Information processing device, information processing program and information processing method

Abstract

To acquire more accurate live load displacement by analyzing a measurement result of a sensor attached to a structure.SOLUTION: The present invention relates to an information processing device for acquiring live load variation of a structure by analyzing a measurement result of a sensor attached to the structure. Then, the information processing device of the invention has: measurement value acquisition means for acquiring a measurement value measured by acceleration measurement means for measuring acceleration of the structure in each time sequence; noise data recording means for regarding a section in which the measurement value acquired by the measurement value acquisition means is a prescribed level or less as a noise section and recording noise data based on the measurement value of the noise section; error factor determination means for performing error factor determination processing for determining a dominant error factor in the live load variation of the structure based on the measurement value from the noise data and spectral information of the measurement value; and variation estimation means for estimating live load variation in the structure on the basis of the measurement value to acquire estimated live load variation by a method corresponding to the error factor determined by the error factor determination means.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing device, an information processing program, and an information processing method, and can be applied to, for example, live load displacement measurement of a measurement target structure by a sensor. [Background technology]

[0002] Currently, efforts are underway to install sensors on bridges to measure their response to external forces, thereby estimating bridge deterioration and fatigue and making inspection and maintenance work more efficient.

[0003] When estimating deterioration and fatigue, the live load displacement, which is the cause of fatigue, is often used as an index. A known method for measuring the live load displacement is to attach an acceleration sensor to the bridge and calculate the displacement response by integrating the acceleration response twice (see Non-Patent Documents 1 and 2).

[0004] Non-Patent Document 1 proposes a method for improving the accuracy of displacement calculation by estimating an integral constant of velocity and displacement when acceleration is measured using a MEMS (Micro Electro Mechanical Systems) acceleration sensor.

[0005] In addition, Non-Patent Document 2 discloses a method for optimizing the cutoff frequency of a low-cut filter used during integration based on the noise spectrum and the signal-to-noise ratio when integrating acceleration data measured by a digital strong-motion seismograph to obtain displacement. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Hidehiko Sekiya, Koichi Yokozeki, Kentaro Kimura, Takuhiro Konishi, and Chiju Miki, "Proposal of a method for calculating displacement response using bridge acceleration records", Journal of the Japan Society of Civil Engineers, Proceedings of the Society of Civil Engineers, A1 (Structural and Earthquake Engineering), 72(1), 61-74, 2016, Japan Society of Civil Engineers [Non-Patent Document 2] Yasuhiro Hayashi, Yutaka Katsukura, Takahide Watanabe, Shunichi Kataoka, Haruhiko Yokota, Teiji Tanaka, "Reliability of Displacement Obtained by Integrating Acceleration Records of Digital Strong Motion Seismographs", Proceedings of the Architectural Institute of Japan, Vol. 419, 1991, pp. 57-66 Summary of the Invention [Problem to be solved by the invention]

[0007] However, the technique described in Non-Patent Document 1 has problems because the dominant factor that reduces accuracy varies depending on the amplitude of the displacement response and the measurement conditions. For example, the method in Non-Patent Document 1 improves accuracy when the deviation of the velocity-displacement integral constant is the dominant error factor, but in other cases (for example, when the sensor's own noise is dominant), accuracy may not improve.

[0008] In addition, in the technology described in Non-Patent Document 2, depending on the commercially available MEMS acceleration sensor, the noise characteristics may vary from one sensor to another or may be temperature and humidity dependent, causing the noise spectrum and S / N ratio to fluctuate, making it difficult to optimize the parameters in advance.

[0009] In view of the above problems, there is a need for an information processing device, an information processing program, and an information processing method that can analyze the measurement results of sensors attached to a structure and obtain live load displacement with higher accuracy. [Means for solving the problem]

[0010] The first invention is characterized by having a measurement value acquisition means for acquiring measurement values ​​measured by an acceleration measurement means for measuring the acceleration of a structure in a time series manner, a noise data recording means for regarding a section in which the measurement value acquired by the measurement value acquisition means is below a predetermined value as a noise section and recording noise data based on the measurement value in the noise section, an error factor determination means for performing an error factor determination process for determining a dominant error factor in the live load mutation of the structure based on the measurement value from the noise data and spectral information of the measurement value, and a mutation estimation means for estimating the live load mutation in the structure based on the measurement value using a method according to the error factor determined by the error factor determination means to acquire an estimated live load mutation.

[0011] The second information processing program of the present invention is characterized in that it causes a computer to function as a measurement value acquisition means for acquiring measurement values ​​measured by an acceleration measurement means for measuring the acceleration of a structure in a time series manner, a noise data recording means for regarding a section in which the measurement value acquired by the measurement value acquisition means is below a predetermined value as a noise section and recording noise data based on the measurement values ​​in the noise section, an error factor determination means for performing an error factor determination process for determining the dominant error factor in the live load mutation of the structure based on the measurement values ​​from the noise data and spectral information of the measurement values, and a mutation estimation means for estimating the live load mutation in the structure based on the measurement values ​​using a method corresponding to the error factor determined by the error factor determination means and acquiring an estimated live load mutation.

