JPH08292252A - System for searching for sound source - Google Patents

Abstract

PURPOSE: To search for a position of a sound source by setting a plurality of microphones at measuring points, obtaining/storing sound pressure data, obtaining an reverse propagation time from a distance of each measuring point to a plurality of reproduction points on a reproduction face, and obtaining for each reproduction point a reproduction formula of the sound pressure data converted by being advanced by the time. CONSTITUTION: When a sound source O is present at a point on a sound source face S, sound pressure waveforms P1 (t), P2 (t), Pn (t) are obtained, for example, at measuring points X1 , X2 , Xn among the measuring points X1 ...Xn on a sound pressure-measuring face K. Measuring times of the waveforms at a reference position (g) are respectively t1 , t2 , tn . Reproduction points R1 ...Rm are set on a reproduction face R and each of distances between the reproduction points and measuring points X1 ...Xn is obtained, thereby to calculate reverse propagation distances r11 ...rn1 to sound pressure P1 (t)...Pn (t). Then, sound pressures P'11 (t), P'21 (t)...P'n1 (t) are obtained by converting inverse propagation times of sound from the measuring points X1 ... to, e.g. the reproduction point R1 . A reproduction formula is operated for the reproduction point R1 . When the reproduction formula is applied to the whole of the reproduction face R, a formula I is obtained. A position of the sound source is searched in this manner.

Description

Detailed Description of the Invention

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound source search system, and more particularly to a device for searching the position and size of a noise source in a vehicle engine or the like.

[0001]

2. Description of the Related Art Conventionally, various methods for searching a sound source have been proposed. One of them has two scanning microphones arranged at a fixed interval of about 6 to 50 mm in front of and behind the sound source. Is installed in the near field of the sound source and the microphone output is obtained while moving both scanning microphones in parallel,
There is a method for obtaining the sound intensity of the sound source (“sound intensity method”).

In this method, the sound source is captured as acoustic energy flowing in a unit time in a unit area, the average value of the sound pressures measured by both scanning microphones is p (t), and the particle velocity (a value proportional to the sound velocity, per unit distance. U (t)
Then, by measuring the vector quantity shown in the following equation (1) as the time average of the product of both, it is found that the sound source is located on the extension line of both scanning microphones where this vector quantity becomes maximum. To do.

[Equation 1] The particle velocity u (t) represents a change in phase.

In such a sound source search method based on the sound intensity method, since the measurement results at the scanning microphone measurement points are independent from each other, there is no correlation. Therefore, when the direction of the sound source is not parallel to the extension line direction of the microphone. There is a problem that a correct search cannot be performed.

On the other hand, a sound source search method based on the "acoustic holography method" has been studied before the sound intensity method described above, and one microphone for scanning a measurement surface at a certain distance from the sound source and another The sound source is searched for by obtaining the sound pressure intensity and phase of the two microphone output signals with a reference (reference) microphone fixed at one appropriate position.

To explain this in a little more detail, in this method, the microphone output at each position is obtained by scanning the microphone M1 on the measurement surface K in the horizontal and vertical directions as schematically shown in FIG. Can be done,
The following reproduction formula (2) showing the back propagation formula of the sound wave which is the basis for obtaining these microphone outputs while considering these microphone outputs and the sound pressure output of the reference microphone,

[Equation 2] Playback surface (sound source surface) R played by the microphone M1
Complex volume velocity amplitude Q at one point R ₁ (x, y, 0)
(X, y, 0) is calculated. However, r in the above equation (2) represents the distance between one point P (x, y, zo) on the measurement surface K and one point R ₁ (x, y, 0) on the reproduction surface R. Note that P (x, y, zo) in the equation (2) represents the complex sound pressure obtained by the microphone M1.

Then, the microphone M1 moves horizontally on the measuring surface K.
The complex volume velocity amplitude Q (x, y, 0) of one point R ₁ (x, y, 0) on the reproduction surface R obtained by the output of each microphone position when scanned in the vertical direction is added, and this reproduction is performed. By taking the point R ₁ at all points on the reproduction surface R and similarly adding all the complex volume velocity amplitudes Q (x, y, 0) from the measurement surface K, the point with the largest amplitude Q is taken as the sound source. To be searched.

FIG. 1 illustrates this point more specifically.
7 and FIG. 18, first, in FIG. 17A, when the reproduction point R ₁ on the reproduction surface R is reproduced as the sound source S, the sound is picked up at the microphone position M1-1 of the microphone M1 as shown in this figure. The reproduced sound and the reproduced sound picked up at the microphone position M1-2 have the same distance between the sound wave propagation path and the back propagation path and have the same phase. As shown, the absolute value of the combined vector of the reproduced vectors (for example, Q1 to Q3) becomes maximum when the reproduction point R ₁ on the reproduction surface R is the sound source S, that is, at the position of the sound source S.

