JP7118626B2 - System, method and program - Google Patents

Description

Embodiments of the present invention relate to systems, methods and programs.

In recent years, industrial equipment used in a wide range of fields has been developed, contributing to industrial development. In order to use such industrial equipment continuously, maintenance management and the like of the industrial equipment are very important.

Therefore, non-contact detection of an abnormality or the like in the industrial equipment at an early stage is performed by, for example, collecting operating sounds of the industrial equipment with a microphone.

Japanese Unexamined Patent Application Publication No. 2008-141718

By the way, for example, by using a plurality of microphones, noise from directions other than a specific direction (directional noise) is suppressed, and the target sound (the operating sound of industrial equipment) from the specific direction is efficiently extracted. It is known.

However, there are cases where the target sound cannot be appropriately extracted depending on the environment in which the above-described multiple microphones are arranged.

Therefore, the problem to be solved by the present invention is to provide a system, method, and program used for properly extracting a target sound.

According to an embodiment, a system is provided that collects sounds emitted from an object using first to n-th microphones (where n is 3 or more) . When the system collects sounds emitted from the object with at least an a-th (a is a natural number from 1 to n) and a b-th (b is a natural number from 1 to n different from a) microphones A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the a-th microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) microphone and the d-th (d is 1 ~n, a natural number different from c) based on the degree of variation of the second distribution of the phase difference between the c-th microphone and the d-th microphone when collecting the sound emitted from the object with the microphone collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or including at least the c-th microphone and the d-th microphone , based on a second goodness of fit; Input means for inputting a selection result as to whether to collect the sound emitted from the object in combination, and collection for collecting the sound emitted from the object using a plurality of microphones in combination based on the input selection result. and means . At least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone. The first goodness of fit is calculated to have a high value when the first distribution of the phase differences is not dispersed. A low value is calculated when the first distribution of the phase difference is scattered. The second goodness of fit is calculated to have a high value when the second distribution of the phase differences does not fluctuate, and a low value is calculated when the second distribution of the phase differences fluctuates. The first distribution includes a distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time. The second distribution includes distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing microphone signals relating to sounds collected by the c-th microphone and the d-th microphone for a certain period of time.

1 is a schematic diagram showing an example of industrial equipment to which the system according to the first embodiment is applied; FIG. FIG. 2 is a diagram showing an example of the hardware configuration of the system; FIG. 2 is a block diagram showing an example of the functional configuration of the system; FIG. The figure which shows an example of a structure of a process part. 4 is a flowchart showing an example of the processing procedure of the system; FIG. 4 is a diagram showing an example of a histogram representing the distribution of phase differences; FIG. 5 is a diagram showing another example of a histogram representing the distribution of phase differences; The figure for demonstrating the abnormality detection system using a system. FIG. 4 is a diagram for explaining a system using a sound source; The schematic diagram which shows an example of the industrial equipment with which the system which concerns on 2nd Embodiment is applied. The figure which shows an example of a structure of a process part. 4 is a flowchart showing an example of the processing procedure of the system;

Embodiments will be described below with reference to the drawings.
(First embodiment)
The system according to the first embodiment is used when collecting (recording) operating sounds of industrial equipment, for example, in order to detect an abnormality (or deterioration, etc.) of the industrial equipment.

FIG. 1 is a schematic diagram showing an example of industrial equipment to which the system according to this embodiment is applied. In FIG. 1, it is assumed that the industrial equipment is an uninterruptible power supply (UPS).

The uninterruptible power supply 1 is a power supply that can supply power even when a power failure or the like occurs, and can stably supply AC power to a load such as a computer system, for example.

A power converter 2 and a fan 3 are provided in the housing of the uninterruptible power supply 1 . The power converter 2 includes a converter, an inverter, and the like. The converter converts, for example, commercial AC power into DC power and outputs the DC power. The DC power output from the converter in this way is stored in a battery or the like (not shown), for example. The inverter converts the DC power from the battery into AC power and supplies it to the load.

The fan 3 is provided to cool the inside of the uninterruptible power supply 1 and operate the uninterruptible power supply 1 normally.

Here, for example, as described above, it is assumed that the operation sound of the fan 3 provided in the housing of the uninterruptible power supply 1 is collected in order to detect an abnormality of the fan 3 . The operation sound of the fan 3 is collected using a microphone (hereinafter simply referred to as a microphone) provided near the fan 3 inside the housing of the uninterruptible power supply 1 .

In this case, since the above-described power conversion unit 2 and the like are provided in the housing of the uninterruptible power supply 1, the microphone not only hears the operating sound of the fan 3, but also, for example, the power conversion unit 2 (converter and inverter). It also collects operating sounds.

For this reason, it is conceivable to provide, for example, two microphones in the housing of the uninterruptible power supply 1 and perform beam forming processing (noise suppression processing) on signals related to sound collected by each of the microphones. According to such beamforming processing, since it is possible to give directivity to sound from a specific direction, noise (directional noise) other than the operation sound of the fan 3 is suppressed, and the noise collected by the microphone is suppressed. The operation sound of the fan 3 can be extracted from the sound.

By the way, the housing of the uninterruptible power supply 1 is made of metal such as iron. The collected sound includes the echo sound of the operation sound of the fan 3 (the sound of the operation sound of the fan 3 echoing on the wall surface of the housing) and the like.

Since such reverberating sounds are input to the microphones from various directions, even if the beamforming process described above is executed, depending on the position where the microphones are installed, directional noise cannot be suppressed due to the effects of the reverberating sounds, etc. As a result, it may not be possible to properly extract the operating sound of the fan 3 .

