Analysis device, analysis method, and program

Abstract

This analysis apparatus comprises: a vibration image generation unit for generating, on the basis of an image obtained by imaging an inspection target, a vibration image obtained by visualizing, for each region of the inspection target on the image, a vibration level at a first natural vibration frequency at which a peak of the vibration level appears; a sound source image generation unit for generating a sound source image that is obtained by visualizing the sound pressure level for each of the regions on the image on the basis of recorded sound data of the inspection target simultaneously recorded with the image; an extraction unit for extracting a first natural vibration frequency included in a predetermined range from a second natural vibration frequency at which a peak of the sound pressure level of noise appears; and an analysis image generation unit for generating an analysis image in which the vibration image corresponding to the extracted first natural vibration frequency overlaps with the sound source image.

Description

[Technical Field]

[0001] The present disclosure relates to an analysis device, an analysis method, and a program. This application claims priority based on Japanese Patent Application No. 2022-036498, filed on March 9, 2022, the contents of which are incorporated herein by reference. [Background technology]

[0002] There are technologies being considered that can ascertain the state of an object to be inspected from images (video) of the object, such as a structure or machine. For example, Patent Document 1 describes a technology that measures the in-plane displacement of a structure, such as a bridge, from images of the structure to inspect the state of the structure. There are also technologies being considered that can measure the vibration state of the object to be inspected (the location where vibration is occurring, the magnitude of vibration, etc.) from the captured images. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2020 / 255231 Summary of the Invention [Problem to be solved by the invention]

[0004] The object of inspection may generate noise due to vibration. In order to take appropriate measures to suppress the noise, it is necessary to identify the vibration that is causing the noise.

[0005] The present disclosure has been made in consideration of such problems, and provides an analysis device, an analysis method, and a program that can visualize the correlation between vibration and noise of an object to be inspected. [Means for solving the problem]

[0006] According to one aspect of the present disclosure, an analysis device includes: a vibration image generation unit that acquires vibration information including a time series of vibration levels of the test object and a distribution of vibration levels for each frequency based on an image of the test object, and generates a vibration image that visualizes the vibration level for each region of the test object on the image at a first natural frequency at which the vibration level peak appears; a sound source image generation unit that acquires noise information including a time series of sound pressure levels of noise generated from the test object and a distribution of sound pressure levels for each frequency based on audio data of the test object recorded simultaneously with the image, and generates a sound source image that visualizes the sound pressure level for each region on the image; an extraction unit that extracts a first natural frequency that is within a predetermined range from a second natural frequency at which the sound pressure level peak of the noise appears based on the vibration information and the noise information; and an analysis image generation unit that generates an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image.

[0007] According to one aspect of the present disclosure, an analysis method includes the steps of: acquiring vibration information, based on an image of the test object, including a time series of vibration levels of the test object and a distribution of vibration levels for each frequency, and generating a vibration image that visualizes the vibration levels for each region of the test object on the image at a first natural frequency at which a peak of the vibration level appears; acquiring noise information, based on audio data of the test object recorded simultaneously with the image, including a time series of sound pressure levels of noise generated from the test object and a distribution of sound pressure levels for each frequency, and generating a sound source image that visualizes the sound pressure levels for each region on the image; extracting, based on the vibration information and the noise information, a first natural frequency that is within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears; and generating an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image.

[0008] According to one aspect of the present disclosure, the program causes an analysis device to perform the following steps: acquiring vibration information including a time series of vibration levels of the test object and a distribution of vibration levels for each vibration frequency based on an image of the test object, and generating a vibration image that visualizes the vibration level for each region of the test object on the image at a first natural frequency at which the vibration level peak appears; acquiring noise information including a time series of sound pressure levels of noise generated by the test object and a distribution of sound pressure levels for each vibration frequency based on audio data of the test object recorded simultaneously with the image, and generating a sound source image that visualizes the sound pressure level for each region on the image; extracting a first natural frequency that is within a predetermined range from a second natural frequency at which the sound pressure level peak of the noise appears based on the vibration information and the noise information; and generating an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image. [Effects of the Invention]

[0009] The analysis device, analysis method, and program according to the present disclosure make it possible to visualize the correlation between the sound source of the noise and vibration of the object of inspection. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a diagram illustrating an overall configuration of an analysis system according to a first embodiment of the present disclosure. [Figure 2] 1 is a block diagram showing a functional configuration of an analysis device according to a first embodiment of the present disclosure. [Figure 3] 5 is a flowchart illustrating an example of processing by the analysis device according to the first embodiment of the present disclosure. [Figure 4] FIG. 2 is a diagram illustrating an example of a vibration image according to the first embodiment of the present disclosure. [Figure 5] FIG. 2 is a diagram illustrating an example of a sound source image according to the first embodiment of the present disclosure. [Figure 6] FIG. 2 is a first diagram for explaining the function of the analysis device according to the first embodiment of the present disclosure. [Figure 7]FIG. 2 is a second diagram for explaining the function of the analysis device according to the first embodiment of the present disclosure. [Figure 8] FIG. 2 is a diagram illustrating an example of an analysis image according to the first embodiment of the present disclosure. [Figure 9] FIG. 10 is a first diagram for explaining the function of an analysis device according to a second embodiment of the present disclosure. [Figure 10] FIG. 2 is a second diagram for explaining the function of the analysis device according to the second embodiment of the present disclosure. [Figure 11] FIG. 10 is a first diagram for explaining the function of an analysis device according to a third embodiment of the present disclosure. [Figure 12] FIG. 10 is a second diagram for explaining the function of the analysis device according to the third embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0011] First Embodiment An analysis system 1 and an analysis device 10 according to a first embodiment of the present disclosure will be described below with reference to FIGS.

