JP6501673B2 - Method of judging deterioration of structure - Google Patents

Description

  The present invention relates to, for example, a method of determining whether a structure such as an inner wall of a bridge or a tunnel is deteriorated.

Conventionally, the sound generated when the surface of a structure is struck with a hammer or the like is used as a method of inspecting whether a concrete structure such as an inner wall of a bridge or tunnel has a non-sounding portion such as a crack or a void. The workers heard that it was judged whether there was a non-health department. However, such a method not only has low working efficiency, but also has the problem that the judgment may differ depending on the worker.
Therefore, a hammer and a microphone are integrally formed, and the sound generated when the vibration is generated is collected by the microphone, and the frequency spectrum of the collected sound and the reference frequency spectrum which is the frequency spectrum of the sound part stored in advance. The method of determining the presence or absence of peeling of a concrete structure is proposed by comparing (refer patent document 1).

JP 2001-249117 A

  However, in the above-described conventional method, since the deterioration state of the structure is determined from the sound pressure level in the specific frequency band, the vibration method, the position of the microphone, and the surrounding conditions are large. It can not be said that it is sufficient as a deterioration determination method.

  The present invention has been made in view of the conventional problems, and an object of the present invention is to provide a method for reliably determining whether or not a structure is deteriorated.

The present invention is a method of determining whether the structure is deteriorated or not from the sound pressure signal of the sound generated by the structure when striking the surface of the structure, and the reference soundness of the structure is determined. From the step of setting the region, the step of collecting the sound pressure signal of the generated sound of the plurality of excitation points in the reference sound region, and the sound pressure signal of the generated sound, the data of the judgment item for each excitation point The step of calculating and constructing a reference data group, the step of calculating the correlation function between the data of the judgment items from the data of the judgment items, and the sound pressure of the generated sound at the measurement point which is the deterioration judgment portion of the structure The step of collecting a signal, the step of calculating measurement data which is data of the determination item from the sound pressure signal of the generated sound at the measurement point, the measurement data and the correlation function using the measurement data Said criteria Calculating the Mahalanobis distance from the data group, and performing deterioration determination of the measurement point using the calculated Mahalanobis distance, wherein the reference sound area is the structure of the structure The height of the generated sound when the surface is vibrated, or a continuous area where the timbre does not change, and the determination items include the sound pressure level of the generated sound, each level of 1/3 octave band, and time series waveform It is characterized in that it is one or more of the variation and the abundance, and the variation and the abundance of the frequency characteristic.
As described above, an area where the sound does not change when the structure to be measured is excited is regarded as a reference sound area, data of the reference sound area is used as a reference data group, and Mahalanobis between the measurement data and the reference data group It is determined whether or not the structure is deteriorated based on the distance d. Therefore, the deterioration of the structure can be determined easily and reliably.
In addition, since the reference data group is data of the structure to be judged to be deteriorated, it is possible to obtain the reference data group at the site, and to obtain the reference data group including the information on the shape and reinforcement of the site structure. Therefore, the deterioration determination can be performed accurately.

Further, in the step of performing the deterioration determination, using a threshold set based on the Mahalanobis distance calculated using data of the generated sound when a plurality of locations of the non-degraded test object are excited and determined in advance. Since the deterioration determination of the measurement point is performed, the accuracy of the deterioration determination can be further enhanced.
Further, in the step of performing the deterioration determination, a threshold set based on the distance of Mahalanobis calculated using data of sound generated when a test body having a cavity inside is excited and determined in advance and determined by cracking The determination of deterioration of the measurement point is performed using one or both of the threshold value set based on the Mahalanobis distance calculated using the data of the sound generated when the test body in which the floating exists is excited. Since this is done, it is also possible to judge the deterioration state of the measurement point.

Further, according to the present invention, in the step of collecting the sound pressure signal of the generated sound at the measurement point, the sound pressure signal of the generated sound is collected by a plurality of microphones and the sound pressure signal input to each of the microphones. Estimating the sound source direction which is the direction of the generated sound from the arrival time difference, photographing the direction of the measurement point or the estimated sound source direction by the photographing means, and determining the deterioration of the measurement point And the data of the estimated sound source direction and the image data which is the data of the photographed image are synthesized to draw a figure indicating the estimated sound source direction and the result of the deterioration determination of the measurement point And D. creating a degraded position image that is an image.
Thus, the position of the non-sounding part can be easily identified without marking the excitation point each time it is determined. In addition, by superimposing a plurality of deterioration position images, the distribution state of the non-sound portion can be easily obtained.