Therefore, in the present embodiment, as shown in FIG. 1, at least three microphones (first microphone 10a, second microphone 10b, and third microphone 10c) are provided in the housing of the uninterruptible power supply 1, and the first microphone 10a, the second microphone 10b and the third microphone 10c.

In FIG. 1, three microphones are provided in the housing of the uninterruptible power supply 1, but the first to n-th microphones (n is a natural number of 3 or more) Just do it. Also, the number of microphones to be selected should be at least two or more.

In the housing of the uninterruptible power supply 1, the first to n-th microphones may be arranged, for example, in an array or two-dimensional plane, or may be three-dimensionally arranged in a three-dimensional space.

In the following description, the operation sound of the fan 3 extracted using the system 10 according to this embodiment is called a target sound, and the fan 3 is called an object. That is, the system 10 according to the present embodiment has a configuration that selects a combination of microphones suitable for extracting the target sound emitted from the object and collects the sound using the selected combination of microphones.

FIG. 2 shows an example of the hardware configuration of the system 10 according to this embodiment. The system 10 is implemented as, for example, a server device or the like, and includes a CPU 11, a nonvolatile memory 12, a main memory 13, a communication device 14, and the like, as shown in FIG.

CPU 11 is a hardware processor that controls the operation of various components within system 10 . The CPU 11 executes various programs loaded into the main memory 13 from the non-volatile memory 12, which is a storage device. The programs executed by the CPU 11 include an operating system, a program for selecting two microphones among the first microphone 10a to the third microphone 10c (hereinafter referred to as a microphone selection program), and the like. The CPU 11 also executes, for example, a basic input/output system (BIOS), which is a program for hardware control.

Communication device 14 is, for example, a device configured to perform wired or wireless communication.

FIG. 3 is a block diagram showing an example of the functional configuration of the system 10 according to this embodiment. As shown in FIG. 3, system 10 includes processing unit 101 and collecting unit 102 .

In this embodiment, part or all of the processing unit 101 and the collecting unit 102 are implemented by causing the CPU 11 to execute the microphone selection program, that is, by software. Part or all of the processing unit 101 and the collecting unit 102 may be implemented by hardware such as an IC (Integrated Circuit), or may be implemented as a combination of software and hardware.

Here, each of the first microphone 10a to the third microphone 10c is provided with the first microphone 10a to the third microphone 10c, for example, input (collect) the sound generated in the housing of the uninterruptible power supply 1, It is a device that converts the input sound into an electric signal (hereinafter referred to as a microphone signal) and outputs it. It is assumed that the microphone signals output from the first microphone 10a to the third microphone 10c are transmitted to the system 10 via a communication device or the like (not shown).

The processing unit 101 uses two microphones suitable for extracting a target sound (for example, the operating sound of the fan 3) based on the microphone signals output from each of the first microphone 10a, the second microphone 10b, and the third microphone 10c. Execute processing for selecting (determining) a combination of two microphones.

Here, the correlation between the two microphones used to extract the target sound is represented by the phase difference between the microphone signals output from each of the microphones (sounds collected by each of the microphones). The processing unit 101 receives a microphone signal having a phase difference that enables highly accurate estimation of the above-described target sound and other noise (e.g., the operation sound of the power conversion unit 2 or the reverberation sound of the operation sound of the fan 3). to select two microphones that output .

The processing unit 101 outputs (transmits) information indicating the selected two microphones (hereinafter referred to as microphone selection information) to the collection unit 102 .

The collecting unit 102 uses the first to third microphones 10a to 10c to collect sounds generated inside the housing of the uninterruptible power supply 1 in which the first to third microphones 10a to 10c are provided. In this case, the collecting unit 102 collects the microphone signals output from the two microphones indicated by the microphone selection information output by the processing unit 101 among the microphone signals output from each of the first microphone 10a to the third microphone 10c. are collected and output.

FIG. 4 shows an example of the configuration of the processing unit 101 shown in FIG. 3 described above. As shown in FIG. 4 , the processing section 101 includes a first selection section 201 , a first calculation section 202 , a second calculation section 203 and a second selection section 204 .

The first selection unit 201 selects (a combination of) any two microphones from the first microphone 10a to the third microphone 10c.

The first calculator 202 calculates the phase difference between the two microphones based on the microphone signals output from the two microphones selected by the first selector 201 .

The second calculation unit 202 calculates a goodness of fit based on (distribution of) the phase difference calculated by the first calculation unit 202 . The degree of suitability calculated by the second calculation unit 202 is a value representing the degree to which the two microphones (combination) selected by the first selection unit 201 are a combination of microphones suitable for extracting the target sound. is.

A second selection unit 204 extracts a target sound from a combination of any two microphones among the first microphone 10a to the third microphone 10c based on the degree of conformity calculated by the second calculation unit 202. Select the appropriate microphone combination for (i.e., the combination of microphones used to collect the sound).

An example of the processing procedure of the system 10 according to the present embodiment will be described below with reference to the flowchart of FIG. Here, processing of the processing unit 101 included in the system 10 will be mainly described. It should be noted that the processing shown in FIG. 5 is performed, for example, before maintenance and inspection of an object (for example, fan 3).

First, the processing unit 101 acquires microphone signals output from each of the first to third microphones 10a to 10c (step S1). It should be noted that the microphone signal acquired in step S1, for example, in a state in which the target (fan 3) is operating in the housing of the uninterruptible power supply 1 shown in FIG. It is a signal related to sound for a certain period of time input by each of the third microphones 10a to 10c.