[0012] (Overall composition) FIG. 1 is a diagram showing the overall configuration of an analysis system according to a first embodiment of the present disclosure. The analysis system 1 is a system for measuring vibrations and noise of an inspection object 9 and outputting an analysis image that enables identification of the noise source location and mechanism (such as the vibration that causes the noise) to be visually confirmed. The inspection object 9 is, for example, a rotating machine such as a motor or a gas turbine. FIG. 1 shows an example in which the inspection object 9 is the rotating shaft of a rotating machine. In addition, in other embodiments, the inspection object 9 may be a rotating blade of a gas turbine or a structure such as a pipe.

[0013] As shown in FIG. 1, the analysis system 1 includes a camera 2, a plurality of microphones 3, and an analysis device .

[0014] The camera 2 captures an image of the inspection object 9. The image D1 (video) captured by the camera 2 is transmitted to the analysis device 10.

[0015] Each of the microphones 3 records a sound (noise) emitted by the test object 9. The sound data D2 recorded by each microphone 3 is transmitted to the analysis device 10.

[0016] The analysis device 10 generates an analysis image that visualizes the vibration and noise state of the inspection object 9 based on the image D1 and the audio recording data D2.

[0017] (Functional configuration of the analysis device) FIG. 2 is a block diagram showing the functional configuration of the analysis device according to the first embodiment of the present disclosure. As shown in FIG. 2, the analysis device 10 includes a processor 11, a memory 12, a storage 13, a communication interface 14, a display device 15, and an input device 16.

[0018] The processor 11 operates according to a predetermined program to function as a vibration image generating unit 110, a sound source image generating unit 111, an extracting unit 112, an analysis image generating unit 113, and an estimating unit 114.

[0019] The vibration image generating unit 110 acquires vibration information including a time series of vibration levels of the inspection object 9 and a distribution of vibration levels for each vibration frequency based on an image D1 obtained by capturing the inspection object 9. The vibration image generating unit 110 also generates a vibration image that visualizes the vibration level for each region of the inspection object on the image D1 at a first natural frequency at which a peak of the vibration level appears.

[0020] The sound source image generation unit 111 acquires noise information including a time series of sound pressure levels of noise generated from the test object 9 and a distribution of sound pressure levels for each frequency based on sound recording data D2 of the test object 9 recorded simultaneously with the image D1. The sound source image generation unit 111 also generates a sound source image that visualizes the sound pressure levels for each region on the image D1.

[0021] The extraction unit 112 extracts, based on the vibration information and the noise information, a first natural frequency that is included within a predetermined range from the second natural frequency at which the peak of the sound pressure level of the noise appears.

[0022] The analysis image generating unit 113 generates an analysis image by superimposing a vibration image corresponding to the first natural frequency extracted by the extracting unit 112 on a sound source image.

[0023] The estimation unit 114 estimates the vibration that is the cause of the noise based on the analysis image.

[0024] The predetermined program executed by the processor 11 is stored in a computer-readable recording medium. Computer-readable recording media include magnetic disks, optical magnetic disks, CD-ROMs, DVD-ROMs, and semiconductor memories. The computer program may be distributed to a computer via a communication line, and the computer that receives the distribution may execute the program. The program may also be a program for implementing some of the above-described functions. Furthermore, the program may be a so-called differential file (differential program) that can realize the above-described functions in combination with a program already stored in the computer system.

[0025] The memory 12 has a memory area necessary for the operation of the processor 11 .

[0026] The storage 13 is a so-called auxiliary storage device, such as a hard disk drive (HDD) or a solid state drive (SSD).

[0027] The communication interface 14 is an interface for transmitting and receiving various signals to and from external devices (such as the camera 2 and microphone 3).

[0028] The display device 15 is a display device that displays the analysis image and the like, and is, for example, a liquid crystal display or an organic EL display.

[0029] The input device 16 is an input device that accepts operations by the user of the analysis device 10, and is, for example, a general mouse, keyboard, touch sensor, or the like.

[0030] (Processing flow of the analysis device) FIG. 3 is a flowchart showing an example of processing of the analysis device according to the first embodiment of the present disclosure. An example of a process in which the analysis device 10 generates an analysis image for analyzing the location and cause of a noise source will be described below with reference to FIG.

[0031] First, the vibration image generating unit 110 acquires vibration information based on the image D1 of the inspection object 9 captured by the camera 2 (step S10). The vibration information includes a time series of vibration levels for each region (e.g., for each pixel) on the image D1 and a distribution of vibration levels for each vibration frequency. First, the vibration image generating unit 110 acquires vibration levels that represent the amount of displacement of the inspection object 9 at each pixel of the image D1 by performing known image processing on the image D1. The vibration image generating unit 110 also performs frequency analysis (e.g., fast Fourier transform) on the time series of vibration levels to obtain a frequency spectrum that represents the distribution of vibration levels for each vibration frequency.

[0032] The vibration image generating unit 110 detects the first natural frequency at which the vibration level reaches a peak based on the frequency spectrum, and generates a vibration image that visualizes the vibration level of each region on the image D1 at the first natural frequency (step S11).