[0012] The third invention of the present invention is an information processing method performed by an information processing device, characterized in that the information processing device has a measurement value acquisition means, a noise data recording means, an error factor determination means and a mutation estimation means, wherein the measurement value acquisition means acquires measurement values ​​measured by an acceleration measurement means that measures the acceleration of a structure for each time series, the noise data recording means regards a section in which the measurement value acquired by the measurement value acquisition means is below a predetermined value as a noise section and records noise data based on the measurement value in the noise section, the error factor determination means performs an error factor determination process to determine a dominant error factor in the live load mutation of the structure based on the measurement value from the noise data and spectral information of the measurement value, and the mutation estimation means estimates the live load mutation in the structure based on the measurement value using a method corresponding to the error factor determined by the error factor determination means to obtain an estimated live load mutation. Effect of the Invention

[0013] According to the present invention, the measurement results of sensors attached to a structure can be analyzed to obtain live load displacement with higher accuracy. [Brief description of the drawings]

[0014] [Figure 1] FIG. 2 is a block diagram showing a functional configuration of a displacement measuring device according to an embodiment. [Diagram 2] 4 is a diagram showing an example of the configuration of data recorded in a noise model recording unit according to the embodiment; FIG. [Diagram 3] 11 is a flowchart illustrating an example of an operation of an error personnel determination unit according to the embodiment. [Figure 4] FIG. 4 is an explanatory diagram showing a displacement and accuracy calculation process (first pattern process) performed by a displacement estimating unit according to the embodiment. [Diagram 5] FIG. 11 is an explanatory diagram showing a displacement and accuracy calculation process (second pattern process) performed by the displacement estimating unit according to the embodiment. [Figure 6] FIG. 11 is an explanatory diagram showing a displacement and accuracy calculation process (third pattern process) performed by the displacement estimating unit according to the embodiment. [Figure 7] FIG. 13 is an explanatory diagram showing a displacement and accuracy calculation process (fourth pattern process) performed by the displacement estimating unit according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] (A) Main embodiment Hereinafter, an embodiment of an information processing device, an information processing program, and an information processing method according to the present invention will be described in detail with reference to the drawings. In this embodiment, an example in which the information processing device, the information processing program, and the information processing method according to the present invention are applied to a displacement measuring device will be described.

[0016] (A-1) Configuration of the embodiment FIG. 1 is a block diagram showing the functional configuration of a displacement measuring device 10 according to this embodiment.

[0017] The displacement measuring device 10 is a device that analyzes measurement value data (hereinafter referred to as “measurement data”) supplied from a measuring unit 20 attached to a structure X, and performs a process of estimating (calculating) the live load displacement in the structure X (hereinafter referred to as “live load displacement estimation process”).

[0018] As the structure X, for example, various structures such as bridges and buildings (structures in which live load deviation may occur) can be applied. In this embodiment example, the following description will be given assuming that the structure X is a bridge.

[0019] First, the configuration of the measurement unit 20 will be described.

[0020] The measurement unit 20 operates at a predetermined period (hereinafter referred to as a “sampling period”) to perform measurements, and supplies the displacement measuring device 10 with measurement data based on the measurement results.

[0021] The measurement unit 20 includes an acceleration sensor 21 and a measurement environment recording unit 22 .

[0022] The acceleration sensor 21 has at least one measurement axis (acceleration measurement axis), and outputs acceleration data measured at a constant sampling period.

[0023] The measurement environment recording unit 22 is a sensor that acquires the temperature and / or humidity (hereinafter referred to as "measurement environment information") when the acceleration sensor 21 performs measurement. That is, the measurement environment recording unit 22 has a temperature sensor and / or a humidity sensor. In this embodiment, the measurement environment recording unit 22 is a sensor that can acquire both the temperature and humidity as the measurement environment information.

[0024] As described above, the measurement unit 20 supplies the displacement measuring device 10 with measurement data including acceleration data (measured acceleration values) based on the results of measurement by the acceleration sensor 21 and measurement environment information at a constant sampling period.

[0025] The measurement unit 20 assigns a time stamp based on the current time to the acquired measurement data and supplies the data to the displacement measuring device 10. The time stamp is recorded based on the time when the acceleration is measured, and may be one time stamp recorded for each sample, or may be one that records the time of the first sample and the sampling period. In other words, the time stamp may be in any format as long as it is possible to know the time when each measurement data was measured. Note that the time stamp may be recorded on the displacement measuring device 10 side (assigned to the acquired measurement data) or on the measurement unit 20 side.

[0026] The means (communication means) for transmitting measurement data from the measuring unit 20 to the displacement measuring device 10 is not limited, and various transmission methods (communication methods) can be applied. For example, the measuring unit 20 and the displacement measuring device 10 may be configured to perform wired communication (e.g., communication using various wired communication means such as a wired LAN or a serial interface) or wireless communication (e.g., wireless communication such as a wireless LAN or 920 MHz band wireless communication).

[0027] Next, an overview of the live load displacement estimation process by the displacement measuring device 10 will be described.

[0028] As described above, when estimating live load displacement from acceleration data based on the techniques described in Non-Patent Documents 1 and 2, errors may occur due to various factors (hereinafter simply referred to as "error factors"). Note that in this specification, among the various factors contained in the measured value, the main error factors that are considered to be dominant may be referred to simply as error factors.

