determination order "2" and the resulting value (1.885) is taken as the estimated natural frequency. The estimated natural frequency is updated each time a new peak appears.

[0106] In Figure 13, the "Interval" column shows the frequency difference between adjacent peaks. For example, if the interval is significantly smaller than the estimated natural frequency, it is likely to be an unwanted peak. If the peak height is small, it is even more likely to be an unwanted peak. Also, if the estimated order deviates significantly from an integer, it is likely to be an unwanted peak.

[0107] If these conditions are met, the peak is not adopted as part of the natural frequency group and the system moves to the next peak. In other words, the natural frequency group derivation unit 304 does not update the estimated natural frequencies. For example, the peaks at 5.17 Hz, 6.57 Hz, and 12.85 Hz have a frequency difference of 1.4 Hz or less from the previous peak, which is significantly smaller than the estimated natural frequencies, and the peak height is also low. Therefore, these peaks are not adopted.

[0108] Subsequently, the same processing described above is performed on the next peak data to estimate its order. For example, the order derivation unit 307 divides the frequency (5.72 in this example) by the estimated natural frequency (1.885 in this example). This value (3.034 in this example) is taken as the estimated order. The estimated order value (3.034 in this example) is rounded, and the resulting integer value (3 in this example) is taken as the determination order by the order derivation unit 307. Next, the natural frequency group derivation unit 304 updates the estimated natural frequency with the value obtained by dividing the frequency "5.72" by the determination order "3" (1.907).

[0109] Figure 14 is an explanatory diagram illustrating the method for estimating the order and deriving the natural frequency group in the embodiment.

[0110] Figure 14 illustrates a case where the degree estimation fails. For explanatory purposes, Figure 14 assumes there are no unnecessary peaks. Also for explanatory purposes, the correct degree is shown in Figure 14.

[0111] The peak frequencies in Figure 14 show an example of data obtained by measuring vibration frequencies using a certain type of cable. The i-th order peak frequency f when using this type of cable is shown. i It is known that this can be calculated using equation (1).

[0112] f i 2 =A × i 2 +B×i 4 …(1) In equation (1), A is i 2 The coefficient of is i 4 This is the coefficient of the equation. For example, in equation (1), when B is 0, the peak intervals are perfectly equal, but when B > 0, the peak intervals widen as the degree i increases (as the value of degree i becomes larger). Furthermore, the tendency for the peak intervals to widen is proportional to the square of the degree. Therefore, the peak intervals expand rapidly.

[0113] Using the data exemplified in Figure 14, we apply the method for deriving the natural frequency group described above using Figures 8 to 11.

[0114] In other words, if we calculate the average value between peaks for each group and derive the estimated natural frequency, when the initial frequency is "5.92317 Hz", the estimated natural frequency (hereinafter also called the "initial estimated natural frequency") is "0.751 Hz" (see Figure 14). Note that the estimated order "7.887044" is rounded to the decision order "8", and the frequency / decision order is "0.740396". Assume that parameter values ​​were obtained similarly for other frequencies.

[0115] Frequency measurements are subtly affected by factors such as wind or vibration sources, resulting in different values ​​for each measurement. Consequently, there is fluctuation in the peak frequency, and the estimated natural frequency may become smaller. Because this does not match the physical characteristics, when updating the estimated natural frequency, a process is added that "if the 'frequency / decision order' becomes smaller than the previous estimated natural frequency, do not change it."

[0116] For example, the frequency / decision order "0.71796Hz" in the second row of Figure 14 is smaller than the estimated natural frequency "0.751Hz" in the first row, so the value of the estimated natural frequency in the second row is not updated. On the other hand, the frequency / decision order "0.771246Hz" in the third row of Figure 14 is larger than the estimated natural frequency "0.751Hz" in the second row, so the value of the estimated natural frequency in the third row is updated to "0.771246Hz".

[0117] Even after applying this processing, the determined order in this example differs from the correct value. This is because the order estimation for the first line (the beginning) frequency "5.92317Hz" failed, and a discrepancy in the order occurred during the estimation process (compared to the correct answer, the order of the third line frequency "9.25495Hz" is off by another 1).

[0118] Regarding the former, the method shown in Figures 8 to 11 calculates the average interval across the entire frequency range. However, because the peak interval changes depending on the frequency, the peak interval in the low-frequency band becomes shorter than the estimated result. Therefore, it is not possible to derive the correct answer "9" from "7.887044" obtained by dividing by the estimated natural frequency. Subsequent estimations fail for all orders because the error persists.