In the following description, among the microphone signals acquired in step S1, the microphone signal output by the first microphone 10a is referred to as the microphone signal of the first microphone 10a for convenience. Similarly, of the microphone signals acquired in step S1, the microphone signal output by the second microphone 10b is the microphone signal of the second microphone 10b, and the microphone signal output by the third microphone 10c is the microphone signal of the third microphone 10c. called a signal.

Here, for each combination of two microphones (hereinafter referred to as a microphone pair) of the plurality of microphones (here, the first microphone 10a to the third microphone 10c) provided in the housing of the uninterruptible power supply 1 , the following steps S2 to S4 are executed.

In this case, the selection unit 201 included in the processing unit 101 selects one microphone pair from the first microphone 10a to the third microphone 10c (step S2). Here, it is assumed that a combination of the first microphone 10a and the second microphone 10b is selected as the microphone pair.

Next, the first calculator 202 acquires the microphone signals of the first microphone 10a and the second microphone 10b selected as the microphone pair in step S2. The first calculator 202 calculates the phase difference between the first microphone 10a and the second microphone 10b based on the acquired microphone signals of the first microphone 10a and the second microphone 10b (step S3).

In step S3, the first calculator 202 divides, for example, the microphone signal of the first microphone 10a and the microphone signal of the second microphone 10b for each unit time, and divides each of the divided microphone signals of the first microphone 10a and A phase difference is calculated based on each of the divided microphone signals of the second microphone 10b. That is, the first calculator 202 can calculate the phase difference for each unit time based on the corresponding microphone signals of the first microphone 10a and the second microphone 10b divided for each unit time. The phase difference can be calculated from, for example, the time difference between the microphone signals (sounds) of the first microphone 10a and the second microphone 10b obtained by a cross-correlation function. As the phase difference, for example, a phase difference between frequency components obtained by calculating an index such as cross spectrum or coherence from Fourier transform between microphone signals may be used.

In the present embodiment, by calculating the phase difference for each unit time in step S3 described above, it is possible to obtain the phase difference distribution of the sounds collected by the first microphone 10a and the second microphone 10b for a certain period of time. .

Next, the second calculator 203 calculates the degree of conformance for the microphone pair (combination of the first microphone 10a and the second microphone 10b) selected in step S2 based on the phase difference distribution calculated in step S3. (Step S4).

Here, FIGS. 6 and 7 show examples of histograms representing the distribution of the phase differences calculated in step S3. The histograms shown in FIGS. 6 and 7 indicate the number of times (frequency) that each phase difference was calculated in step S3.

For example, when the phase difference is calculated based on one of the corresponding microphone signals of the first microphone 10a and the second microphone 10b divided for each unit time, 1 is added to the number of times the phase difference is calculated. be done. The histograms shown in FIGS. 6 and 7 can be generated by performing such processing each time the phase difference is calculated for each of the microphone signals divided by unit time.

FIG. 6 shows an example in which the phase difference distribution calculated in step S3 is relatively uniform. Specifically, in the example of the histogram shown in FIG. 6, the number of times the phase difference was calculated to be 0 degrees and the number of times the phase difference was calculated to be -150 degrees were the number of times other phase differences were calculated. It has been shown that there are more compared to According to this, for example, when the first microphone 10a and the second microphone 10b are arranged in front of the object (fan 3), the microphone signal for which the phase difference is calculated to be 0 degrees is the signal related to the target sound. It can be estimated that the microphone signal for which −150 degrees is calculated as the phase difference is a signal related to sound (directional noise) from a direction different from the target sound. In other words, when the phase difference distribution shown in FIG. 6 is obtained, the (combination of) the first microphone 10a and the second microphone 10b suppresses the directional noise and appropriately extracts the target sound. It can be said that this is a combination of microphones with a high degree of accuracy in estimating the correlation between microphones.

On the other hand, FIG. 7 shows an example in which the phase difference distribution calculated in step S3 is relatively uneven. Specifically, the histogram example shown in FIG. 7 indicates that the number of times each phase difference is calculated is approximately the same. According to this, compared with the example shown in FIG. 6, the microphone signal related to the target sound or directional noise cannot be estimated. As shown in FIG. 7, the factors that cause the phase difference distribution to vary are that the first microphone 10a and the second microphone 10b are affected not only by the target sound and directional noise input from a specific direction, but also It is provided at a position where it is easy to input reverberation sound from (that is, it is easily affected by reverberation sound). That is, when the phase difference distribution as shown in FIG. 7 is obtained, the first microphone 10a and the second microphone 10b (combination) can suppress directional noise and appropriately extract the target sound. It can be said that this is a combination of microphones in which the estimation accuracy of the correlation between microphones is not so high.

Therefore, in step S4 described above, the degree of variation in the distribution of the phase difference calculated in step S3 is used as an index, and when the distribution of the phase difference is not varied, a high degree of conformity is calculated, and the phase difference It is assumed that a low fitness is calculated when the distribution of is scattered. The degree of conformity (the degree of variation in the phase difference distribution) may be calculated using the variance and average value in the histogram described above, or may be calculated using an amount of information such as the Kullback-Leibler divergence. good too.

When the process of step S4 is executed, it is determined whether or not the processes of steps S2 to S4 have been executed for all microphone pairs among the plurality of microphones (step S5).

If it is determined that the process has not been executed for all microphone pairs (NO in step S5), the process returns to step S2 and the process is repeated. In the process of step S2 in this case, a microphone pair different from the already selected microphone pair is selected.

Note that the microphone pair in this embodiment is a combination of two microphones, but a different microphone pair (a combination of microphones) may be a combination in which at least one of the two microphones of the microphone pair is different. That is, when the first microphone 10a and the second microphone 10b have already been selected as a microphone pair as described above, in the processing of step S2 that is executed again, for example, the combination of the first microphone 10a and the third microphone 10c Alternatively, a combination of the second microphone 10b and the third microphone 10c may be selected as a microphone pair.