[0029] The error factors that occur during the live load displacement estimation process can be broadly divided into the following four categories.

[0030] The first error factor (first pattern) is "sensor noise." For example, the components of noise (e.g., displacement other than sound (vibration) indicating the live load displacement (live load response) to be measured; hereinafter referred to as "sensor noise") included in the acceleration data measured by the acceleration sensor 21 change depending on the individual difference of the acceleration sensor 21 and the measurement environment (e.g., temperature, humidity, etc.) of the place where the acceleration sensor 21 is placed. In particular, when an inexpensive sensor is used as the acceleration sensor 21, there is a risk that the individual difference will become large and the sensor noise components will also change significantly. Depending on the nature of the sensor noise, an error will occur in the live load displacement obtained by the live load displacement estimation process.

[0031] The second error cause (second pattern) is "distortion due to integration." In the process of live load displacement estimation, noise components and components unnecessary for integration (e.g. gravity components, sensor noise components) are removed from the acceleration displacement using a frequency filter (e.g. low-cut filter or high-cut filter), but there is a risk that the vibration components that should be measured from the displacement of the acceleration data will also be removed by the frequency filter processing. In such a case, distortion (error in the live load displacement due to distortion) occurs during the integration process performed in the process of live load displacement estimation, and an error also occurs in the live load displacement found by the live load displacement estimation process.

[0032] The third error factor (third pattern) is "error in the integral constant." In the integration process performed in the process of live load displacement estimation, it is necessary to estimate a suitable integral constant, but if there is an error in this integral constant, naturally an error will occur in the live load displacement found by the live load displacement estimation process. For example, when the vibration component contained in the acceleration data is sufficiently larger than the noise component, it becomes difficult to find an accurate integral constant, and the error originating from the integral constant becomes prominent in the live load displacement found by the live load displacement estimation process (the error in the integral constant becomes dominant).

[0033] A fourth error cause (fourth pattern) is "abnormality in the measurement environment." For example, depending on the installation environment of the acceleration sensor 21, the acceleration sensor 21 itself may shake (e.g., due to loose installation or contact with a living thing). If there is an abnormality in the measurement environment, an error corresponding to the abnormality will occur in the live load displacement obtained by the live load displacement estimation process.

[0034] As described above, in the live load displacement estimation process performed in the displacement measuring device 10, the above four error factors are considered. Therefore, the error factor determination unit 11 performs a process (hereinafter referred to as "error factor determination process") of determining (analyzing) an error factor (for example, a dominant factor affecting the error among the above first to fourth patterns) from the acceleration data, and notifies the displacement estimation unit 13 of the result. In this embodiment, the error factor determination unit 11 outputs any one of the first pattern (sensor noise), the second pattern (distortion due to integration), the third pattern (error in the integration constant), or the fourth pattern (abnormality in the measurement environment) as a result of the error factor determination process. The output format of the error factor determination process result by the error factor determination unit 11 is not limited, and for example, it may simply output an identifier (for example, any one of values ​​1 to 4) corresponding to any one of the first to fourth patterns. Then, the displacement estimation unit 13 changes the procedure of the live load displacement estimation process to a suitable content according to the error factor determined by the error factor determination process. The details of the live load displacement estimation process performed by the displacement estimation unit 13 will be described later.

[0035] Next, the internal configuration of the displacement measuring device 10 will be described with reference to FIG.

[0036] The displacement measuring device 10 includes an error factor determining unit 11, a noise model recording unit 12, a displacement estimating unit 13, and a measurement data acquiring unit 14.

[0037] The displacement measuring device 10 may be configured entirely from hardware (e.g., a dedicated chip, etc.), or may be configured partially or entirely from software (computer program). For example, the displacement measuring device 10 may be configured by installing a program (including the information processing program according to the embodiment) in a computer having a memory, a processor, etc.

[0038] The measurement data acquisition unit 14 acquires the measurement data (acceleration data and measurement environment information) and the time stamp supplied from the measurement unit 20 , and supplies them to the displacement estimation unit 13 .

[0039] The error factor determination unit 11 performs error factor determination processing based on each data (acceleration data, timestamp, and measurement environment information) supplied from the measurement data acquisition unit 14, and supplies each data (acceleration data, timestamp, and measurement environment information) together with the result of the error factor determination processing to the displacement estimation unit 13. The error factor determination unit 11 also uses information recorded in the noise model recording unit 12 during the error factor determination processing. The error factor determination unit 11 supplies the result of the error factor determination processing (for example, dominant error factors and, if necessary, noise spectrum data) to the displacement estimation unit 13. The error factor determination unit 11 supplies data of a section of the acceleration data where noise components are dominant to the noise model recording unit 12 together with the timestamp and measurement environment information.

[0040] In the noise model recording unit 12, data of a spectrum (hereinafter also referred to as "noise spectrum") obtained by converting data of a section of acceleration data in which noise components are dominant into a frequency domain is recorded as "noise data". In the noise model recording unit 12, the noise data is recorded after being classified according to measurement environment information (combination of temperature and humidity). The noise model recording unit 12 converts the supplied acceleration data into a frequency domain to obtain it as noise data, and records the obtained noise data as data of a class according to the measurement environment information. When multiple noise data classified into the same class are registered, the noise model recording unit 12 may register only the latest noise data, or may register an average of multiple noise data (noise spectrum data).