On the other hand, when the reproduction point R ₁ is reproduced at a position on the reproduction surface R slightly deviated from the sound source S as shown in FIG. 17 (b), the propagation path at the microphone position M1-1 as shown in the figure. The back propagation path at the time of reproduction becomes longer, and at the microphone position M1-2, the back propagation path at the time of reproduction becomes shorter than the propagation path.

Therefore, there is a phase difference between the reproduced sound picked up at the microphone position M1-1 and the reproduced sound picked up at the microphone position M1-2.
As shown in (b), the combined vector (Q1 to Q3) at the reproduction point R ₁ is smaller than the combined vector shown in FIG.

By thus setting a large number of reproduction points R ₁ on the reproduction surface R, when the reproduction point corresponds to the sound source position, the maximum combined vector is taken as shown in FIG. Since a small composite vector is obtained at the position of (b) in the same figure, the reproduction point in the former case is searched as a sound source.

[0011]

In the case of a sound source search using the acoustic holography method as described above, the sound pressure data is converted into a specific frequency f ₀ as shown in the above equation (2). The wave number k at (ω = 2πf ₀ ) is used and is formulated in the frequency domain.

For this reason, the measured sound pressure data must be obtained in the frequency domain from the time of measurement, or it must be converted into frequency data by a technique such as fast Fourier transform (FFT) after measurement in the time domain. Caused the following problems.

The fast Fourier transform is based on the premise that the sound pressure data is temporally stationary.
It cannot be applied to a sound wave whose frequency component changes with the passage of time. That is, it cannot be applied to search for a sound wave of a transient phenomenon.

Since the above-mentioned sound source reproduction formula (2) is applied to each frequency component, it is not possible to search based on the difference in characteristics of sound waves such as the difference in tone color of musical instruments. For example, in an orchestra. It is not possible to detect only the sound of the violin from and detect its position. This means that the position where the sound wave is generated cannot be detected for the abnormal noise that is a problem of the sound of the engine or the vehicle.

Recently, a sound source search method based on the "near-field acoustic holography method" has been studied. This method exists only in a mode in which a sound wave diffuses far from the sound source and in the vicinity of the sound source, and is attenuated at a short distance. The latter sound wave is also included in the search target, paying attention to the fact that there is a mode
In this case, the distance between the sound source and the measurement surface is 1 / 2λ to 1λ
If the sound source is a heat generating device (for example, an engine manifold), the microphone cannot be installed at a close distance, which is not suitable for measurement.

Further, as an application example of the sound intensity method, in addition to two scanning microphones, one reference microphone is used to detect the particle velocity of the sound wave propagating rearward from the measurement surface. There is a "proximity particle velocity method" that predicts the directivity and sound pressure level of the sound source. However, this method does not perform sound source search and cannot be used.

Therefore, according to the present invention, by improving the acoustic holography method as described above, reproduction calculation is performed using the sound wave waveform on the time axis measured on the measurement surface, and thus the sound source can be searched. With the goal.

[0018]

[Means for Solving the Problems]

[1] In order to achieve the above object, in the sound source search method according to the present invention, a plurality of microphones arranged at measurement points for the sound source and sound pressure data measured at each measurement point by each microphone are input and stored. Then, the reproduction formula of the sound pressure data converted by advancing the back propagation time obtained from each distance to a plurality of reproduction points arbitrarily set on the predetermined reproduction surface of each measurement point is calculated for each reproduction point. Arithmetic means for searching the position of the sound source.

First, the concept of the sound source search method according to the present invention [1] will be described with reference to FIGS. 1 to 4.

FIG. 1 is a diagram showing the propagation state of a sound wave on the time axis. When a sound source O having a sound pressure waveform as shown in FIG. 2A exists at a certain point on the sound source surface S. , Of the measurement points X _1, X _2, ..., X _n on the sound pressure measurement surface K, for example, X _1, X ₂ ,
The sound pressure waveforms P ₁ (t) _, P ₂ (t) _, P _n (t) (these are processed as data) as shown in FIGS. 2B to 2D are obtained for X _n . Suppose The reference position of the sound pressure waveform is g.

These sound pressure waveforms P ₁ (t) _, P ₂ (t) _, P _n (t)
Since the measurement time at the position g where the criteria are respectively as shown _{t 1, t 2, t n} , respectively from the time t ₀ the wave is generated from the sound source _{O Δt 1 (= t 1 -t} 0 ), Δt ₂ (= t _2-
t ₀ ), Δt _n = t _n -t ₀ ).