[0119] In equation (1), in environments where B > 0, the estimated natural frequency increases monotonically with increasing order. In real environments, fluctuations exist in the peak frequency, so the estimated natural frequency may temporarily decrease, but this is unlikely to continue.

[0120] In other words, if, after the order estimation is complete, [frequency / decision order] < [(pre-update) estimated eigenfrequency]...equation (2), then there is a high probability that the order estimation has failed.

[0121] Therefore, if equation (2) holds true, the degree derivation unit 307 changes the value of the degree and recalculates the degree estimation. The verification of the degree estimation and the recalculation of the degree estimation will be explained using Figures 4, 14, and 15.

[0122] In S111 of Figure 4, as described above, the order derivation unit 307 divides the frequency of each peak data by its natural frequency. If the obtained value (quotient) is close to an integer, it is determined to be a valid peak, and the order derivation unit 307 estimates the order by rounding the value. The order derivation unit 307 estimates the order for all frequencies.

[0123] Once the order estimation is complete, the order derivation unit 307 determines whether the condition [frequency / decision order] < [(pre-update) estimated natural frequency] in equation (2) holds true (S112).

[0124] Then, if [frequency / decision order] < [(previous) estimated natural frequency] is true (S112 / YES), the order derivation unit 307 changes the value of the order, and the natural frequency group derivation unit 304 recalculates the estimation of the estimated natural frequency (S113).

[0125] Figure 4 illustrates a case where the process moves from S113 to S111. This is an example of a process intended to recalculate the estimated order and the estimated natural frequency when equation (2) is satisfied. However, the process is not limited to moving to S111; it may also return to the steps necessary for changing the order during the derivation process of the natural frequency group. Alternatively, the process may move to the step of acquiring new measurement data.

[0126] Figure 15 is an explanatory diagram illustrating the method for estimating the order and deriving the natural frequency group in the embodiment.

[0127] Figure 15 shows the results after the order derivation unit 307 changes the order value and the natural frequency group derivation unit 304 recalculates the estimation of the natural frequency group. Here again, for ease of explanation, the correct order is shown.

[0128] The following shows an example of how the order derivation unit 307 changes the value of the order.

[0129] For example, the order derivation unit 307 increases the value of the determination order of the first data (the first row in Figures 14 and 15), which is "5.92317", by one unit, from "8" to "9". As a result, the value of the order changes, and the frequency / determination order becomes "0.65813", and the estimated natural frequency is updated to "0.65813".

[0130] For the frequencies in the second row and beyond, the method shown in Figures 8 to 11 is repeated. For example, the estimated order of the second row frequency "7.89756" becomes "12.000". Rounding this to an integer, the order derivation unit 307 estimates a determination order of "12". The value of frequency / determination order for the second row becomes "0.65813", which is the same as the estimated natural frequency value for the first row. In other words, the process of "not changing the value if 'frequency / determination order' is smaller than the previously estimated natural frequency" means that the estimated natural frequency for the second row is not updated.

[0131] For the third row, the estimated order for the frequency "9.25495" is "14.062". Rounding this to an integer value, the order derivation unit 307 estimates the determination order to be "14". The value of frequency / determination order for the third row becomes "0.661068", which is greater than the estimated natural frequency value for the second row, so the estimated natural frequency value for the third row is updated to "0.661068". The same process is performed for other frequencies to determine whether or not to update the estimated natural frequency.

[0132] In this way, the order derivation unit 307 forcibly increases the order value of the leading data by "1", further reducing the frequency / decision order value. However, the initial estimated natural frequency is obtained from the entire frequency band, and the natural frequency at low frequencies will be smaller than the initial estimated natural frequency.

[0133] Therefore, it is reasonable for the order derivation unit 307 to significantly change the value of the order and re-estimate the natural frequency. In other words, as in the example above, it is effective for the order derivation unit 307 to increase the value of the order by "1". Also, although this example illustrates the case where the order derivation unit 307 increases the value of the order, it is also effective for the order derivation unit 307 to decrease the value of the order.

[0134] Furthermore, in the example shown in Figure 10, it can be seen that after the order derivation unit 307 forcibly changes the order value of the first data, the estimated natural frequency is updated each time, and the value of the estimated natural frequency increases. From this, it can be seen that if the order is estimated correctly, the peak spacing will widen for most peak frequencies. It can also be seen that all order estimation results match the correct answer.