  The summary of the invention does not enumerate all necessary features of the present invention, and a subcombination of these feature groups can also be an invention.

It is a figure showing composition of a degradation judging system of a concrete structure concerning this embodiment. It is a flowchart which shows the creation method of the degradation determination method diagnostic image which concerns on this Embodiment. It is a figure which shows an example of the detection rod which is an excitation means. It is a figure which shows the calculation method of each level of a 1/3 octave band. It is a figure which shows a reference data group and the normalized reference data group. It is a figure which shows the example of distribution of the distance of Mahalanobis of the test object which has a space | gap part. It is a figure which shows the extraction method of the abundance and the variation in a frequency spectrum. It is a figure showing composition of a degradation judging system provided with the creation function of a degradation position picture. It is a figure which shows the structure of a sound / imaging | video extraction unit. It is a figure which shows the example of arrangement | positioning of a microphone. It is a figure which shows the production method of a degradation position distribution image. It is a figure which shows an example of a degradation position distribution image.

FIG. 1 is a view showing the configuration of a deterioration determination system for a concrete structure (hereinafter referred to as a deterioration determination system) 1 according to the present embodiment, wherein the deterioration determination system 1 comprises a hammer 2 as vibration means and sound collection A microphone 3 as a means, a deterioration determination device 4 and a display device 5 are provided.
As shown in the figure, in this example, the worker W (only the arm is described) is the surface of the wall of a building (hereinafter referred to as the wall 6) as a concrete structure which is a target of degradation diagnosis. The sound pressure signal of the sound generated by the wall portion 6 is detected by collecting with the microphone 3 a sound generated when the hammer 2 strikes and vibrates (hereinafter referred to as a generated sound).
The deterioration determination device 4 includes a sound data capture unit 41, a storage unit 42, a statistical data generation unit 43, a correlation matrix generation unit 44, and a deterioration determination unit 45.
Each unit from the storage unit 42 to the deterioration determination unit 45 of the deterioration determination device 4 includes, for example, software of a personal computer and a memory.
The sound data acquisition means 41 includes an amplifier 41a and an A / D converter 41b.
The amplifier 41 a removes high frequency noise components from the sound pressure signal measured by the microphone 3 and amplifies the sound pressure signal. The A / D converter 41 b A / D converts the amplified sound pressure signal, and sends sound pressure waveform data, which is the A / D converted sound pressure signal, to the storage unit 42.
The storage unit 42 includes a reference generated sound storage unit 42 a and a measurement generated sound storage unit 42 b.
The reference generation sound storage unit 42a stores sound pressure waveform data of an excitation point in a reference sound area to be described later, and the measurement generation sound storage unit 42b outputs sound pressure waveform data of an excitation point for determining deterioration (hereinafter referred to as measurement sound Store data).
When it is determined that the point in the reference sound area is deteriorated, the sound pressure waveform data of the excitation point in the reference sound area that has been excited again is stored in the measurement generated sound storage unit 42b as measurement sound data.