[0041] FIG. 2 is a diagram showing an example of the configuration of data recorded in the noise model recording unit 12. As shown in FIG.

[0042] The data shown in FIG. 2 shows an example in which noise data is registered for each class determined by a combination of humidity and temperature. In the data shown in FIG. 2, humidity is divided into classes of 0-10%, 10-20%, and so on, in increments of 10% from 0% (grid width). Also, in the data shown in FIG. 2, temperature is divided into classes in increments of 10°C from 0°C (grid width). For example, in the data shown in FIG. 2, noise data D11 is registered for a class of [temperature 0-10°C, humidity 0-10°C]. Also, in the data shown in FIG. 2, noise data D12 is registered for a class of [temperature 0-10°C, humidity 10-20°C]. Note that the grid width for classifying data is not limited, and various settings can be applied.

[0043] When the error factor determination unit 11 supplies the result of the error factor determination process and each data (acceleration data, time stamp, and measurement environment information) to the displacement estimation unit 13, the displacement estimation unit 13 performs a process (live load displacement estimation process) of estimating (calculating) the live load displacement based on the acceleration data in a procedure according to the result of the error factor determination process. The displacement estimation unit 13 has an accuracy calculation unit 131, a frequency filter 132, and an integrator 133.

[0044] The accuracy calculation unit 131 acquires and outputs information (hereinafter referred to as "accuracy information") including an error value (hereinafter referred to as "error amount") based on the pattern of the error factor notified from the error factor determination unit 11 or a label indicating low reliability (hereinafter referred to as "low reliability label"). When the accuracy calculation unit 131 cannot calculate the error amount due to an abnormality in the measurement environment, it outputs a low reliability label indicating low accuracy (reliability) as the accuracy information. The displacement estimation unit 13 outputs the accuracy information (error amount or low reliability label) as additional data together with the live load displacement. The method by which the displacement estimation unit 13 outputs the processing result (live load displacement and additional data) is not limited, and various methods can be applied. For example, the displacement estimation unit 13 may simply write the processing result to a data recording medium, may transmit the processing result to another device by communication (for example, transmit by various telegrams), or may output the processing result from various output devices (for example, displays, printers, etc.).

[0045] If necessary, the displacement estimation unit 13 performs live load displacement estimation processing by using noise data (noise spectrum data) recorded in the measurement environment recording unit 22. The displacement estimation unit 13 uses a frequency filter 132 and an integrator 133 when performing live load displacement estimation processing, if necessary.

[0046] The frequency filter 132 performs filtering processing of the input signal as a high pass filter (HPF) or a low pass filter (LFP). When operating as an HPF, the frequency filter 132 outputs a signal in which frequency components of the input signal that are less than a cutoff frequency (or a preset default value) specified by the displacement estimation unit 13 are reduced. When operating as an LPF, the frequency filter 132 outputs a signal in which frequency components of the input signal that are equal to or greater than a cutoff frequency (or a preset default value) specified by the displacement estimation unit 13 are reduced.

[0047] The integrator 133 performs integration in the time direction by referring to the acceleration data and the time stamp. The integrator 133 may calculate the integral value by piecewise quadrature by obtaining the time width between samples of each acceleration data from the time stamp, or may perform another calculation that achieves a similar result.

[0048] (A-2) Operation of the embodiment Next, the operation of the displacement measuring device 10 according to this embodiment will be described.

[0049] When the measurement data acquisition unit 14 receives the measurement data (including acceleration data and measurement environment information) together with the time stamp from the measurement unit 20, the measurement data acquisition unit 14 supplies the data to the displacement estimation unit 13.

[0050] When the acceleration data for a predetermined period (for example, one frame period when converted to the frequency domain) is accumulated, the error factor determination unit 11 performs an error factor determination process on the acceleration data for the predetermined period (hereinafter, the acceleration data for the predetermined period is referred to as "processing target data"). The displacement estimation unit 13 supplies the processing target data (including corresponding measurement environment information and time stamp information) to the displacement estimation unit 13 together with the result of the error factor determination process. At this time, if the processing target data (acceleration data) does not include a live load response, the error factor determination unit 11 updates the value of the measurement environment information in the noise model recording unit 12, and if the processing target data (acceleration data) includes a live load response, classifies the dominant error factor in displacement estimation into one of the first to fourth patterns, and obtains the classification result as the result of the error factor determination process.

[0051] When the displacement estimation unit 13 is supplied with the data to be processed together with the result of the error factor determination process, it performs a process of estimating (calculating) the live load displacement based on the data to be processed (acceleration data) in a procedure according to the result of the error factor determination process (live load displacement estimation process) and a process of acquiring accuracy information (error amount or low reliability label).The displacement estimation unit 13 then outputs the accuracy information (error amount or low reliability label) as additional data together with the estimated live load displacement.

[0052] Next, the error factor determination process performed by the error factor determination unit 11 will be described in detail.