[0033] FIG. 4 is a diagram illustrating an example of a vibration image according to the first embodiment of the present disclosure. As shown in Fig. 4, the vibration image generating unit 110 detects peaks of the vibration level based on a frequency spectrum SP1 included in vibration information D3 acquired from image D1. In the example of Fig. 4, four peaks P1a, P1b, P1c, and P1d appear in frequency spectrum SP1. The frequencies f1a, f1b, f1c, and f1d corresponding to these peaks are first natural frequencies that indicate the vibration mode of the inspection object 9. Furthermore, the vibration image generating unit 110 generates vibration images D4 for each of the first natural frequencies f1a, f1b, f1c, and f1d.

[0034] For example, the vibration image D4 is an image obtained by performing image processing (vibration amplification processing) on the image D1 to amplify the displacement amount of each pixel by X times. As a result, the vibration image generating unit 110 can obtain a vibration image in which the state of vibration at each time (vibration level of each pixel) can be visually recognized, even when minute vibrations that are difficult to detect with the naked eye occur in the inspection object 9.

[0035] The vibration image D4 may also be a contour diagram showing the distribution of vibration levels at the first natural frequency, or a vibration vector image showing vectors representing the direction and magnitude of vibration of each pixel at the first natural frequency.

[0036] The sound source image generation unit 111 acquires noise information based on the sound recording data D2 recorded by each of the multiple microphones 3 (step S12). The noise information includes a time series of sound pressure levels for each region (each pixel) of the image D1 and a distribution of sound pressure levels for each frequency. Specifically, the sound source image generation unit 111 acquires a time series of sound pressure levels of noise for each pixel on the image D1 based on the sound pressure difference, time difference, etc. of the noise recorded in each sound recording data. The sound source image generation unit 111 also performs frequency analysis (e.g., fast Fourier transform) on the time series of sound pressure levels to obtain a frequency spectrum that represents the distribution of sound pressure levels for each frequency.

[0037] The sound recording data D2 is recorded simultaneously with the image D1. A trigger signal indicating the timing for starting analysis may be recorded in the image D1 and the sound recording data D2. In this case, the sound source image generation unit 111 synchronizes the image D1 and the sound recording data D2 based on the trigger signal.

[0038] The sound source image generating unit 111 detects the second natural frequency at which the vibration level peaks based on the frequency spectrum, and generates a sound source image that visualizes the sound pressure level of each region on the image D1 at the second natural frequency (step S13).

[0039] FIG. 5 is a diagram showing an example of a sound source image according to the first embodiment of the present disclosure. As shown in Fig. 5, the sound source image generation unit 111 detects peaks in the sound pressure level based on the frequency spectrum SP2 included in the noise information D5 acquired from the sound recording data D2. In the example of Fig. 5, two peaks P2a and P2b appear in the frequency spectrum SP2. The frequencies f2a and f2b corresponding to these peaks are second natural frequencies related to the noise of the test object 9. The sound source image generation unit 111 also generates sound source images D6 for each of the second natural frequencies f2a and f2b. The sound source images D6 are, for example, contour diagrams representing the distribution of the sound pressure level at the second natural frequencies.

[0040] Next, the extraction unit 112 extracts natural frequencies that approximately match between the structural system and the acoustic system based on the vibration information D3 and the noise information D5. Specifically, the extraction unit 112 compares the frequency spectra SP1 and SP2 of the vibration and the noise, and extracts natural frequencies at which peaks appear in both the vibration level and the sound pressure level (step S15).

[0041] FIG. 6 is a first diagram for explaining the function of the analysis device according to the first embodiment of the present disclosure. According to the frequency spectrum SP1 of the vibration illustrated in FIG. 6, the first natural frequencies at which the vibration level peaks P1a to P1d appear are frequencies f1a to f1d. According to the frequency spectrum SP2 of the noise illustrated in FIG. 6, the second natural frequencies at which the sound pressure level peaks P2a to P2b appear are frequencies f2a and f2b. The extraction unit 112 extracts, from among the multiple first natural frequencies of the vibration, first natural frequencies that fall within a predetermined range from the second natural frequency of the noise as natural frequencies at which the vibration level and sound pressure level both increase. In the example of FIG. 6, the predetermined range R centered on the second natural frequency f2b includes only one first natural frequency f1c, at which the vibration level peak P1c appears. In this case, the extraction unit 112 extracts the first natural frequency f1c.

[0042] FIG. 7 is a second diagram for explaining the function of the analysis device according to the first embodiment of the present disclosure. 7, a plurality of vibration level peaks P1b, P1c, and P1d appear in a predetermined range R centered on the second natural frequency f2b. In this case, the extraction unit 112 extracts a plurality of first natural frequencies f1b, f1c, and f1d corresponding to these peaks P1b, P1c, and P1d.

[0043] 6 and 7, no vibration level peak appears within a predetermined range from the second natural frequency f2a where the sound pressure level peak P2a appears. In this case, the extraction unit 112 determines that the noise indicated by this peak P2a has no correlation with vibration (for example, it is a sound unrelated to vibration, such as wind noise), and does not extract the first natural frequency for this second natural frequency f2a.

[0044] Next, the analysis image generating unit 113 generates an analysis image by superimposing the sound source image D6 on the vibration image D4 corresponding to the first natural frequency extracted by the extracting unit 112 (step S15).

[0045] FIG. 8 is a diagram illustrating an example of an analysis image according to the first embodiment of the present disclosure. If only one first natural frequency is extracted in step S14, the analysis image generation unit 113 generates one analysis image D7 by superimposing a sound source image D6 on a vibration image D4 corresponding to this first natural frequency. Also, as in the examples of FIGS. 7 and 8, if multiple first natural frequencies are extracted in step S14, the analysis image generation unit 113 generates multiple analysis images D7 by superimposing a sound source image D6 on each vibration image D4 corresponding to the first natural frequency. Note that, while the example of FIG. 8 illustrates an example in which a contour diagram is used as the vibration image D4, the present invention is not limited to this. In other embodiments, a vibration amplified video or a vibration vector image may be used as the vibration image D4.