[0135] Furthermore, the choice of at which to change the order is important. For example, the order derivation unit 307 changes the order at the first peak frequency where equation (2) holds, or at the peak frequencies immediately before or after it (for example, the peak frequency immediately preceding the first peak frequency). This is expected to correct subsequent order estimation failures by changing the order at frequencies where order estimation is thought to have failed.

[0136] The timing for changing the estimated order may be, for example, by changing the order of all peak frequencies from the beginning of the frequency data obtained by measurement. In other words, the order of all frequency data may be changed. Alternatively, for example, the change may be made at a peak frequency a predetermined number of times before the peak frequency at which equation (2) holds.

[0137] Regarding recalculations to change the estimated order, it is not easy to determine whether or not to change the order of the leading (initial estimated natural frequency). One method is for the natural frequency group derivation unit 304 and the order derivation unit 307 to perform the order estimation process multiple times, and then select the result with the most peaks where equation (2) does not hold. In other words, the order derivation unit 307 changes the leading order using the result with the fewest failures in order estimation, thereby correcting the failures in the estimated order.

[0138] Furthermore, by changing the order, it is necessary to stop the search when the estimated natural frequency deviates significantly from the initial estimated natural frequency. In other words, while there is a characteristic that the peak spacing widens as the frequency increases, it is necessary to stop the search for the order when this widening of the peak spacing exceeds a threshold. For example, one can set a threshold that indicates the frequency at which the natural frequency deviates from the initial estimated natural frequency, and stop the search for the order when the estimated natural frequency deviates beyond that threshold.

[0139] As described above, by performing steps S112 and S113 in Figure 4, the failure of order estimation related to the estimated natural frequency can be corrected by taking into account the characteristic that the peak spacing widens as the frequency increases. As a result, the natural frequency can be estimated with high accuracy, and furthermore, minute changes in physical values ​​such as tension can be captured. In other words, minute changes in physical values ​​(e.g., tension) for structural integrity and anomaly detection can be captured, and subtle anomalies in the structure can be detected.

[0140] After processing S112 and S113 in Figure 4, the order estimation process is repeated one or more times. For example, when a result is obtained in which equation (2) does not hold, or when the frequencies in which equation (2) holds are in a high-frequency range that does not contribute to the estimation of the natural frequencies, the system determines that the estimation of the estimated natural frequencies is consistent, and at that time, the natural frequency group derivation unit 304 estimates the natural frequencies.

[0141] Subsequently, the determination unit 305 uses the natural frequencies estimated by the natural frequency group derivation unit 304 to determine the tension acting on the cable.

[0142] If the tension value deviates significantly from the range of values ​​obtained in past measurements, the determination unit 305 may issue a warning. When issuing a warning, the natural frequency group derivation unit 304 may re-estimate the natural frequencies, or the determination unit 305 may re-estimate the tension value, and if the tension value still deviates, the determination unit 305 may issue a warning. When issuing a warning, the degree of abnormality of the structure, which is determined based on the tension value and threshold, may also be output (including a display concept) along with the warning.

[0143] This example illustrates a case where it is determined whether the tension acting on the cable is normal. Therefore, the determination unit 305 calculates the tension using the estimated natural frequency and determines whether the value of that tension is a normal value.

[0144] However, this is not the only way to assess structural abnormalities. Deterioration of structures can also be determined using natural frequencies, and any method of determining structural abnormalities using natural frequencies can be broadly applied. In any case, when monitoring the abnormalities or soundness of a large number of structures, it is necessary to pre-set the natural frequencies of each structure, but according to this embodiment, estimated natural frequencies can be used without pre-setting them.

[0145] (A-3) Effects of the Embodiment As described above, this embodiment is expected to provide the following effects.

[0146] According to this embodiment, the order of the natural frequency can be estimated from the vibration frequency measured by the sensor, and the natural frequency can be automatically estimated. If it is suspected that there is a failure in the estimation of the order, according to this embodiment, the order can be changed to set the correct order.

[0147] For example, compared to a method that pre-specifies the frequency range for each order, this method has the advantage of being able to estimate the natural frequencies even when sudden events such as accidents occur and the natural frequencies change significantly.

[0148] Furthermore, for example, even if the interval widens as the degree increases, the correct degree can be automatically set.