On the other hand, assume a case where it is determined that processing has been performed for all microphone pairs (YES in step S5). In this case, the processing of steps S2 to S4 has been performed for all the above-described microphone pairs.

The second selection unit 204 selects one microphone pair out of all the microphone pairs based on the degree of conformity for each microphone pair calculated in step S4 (step S6). In this step S6, for example, the microphone pair for which the highest compatibility is calculated is selected.

In other words, for example, when the highest matching degree is calculated for a microphone pair (combination) of the first microphone 10a and the second microphone 10b, the second selection unit 204 selects the first microphone 10a and the second microphone 10b. A combination of the microphones 10b is selected to collect the sound emitted from the object (that is, the sound generated inside the housing including the target sound).

When the process of step S6 is executed, the second selection unit 204 outputs microphone selection information indicating the microphone pair selected in step S6 (that is, the selection result of the second selection unit 204) to the collection unit 102 ( step S7).

As described above, the processing shown in FIG. 5 is executed before maintenance and inspection of the target object. to collect sounds. In other words, when the microphone selection information indicates the microphone pair of the first microphone 10a and the second microphone 10b (that is, when the microphone pair of the first microphone 10a and the second microphone 10b is selected in step S6), the collection unit 102 collects the microphone signal of the first microphone 10a and the microphone signal of the second microphone 10b among the microphone signals output by each of the first microphone 10a to the third microphone 10c. The microphone signal collected by the collection unit 102 is output for detecting anomalies in the object.

As described above, in this embodiment, a combination of a plurality of microphones from among the first to n-th microphones is selected according to the results of sound collection by the first to n-th microphones. That is, in the present embodiment, the distribution of the phase difference between the a-th microphone and the b-th microphone when the sound emitted from the object is collected by at least the a-th microphone and the b-th microphone, and the distribution of the phase difference between at least the c-th microphone and the and the distribution of the phase difference between the c-th microphone and the d-th microphone when collecting the sound emitted from the object with the d-th microphone, and the combination including at least the a-th microphone and the b-th microphone Select whether to collect sound emitted from an object or to collect sound emitted from a combined object including at least the cth and dth microphones.

The above a is a natural number from 1 to n, and b is a natural number from 1 to n that is different from a. Also, c is a natural number from 1 to n, and d is a natural number from 1 to n that is different from c. Also, the combination of the a-th microphone and the b-th microphone is a combination in which at least one microphone is different from the combination of the c-th microphone and the d-th microphone. That is, at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone.

With the above-described configuration, in the present embodiment, for example, at least three or more microphones provided for collecting sound emitted from the object are calculated for each combination of two microphones. Based on the phase difference distribution It is possible to select the optimum combination of microphones (a combination of microphones suitable for extracting the target sound).

Further, the distribution of the phase difference between the multiple microphones may be calculated for each frequency band of the microphone signal, and the optimum combination of microphones may be selected for each frequency band. That is, whether the sound emitted from the object is collected by the combination of the a-th microphone and the b-th microphone or the sound emitted by the object is collected by the combination of the c-th microphone and the d-th microphone is determined by, for example, , is selected based on the distribution of the phase difference in the first frequency band between the a-th microphone and the b-th microphone and the distribution of the phase difference in the first frequency band between the c-th microphone and the d-th microphone, and A further selection may be made based on a distribution of phase differences in a second frequency band between microphones a and b and a distribution of phase differences in said second frequency band between microphones c and d.

Furthermore, the distribution of phase differences between multiple microphones may be calculated for each period during which microphone signals are collected, and an optimum combination of microphones may be selected for each period. That is, whether the sound emitted from the object is collected by the combination of the a-th microphone and the b-th microphone or the sound emitted by the object is collected by the combination of the c-th microphone and the d-th microphone depends on, for example, the a-th microphone. selected based on the distribution of the phase difference in the first period between the microphone and the b-th microphone and the distribution of the phase difference in the first period between the c-th microphone and the d-th microphone, and A further selection may be made based on the distribution of the phase difference in the second period between the microphones and the distribution of the phase difference in the second period between the c-th microphone and the d-th microphone.

Moreover, the distribution of the phase difference between the a-th microphone and the b-th microphone described above is the phase difference for each frequency bin obtained by dividing the microphone signal related to the sound for a certain period of time collected by the a-th microphone and the b-th microphone. It may be distribution of appearance frequency. Similarly, the distribution of the phase difference between the c-th microphone and the d-th microphone is the appearance of the phase difference for each frequency bin obtained by dividing the microphone signal related to the sound collected by the c-th microphone and the d-th microphone for a certain period of time. It may be a frequency distribution.

Note that the distribution of the phase differences between the multiple microphones in the present embodiment indicates the correlation between the multiple microphones, and may be anything as long as it relates to the phase differences of the sounds input to the multiple microphones. Examples include coherence, distribution of time differences and phase differences of sounds input to a plurality of microphones, and the like. Further, these may be calculated based on the phase difference calculated per unit time, the phase difference calculated over a specific period, the phase difference calculated for each frequency component, the phase difference calculated over a specific or all frequency band It may be calculated based on the calculated phase difference, or may be calculated by combining these calculation results.

In the present embodiment, it is assumed that the object and a plurality of microphones are provided inside the housing as described above. By selecting a combination of small microphones, it is possible to collect sound with a combination of microphones provided at positions less affected by sounds echoed off the walls of the housing.
Also, in the present embodiment, two microphones are selected from a plurality of microphones, but the number of microphones to be selected may be three or more.