[0053] FIG. 3 is a flowchart showing details of the error factor determination process performed by the error factor determination unit 11.

[0054] First, the error factor determination unit 11 determines whether the data to be processed (acceleration data) includes a value exceeding the threshold A (S101), and if the data includes a value exceeding the threshold A, the process proceeds to step S102, which will be described later, and if not, the process proceeds to step S110, which will be described later. Any value can be applied to the threshold A. For example, the threshold A may be a value that is set in advance based on noise characteristics on the specification (spec sheet) of the acceleration sensor 21, or may be a suitable value obtained by experiment or the like. If the data to be processed (acceleration data) includes a value exceeding the threshold A, it can be determined that the data to be processed (acceleration data) includes a response to a live load, and if not, it can be determined that the data to be processed (acceleration data) does not include a response to a live load (i.e., noise components are dominant).

[0055] If the data to be processed (acceleration data) does not include a value exceeding the threshold A, the data to be processed (acceleration data) is converted to the frequency domain to obtain noise data (noise spectrum data) (S110), and the data is added and registered in the noise model recording unit 12 (S111), and processing of the data to be processed is terminated. At this time, the noise model recording unit 12 registers the noise data by linking it to a class according to the value of the measurement environment information corresponding to the data to be processed. At this time, the class to be linked to the newly registered noise data by the noise model recording unit 12 may be, for example, a class according to the measurement environment information of an arbitrary timestamp (for example, the first or last timestamp) of the data to be processed, or may be a class according to the average value of the measurement environment information corresponding to each timestamp (time series) of the data to be processed.

[0056] If it is determined in step S101 described above that the data to be processed (acceleration data) contains a value that exceeds threshold A, the error factor determination unit 11 determines whether or not the data to be processed (acceleration data) corresponds to an abnormality in the measurement environment (fourth pattern) based on whether the data to be processed (acceleration data) contains a value that exceeds threshold B (wherein the relationship of threshold B > threshold A is satisfied) (S102), and if it is determined that the data to be processed (acceleration data) corresponds to an abnormality in the measurement environment (fourth pattern), the process proceeds to step S106 described later, and if not, the process proceeds to step S104 described later.

[0057] If it is determined in step S102 described above that the data to be processed (acceleration data) corresponds to an abnormality in the measurement environment (fourth pattern), the error factor determination unit 11 acquires the fourth pattern (abnormality in the measurement environment) as a result of the error factor determination process (S106), and terminates the process.

[0058] In addition, the error factor determination unit 11 may determine whether the data to be processed (acceleration data) corresponds to an abnormality in the measurement environment (fourth pattern) by determining whether the rate of change of acceleration (a value obtained by first-order differentiation of the acceleration data output from the measurement unit) includes a value exceeding threshold value B'. Any value may be applied as threshold value B and threshold value B'. For example, threshold value B and threshold value B' may be values ​​that are set in advance based on the noise characteristics and measurement range value of acceleration sensor 21, or may be suitable values ​​obtained by experiment or the like.

[0059] If it is determined in step S102 that the data to be processed (acceleration data) does not correspond to the measurement environment abnormality (fourth pattern), the error factor determination unit 11 converts the data to be processed (acceleration data) from the time domain to the frequency domain to obtain an "acceleration spectrum" (S103). When converting the acceleration data to the frequency domain in the displacement measuring device 10, various conversion means such as FFT (Fast Fourier Transform) can be applied.

[0060] Next, the error factor determination unit 11 obtains the difference between the value of the peak component in the obtained acceleration spectrum and the value of the noise data (noise spectrum) obtained from the noise model recording unit 12 (the value of the frequency at which the acceleration spectrum has a peak value), and checks whether the difference is equal to or greater than a certain threshold C set in advance (S104). If the difference is equal to or greater than the threshold C, the process proceeds to step S107, which will be described later, and if not, the process proceeds to step S105, which will be described later. At this time, the error factor determination unit 11 obtains noise data of a class according to the measurement environment information corresponding to the data to be processed (acceleration data). If the error factor determination unit 11 determines that the data to be processed (acceleration data) falls under the third pattern (integral constant error), the process proceeds to step S107, which will be described later, and if not, the process proceeds to step S105, which will be described later. Note that any value can be applied to the value of the threshold C. The threshold C may be, for example, a suitable value obtained by an experiment or the like.

[0061] In step S104, the class of noise data acquired by the error factor determination unit 11 may be, for example, a class according to the measurement environment information of an arbitrary timestamp (for example, the first or last timestamp) of the data to be processed, or may be a class according to the average value of the measurement environment information corresponding to each timestamp (time series) of the data to be processed. When the error factor determination unit 11 attempts to acquire noise data of a class according to the measurement environment information corresponding to the data to be processed (acceleration data), if the noise data of the class does not exist (if it is not registered), the error factor determination unit 11 may apply preset noise data (for example, that preset as default noise data), or may acquire noise data from the noise model recording unit 12 whose measurement environment information is close to the class to be acquired (for example, a class corresponding to an adjacent cell in a matrix as shown in FIG. 2). Here, the noise data (noise spectrum) acquired by the error factor determination unit 11 is referred to as "target noise data" or "target noise spectrum". Furthermore, in step S104, when the error factor determination unit 11 compares the value of the peak component in the acceleration spectrum with the value of the target noise data (target noise spectrum), it may compare data obtained by smoothing the target noise data (target noise spectrum) with a Hanning window. In this comparison, the error factor determination unit 11 may compare the values ​​of the frequencies at which the peaks occur, or may extract, for example, a spectrum that is the full width at half maximum of the peak of the acceleration data, and compare the area with that of the target noise data (target noise spectrum) in the same frequency band.