[0149] Furthermore, according to this embodiment, by automatically removing unnecessary peaks, it becomes possible to estimate natural frequencies without scrutinizing the data content. By eliminating the effort required to verify this data, it becomes possible to perform more detailed measurements at a lower cost.

[0150] (B) Other embodiments Although various modified embodiments have been mentioned in the embodiments described above, the present invention can also be applied to the following modified embodiments.

[0151] (B-1) When implementing the method for deriving natural frequencies described in the above embodiment, the following points should be noted, although they are not essential.

[0152] Since the frequency characteristics of the sensor need to be considered, it is desirable to limit the range of peak frequencies in Figure 8(A). For example, it is desirable to use peaks that appear in a specific frequency range, or to not use peaks that appear below or above a specific frequency.

[0153] Since the noise characteristics of the sensor must be taken into consideration, when selecting the peaks used in Figure 8(A), it is desirable not to use peaks with a height below a certain level.

[0154] As the frequency increases, the peak spacing widens, so it is desirable not to use differences exceeding a certain value in Figure 10(A). This is because using differences between distant peaks will prevent the natural frequency from being accurately determined.

[0155] (B-2) Accelerometers can generally acquire data from the x, y, and z axes simultaneously. The direction in which the natural vibration of the cable appears on the three axes will vary depending on the installation conditions. Also, not all peaks will appear in a single direction. For example, a peak at four times the natural frequency may appear on the X axis, and a peak at five times the natural frequency may appear on the Y axis. Therefore, it is also possible to use a configuration that integrates and utilizes data from multiple axes and sensors.

[0156] (B-3) By using the results of multiple measurements together, it is possible to estimate the natural frequency with greater accuracy and to adopt a configuration that allows for more reliable determination.

[0157] For example, peaks in frequency bands that only appear occasionally are actively removed. Also, for example, measurements where the peak height is low compared to preceding and succeeding measurements are not used.

[0158] (B-4) It is also possible to use a configuration that combines measurement results from multiple sensors. For example, measurements can be taken simultaneously with multiple sensors, and only the frequency peaks measured in common can be used. Alternatively, multiple sensors can be mounted at different locations, and only the frequency peaks measured in common can be used.

[0159] For example, suppose we measure the vibration frequency of a structure to be monitored using multiple sensors, and a failure in order estimation is confirmed based on the measurement data from one of the sensors, but no failure in order estimation is confirmed based on the measurement data from the other sensors. In that case, we may perform order estimation and natural frequency estimation using the results from the sensors other than the one whose order estimation failed. In such cases, we may decide in advance the number or proportion of sensors that will not be used.

[0160] Furthermore, it is conceivable to adopt a configuration that includes a function to calculate the degree of deviation from the conditions in the estimation of natural frequencies, the estimation of the order of each peak, and the decision of whether or not to adopt them, and to output the reliability of the estimation. [Explanation of Symbols]

[0161] 1...Monitoring system, 10...Master unit, 20...Sensor equipment, 30...Monitoring device, 201...Communication unit, 202...Control unit, 203...Accelerometer, 204...Schedule determination unit, 205...Timer unit, 206...Clock unit, 207...Data storage unit, 301...Communication unit, 302...Control unit, 303...Peak data analysis unit, 304...Natural frequency group derivation unit, 305...Determination unit, 306...Data storage unit.

Claims

1. A monitoring device that monitors an abnormality in a structure supported by a cable, based on the tension of the cable estimated from the natural frequency based on the vibration waveform of the cable, A storage means for storing vibration frequency spectrum data based on vibration waveforms measured by a detection means provided on the cable being monitored, A peak data analysis means for deriving peak data including the peak frequency and peak height values ​​for each peak appearing in each of the plurality of vibration frequency spectrum data stored in the storage means, A natural frequency group derivation means that, based on the multiple peak data for each peak derived by the peak data analysis means, rearranges the peak data in descending order of peak height and derives the peak frequencies of the peaks that are relatively high in peak height and appear at equal reproducibility as the estimated natural frequencies of the monitored target. An order derivation means that derives an integer value as the estimated order of a peak that represents an integer multiple of the natural frequency, obtained by dividing the peak frequency of the peak data by the estimated natural frequency and taking the quotient. Equipped with, The aforementioned degree derivation means, If the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency, the estimated order of the peak of the peak data is changed. A monitoring device characterized by the following features.