Here, the microphone signals collected and output by the system 10 according to the present embodiment (that is, the microphone signals output from the two selected microphones) are used to detect anomalies in the target object.

An abnormality detection system using the system 10 according to the present embodiment will be described below with reference to FIG. The object is assumed to be the fan 3 provided inside the housing of the uninterruptible power supply 1 as described above. As shown in FIG. 8, the anomaly detection system includes a system 10, a signal processing section 20, and an anomaly detection section 30 according to this embodiment.

Although the signal processing unit 20 and the abnormality detection unit 30 are provided outside the system 10 in FIG. 8 , they may be provided inside the system 10 .

The system 10 collects microphone signals output from (a combination of) two microphones selected by executing the processing shown in FIG.

The signal processing unit 20 receives a microphone signal collected by the system 10 (collecting unit 102) and performs beam forming processing on the microphone signal.

Here, the microphone signals output from the two microphones selected in the system 10 have little variation in the distribution of the phase difference, as shown in the histogram of FIG. It is possible to estimate with accuracy. Therefore, the signal processing unit 20 can extract (a signal related to) the operating sound of the fan 3 with directional noise suppressed from the microphone signal collected in the system 10 by beamforming processing.

The abnormality detection unit 30 detects an abnormality of the fan 3 based on the operation sound of the fan 3 extracted by the signal processing unit 20 . Specifically, a signal related to, for example, normal operation sound of the fan 3 is held in advance inside the abnormality detection unit 30, and the abnormality detection unit 30 detects the previously held signal (normal operation sound of the fan 3). ) and the signal extracted by the signal processing unit 20 (the current operating sound of the fan 3 ), the abnormality of the fan 3 can be detected. It should be noted that the detection processing described here is an example, and other processing can be executed if it is possible to detect an abnormality of the fan 3 using the operation sound of the fan 3 extracted by the beam forming processing described above. I don't mind if you do.

According to the anomaly detection system described above, by using the microphone signal output from the microphone selected by the system 10 (that is, the microphone suitable for extracting the target sound), the anomaly of the object is detected with high accuracy. It becomes possible to

In the present embodiment, the process shown in FIG. 5 (that is, the process of selecting the optimum microphone combination) was described as being executed before maintenance and inspection of the target object. In some cases, the object cannot be activated.

In such a case, for example, as shown in FIG. 9, the sound source 4 for calibration arranged in the housing of the uninterruptible power supply 1 instead of the target object (fan 3) emits sound. may be executed.

In general, since the wavelength differs depending on the frequency band, the influence of the reverberation also differs depending on the frequency band of the sound emitted from the object. For this reason, as the sound source 4 described above, one whose frequency band can be adjusted so as to emit sound corresponding to the sound emitted from the object (sound in the same frequency band as the sound emitted from the object) is used. is preferred. The sound source 4 may be, for example, a speaker that reproduces a sweep sound whose frequency changes. It may be something that can reproduce sound.

Also, noise such as the operation sound of the power conversion unit 2 may be generated using another sound source. In the case of converters and inverters, a sound source that can reproduce operating sounds that change according to their switching frequencies may be used.

As described above, by collecting the sound emitted from the sound source 4 arranged in the housing of the uninterruptible power supply 1 with a plurality of microphones, even if the object is not in operation, It is possible to select the optimum combination of microphones according to the situation.

In the present embodiment, the system 10 is implemented as a remote server device or the like (implemented on a server), but the system 10 is implemented integrally with a plurality of microphones. configuration may be used. In this case, the system 10 may be installed inside the housing of the uninterruptible power supply 1 . Moreover, although the system 10 has been described as including the processing unit 101 and the collection unit 102, at least one of the processing unit 101 and the collection unit 102 may be provided outside the system 10, for example.

Further, in the present embodiment, the object is described as being the fan 3 provided in the uninterruptible power supply 1, but the object is other parts provided in the housing of the uninterruptible power supply 1 (mechanically driven component) or the like.

In addition, in the present embodiment, the case where the target sound and the plurality of microphones are provided in the housing is affected by reverberant sound when collecting the target sound has been described, but reverberant sound and ambient noise exist. Examples of the environment include collecting operating sounds of production equipment covered with acrylic plates in a factory, or collecting sounds emitted from equipment such as pumps installed in the basement. The system 10 according to this embodiment can also be applied in such an environment.

Note that the target object in this embodiment may be parts of other devices (for example, MFPs, computers, robots, drones, etc.). Moreover, the system 10 according to the present embodiment may be used when collecting sounds (target sounds) emitted from objects that are not provided in the housing.

(Second embodiment)
Next, a second embodiment will be described. FIG. 10 is a schematic diagram showing an example of industrial equipment to which the system according to this embodiment is applied. In addition, in the present embodiment, as in the first embodiment described above, it is assumed that the industrial equipment is an uninterruptible power supply. Therefore, in FIG. 10, the same reference numerals are given to the same parts as in FIG. 1, and detailed description thereof will be omitted. Here, mainly different parts from FIG. 1 will be described. Also, since the hardware configuration and functional configuration of the system according to the present embodiment are as described above with reference to FIGS. 2 and 3, they will be described with reference to FIGS.

Here, in the above-described first embodiment, the system 10 selects a combination of two microphones suitable for extracting the target sound, for example, before maintenance and inspection of the uninterruptible power supply 1, and at the time of the maintenance and inspection A microphone signal output from the selected microphone is collected. According to the microphone signal collected in this way, the influence of reverberant sound or the like generated inside the housing of the uninterruptible power supply 1 is small, so the target sound can be appropriately extracted.