[0062] In the above-mentioned step S104, if the difference is equal to or greater than the threshold C, the error factor determination unit 11 determines that the data to be processed (acceleration data) corresponds to the third pattern (integral constant error), obtains the third pattern (integral constant error) as the result of the error factor determination process (S107), and terminates the process.

[0063] On the other hand, in the above step S104, if the difference is less than the threshold C, the error factor determination unit 11 judges that the data to be processed (acceleration data) corresponds to the second pattern (distortion due to integration) if the frequency (hereinafter referred to as "peak frequency") corresponding to the peak component in the acceleration spectrum is equal to or less than the threshold D [Hz], and proceeds to step S108 described later, otherwise proceeds to step S109 described later. The value of the threshold D is not limited to the above example, and any value can be applied. The threshold D may be, for example, 1 Hz, or any other suitable value obtained by experiment or the like. The value of the threshold D is desirably set to the cutoff frequency of the HPF used in the displacement estimation unit 13 described later.

[0064] In the above-mentioned step S105, if the peak frequency is equal to or lower than the threshold value D, the second pattern (distortion due to integration) is acquired as a result of the error factor determination process (S108), and the process ends.

[0065] Furthermore, in the above-mentioned step S105, if the peak frequency is higher than the threshold D, the first pattern (sensor noise) is acquired as a result of the error cause determination process (S109), and the process ends.

[0066] As described above, the error factor determination unit 11 performs the error factor determination process. Then, the displacement estimation unit 13 performs a process of acquiring live load displacement estimation and accuracy information based on the error factor notified from the error factor determination unit 11. Specifically, the error factor determination unit 11 classifies the pattern of the error factor considering whether or not a value exceeding the threshold B is included in the data to be processed. Specifically, the error factor determination unit 11 classifies the pattern of the error factor considering whether or not a value exceeding the threshold B (or a component exceeding the threshold B') is included in the data to be processed. In addition, the error factor determination unit 11 classifies the pattern of the error factor considering the comparison result (whether or not the difference exceeds the threshold C) between the value of the peak component in the acquired acceleration spectrum and the value of the noise data (noise spectrum) acquired from the noise model recording unit 12. Furthermore, the error factor determination unit 11 classifies the pattern of the error factor considering whether or not the peak frequency is equal to or lower than the threshold D [Hz].

[0067] Next, a specific example of the process of the displacement estimation unit 13 according to the result of the error factor determination process will be described.

[0068] Here, as an example, we will explain the case where the displacement estimation unit 13 takes an output format such as "maximum displacement XX mm + error amount YY %" based on the ratio of the maximum value among the calculated live load displacement values ​​and the calculated error amount.

[0069] [First pattern (sensor noise)] First, the process of the displacement estimation unit 13 when the result of the error factor determination process is the first pattern (sensor noise) will be described with reference to FIG.

[0070] [Live Load Displacement] As shown in FIG. 4(a), when the error factor is the first pattern (sensor noise), the displacement estimation unit 13 applies an HPF with a predetermined cutoff frequency (the frequency of the above threshold D) to the data to be processed (acceleration data), and then obtains the live load displacement by performing second-order integration using the integrator 133.

[0071] [Accuracy Information] In addition, as shown in Fig. 4(b), in the case of the first pattern (sensor noise), the accuracy calculation unit 131 acquires target noise data (target noise spectrum) from the noise model recording unit 12, applies an HPF with a predetermined cutoff frequency (the frequency of the above threshold D, as in the case of calculating the live load displacement) to the acquired target noise data (target noise spectrum), and then performs a second-order integration in the frequency domain using the integrator 133. Then, the accuracy calculation unit 131 restores the result of the second-order integration to noise data in the time domain by a discrete Fourier transform, and then calculates the standard deviation of the restored noise data. Then, the accuracy calculation unit 131 acquires the calculated standard deviation value as an error amount (accuracy information).

[0072] [Second pattern (distortion due to integration)] Next, the process of the displacement estimation unit 13 when the result of the error factor determination process is the second pattern (distortion due to integration) will be described with reference to FIG.

[0073] [Live Load Displacement] As shown in Fig. 5(a), in the case of the second pattern (distortion by integration), the displacement estimation unit 13 obtains the optimal cutoff frequency f [Hz] of the HPF to be applied to the data to be processed (acceleration data). In this case, the displacement estimation unit 13 may optimize f [Hz] by, for example, integrating the components of the acceleration spectrum below f Hz and the components of the target noise data (target noise spectrum) obtained from the noise model recording unit 12 above f Hz in the frequency domain, and obtaining f [Hz] when the ratio of the respective integral values ​​is maximized. Then, the displacement estimation unit 13 applies the HPF with the cutoff frequency f [Hz] to the data to be processed (acceleration data), and then performs second-order integration by the integrator 133 to obtain the live load displacement.