2. The natural frequency group derivation means is The multiple peak data are sorted in ascending order of their peak frequencies, the difference between the peak frequencies of two of the peak data is taken, and a first difference value representing the peak interval of all peaks is derived. Furthermore, the first difference values ​​between all peaks are sorted in ascending order, and the difference between two adjacent first difference values ​​is taken in order to derive a second difference value that indicates the order boundary of the natural frequency. The first difference value and the second difference value are treated as a pair of data, and all of the aforementioned pairs of data are sorted in ascending order of the second difference value to identify those with relatively large second difference values. All pairs of data are sorted in ascending order of the first difference value, and integer group numbers are sequentially assigned to all pairs of data, with the pairs having the relatively larger second difference value serving as separators. The group numbers are used as the order of the peaks, and the quotient obtained by dividing the average of the first difference values ​​of the same group number by the group number is estimated as the estimated natural frequency. The monitoring device according to feature 1.

3. The monitoring device according to claim 2, further comprising a determination unit that uses the estimated natural frequencies derived by the natural frequency group derivation means to derive the tension of the cable using an existing relational expression, and issues a warning if the tension value deviates from a predetermined range based on the tension value and a threshold value.

4. A monitoring system that monitors an abnormality in a structure supported by a cable, based on the tension of the cable estimated from the natural frequency based on the vibration waveform of the cable, The detection means provided on the cable to be monitored, A storage means for storing vibration frequency spectrum data based on the vibration waveform measured by the detection means, A peak data analysis means for deriving peak data including the peak frequency and peak height values ​​for each peak appearing in each of the plurality of vibration frequency spectrum data stored in the storage means, A natural frequency group derivation means that, based on the multiple peak data for each peak derived by the peak data analysis means, rearranges the peak data in descending order of peak height and derives the peak frequencies of the peaks that are relatively high in peak height and appear at equal reproducibility as the estimated natural frequencies of the monitored target. An order derivation means that derives an integer value as the estimated order of a peak that represents an integer multiple of the natural frequency, obtained by dividing the peak frequency of the peak data by the estimated natural frequency and taking the quotient. Equipped with, The aforementioned degree derivation means, If the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency, the estimated order of the peak of the peak data is changed. A monitoring system characterized by the following features.

5. A monitoring method in which a structure supported by a cable is the object of monitoring, and abnormalities of the object of monitoring are monitored from the tension of the cable estimated from the natural frequency based on the vibration waveform of the cable, The storage means stores vibration frequency spectrum data based on vibration waveforms measured by detection means provided on the cable being monitored. The peak data analysis means derives peak data including the peak frequency and peak height values ​​for each peak appearing in each of the plurality of vibration frequency spectrum data stored in the storage means. The natural frequency group derivation means sorts the peak data in descending order of peak height based on the multiple peak data for each peak derived by the peak data analysis means, and derives the peak frequencies of the peaks that have relatively high peak heights and appear at equal reproducibility as the estimated natural frequencies of the monitored target. The degree derivation method is, The integer value obtained by dividing the peak frequency of the aforementioned peak data by the estimated natural frequency is derived as the estimated order of the peak, which represents an integer multiple of the natural frequency. If the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency, the estimated order of the peak of the peak data is changed. A monitoring method characterized by the following features.

6. A monitoring program that monitors an abnormality in a structure supported by a cable, based on the tension of the cable estimated from the natural frequency based on the vibration waveform of the cable, A computer having storage means for storing vibration frequency spectrum data based on vibration waveforms measured by detection means provided on the cable being monitored, A peak data analysis means for deriving peak data including the peak frequency and peak height values ​​for each peak appearing in each of the plurality of vibration frequency spectrum data stored in the storage means, A natural frequency group derivation means that, based on the multiple peak data for each peak derived by the peak data analysis means, rearranges the peak data in descending order of peak height and derives the peak frequencies of the peaks that are relatively high in peak height and appear at equal reproducibility as the estimated natural frequencies of the monitored target. An order derivation means that derives an integer value as the estimated order of a peak that represents an integer multiple of the natural frequency, obtained by dividing the peak frequency of the peak data by the estimated natural frequency and taking the quotient. and make it work The aforementioned degree derivation means, If the integer value obtained by dividing the peak frequency of the peak data by the estimated order of the peak is smaller than the value of the estimated natural frequency, the estimated order of the peak of the peak data is changed. A monitoring program characterized by the following features.

Images