The statistical data creation unit 43 includes a sound pressure waveform extraction unit 43a, a frequency spectrum calculation unit 43b, and a 1/3 octave band value calculation unit 43c.
When generating the correlation matrix, the sound pressure waveform extraction unit 43a extracts the sound pressure waveform data for each excitation point stored in the reference generated sound storage unit 42a, and performs measurement when determining deterioration of the measurement point. The sound pressure waveform data of the excitation point stored in the generated sound storage unit 42b is extracted.
The frequency spectrum calculation unit 43 b performs FFT processing on the extracted sound pressure waveform data to calculate a frequency spectrum.
The 1/3 octave band value calculation unit 43c obtains each level of the 1/3 octave band for each excitation point P using a rectangular filter for each frequency spectrum.
Assuming that the excitation point in the reference sound area is P _i (i = 1 to n) and the center frequency of the 1/3 octave band is f _j (j = 1 to k), the excitation point P in the reference sound area is Each level of the 1/3 octave band in _i is the data (reference data y _ij ) constructing the reference data group, and each level of the 1/3 octave band in the excitation point P _r in the degradation judgment portion is the measurement data y _rj is there.
Hereinafter, the vibration point P _r in the deterioration determination portion that measurement point P _r.
The correlation matrix generation means 44 generates a correlation matrix R between each level of the 1/3 octave band using reference data y _ij which is data of a reference data group.
The deterioration determination unit 45 includes a Mahalanobis distance calculation unit 45a and a determination unit 45b.
Mahalanobis distance calculating unit 45a includes a respective level y _rj 1/3 octave band is the data of the measuring points P _r, by using the correlation matrix R, the data and the reference data group generated sound at the measurement point P _r Find the distance D ² of Mahalanobis.
Method will be described later calculates the production methods and the Mahalanobis distance D ² of the correlation matrix.
The determination unit 45 b determines whether the measurement point P _r is a healthy part by comparing the Mahalanobis distance D ² with a preset threshold.
The display device 5 displays the determination result of the deterioration determination means 45 on a display screen 5 a such as a display.

Next, the deterioration determination method will be described with reference to the flowchart of FIG.
First, a reference sound area is set (step S11).
The reference sound area refers to a continuous area in which the height (or timbre) of the generated sound does not change when striking and vibrating the surface of the wall 6 to be judged to be deteriorated. Hereinafter, tapping sound data of the excitation point in the reference sound area is referred to as sound part data.
Since the sound characteristics of the hitting sound are different between the sound part and the non sound part, it is determined whether the vibration point is the sound part by examining the sound pressure level of one or more specific frequency regions set in advance. it can.
In addition, since the height (or timbre) of the generated sound is obviously different between the hitting sound in the sound part and the hitting sound in the non sound part, the worker W determines whether the vibration point is a sound part or not. Also good.
Although a hammer 2 may be used to hit the wall 6, as shown in FIG. 3 (a), a plurality of protrusions 7q are formed on the outer periphery that collide with the surface (wall 6) of the inspection object. An impact is continuously given to the surface by rolling the rotary head 7a along the inspection surface with the detection rod 7 provided with the rotary head 7a and the head support bar 7b rotatably supporting the rotary head 7a. If sound is generated, data of the sound part can be collected efficiently.

After the setting of the reference sound area is completed, the background noise data and the sound area data are acquired (step S12), and the acquired sound area data (sound pressure waveform data of a plurality of excitation points) can be used. It is determined whether it is data (step S13).
If it is determined in step S13 that (sound pressure level of data of sound part) <(sound pressure level of dark noise data) +10 dB, a warning “S / N ratio is insufficient” is given. It emits (step S14). The background noise includes, for example, the noise of a large truck passing near the measurement point, etc. In such a case, after leaving a certain amount of time, the process returns to step S12, and the background noise data and the sound part You can recover the data.
If it is determined in step S13 that (sound pressure level of data of sound part)) (sound pressure level of dark noise data) +10 dB, the process proceeds to step S15, and reference data for calculating a correlation matrix Build a group.