[0046] The analysis image generation unit 113 displays the generated analysis image D7 on the display device 15. A user of the analysis device 10 can confirm the location of the noise source and the vibration that causes the noise by referring to the analysis image D7 displayed on the display device 15. When the analysis image generation unit 113 generates multiple analysis images D7, it may display them side by side on the display device 15 so that they can be compared, or it may switch between the analysis images D7 in accordance with an operation by the user of the analysis device 10. Furthermore, the analysis image generation unit 113 may display, together with the analysis image D7, information such as the frequency spectra SP1 and SP2 of the vibration and noise, and the first and second natural frequencies used in the analysis image D7.

[0047] As shown in the example of FIG. 7, when there are multiple first natural frequencies f1b, f1c, and f1d close to the second natural frequency f2b, it is difficult to identify which vibration (vibration mode) of the first natural frequency is causing the noise indicated by the second natural frequency f2b. However, as shown in FIG. 8, by overlaying a sound source image D6 on each vibration image D4 of the first natural frequencies f1b, f1c, and f1d, it becomes easy to compare the location of vibration due to each vibration mode with the noise source location. In the example of FIG. 8, the location where the vibration of the first natural frequency f1b occurs among the three first natural frequencies is approximately the same as the noise source location (the location where the sound pressure level is maximum). Furthermore, the locations where the vibrations of the other first natural frequencies f1c and f1d occur are far from the noise source location. Therefore, the user of the analysis device 10 can infer that the noise is caused by the vibration of the first natural frequency f1b.

[0048] In addition, there are cases where vibrations propagate from a vibration position (vibration source) to other locations, generating noise from locations distant from the vibration position. Taking such cases into consideration, the analysis image generation unit 113 may, for example, perform a correlation analysis of vibration data between two points between one vibration and another vibration for each of the multiple vibrations represented by the first natural frequency extracted by the extraction unit 112, and display the analysis results together with the analysis image D7. The user refers to the analysis results of the correlation analysis between each vibration along with the analysis image D7 and selects one or more vibrations estimated to be the cause of the noise. For example, in the analysis image D7 shown in FIG. 8, the vibration position of the first natural frequency f1b substantially coincides with the sound source position of the noise, and the correlation between the first natural frequency f1b and the first natural frequency f1c is greater than a predetermined value. In this case, the user estimates that the two vibrations, the first natural frequency f1b and the first natural frequency f1c, are candidates for the cause of the noise based on the analysis image D7 and the analysis results of the correlation analysis.

[0049] Furthermore, the estimation of the cause of the noise may be performed automatically by the analysis device 10, rather than by a user visually inspecting the analysis image D7. Specifically, the estimation unit 114 compares the areas (pixels) in each analysis image D7 where the sound pressure level and vibration level are maximum, and estimates whether the vibration of the first natural frequency shown in each analysis image D7 is the cause of the noise. If the distance between the area in the analysis image D7 where the sound pressure level is maximum (sound source position) and the area in the analysis image D7 where the vibration level is maximum (vibration position) is equal to or less than a threshold, the estimation unit 114 estimates that the vibration of the first natural frequency shown in this analysis image D7 is the cause of the noise. On the other hand, if the distance between the sound source position and the vibration position exceeds the threshold, the estimation unit 114 estimates that the vibration shown in this analysis image D7 is not the cause of the noise. The area where the sound pressure level is maximum is the area where the maximum sound pressure level is detected, or the area where a value up to X% (e.g., 90%) of the maximum sound pressure level is detected. The same applies to the area where the vibration level is maximum.

[0050] Furthermore, when the analysis image generation unit 113 performs a correlation analysis of each vibration with other vibrations, the estimation unit 114 may further estimate the vibration causing the noise based on the analysis results. For example, in the analysis image D7 shown in FIG. 8, the vibration position of the first natural frequency f1b substantially coincides with the noise source position, and the correlation between the first natural frequency f1b and the first natural frequency f1c is greater than a preset value. In this case, the estimation unit 114 may estimate that the two vibrations represented by the first natural frequency f1b and the first natural frequency f1c are candidates for the cause of the noise based on the analysis image D7 and the analysis results of the correlation analysis. This makes it possible to prevent the vibration position from being excluded from the candidates for the cause of the noise when noise occurs at a position distant from the vibration position.

[0051] The estimation unit 114 displays on the display device 15 an analysis image D7 showing the vibration estimated to be the cause of the noise together with the estimation result.

[0052] If the estimation unit 114 determines that the distance between the sound source position and the vibration position for multiple analysis images D7 is equal to or less than a threshold, it may estimate the multiple vibrations shown in these analysis images D7 as candidates for the cause of the noise. In this case, the estimation unit 114 may display the multiple estimation results and analysis images D7 on the display device 15 to prompt the user to check them. By narrowing down the candidates and displaying them in this way, the user can reduce the effort required to check all analysis images D7.

[0053] (Action, effect) As described above, the analysis device 10 according to this embodiment includes a vibration image generation unit 110 that generates a vibration image D4 that visualizes the vibration level for each region of the test object 9, a sound source image generation unit 111 that generates a sound source image D6 that visualizes the sound pressure level for each region of the test object 9, an extraction unit 112 that extracts a first natural frequency at which peaks of the vibration level and sound pressure level commonly appear, and an analysis image generation unit 113 that generates an analysis image in which the vibration image D4 corresponding to the extracted first natural frequency and the sound source image D6 are superimposed.