Then, these delay times Δt₁, Δt₂, Δ
t_nIs sound source O and measurement point X_1,X₂, X _nVirtual distance r₁,
r₂, r_nAre each divided by the sound velocity c
It

Next, as shown in FIG. 3, reproduction points R _1, R _2, ..., R _m are set on the reproduction surface R. As a result, the measurement surface K
At the measurement points X _1, X _2, ..., X _n and the reproduction points R _1, R _{2 ,} .
The distance to each R _m can be uniquely determined.

For example, the reproduction point R₁For measurement point X_1,
X_2,…, X_nDistance to each of₁₁, r _{twenty one}, r_31,…, R_n1But
Both coordinates and the distance between the preset reproduction plane R and the measurement plane K
Distance Z₀Easily obtained from the playback point R₁Smell
Measurement point X when performing regeneration calculation_1,X_2,…, X_nEach of
Sound pressure P measured at₁(t)_,P₂(t)_,…_,P_nsound for (t)
Corresponds to the back propagation distance of.

Therefore, these back propagation distances r ₁₁ , r ₂₁ , r
By reaching _31, ..., r _n1 by the sound velocity c, the measurement point X _1, X _2, ..., X _n reaches the reproduction point R ₁ (delay).
Time is required.

Assuming now that the reproduction point R ₁ shown in FIG. 3 corresponds to the position of the sound source O in FIG. 1, the time to reach the distance r ₁₁ from the measurement point X ₁ to the reproduction point R ₁ is , Exactly the delay time Δt ₁ caused by the distance r _{1 in} FIG. ₁ (FIG. 2B)
Reference).

Therefore, with respect to the propagation of the sound wave from the measurement point X ₁ to the reproduction point R ₁ , it suffices to proceed in the opposite direction by the delay time Δt ₁ .

This state is shown in FIG. 4, and when the sound pressure waveform P ₁ (t) at the measurement point X ₁ shown in FIG. 4A is used as a reference, as shown in FIG. It can be seen that the process is advanced by time Δt ₁ .

In this way, the measurement points X _1, X _2, ..., X _n
Sound pressures P ₁₁ ^' (t) _, P ₂₁ ^' (t) _, ... _, P _n1 ^' (t) obtained by converting the respective back propagation times of the sounds from the sound source to the reproduction point R ₁ are represented by the following equations. It

(Equation 3) Therefore, the reproduction formula for the reproduction point R ₁ is given by the following formula.

[Equation 4] When this is executed on the entire reproduction surface R, the following equation is obtained.

(Equation 5) In the above equations (4) and (5), the sound pressure P ₁ ^' (t) _, P
₂ ^' (t) _, ... _, P _n ^' (t) are respectively r ₁₁ , r ₂₁ , r _31, ..., r _n1
The reason for dividing by is to adjust the amplitude change due to the difference in back propagation distance (to restore the original amplitude by the attenuated amplitude).

The principle of sound source search in this method is shown in FIG.
This will be described with reference to (a), (b) and FIGS. First, when the reproduction point and the sound source point match as shown in FIG.
The reproduced waveform based on the equation (4) has a different amplitude but a uniform time (FIG. 5 (a)), and the synthesized waveform is as shown in FIG. 5 (b).

When a reproduction point other than the sound source point is reproduced, the actual sound wave propagation distance differs from the back propagation distance used in the reproduction calculation, as shown in FIG. 17 (b). Therefore, the back propagation equation (4)
The waveforms reproduced by are not aligned in time as shown in FIG.

The waveform obtained by adding and synthesizing the waveforms is as shown in FIG. 9D, and the magnitude Pb thereof is always smaller than Pa in FIG.

As a result, the reproduction value becomes maximum at the position of the sound source.

In this way, the above equations (4) and (5)
Is a simple process of moving the measured sound pressure P (t) on the time axis for the back propagation time Δt in the reproduction calculation, and since the time-frequency conversion is not necessary at all, the processing speed is also increased.

In the reproduction formula thus obtained, by specifying the time t, the reproduction point giving the largest value at that time can be set as the sound source, or by the method as described with reference to FIG. You may search for a sound source.

[2] In the present invention, at least one scanning microphone arranged with respect to a sound source having periodicity,
A means for moving the scanning microphone, a reference microphone, and sound pressure data and reference sound pressure data measured at each measurement point by the scanning microphone and the reference microphone which are moving are input and stored, respectively, Each back propagation time is calculated from each distance to a plurality of playback points arbitrarily set on a predetermined playback surface, and the measured sound pressure data and the reference sound pressure data corrected by advancing by the back propagation time are sequentially After frequency conversion in a predetermined time window that is shifted at regular intervals, the reference sound pressure data is used to standardize the phase of the measured sound pressure data, and the phase standardized sound pressure data is reproduced by a reproduction formula. And a calculation unit that searches for the position of the sound source by obtaining and integrating each reproduction point.

That is, in the case of the above [1], microphones for measuring the sound pressure from the sound source are individually arranged at the measurement points X _1, X _2, ..., X _n on the measurement surface K. However, in this case, many microphones are required, and the measurement equipment must be large-scale.