As described above, the reference data group is the value y _ij (j = 1 to k) of k determination items which are variables of the sound pressure waveform data obtained for each of the excitation points P _i (i = 1 to n). It consists of
Judgment items include sound pressure level, each level of 1/3 octave band, amount of change of time axis waveform and amount of change, amount of change of frequency characteristic and amount of exist, difference between position of excitation point and direction of sound source It can be mentioned.
In this example, each level of the 1/3 octave band is used as a determination item, and a reference data group for calculating a correlation matrix is obtained by obtaining each level of the 1/3 octave band of the acquired healthy part data. To construct.
Specifically, as shown in FIG. 4A, the time-series waveform of the tapping sound in the reference sound area, which is sound area data, is cut out at 125 msec (sampling frequency 16384 Hz, data length 2048) intervals, and each time After the series waveform is Fourier-transformed to obtain a frequency spectrum as shown in FIG. 4B, each level of 1/3 octave band is determined using a rectangular filter for each frequency spectrum.
In this example, the frequency range to be targeted in Fourier transform is set to 890 Hz to 7070 Hz in consideration of the influence of the road traffic, etc. Therefore, as shown in FIG. The central frequency f _j (j = 1 to k) of the above becomes 1000 Hz, 1250 Hz, 1600 Hz, 2000 Hz, 2500 Hz, 3150 Hz, 4000 Hz, 5000 Hz, 6300 Hz. That is, the number of determination items (hereinafter referred to as the item number k), which is the number of variables constituting the reference data group, is k = 9.
Further, in this example, the number n of data in the sound part, which is the number of data for constructing the reference data group, is n = 20. In multivariate analysis, it is desirable for the number of data to be large, but since the correlation between variables is high in the sound part, the number of data may be n> k.
As described above, assuming that the data number which is the acquisition position of the tapping sound is i (i = 1 to n) and the center frequency f _j which is the determination item is j (j = 1 to k), the reference data group is n × It is constructed from k reference data y _ij .

Next, the reference data y _ij is normalized to calculate a correlation matrix (step S16).
Specifically, as shown in FIG. 5, after obtaining the average m _j and the standard deviation sigma _j for each determination item reference data y _ij, reference data y _ij by using the mean m _j and the standard deviation sigma _j Normalize Assuming that the normalized data is Y _ij , Y _ij = (y _ij −m _j ) / σ _j . The mean of the normalized data Y _ij is zero and the standard deviation is one.
Next, the correlation coefficient r _pq between Y _ip and Y _iq is calculated from the normalized data Y _{ij using} the following equation 1 to obtain the following equation 2. As shown, a k × k correlation matrix R with 1 diagonal element and r _pq elements in p rows and q columns is calculated.

ここで、ｋは項目数である。 After the calculation of the correlation matrix is completed, the generated sound at the measurement point P _r which is the excitation point at the deterioration determination location is acquired, and the background noise data at the time of measurement is acquired (step S17).
In addition, as degradation determination location, it may be except the reference | standard sound area of the wall part 6, and the part in a reference | standard sound area may be included.
Next, it is determined whether the acquired measurement data is usable data (step S18). If it is determined in step S 18 that (sound pressure level of measurement data) <(sound pressure level of background noise data) +10 dB, a warning “S / N ratio is insufficient” is issued (step S18) At the same time as step S19), after a certain amount of time has passed, the process returns to step S17 and the measurement data and the background noise data are read again.
If it is determined in step S18 that (sound pressure level of measurement data) ≧ (sound pressure level of background noise data) +10 dB, the process proceeds to step S20, and each of 1/3 octave bands of generated sound is generated. Measured data y _rj (j = 1 to k) which is a level is calculated.
Next, the calculated measurement data y _rj (j = 1 to k) is normalized, and each level Y _rj (j = 1 to k) of the 1/3 octave band which is the normalized measurement data and the step S16 Using the inverse matrix R ^{−1 of the} calculated correlation matrix R, the distance D ² of Mahalanobis between the measurement data at the measurement point P _r and the reference data group is determined (step S 21).
The Mahalanobis distance D ² can be obtained using the equation shown in the following [Equation 3].

Here, k is the number of items.

Next, deterioration determination of the measurement point P _r is performed using the Mahalanobis distance D ² obtained in step S21 and the threshold value K set in advance (step S22).
When D ² > K, it is determined that the measurement point P _r is a non-healthy part, and when D ² ≦ K, it is determined that the measurement point P _r is a health part. In this example, K = 4.
After the determination, the determination result is displayed on the display screen 5a such as a display (step S23).
In step S24, it is determined whether the measurement has ended.
If the measurement is not completed, the process returns to step S17 to measure the sound generated at the next measurement point _{Pr + 1} , and when the measurement is completed, the present process is ended.
If it is determined that the measurement point P _r is a non-healthy part, the worker W may mark the measurement point P _r with a spray or the like. As a result, not only the position of the non-healthy part but also the distribution of the non-healthy part in the wall 6 can be known.