[0054] With this configuration, the analysis device 10 can provide the user with an analysis image D7 that enables visual recognition of the correlation between the strength of vibration and noise, the correlation between the vibration position and the position of the noise source, etc. Furthermore, the user can easily estimate the vibration that is causing the noise by referring to the analysis image D7, and can therefore take appropriate measures against the noise at an early stage.

[0055] Furthermore, the extraction unit 112 extracts a plurality of first natural frequencies that are included within a predetermined range from the second natural frequency, and the analysis image generation unit 113 generates a plurality of analysis images D7 that correspond to the plurality of first natural frequencies, respectively.

[0056] In this way, when there are multiple vibration modes (first natural frequencies) close to the second natural frequency of the noise, the analysis device 10 can generate an analysis image D7 for each vibration mode. Furthermore, even if the user cannot narrow down the vibration causing the noise to one based on the natural frequency, the user can accurately estimate the vibration causing the noise by checking the correlation between the vibration position and the noise source position in the analysis image D7.

[0057] The analysis device 10 further includes an estimation unit 114 that automatically estimates the vibration that causes the noise based on the analysis image D7.

[0058] In this way, analysis device 10 can make the user recognize the vibration that is estimated to be the cause of the noise. Furthermore, even if multiple analysis images D7 are generated, analysis device 10 can reduce the effort required for the user to check all analysis images D7 by presenting to the user the analysis image D7 that corresponds to the vibration estimated to be the cause.

[0059] Furthermore, if the distance between the vibration position and the sound source position in analysis image D7 is equal to or less than a threshold, estimation unit 114 estimates that the vibration of the first natural frequency shown in this analysis image D7 is the cause of the noise. Furthermore, if the distance between the vibration position and the sound source position in analysis image D7 exceeds a threshold, estimation unit 114 estimates that the vibration of the first natural frequency shown in this analysis image D7 is not the cause of the noise.

[0060] In this way, the analysis device 10 can accurately estimate whether or not the vibration in each analysis image D7 is the cause of the noise, based on the correlation between the vibration position and the sound source position shown in the analysis image.

[0061] <Second embodiment> Next, an analysis system 1 and an analysis device 10 according to a second embodiment of the present disclosure will be described with reference to FIGS. Components common to the first embodiment are given the same reference numerals and detailed description thereof will be omitted.

[0062] In the first embodiment, the extraction unit 112 extracts the first natural frequency that is included within a predetermined range from the second natural frequency of the noise. In contrast, the extraction unit 112 according to the present embodiment further extracts the first natural frequency when a frequency that is a predetermined multiple (n times) of the first natural frequency is included within a predetermined range from the second natural frequency.

[0063] FIG. 9 is a first diagram for explaining the function of the analysis device according to the second embodiment of the present disclosure. As shown in Fig. 9, the inspection object 9 may collide with structures on both sides due to vibration, generating noise. In this case, noise occurs twice during one cycle of vibration (one cycle of the first natural frequency). In other words, noise occurs at twice the frequency of the first natural frequency of vibration.

[0064] FIG. 10 is a second diagram for explaining the function of the analysis device according to the second embodiment of the present disclosure. Fig. 10 shows frequency spectra SP1 and SP2 of the vibration and noise exemplified in Fig. 9. In the example of Fig. 10, among the multiple peaks of the vibration level, a frequency f1c' that is twice the first natural frequency f1c at which peak P1c appears is assumed to be included within a predetermined range R of a second natural frequency f2b of the noise. Furthermore, it is assumed that the other first natural frequencies f1a, f1b, and f1d are not included within the predetermined range R of the second natural frequency even when multiplied by n. In this case, the extraction unit 112 extracts only the first natural frequency f1c as a natural frequency correlated with the second natural frequency f2b of the noise.

[0065] In this way, the analysis device 10 can generate an analysis image D7 that shows candidates for vibrations that cause noise.

[0066] <Third embodiment> Next, an analysis system 1 and an analysis device 10 according to a third embodiment of the present disclosure will be described with reference to FIGS. Components common to the first embodiment are given the same reference numerals and detailed description thereof will be omitted.

[0067] In the first embodiment, it has been described that one analysis image D7 is generated by superimposing the sound source image D6 on the vibration image D4 corresponding to the first natural frequency. Also, it has been described that when there are multiple first natural frequencies, the analysis image generation unit 113 generates multiple analysis images D7 by superimposing the sound source image D6 on each vibration image D4 corresponding to the first natural frequency. It has been explained that the analysis device 10 according to the second embodiment can generate an analysis image D7 that represents candidates for vibrations that cause noise, taking into account the possibility that noise may occur at a period that is n times the period of the vibration. In contrast to these, the analysis image generation unit 113 according to this embodiment displays the frequency spectra of the vibration and noise used to create the analysis image in an overlapping manner, and displays the analysis image related to the peak of the frequency spectrum specified by the user.

[0068] FIG. 11 is a first diagram for explaining the function of the analysis device according to the third embodiment of the present disclosure. The analysis image generating unit 113 displays a frequency spectrum in which a frequency spectrum SP3 of vibration contained in vibration information D3 acquired from image D1 and a frequency spectrum SP4 of noise contained in noise information D5 acquired from sound recording data D2 are superimposed, with the horizontal axis representing the vibration frequency, the first vertical axis representing the vibration level, and the second vertical axis representing the sound pressure level, as shown in Fig. 11. Note that the display layout in Fig. 11 is an example and is not limited to this.