Note that once a combination of two microphones is selected in the system 10 as described above, sound can be collected using the two microphones at the time of subsequent maintenance and inspection. Such a process of selecting a combination of microphones does not have to be repeatedly executed.

However, for example, if the environment (temperature, humidity, etc.) in the housing of the uninterruptible power supply 1 changes, the speed of sound emitted in the housing may also change. The combination of microphones that has been changed is not necessarily the optimum combination even in the environment after the change.

Therefore, in the system 10 according to the present embodiment, as shown in FIG. 10, for example, a sensor 10d is provided in the housing of the uninterruptible power supply 1 for detecting the state of the environment (surrounding environment) in the housing. Assume that it has a configuration for executing a process of selecting a combination of microphones again based on the (state of) environment detected by 10d.

Note that the sensor 10d in the present embodiment includes, for example, a sensor (temperature sensor and humidity sensor) for detecting the temperature and humidity in the housing, but the environment around the sensor 10d (that is, inside the housing) is detected. Other sensors may be used as long as they can be used. Further, the number of sensors 10d provided in the housing of the uninterruptible power supply 1 may be one or plural.

FIG. 11 shows an example of the configuration of the processing unit 101 included in the system 10 according to this embodiment. As shown in FIG. 11 , the processing unit 101 in this embodiment includes the acquisition unit 205 and the determination including section 206 .

Acquisition unit 205 acquires a signal related to the environment detected by sensor 10d shown in FIG. 10 (hereinafter referred to as a sensor signal).

The determination unit 206 determines whether the environment inside the housing has changed based on the sensor signal acquired by the acquisition unit 205 . A determination result by the determination unit 206 is output to the second calculation unit 203 .

An example of the processing procedure of the system 10 according to the present embodiment will be described below with reference to the flowchart of FIG. 12 . It is assumed that the processing shown in FIG. 12 is executed, for example, during or before maintenance inspection of the uninterruptible power supply 1 (fan 3).

Here, in the present embodiment, it is assumed that the processing shown in FIG. 5 described above (hereinafter referred to as microphone selection processing) has been executed in the past. When this microphone selection process is executed, microphone selection information is output from the processing unit 101 , and the microphone selection information is held in the processing unit 101 .

For example, when maintenance and inspection of the uninterruptible power supply 1 is started, the acquisition unit 205 acquires a sensor signal from the sensor 10d (step S11). Here, it is assumed that the sensor signal acquired in step S11 is a signal relating to the temperature and humidity inside the housing. That is, the acquisition unit 205 can acquire the current temperature and humidity in the housing based on the sensor signal acquired in step S11.

Next, the determination unit 206 determines whether the environment inside the housing of the uninterruptible power supply 1 has changed based on the sensor signal acquired in step S11 (step S12).

Here, it is assumed that the determination unit 206 holds sensor signals related to the environment (temperature and humidity) within the housing when the microphone selection process was executed in the past, for example. That is, the determination unit 206 can acquire the temperature and humidity inside the housing when the microphone selection process was performed in the past based on the sensor signal held in the determination unit 206 .

In this case, the determination unit 206 determines based on the temperature and humidity (that is, the current temperature and humidity in the housing) acquired based on the sensor signal acquired in step S11 and the sensor signal held in the determination unit 206 (that is, the temperature and humidity in the housing when the microphone selection process was executed in the past) obtained by , it is determined that the environment within the housing has changed.

If it is determined that the environment has changed (YES in step S12), microphone selection processing is executed (step S13). Since this microphone selection process is the process shown in FIG. 5, detailed description thereof will be omitted here.

On the other hand, if it is determined that the environment has not changed (NO in step S12), the microphone selection process is not executed, and the process of FIG. 12 ends.

As described above, in this embodiment, when it is determined that the environment inside the housing has changed based on the sensor signal (that is, when the change in the environment is detected by the sensor 10d), the microphone is selected again. Processing is executed, and sound is collected by (a combination of) two microphones selected in the microphone selection processing during maintenance and inspection.

On the other hand, if it is determined that the environment inside the housing has not changed, the two microphones of the combination indicated by the microphone selection information held in the processing unit 101 (that is, the microphone selection performed in the past) Sound is collected using two microphones selected in the process).

According to such a configuration, it is possible to reselect the optimum combination of microphones according to changes in the environment inside the housing of the uninterruptible power supply 1 . Moreover, when there is no change in the environment, the microphone selected in the microphone selection process performed in the past can be used to collect sound, so the amount of processing in the system 10 can be reduced.

In the present embodiment, environmental changes based mainly on the temperature and humidity inside the housing have been described. , the microphone selection process is executed again. In order to detect such environmental changes, a sensor other than the temperature sensor and the humidity sensor described above may be provided in the housing as the sensor 10d.

Furthermore, in the present embodiment, the sensor 10d is provided in the housing of the uninterruptible power supply 1, but if the sensor 10d cannot be provided in the housing, the sensor 10d is It may be provided outside the housing.

Also, in the present embodiment, the microphone selection process is executed based on changes in the environment. Therefore, the microphone selection process may be executed again (that is, calibration may be performed) when aging deterioration of the microphone is detected. Furthermore, the microphone selection process may be executed when the object is replaced due to aged deterioration of the object or the like.

As described above, according to the present embodiment, it is possible to perform maintenance and inspection of equipment while changing the combination of microphones to be selected (that is, the optimum combination of microphones) in real time.

In addition, for example, it is also possible to have a configuration in which an optimum combination of microphones is selected in advance for each environment (temperature and humidity) in the housing, and the combination of microphones is switched according to the environment detected by the sensor 10d. is.