[Example of experiment]
FIG. 6 is a view showing a distribution example of the Mahalanobis distance D ² of the generated sound generated when a rectangular solid concrete test body having a void and a float is vibrated, and the void is used as a reference data group And the data of the sound generated when a sound part which is separated from the floating part by a predetermined distance is excited are used.
In the figure, the horizontal axis is the excitation position, the vertical axis is the Mahalanobis distance D ² , and for D ² > 100, D ² = 100.
As apparent from FIG. 6, in the gap portion and the floating portion, it can be seen that the Mahalanobis distance D ² is very large, far away measurement data from the reference data set.
Thus, it was confirmed that it is possible to perform the deterioration determination of the concrete structure by using the Mahalanobis distance D ^2.

  As mentioned above, although this invention was demonstrated using embodiment and an experiment example, the technical scope of this invention is not limited to the range as described in the said embodiment. It is obvious to those skilled in the art that various changes or modifications can be added to the above embodiment. It is also apparent from the scope of the claims that the embodiments added with such alterations or improvements can be included in the technical scope of the present invention.

For example, although the deterioration state of the wall portion 6 of the building is determined in the above embodiment, the present invention is also applicable to other concrete structures such as the inner wall of a bridge or tunnel and further the frame of the building. . Further, the present invention can be applied not only to concrete structures but also to deterioration diagnosis of structures consisting of mortar, tiles and the like.
Further, in the above embodiment, the worker W strikes the surface of the wall portion 6 with the hammer 2 to vibrate, but for example, the hammer, the means for rotating the hammer, and the hammer are moved up and down. The automatic excitation apparatus is provided with means and traveling means, and the position of the excitation point at which the surface of the wall portion 6 is struck and excited can be easily recorded using this automatic excitation apparatus.

Further, in the above embodiment, although each level of 1/3 octave band is used as a determination item, sound pressure level, change amount and amount of time axis waveform, change amount and amount of frequency characteristic, etc. are determination items. It is also good.
For example, in the case where the change amount and the abundance of the frequency characteristic are used as determination items, the sound pressure signal of the sound part data is obtained by Fourier transform with 16384 Hz sampling frequency and 2048 data lengths (8 Hz frequency resolution) A frequency spectrum may be used. Also in this case, it is preferable to set the frequency range of the frequency spectrum of the hitting sound to be 890 Hz to 7070 Hz, in which the amount of change and the amount of presence are obtained in consideration of the influence of road traffic noise and the like.
The amount of change and the amount of presence are used to extract the characteristics of the frequency spectrum, and as shown in FIG. 7, a plurality of sample lines parallel to the frequency axis (horizontal axis) are set, and the frequency characteristic value is a sample The count number of the frequency larger than the value of the line is the abundance, and the count number of the frequency at which the frequency characteristic value crosses the sample line is the variation.
For example, if the number of sample lines is seven, the number of measurement items is k = 7 × 2 = 14. Therefore, the number of data in the sound part may be n> 14.
When the change amount and the abundance of the frequency characteristic are used as determination items, the change amount and the abundance of the 1/3 octave band may be used instead of the change amount and the abundance of the frequency spectrum.
Further, as the sound pressure level, a dB combined value of 1/3 octave band may be used.
In addition, when using a sound pressure level as a determination item, it is needless to say that it is necessary to use with another determination item.

In the above embodiment, the threshold value for determining deterioration is K = 4. However, the Mahalanobis distance calculated using data y _sj obtained by vibrating a plurality of locations of the non-degraded test object determined in advance. The deterioration determination of the measurement point may be performed using the threshold value K _s set based on D _s ² . The threshold K _s, for example, the average value D _sm ² of D _s ^2, or, to a maximum value D _smax ² of D _s ^2, may be such as used multiplied by the coefficient is a> 0.
Also, using a threshold value K _v set based on the distance D _v of the Mahalanobis calculated using data y _{v j} of the generated sound when a test body having a cavity inside is determined and determined in advance, measurement points are used. It may be determined whether or not the degradation state of is hollow.
In addition, using the threshold value K _f set based on the Mahalanobis distance D _f calculated using the data y _fj of the generated sound when the test body in which there is lift due to cracking is excited, the deterioration state of the measurement point is It is also possible to determine whether it is a float.