[0069] There are two possible reasons for overlaying two frequency spectra to display a frequency spectrum with a wider frequency range: Regarding vibration, when the frequency is low, the displacement of the test object is large and it is easy to judge the vibration. When the frequency is high, the displacement of the test object is small and it can be difficult to judge the vibration. Regarding sound pressure, at low frequencies, it can be difficult to pinpoint the source of the noise as the wavelength gets longer. At high frequencies, the wavelength is short, making it easier to pinpoint the source of the noise. That is, the lower the frequency of the object being inspected, the easier it is to determine the vibration, and the higher the frequency, the easier it is to identify the source of the noise. As described above, since the detectable ranges for vibration and sound pressure are different, it is possible that, depending on the region of each frequency spectrum, there may be a lack of information on either the vibration frequency spectrum or the noise frequency spectrum, as in the ranges Y1 and Y2 shown in Figure 11. For example, Y1 has a vibration frequency spectrum SP3, but lacks information on a noise frequency spectrum SP4. For example, Y2 has noise frequency spectrum SP4, but lacks information on vibration frequency spectrum SP3. This makes it possible to preferably identify the location of vibration or noise even in frequency ranges (shown as ranges Y1 and Y2) outside range Z where the vibration frequency spectrum SP3 and the noise frequency spectrum SP4 overlap.

[0070] The test object 9 may have a peak in vibration level or sound pressure level in ranges Y1 and Y2. As shown in Fig. 11, in one example according to this embodiment, a peak in sound pressure level is observed on the high frequency side (range Y2) and a peak in vibration level is observed on the low frequency side (range Y1).

[0071] When generating analysis image D7 (see FIG. 8), the analysis image generation unit 113 of this embodiment takes into account that peaks in vibration level or sound pressure level may exist in ranges Y1 and Y2, and displays the frequency spectrum in a frequency range including SP3 and SP4.

[0072] FIG. 12 is a second diagram for explaining the function of the analysis device according to the third embodiment of the present disclosure.

[0073] The user specifies a peak of the vibration level or sound pressure level from the frequency spectrum displayed by the analysis image generation unit 113. The analysis image generation unit 113 additionally displays an image related to the peak of the specified frequency spectrum together with the frequency spectrum.

[0074] For example, when peak P4b of SP4 is specified in range Z where the frequency spectra of SP3 and SP4 overlap, only one first natural frequency f3e, where peak P3e of the vibration level appears, is included in a predetermined range S centered on second natural frequency f4b. In this case, since there is only one first natural frequency extracted in step S14, the analysis image generation unit 113 additionally displays one analysis image D7 in which sound source image D6 at f4b is superimposed on vibration image D4 corresponding to this first natural frequency. When the peak P3e of SP3 is designated, if there is an analysis image D7 related to the vibration image D4 corresponding to this first natural frequency f3e, the analysis image generating unit 113 displays one additional analysis image D7.

[0075] Similarly, when peak P4d of SP4 is specified in range Z where the frequency spectra of SP1 and SP2 overlap, first natural frequencies f3h, f3i, and f3j, where vibration level peaks (P3h, P3i, P3j) appear, exist in a predetermined range T centered on the second natural frequency f4d. In this case, because there are multiple first natural frequencies extracted in step S14, the analysis image generation unit 113 additionally displays multiple analysis images D7 in which a sound source image D6 at f4d is superimposed on each vibration image D4 corresponding to the first natural frequency. When peaks P3h, P3i, or P3j of SP3 are specified, if there are multiple analysis images D7 associated with each of the vibration images D4 corresponding to the first natural frequencies (f3h, f3i, f3j), the analysis image generation unit 113 additionally displays those analysis images D7.

[0076] Furthermore, when peak P3c of SP3 is specified in range Z where the frequency spectra of SP3 and SP4 overlap, if there is no analysis image D7 related to vibration image D4 corresponding to the first natural frequency f3c, the analysis image generation unit 113 additionally displays vibration image D4. The same applies when P3d, P3f, or P3g is specified.

[0077] Furthermore, when peak P4a of SP4 is specified in range Z where the frequency spectra of SP3 and SP4 overlap, if there is no peak of the vibration level within a predetermined range centered on the second natural frequency f4a, analysis image D7 will not be additionally displayed. Analysis image generation unit 113 additionally displays sound source image D6. The same applies when P4c, P4e, P4f, or P4g is specified. Furthermore, when a certain first natural frequency is multiplied by n and the result is included within a predetermined range of the second natural frequency f4a, the analysis device 10 can additionally display an analysis image D7 that represents candidates for vibrations that may be causing noise associated with the certain first natural frequency and the second natural frequency f4a. In this case, the user may set whether to display the analysis image D7 representing the candidates.

[0078] When an arbitrary natural frequency is designated within ranges Y1 and Y2 where the frequency spectra of SP3 and SP4 do not overlap, the analysis image generating unit 113 additionally displays the next image together with the frequency spectrum. When peak P3a or P3b of SP3 is specified, if the peak of SP4 does not exist in range Y1, the analysis image generating unit 113 additionally displays vibration image D4. When peak P4e, P4f, or P4g of SP4 is specified, if the peak of SP1 does not exist in range Y2, analysis image generating unit 113 additionally displays sound source image D6.

[0079] Furthermore, in one example according to this embodiment, by displaying a wide frequency range, if the peak of SP3 is included within a predetermined range of one of the second natural frequencies as a result of multiplying the first natural frequency present in range Y1 by n, an analysis image D7 can be generated that shows candidate vibrations that may cause noise and are related to the first natural frequency and the second natural frequency.