According to at least one embodiment described above, it is possible to provide a system, method, and program used to appropriately extract a target sound.

While several embodiments of the invention have been described, these embodiments have been presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and modifications can be made without departing from the scope of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, as well as the scope of the invention described in the claims and equivalents thereof.

DESCRIPTION OF SYMBOLS 1... uninterruptible power supply, 2... power conversion part, 3... fan, 4... sound source, 10... system, 10a-10c... microphone, 10d... sensor, 11... CPU, 12... non-volatile memory, 13... main memory, DESCRIPTION OF SYMBOLS 14... Communication device 20... Signal processing part 30... Abnormality detection part 101... Processing part 102... Collection part 201... First selection part 202... First calculation part 203... Second calculation part 204... 2nd selection part, 205... Acquisition part, 206... Judgment part.

Claims (22)

A system that collects sounds emitted from an object using first to n-th microphones (n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone an input means for inputting a selection result as to whether to collect emitted sounds;
collecting means for collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The system, wherein the second distribution includes a distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing the microphone signals related to sounds collected by the c-th microphone and the d-th microphone for a certain period of time. A system that collects sounds emitted from an object using first to n-th microphones (n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone an input means for inputting a selection result as to whether to collect emitted sounds;
collecting means for collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
based on a third distribution of the phase difference in the first frequency band between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first frequency band between the c-th microphone and the d-th microphone is selected by
based on a fifth distribution of the phase difference in the second frequency band between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second frequency band between the c-th microphone and the d-th microphone system to be further selected. A system that collects sounds emitted from an object using first to n-th microphones (n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone an input means for inputting a selection result as to whether to collect emitted sounds;
collecting means for collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. The system that collects is selected when a change in the environment is detected by the sensor. A system that collects sounds emitted from an object using first to n-th microphones (n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone an input means for inputting a selection result as to whether to collect emitted sounds;
collecting means for collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
Selection based on a third distribution of the phase difference in the first period between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first period between the c-th microphone and the d-th microphone is,
further based on a fifth distribution of the phase difference in the second period between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second period between the c-th microphone and the d-th microphone system selected. A system that selects a combination of a plurality of microphones from among the first to n-th microphones (where n is 3 or more) according to the results of collecting sounds emitted from an object by the first to n-th microphones,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting the sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or collecting the sound emitted from the object with a combination including at least the c-th microphone and the d-th microphone a processing unit that selects whether to collect sounds emitted from objects,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The system, wherein the second distribution includes a distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing the microphone signals related to sounds collected by the c-th microphone and the d-th microphone for a certain period of time. A system that selects a combination of a plurality of microphones from among the first to n-th microphones (where n is 3 or more) according to the results of collecting sounds emitted from an object by the first to n-th microphones,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting the sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or collecting the sound emitted from the object with a combination including at least the c-th microphone and the d-th microphone a processing unit that selects whether to collect sounds emitted from objects,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
based on a third distribution of the phase difference in the first frequency band between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first frequency band between the c-th microphone and the d-th microphone is selected by
based on a fifth distribution of the phase difference in the second frequency band between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second frequency band between the c-th microphone and the d-th microphone system to be further selected. A system that selects a combination of a plurality of microphones from among the first to n-th microphones (where n is 3 or more) according to the results of collecting sounds emitted from an object by the first to n-th microphones,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting the sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or collecting the sound emitted from the object with a combination including at least the c-th microphone and the d-th microphone a processing unit that selects whether to collect sounds emitted from objects,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. The system that collects is selected when a change in the environment is detected by the sensor. A system that selects a combination of a plurality of microphones from among the first to n-th microphones (where n is 3 or more) according to the results of collecting sounds emitted from an object by the first to n-th microphones,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting the sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or collecting the sound emitted from the object with a combination including at least the c-th microphone and the d-th microphone a processing unit that selects whether to collect sounds emitted from objects,
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
Selection based on a third distribution of the phase difference in the first period between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first period between the c-th microphone and the d-th microphone is,
further based on a fifth distribution of the phase difference in the second period between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second period between the c-th microphone and the d-th microphone system selected. The first distribution is calculated by collecting sound from a sound source that emits a sound corresponding to the sound emitted from the object with at least the a-th microphone and the b-th microphone,
The system according to any one of claims 1 to 8, wherein said second distribution is calculated by collecting sound from said sound source with at least said cth microphone and said dth microphone. The system according to any one of claims 1 to 8, wherein the object and the first to nth microphones are arranged within a housing. A method of collecting sounds emitted from an object using first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone inputting a selection of whether to collect emitted sounds;
collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result ;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The second distribution includes a frequency distribution of phase differences for each frequency bin obtained by dividing microphone signals relating to sounds collected by the c-th microphone and the d-th microphone for a certain period of time. A method of collecting sounds emitted from an object using first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone inputting a selection of whether to collect emitted sounds;
collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
based on a third distribution of the phase difference in the first frequency band between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first frequency band between the c-th microphone and the d-th microphone is selected by
based on a fifth distribution of the phase difference in the second frequency band between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second frequency band between the c-th microphone and the d-th microphone method to be further selected. A method of collecting sounds emitted from an object using first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone inputting a selection of whether to collect emitted sounds;
collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. How to collect is the method chosen when a change in the environment is detected by the sensor. A method of collecting sounds emitted from an object using first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone inputting a selection of whether to collect emitted sounds;
collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
Selection based on a third distribution of the phase difference in the first period between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first period between the c-th microphone and the d-th microphone is,
further based on a fifth distribution of the phase difference in the second period between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second period between the c-th microphone and the d-th microphone method selected. A method for selecting a combination of a plurality of microphones from among the first to n-th microphones according to the results of collecting sounds emitted from an object by the first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone; , a natural number different from c) a second goodness of fit based on the degree of variation of a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone; , and
Based on the first goodness of fit and the second goodness of fit, a combination including at least the a-th microphone and the b-th microphone is used to collect sounds emitted from the object, or at least the c-th microphone and the selecting whether to collect sound emanating from the object in combination with a d th microphone;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The second distribution includes a frequency distribution of phase differences for each frequency bin obtained by dividing microphone signals relating to sounds collected by the c-th microphone and the d-th microphone for a certain period of time. A method for selecting a combination of a plurality of microphones from among the first to n-th microphones according to the results of collecting sounds emitted from an object by the first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone; , a natural number different from c) a second goodness of fit based on the degree of variation of a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone; , and
Based on the first goodness of fit and the second goodness of fit, a combination including at least the a-th microphone and the b-th microphone is used to collect sounds emitted from the object, or at least the c-th microphone and the selecting whether to collect sound emanating from the object in combination with a d th microphone;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
based on a third distribution of the phase difference in the first frequency band between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first frequency band between the c-th microphone and the d-th microphone is selected by
based on a fifth distribution of the phase difference in the second frequency band between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second frequency band between the c-th microphone and the d-th microphone method to be further selected. A method for selecting a combination of a plurality of microphones from among the first to n-th microphones according to the results of collecting sounds emitted from an object by the first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone; , a natural number different from c) a second goodness of fit based on the degree of variation of a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone; , and
Based on the first goodness of fit and the second goodness of fit, a combination including at least the a-th microphone and the b-th microphone is used to collect sounds emitted from the object, or at least the c-th microphone and the selecting whether to collect sound emanating from the object in combination with a d th microphone;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. How to collect is the method chosen when a change in the environment is detected by the sensor. A method for selecting a combination of a plurality of microphones from among the first to n-th microphones according to the results of collecting sounds emitted from an object by the first to n-th microphones (where n is 3 or more),
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone; , a natural number different from c) a second goodness of fit based on the degree of variation of a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone; , and
Based on the first goodness of fit and the second goodness of fit, a combination including at least the a-th microphone and the b-th microphone is used to collect sounds emitted from the object, or at least the c-th microphone and the selecting whether to collect sound emanating from the object in combination with a d th microphone;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The sound emitted from the object is collected by a combination including at least the a-th microphone and the b-th microphone, or the sound emitted from the object is collected by a combination including at least the c-th microphone and the d-th microphone. whether to collect
Selection based on a third distribution of the phase difference in the first period between the a-th microphone and the b-th microphone and a fourth distribution of the phase difference in the first period between the c-th microphone and the d-th microphone is,
further based on a fifth distribution of the phase difference in the second period between the a-th microphone and the b-th microphone and a sixth distribution of the phase difference in the second period between the c-th microphone and the d-th microphone method selected. The first distribution is calculated by collecting sound from a sound source that emits a sound corresponding to the sound emitted from the object with at least the a-th microphone and the b-th microphone,
A method according to any one of claims 11 to 18, wherein said second distribution is calculated by collecting sound from said sound source with at least said cth microphone and said dth microphone. The method according to any one of claims 11 to 18, wherein said object and said first to nth microphones are arranged in a housing. A program for collecting sounds emitted from an object using first to n-th microphones (where n is 3 or more),
to the computer,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone, and at least the c-th (c is a natural number from 1 to n) and the d-th (d is from 1 to n) a natural number different from c) a second goodness of fit based on the degree of variation in a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone and collecting sound emitted from the object with a combination including at least the a-th microphone and the b-th microphone, or from the object with a combination including at least the c-th microphone and the d-th microphone inputting a selection of whether to collect emitted sounds;
collecting sounds emitted from the object using a plurality of microphones combined based on the input selection result;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The program, wherein the second distribution includes distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing microphone signals relating to sounds collected by the c-th microphone and the d-th microphone for a certain period of time. A program for selecting a combination of a plurality of microphones from among the first to n-th microphones (where n is 3 or more) according to the results of collecting sounds emitted from an object by the first to n-th microphones. hand,
to the computer,
The a-th (a is a natural number from 1 to n) microphone and the b-th (b is a natural number from 1 to n different from a) microphone when collecting the sound emitted from the object. A first goodness of fit based on the degree of variation in the first distribution of the phase difference between the microphone and the b-th microphone; , a natural number different from c) a second goodness of fit based on the degree of variation of a second distribution of the phase difference between the c-th microphone and the d-th microphone when the sound emitted from the object is collected with a microphone; , and
Based on the first goodness of fit and the second goodness of fit, a combination including at least the a-th microphone and the b-th microphone is used to collect sounds emitted from the object, or at least the c-th microphone and the selecting whether to collect sound emitted from the object in combination including a d th microphone;
at least one of the a-th microphone and the b-th microphone is different from both the c-th microphone and the d-th microphone;
The first goodness of fit is calculated as a high value when the first distribution of the phase difference is not varied, and is calculated as a low value when the first distribution of the phase difference is varied,
For the second goodness of fit, a high value is calculated when the second distribution of the phase difference is not varied, and a low value is calculated when the second distribution of the phase difference is varied,
The first distribution includes a distribution of appearance frequencies of phase differences for each frequency bin obtained by dividing microphone signals related to sounds collected by the a-th microphone and the b-th microphone for a certain period of time,
The program, wherein the second distribution includes distribution of frequency of occurrence of phase differences for each frequency bin obtained by dividing microphone signals relating to sounds collected by the c-th microphone and the d-th microphone for a certain period of time.

Images