[0074] [Accuracy Information] In the case of the second pattern (distortion due to integration), by applying an HPF with a cutoff frequency f [Hz] to the data to be processed (acceleration data), the removed vibration component (amount of vibration component to be measured) becomes a cause of error (distortion). Therefore, here, the accuracy calculation unit 131 calculates the amount of error based on the amount of the removed vibration component. Specifically, as shown in FIG. 5(b), the accuracy calculation unit 131 applies an LPF with a cutoff frequency f [Hz] to the data to be processed (acceleration data), and then obtains the ratio of the displacement obtained by second-order integration by the integrator 133 to the area of ​​the live load displacement obtained in FIG. 5(a) as the amount of error.

[0075] [Third pattern (error in integral constant)] Next, the process of the displacement estimating unit 13 when the result of the error cause determination process is the third pattern (error in the integral constant) will be described with reference to FIG.

[0076] [Live Load Displacement] In the case of the third pattern (error of integral constant), the displacement estimation unit 13 extracts acceleration data (acceleration data from the start time to the end time) between the time when the vibration due to the live load response starts (hereinafter referred to as "start point") and the time when the vibration due to the live load response ends (hereinafter simply referred to as "end point") from the data to be processed (acceleration data) as shown in FIG. 6(a) (hereinafter, this extracted data will be referred to as "extracted section data"). The start point and end point times may be determined assuming the time when the amplitude of the acceleration is maximum and the state of excitation of a representative vibration source. For example, when the structure X to be measured is a bridge, the data before and after the time when the amplitude of the acceleration is maximum may be extracted in consideration of the passing speed of the vehicle causing the excitation and the length of the bridge.

[0077] Then, the displacement estimation unit 13 performs a discrete Fourier transform on the extracted section data, removes the DC component, and performs second-order integration to obtain the live load displacement value (integration may be performed in the time domain, but the integral constant is determined so that the start and end values ​​of the displacement value are the same).The displacement estimation unit 13 then adds an offset so that the start value of the live load displacement value becomes 0.

[0078] [Accuracy Information] When the live load displacement is calculated as described above, the integral constant is set so that the initial values ​​of the velocity and displacement are 0. However, the measurement object is vibrating and the initial values ​​of the velocity and displacement are not 0, so the difference between them is an error. Therefore, in the case of the third pattern (error in the integral constant), the accuracy calculation unit 131 integrates the sections before and after the section cut out in the estimation process of the live load displacement described above ("before the start point" and "after the end point") as shown in FIG. 6(a) in order to estimate the distribution of the initial value. At this time, the accuracy calculation unit 131 performs a second-order integration after removing low-frequency components using an HPF with a cutoff frequency (for example, the frequency corresponding to the above threshold D: 1 Hz) (because the amplitude is small in the non-vibrating section and noise cannot be ignored). Then, the accuracy calculation unit 131 calculates the standard deviation of the series obtained by the second-order integration and obtains the value as the amount of error.

[0079] [Fourth pattern (abnormality in the measurement environment)] [Live Load Displacement] As shown in Figure 7(a), in the case of the fourth pattern (abnormality in the measurement environment), the displacement estimation unit 13 acquires the live load displacement by processing similar to that for the first pattern (sensor noise) (the processing in Figure 4(a) described above).

[0080] [Accuracy Information] As shown in Fig. 7(b), in the case of the fourth pattern (abnormality of the measurement environment), the accuracy calculation unit 131 outputs a label (low reliability label) indicating that the reliability of the data is low, without quantifying the amount of error. In the case of the fourth pattern (abnormality of the measurement environment), the accuracy calculation unit 131 may obtain the amount of error by any of the processes of the first to third patterns, but since the reliability of the amount of error itself is also low, it is preferable to add a low reliability label without obtaining the amount of error.

[0081] In addition, in Figures 4 to 7, an example of a procedure for live load displacement estimation processing suitable for the first to fourth patterns is shown, but the procedure shown in Figures 4 to 7 is just an example, and the procedure is not limited to the above example as long as it improves accuracy and reliability against error factors of the first to fourth patterns.

[0082] (A-3) Effects of the embodiment According to this embodiment, the following effects can be obtained.

[0083] In the displacement measuring device 10 of this embodiment, the optimal live load displacement estimation method is applied depending on the error factor (dominant noise factor) determined by the error factor determination unit 11, thereby improving the accuracy in estimating the live load displacement.

[0084] Furthermore, in the displacement measuring device 10 of this embodiment, the accuracy calculation unit 131 outputs the amount of error (estimated amount of error), making it easier to quantitatively grasp the reliability of the estimated live load deviation.

[0085] (B) Other embodiments The present invention is not limited to the above-described embodiments, and may include modified embodiments such as those exemplified below.

[0086] (B-1) In the displacement measuring device 10 of the above embodiment, the accuracy calculation unit 131 is provided, but the accuracy calculation unit 131 may be omitted and the accuracy information may not be output.