[0080] In this way, analysis device 10 can provide the user with a frequency spectrum with a wider frequency range and an image related to the peak of the frequency spectrum. When a peak in the vibration level or sound pressure level is observed in a range where the frequency spectra of vibration and noise do not overlap, the user can refer to the image related to the peak and infer the cause of the noise or vibration.

[0081] As described above, several embodiments according to the present disclosure have been described, but all of these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope of the invention and its equivalents as defined in the claims, as well as in the scope and spirit of the invention.

[0082] <Additional Notes> The analysis device, analysis method, and program described in the above-described embodiments can be understood, for example, as follows.

[0083] (1) According to a first aspect of the present disclosure, an analysis device (10) includes: a vibration image generation unit (110) that acquires vibration information including a time series of vibration levels of the inspection object (9) and a distribution of vibration levels for each vibration frequency based on an image of the inspection object (9), and generates a vibration image that visualizes the vibration level for each region of the inspection object (9) on the image at a first natural frequency at which a peak of the vibration level appears; a sound source image generation unit (111) that acquires noise information including a time series of sound pressure levels of noise generated from the inspection object (9) and a distribution of sound pressure levels for each vibration frequency based on sound recording data of the inspection object (9) recorded simultaneously with the image, and generates a sound source image that visualizes the sound pressure level for each region on the image; an extraction unit (112) that extracts, based on the vibration information and the noise information, a first natural frequency that is included within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears; and an analysis image generation unit (113) that generates an analysis image by superimposing a vibration image corresponding to the extracted first natural frequency and the sound source image.

[0084] With this configuration, the analysis device can provide the user with an analysis image that allows visual recognition of the correlation between the strength of vibration and noise, the correlation between the position of vibration and the position of the noise source, etc. Furthermore, the user can easily estimate the vibration that is causing the noise by referring to the analysis image, allowing them to take appropriate measures against the noise at an early stage.

[0085] (2) According to a second aspect of the present disclosure, in the analysis device (10) according to the first aspect, the extraction unit (112) extracts a plurality of first natural frequencies that fall within a predetermined range from the second natural frequency, and the analysis image generation unit generates a plurality of analysis images that correspond to each of the plurality of first natural frequencies.

[0086] In this way, when there are multiple vibration modes (first natural frequencies) close to the second natural frequency of the noise, the analysis device can generate an analysis image for each vibration mode. Furthermore, even if the user cannot narrow down the vibration causing the noise to one based on the natural frequency, the user can accurately estimate the vibration causing the noise by checking the correlation between the vibration position and the noise source position in the analysis image.

[0087] (3) According to a third aspect of the present disclosure, in the analysis device (10) according to the first or second aspect, the extraction unit (112) extracts the first natural frequency when the frequency obtained by multiplying the first natural frequency by a predetermined factor is included within a predetermined range from the second natural frequency.

[0088] In this way, the analysis device can generate an analysis image that shows candidates for vibrations that cause noise, taking into account the possibility that noise will occur at a cycle that is a predetermined multiple of the vibration cycle.

[0089] (4) According to a fourth aspect of the present disclosure, the analysis device (10) according to any one of the first to third aspects further includes an estimation unit (114) that estimates the vibration that causes the noise based on the analysis image.

[0090] In this way, the analysis device can make the user aware of the vibration that is estimated to be the cause of the noise. Furthermore, even if multiple analysis images are generated, the analysis device can reduce the user's effort to check all of the analysis images by presenting the analysis image that corresponds to the vibration that is estimated to be the cause of the noise.

[0091] (5) According to a fifth aspect of the present disclosure, in the analysis device (10) according to the fourth aspect, when the distance between the area in the analysis image where the sound pressure level is maximum and the area in the analysis image where the vibration level is maximum is equal to or less than a threshold, the estimation unit (114) estimates that the vibration of the first natural frequency shown in the analysis image is the cause of the noise.

[0092] By doing this, the analysis device can accurately estimate that the vibration in this analysis image is the cause of the noise, based on the correlation between the vibration position shown in the analysis image and the sound source position.

[0093] (6) According to a sixth aspect of the present disclosure, in the analysis device (10) according to the fourth or fifth aspect, if the distance between the area in the analysis image where the sound pressure level is maximum and the area in the analysis image where the vibration level is maximum exceeds a threshold, the estimation unit (114) estimates that the vibration of the first natural frequency shown in the analysis image is not the cause of the noise.

[0094] By doing this, the analysis device can accurately estimate, from the correlation between the vibration position shown in the analysis image and the sound source position, that the vibration in this analysis image is not the cause of the noise.

[0095] (7) According to a seventh aspect of the present disclosure, in the analysis device (10) according to the first or sixth aspect, the analysis image generation unit (113) displays the frequency spectra of vibration and noise superimposed on the same axis, and displays an image related to the peak of a specified frequency spectrum.

[0096] In this way, the analysis device can provide the user with a frequency spectrum with a wider frequency range and an image related to the peak of the frequency spectrum. Furthermore, the object of inspection may have a peak in vibration level or sound pressure level in a range where the frequency spectra of vibration and noise do not overlap. The user can refer to the frequency spectrum and the image related to the peak of the frequency spectrum, even in a range where the frequency spectra of vibration and noise do not overlap, to infer the cause of the noise or vibration.