Further, in the above embodiment, the worker W marks the measurement point P _r determined to be the non-healthy part, but the sound pressure signal of the generated sound is collected by the plurality of microphones M1 to M5. Then, from the arrival time difference of the sound pressure signal input to each of the microphones M1 to M5, the direction (hereinafter referred to as the sound source direction) of the generated sound when viewed from the microphones M1 to M5 is estimated, and the estimated sound source direction is photographed Then, the degradation point determination result of the measurement point, the data of the estimated sound source direction, and the image data of the captured sound source direction are synthesized, and a figure indicating the estimated sound source direction and the degradation determination result of the measurement point is drawn If the created image (deteriorated position image) is created, the position of the non-healthy part can be easily identified. In addition, by superimposing a plurality of deterioration position images, the distribution state of the non-sound portion can be easily obtained.

FIG. 8 is a view showing the configuration of a deterioration determination system 1G having a function of creating a deterioration position image, and the deterioration determination system 1G includes a hammer 2 as an excitation means, a sound / image collecting unit 10, and a deterioration determination device 4G and the display device 5 are provided. The same reference numerals are applied to the same components as the deterioration determination system 1 of the first embodiment.
As shown in FIG. 9, the sound / image collecting unit 10 includes a sound collecting unit 11 having a plurality of microphones M1 to M5, a CCD camera (hereinafter referred to as a camera) 12 as a photographing unit, and a microphone fixing unit 13. , Camera support 14 and a support 15.
The microphones M1 to M5 respectively measure the sound pressure signal of the generated sound when striking the measurement point P _r on the surface of the wall portion 6, and the camera 12 captures an image around the measurement point P _r .
Microphones M1 to M5 are installed on the microphone fixing unit 13, the camera 12 is installed on the camera support 14, and the microphone fixing unit 13 and the camera support 14 are connected by three columns 15. That is, the sound collecting means 11 and the camera 12 as the photographing means are integrated to constitute the sound / image collecting unit 10.
In this example, the sound / image collecting unit 10 is installed such that the shooting direction of the camera 12 is the measurement point P _r on the surface of the wall 6.

  FIG. 10 is a view showing an arrangement example of the microphones M1 to M5. In this example, four microphones M1 to M4 are respectively arranged at predetermined intervals on two straight lines (here, x axis and y axis) orthogonal to each other. 4 to arrange two microphone pairs (M1, M3) and microphone pairs (M2, M4) arranged in L, and use the fifth microphone M5 as a bottom surface formed by the microphones M1 to M4. It was placed at the apex of the pyramid. Thus, four microphone pairs (M5, M1) to (M5, M4) are further configured.

As shown in FIG. 8, the degradation determination device 4G includes a sound data capture unit 41, a video data capture unit 51, a storage unit 42G, a statistical data creation unit 43, a correlation matrix generation unit 44, and a degradation determination. A means 45, a sound source direction estimation means 52, a degradation position image creation means 53, and a degradation position distribution image creation means 54 are provided. The same symbols as in the degradation determination system 1 of the first embodiment, such as the sound data capture means 41, are used.
The sound data acquisition means 41 removes high frequency noise components from the sound pressure signals measured by the respective microphones M1 to M5, amplifies these sound pressure signals, and A / D converts the amplified sound pressure signals. The obtained sound pressure waveform data is sent to the storage means 42G.
The video data capture means 51 receives a video signal of a video centered on the measurement point P _r captured by the camera 12, and measures position image data as image data obtained by A / D converting this video signal. To storage means 42G.
The shooting direction may be the sound source direction estimated by the sound source direction estimating means 52. However, there is no significant difference between the sound source direction estimated as the measurement point P _r and the case where the camera field of view is extremely narrow. Therefore, as in this example, using the sound / video sampling unit 10 It is more efficient to collect video and images simultaneously.

The storage unit 42G stores a reference generated sound storage unit 42a that stores sound portion data that is data of a reference sound area, a measurement generated sound storage unit 42b that stores measurement sound data, and an image storage unit that stores measurement position image data. And 42c.
The statistical data creation unit 43 extracts sound pressure waveform data stored in the storage unit 42G, and creates reference data y _ij and measurement data y _rj .
The correlation matrix generation unit 44 generates a correlation matrix using the reference data y _ij .
The deterioration determination means 45 determines the Mahalanobis distance D ² between the data of the generated sound at the measurement point P _r and the reference data group using the measurement data y _rj and the correlation matrix, and sets in advance the Mahalanobis distance D ² The threshold value is compared with the threshold value to determine whether the measurement point P _r is a healthy part, and the result is sent to the degradation position image creating means 53.