[0087] (B-2) In step S104 (judgment based on the amplitude of the vibration component) shown in FIG. 3, the error factor judgment unit 11 has described a method of making a judgment by comparing the peak component of the acceleration spectrum with the noise spectrum. However, when the vibration frequency is over a wide band, the value of the live load displacement may be large enough to ignore the noise even though the peak component is not prominent. In order to make a correct judgment even in such a case, in step S104, the error factor judgment unit 11 may calculate a displacement estimation value in which the influence of low-frequency noise is reduced using the displacement estimation method in step S101 (when the sensor noise characteristic is dominant), and if the value is equal to or greater than a preset threshold, the process may proceed to step S107 to judge that the error factor is pattern 3 (integral constant error), and if not, the process may proceed to step S105.

[0088] (B-3) In the above embodiment, the displacement measuring device 10 and the measuring unit 20 are described as separate devices (hardware). However, the displacement measuring device 10 itself may be configured to include the measuring unit 20.

[0089] (B-4) In the displacement measuring device 10 of the above embodiment, the data to be processed is classified into one of four error factors (first to fourth pattern error factors) and a live load displacement estimation process suitable for the classified error factor is performed, but the combination and number of corresponding error factors are not limited to the above four. For example, the displacement measuring device 10 may process the data by excluding any one or two error factors from the first to fourth pattern error factors, or may process the data by adding other error factors to the first to fourth pattern error factors. [Explanation of symbols]

[0090] 10... Displacement measuring device, 11... Error factor determination unit, 12... Noise model recording unit, 13... Displacement estimation unit, 14... Measurement data acquisition unit, 20... Measurement unit, 21... Acceleration sensor, 22... Measurement environment recording unit, 131... Accuracy calculation unit, 132... Frequency filter, 133... Integrator, X... Structure

Claims

1. a measurement value acquiring means for acquiring measurement values ​​measured by an acceleration measuring means for measuring the acceleration of the structure in time series; a noise data recording means for recording noise data based on the measurement value of the noise section, the noise section being determined as a section in which the measurement value acquired by the measurement value acquiring means is equal to or smaller than a predetermined value; an error factor determination means for performing an error factor determination process for determining a dominant error factor in the live load displacement of the structure based on the measurement value from the noise data and spectrum information of the measurement value; a mutation estimation means for estimating a live load mutation in the structure based on the measurement value by a method according to the error factor determined by the error factor determination means, and acquiring an estimated live load mutation; 13. An information processing device comprising:

2. a measurement environment information acquisition unit that acquires the measurement environment information recorded in a measurement environment information recorder that records measurement environment information at the time of measurement by the acceleration measurement unit, the noise data recording means classifies and records the noise data for each class according to the measurement environment information; The error factor determination means acquires the noise data of the class corresponding to the measurement value and uses the noise data in the error factor determination process.

2. The information processing apparatus according to claim 1,

3. 3. The information processing apparatus according to claim 2, wherein the measurement environment information includes temperature and / or humidity at the time of measurement by the acceleration measuring means.

4. The information processing device according to claim 1 , further comprising an accuracy information acquisition means for performing an accuracy information acquisition process for acquiring error information related to an error of the estimated live load deviation based on the error factor determined by the error factor determination means.

5. 2 . The information processing apparatus according to claim 1 , wherein the error factor determination means classifies the pattern of the error factor by taking into consideration whether the measured value includes a value equal to or greater than a first threshold value.

6. 6. The information processing apparatus according to claim 5, wherein the error factor determination means classifies the pattern of the error factor in consideration of a comparison result between the spectrum of the measurement value and the spectrum of the noise section.

7. 7. The information processing apparatus according to claim 6, wherein the error factor determination means classifies the pattern of the error factor taking into consideration whether or not a peak frequency of the spectrum of the measured value is lower than a second threshold value.

8. Computer, a measurement value acquiring means for acquiring measurement values ​​measured by an acceleration measuring means for measuring the acceleration of the structure in time series; a noise data recording means for recording noise data based on the measurement value of the noise section, the noise section being determined as a section in which the measurement value acquired by the measurement value acquiring means is equal to or smaller than a predetermined value; an error factor determination means for performing an error factor determination process for determining a dominant error factor in the live load displacement of the structure based on the measurement value from the noise data and spectrum information of the measurement value; a mutation estimation means for estimating a live load mutation in the structure based on the measurement value by a method according to the error factor determined by the error factor determination means, and acquiring an estimated live load mutation; and causing the information processing program to function as described above.

9. An information processing method performed by an information processing device, The information processing device includes a measurement value acquiring means, a noise data recording means, an error factor determining means, and a variation estimating means, The measurement value acquisition means acquires measurement values ​​measured by an acceleration measurement means that measures the acceleration of the structure in time series, the noise data recording means regards a section in which the measurement value acquired by the measurement value acquiring means is equal to or less than a predetermined value as a noise section, and records noise data based on the measurement value of the noise section; The error factor determination means performs an error factor determination process to determine a dominant error factor in the live load displacement of the structure based on the measurement value from the noise data and spectrum information of the measurement value, The variation estimation means estimates a live load variation in the structure based on the measurement value by a method according to the error factor determined by the error factor determination means, and obtains an estimated live load variation.

23. An information processing method comprising:

Images