[0097] (8) According to an eighth aspect of the present disclosure, an analysis method includes the steps of: acquiring vibration information including a time series of vibration levels of the inspection object (9) and a distribution of vibration levels for each vibration frequency based on an image of the inspection object (9); generating a vibration image that visualizes the vibration level for each region of the inspection object (9) on the image at a first natural frequency at which a peak of the vibration level appears; acquiring noise information including a time series of sound pressure levels of noise generated from the inspection object (9) and a distribution of sound pressure levels for each vibration frequency based on sound recording data of the inspection object (9) recorded simultaneously with the image; generating a sound source image that visualizes the sound pressure level for each region on the image; extracting a first natural frequency that is within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears based on the vibration information and the noise information; and generating an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image.

[0098] (9) According to a ninth aspect of the present disclosure, the program causes an analysis device to execute the following steps: acquiring vibration information including a time series of vibration levels of the inspection object (9) and a distribution of vibration levels for each vibration frequency based on an image of the inspection object (9), and generating a vibration image that visualizes the vibration level for each region of the inspection object (9) on the image at a first natural frequency at which a peak of the vibration level appears; acquiring noise information including a time series of sound pressure levels of noise generated from the inspection object (9) and a distribution of sound pressure levels for each vibration frequency based on sound recording data of the inspection object (9) recorded simultaneously with the image, and generating a sound source image that visualizes the sound pressure level for each region on the image; extracting a first natural frequency that is included within a predetermined range from the second natural frequency at which a peak of the sound pressure level of the noise appears based on the vibration information and the noise information; and generating an analysis image in which the vibration image corresponding to the extracted first natural frequency and the sound source image are superimposed. [Industrial Applicability]

[0099] According to the above-described aspect, it is possible to visualize the correlation between the sound source of the noise and the vibration of the test object. [Explanation of symbols]

[0100] 1. Analysis system 2 Cameras 3 microphones 9. Inspection Subjects 10 Analysis device 11 processors 110 Vibration image generation unit 111 Sound source image generation unit 112 Extraction part 113 Analysis image generation unit 114 Estimation Department 12 Memory 13. Storage 14 Communication Interface 15 Display device 16 Input Devices

Claims

1. a vibration image generating unit that acquires vibration information including a time series of vibration levels of the inspection object and a distribution of vibration levels for each vibration frequency based on an image of the inspection object, and generates a vibration image that visualizes the vibration levels for each region of the inspection object on the image at a first natural frequency at which a peak of the vibration level appears; a sound source image generating unit that acquires noise information including a time series of sound pressure levels of noise generated from the test object and a distribution of sound pressure levels for each frequency based on sound recording data of the test object recorded simultaneously with the image, and generates a sound source image that visualizes the sound pressure levels for each region on the image; an extracting unit that extracts a first natural frequency that is included within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears, based on the vibration information and the noise information; an analysis image generating unit that generates an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image; An analysis device comprising:

2. the extraction unit extracts a plurality of the first natural frequencies that are included within a predetermined range from the second natural frequency, the analysis image generation unit generates a plurality of analysis images corresponding to the plurality of first natural frequencies, respectively. The analysis device according to claim 1 .

3. the extraction unit extracts the first natural frequency when a frequency obtained by multiplying the first natural frequency by a predetermined number is included within a predetermined range from the second natural frequency. The analysis device according to claim 1 or 2.

4. further comprising an estimation unit that estimates vibrations that cause the noise based on the analysis image. The analysis device according to any one of claims 1 to 3.

5. the estimation unit estimates that vibration of a first natural frequency shown in the analysis image is the cause of the noise when a distance between a region in the analysis image where the sound pressure level is maximum and a region in the analysis image where the vibration level is maximum is equal to or less than a threshold. The analysis device according to claim 4.

6. the estimation unit estimates that vibration of the first natural frequency shown in the analysis image is not the cause of the noise when a distance between a region in the analysis image where the sound pressure level is maximum and a region in the analysis image where the vibration level is maximum exceeds a threshold. The analysis device according to claim 4 or 5.

7. the analysis image generation unit displays the frequency spectra of vibration and noise superimposed on the same axis, and displays an image related to a specified peak of the frequency spectrum. The analysis device according to claim 1 or 6.

8. acquiring vibration information including a time series of vibration levels of the inspection object and a distribution of vibration levels for each vibration frequency based on an image of the inspection object, and generating a vibration image that visualizes the vibration levels for each region of the inspection object on the image at a first natural frequency at which a peak of the vibration level appears; a step of acquiring noise information including a time series of sound pressure levels of noise generated from the test object and a distribution of sound pressure levels for each frequency based on sound recording data of the test object recorded simultaneously with the image, and generating a sound source image that visualizes the sound pressure levels for each region on the image; extracting a first natural frequency that is included within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears, based on the vibration information and the noise information; generating an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image; An analysis method having the following.

9. acquiring vibration information including a time series of vibration levels of the inspection object and a distribution of vibration levels for each vibration frequency based on an image of the inspection object, and generating a vibration image that visualizes the vibration levels for each region of the inspection object on the image at a first natural frequency at which a peak of the vibration level appears; a step of acquiring noise information including a time series of sound pressure levels of noise generated from the test object and a distribution of sound pressure levels for each frequency based on sound recording data of the test object recorded simultaneously with the image, and generating a sound source image that visualizes the sound pressure levels for each region on the image; extracting a first natural frequency that is included within a predetermined range from a second natural frequency at which a peak of the sound pressure level of the noise appears, based on the vibration information and the noise information; generating an analysis image by superimposing the vibration image corresponding to the extracted first natural frequency and the sound source image; A program that causes the analysis device to execute the above.

Images