ここで、到達時間差Ｄ_ijは、マイクロフォンＭ_iに到達する音圧信号と、このマイクロフォンＭ_iに対して対となるマイクロフォンＭ_jに到達する音圧信号との時間差であり、この対となる２つのマイクロフォンＭ_i及びマイクロフォンＭ_jに入力される信号のクロススペクトルＰ_ij（ｆ）を求め、更に、対象とする前記周波数ｆの位相角情報Ψ（ｒａｄ）を用いて、以下の［数５］に示す式（３）により算出される。

なお、音源方向は周波数毎に計測する。 The sound source direction estimation means 52 performs frequency analysis of sound pressure waveform data (measurement data) collected by the microphones M1 to M5 stored in the storage means 42G by FFT, and the respective microphones M1 to M5 are separated for each frequency f. The phase difference is determined, and from the determined phase difference, the direction of the generated sound which is the direction of the sound source is estimated for each frequency, and the horizontal angle θ and elevation angle φ of the highest frequency of the sound pressure level are measured points P _r Are sent to the degradation position image creating means 53 as the sound source direction (θ _r , φ _r ) of the generated sound when striking the.
In this example, the horizontal angle θ and the elevation angle φ, which are the sound source direction, are determined using the arrival time difference D _ij that is a physical quantity proportional to the phase difference, instead of the phase difference.
Specifically, the sound pressure waveform data is frequency analyzed by FFT, the phase difference between the microphones M1 to M5 is determined for each frequency f, and the direction of the sound source is estimated for each frequency from the determined phase difference. Do. In the present embodiment, the horizontal angle θ and the elevation angle φ, which are sound source directions, are determined using the arrival time difference D _ij that is a physical quantity proportional to the phase difference, instead of the phase difference.
The horizontal angle θ and the elevation angle φ can be expressed by Equations (1) and (2) shown in the following [Equation 4].

Here, the arrival time difference D _ij is a time difference between the sound pressure signal arriving at the microphone M _i and the sound pressure signal arriving at the microphone M _j that is paired with the microphone M _i . The cross spectrum P _ij (f) of the signal input to one microphone M _i and the microphone M _j is determined, and further, using the phase angle information Ψ (rad) of the frequency f as an object, the following [Equation 5] It is calculated by the equation (3) shown in

The sound source direction is measured for each frequency.

In the degradation position image creation means 53, the data (θ _r , φ _r ) of the sound source direction at the measurement point P _r estimated by the sound source direction estimation means 52, the measurement position image data stored in the storage means 42G, and the deterioration determination The sound source position image G _r is generated by combining the judgment result of the means 45 and drawing a figure (in this case, indicated by a cross) indicating the sound source direction and the judgment result in the image as shown in FIG. . The horizontal axis of the sound source position image _Gr is the horizontal angle θ _r and the vertical axis is the elevation angle φ _r .
In this example, a × mark only if the measuring point P _r is a non-sound portion, when the measurement point P _r is a non-sound portion, without drawing the figure, the measurement position image sound source position image G _r And
The sound source position image _Gr is subsequently generated for each measurement point _Pr , and is sequentially sent to the deterioration position distribution image creating means 54.

Degraded positional distribution image creation means 54, a plurality of sound source position image G _r; was synthesized (r = 1 to N, N the number of measurement points), to create a degraded positional distribution image G _T.
In this example, as shown in FIG. 11, measurement points P _r are set such that the distance between adjacent measurement points P _r and P _{r + 1} falls within the image field of the camera 12. Then, from the measurement position image data stored in the storage unit 42G, two deterioration position images G _r and G _{r + 1 in} which the same part is shown are read out, and these two deterioration position images G _r and G _{r +} By superimposing ₁ , degradation position image G _r in which the degradation determination result of two measurement points P _r and P _{r + 1} is drawn and degradation position distribution image G _r in which the degradation determination result of measurement point P _r is drawn _{, r + 1} are synthesized.
Incidentally, as the image data for two of the same portion is captured, for example, a deterioration position image G _r taken at time t _r, may be selected and the image G _{r + 1} taken at time t _{r + 1.}
By synthesizing such operation was repeated N number of degraded position image G ₁ ~G _N, as shown in FIG. 12, the × mark only the unhealthy part to obtain a drawn degraded positional distribution image G _T As a result, it is possible to easily grasp the state of deterioration of the wall 6 as a whole.
The display device 5 displays the degradation position image G _r and the degradation position distribution image G _T as the determination result of the degradation determination means 45 on the display screen 5 a such as a display.

Note that the sound / video signal of the sound generated when the sound / image collecting unit 10 is mounted on a unit moving means provided with raising / lowering means and traveling means and striking the measurement point P _r on the surface of the wall 6 If the coordinates (x _c , y _c , z _c ) of the camera 12 at the time of shooting are stored as a configuration for shooting an image centered on the measurement point P _r , data of the sound source direction (θ _r , φ _r ) a camera coordinate _{(x c, y c, z} c) from a, it is possible to identify the sound source position, it is possible to obtain a more accurate deterioration position distribution image G _T.

1 Concrete Structure Deterioration Judgment System, 2 Hammers, 3 Microphones,
4 deterioration judgment device, 5 display devices, 6 walls,
41 sound data acquisition means, 42 storage means, 43 statistical data generation means,
44 correlation matrix generation means, 45 deterioration determination means.

Claims (4)

It is a method of determining whether the structure is deteriorated or not from the sound pressure signal of the sound generated by the structure when striking the surface of the structure,
Setting a reference sound area of the structure;
Collecting sound pressure signals of sounds generated at a plurality of excitation points in the reference sound area;
Calculating data of a determination item for each excitation point from the sound pressure signal of the generated sound to construct a reference data group;
Calculating a correlation function between the data of the determination items from the data of the determination items;
Collecting the sound pressure signal of the generated sound at the measurement point which is the degradation determination point of the structure;
Calculating measurement data, which is data of the determination item, from a sound pressure signal of the generated sound at the measurement point;
Calculating the Mahalanobis distance between the measurement data and the reference data group using the measurement data and the correlation function;
Performing the deterioration determination of the measurement point using the calculated Mahalanobis distance;
Equipped with
The reference sound area is the height of the generated sound when the surface of the structure is vibrated, or a continuous area in which the timbre does not change,
The judgment items include the sound pressure level of the generated sound, each level of 1/3 octave band, change amount and presence amount of time series waveform, change amount and presence amount of frequency characteristic, or a plurality thereof A method of judging deterioration of a structure characterized by In the step of performing the deterioration determination,
Using the threshold value set based on the Mahalanobis distance calculated using the data of the sound generated when a plurality of locations of the non-degraded test body determined in advance are vibrated, the degradation point of the measurement point is determined The method for determining deterioration of a structure according to claim 1, characterized in that In the step of performing the deterioration determination,
The threshold value set based on the distance of Mahalanobis calculated using the data of the sound generated when the test body with the cavity inside was determined beforehand was used to determine the threshold value and the test body in which the floating due to cracking existed The deterioration determination of the measurement point is performed using one or both of the threshold value set based on the distance of Mahalanobis calculated using the data of the generated sound at that time, or both of them. A method of judging deterioration of a structure according to item 2. In the step of collecting the sound pressure signal of the generated sound at the measurement point,
While collecting a sound pressure signal of the generated sound by a plurality of microphones,
Estimating a sound source direction which is a direction of the generated sound from a difference in arrival time of the sound pressure signal input to each of the microphones;
Shooting an image of the measurement point direction or the estimated sound source direction by a photographing unit;
The degradation determination result of the measurement point, the data of the estimated sound source direction, and the image data that is the data of the captured image are combined, and the estimated sound source direction and the degradation determination result of the measurement point The method for determining deterioration of a structure according to any one of claims 1 to 3, further comprising the step of creating a deterioration position image which is an image in which a graphic showing "is drawn."

Images