Therefore, the present invention [2] provides a sound source search method in which the scanning microphone is moved on the measurement surface K by the moving means.

Here, the actual tone color of the sound wave is determined by the temporal change in the intensity of the frequency component of the sound wave, and the phase relationship of each frequency component of the sound wave does not affect the tone color.

Therefore, even if the sound waves of the same tone color are heard by hearing with the ear, the phase relationships of the frequency components of the sound waves are not always the same.

In the case of simultaneously measuring all the measuring points according to the above [1], the estimated sound wave converted from the data measured at each measuring point by the back propagation time is of course the same sound wave and has a constant phase relationship. is there.

However, when the moving means sequentially measures at each measuring point, there is no guarantee that the phase relationship is constant, and the processing as acoustic holography cannot be performed as it is.

Therefore, in the present invention [2], the sound source must have periodicity, and the sound pressure measured at each moving measurement point using a reference (reference) microphone for dividing the measurement interval. It is necessary to standardize the phase of data.

FIG. 6 shows a calculation processing flow of the calculation means in the sound source search system according to the present invention [2]. First, the measured sound pressure data S (t) is obtained from the scanning microphone and is referred to together (step S1). The reference sound pressure data U (t) is obtained from the microphone (step S2).

Then, these measured sound pressure data S (t)
And the propagation time of the reference sound pressure data U (t) are corrected (steps S3 and S4).

This is shown in FIG. 1 in the above-mentioned present invention [1].
~ As described with reference to Fig. 4, the measured sound pressure data S (t)
In the case of, the back propagation time Δt _s is calculated, and the reference sound pressure data U
In the case of (t), the back propagation time Δt _r is calculated, and the back propagation time is converted as shown in the above equation (3) to calculate the sound pressure data S (t + Δt _S ) and U (t + Δ).
t _r ).

It should be noted that, as shown in the above equation (4), the attenuation of the amplitude due to the difference in the back propagation distance may be taken into consideration.

Since these sound pressure data S (t + Δt _S ) and U (t + Δt _r ) are data in the time domain, the phase of the sound wave is not considered at all, and therefore frequency conversion (for example, FFT) is performed. To extract sample data (unit time width) t _n and element data (frequency / amplitude / phase) subdivided into unit frequency widths (steps S5, S).
6).

FIG. 7 shows a three-dimensional display example of such element data (sound pressure spectrum), and the spectrum at time t _n is expressed by the following equation.

(Equation 6) Therefore, the sound pressure data S (k) and U (k) obtained in these steps S5 and S6 are respectively expressed by the following equations.

(Equation 7)

(Equation 8) Note that such a three-dimensional display of FIG. 7 is displayed at time t _n, frequency f
_When displayed using _{k and} amplitude | P _k |, the sonargram shown in FIG. 8 is obtained.

That is, in steps S5 and S6, not only the frequency conversion is performed, but the original sound wave T as shown in FIG.
Is a constant time window (t1 to t) having three unit time widths t.
3, t2 to t4, t3 to t5, ...
1, frequency conversion is performed on T1, T2 and T3.
The frequencies f1 and f divided by the frequency width f as shown in 0
2, f3, ..., F8 can be represented by respective intensities (line widths).

When frequency conversion is performed in a predetermined time window that is shifted every unit time in this way, unlike the conventional example in which FFT frequency conversion is simply performed, there is no loss of information on the temporal change of the frequency component of the sound wave. The sonargram display as shown in FIGS. 8 and 10 can be performed, and the sound source sound wave can be estimated.

It should be noted that such three-dimensional display or sonargram display itself is not essential to the present invention, and is only a preferable mode that is effective when visually observed at this stage.

Next, the sound pressure data S (k) and U which are frequency-converted in this way and are expressed by the above equations (7) and (8) are obtained.
Phase standardization (phase correction) of (k) is performed (step S
7).

The phase-corrected sound pressure data at this time is
It is given by the following formula.

[Equation 9] FIG. 11 shows the sound pressure data S ′ represented by the above equation (9).
It is a complex representation of (k) and is standardized to the phase indicated by the dotted line.

Thereafter, the calculating means obtains and integrates the sound pressure data whose phase has been corrected as described above for all the measurement points where the scanning microphone is moved by the moving means (step S8), and the integrated result is obtained. It is expressed as the following equation.

[Equation 10] This integration result is as described in connection with FIG. 5 in the present invention [1], and when the reproduction point is not the sound source point,
As shown in FIG. 12A, when the reproduction point is the sound source point, it is understood that the value becomes larger than that in the case of FIG. 12A as shown in FIG. Is the sound source point.

The above integration result is displayed in the frequency domain as it is in a three-dimensional display as shown in FIG. 8 (sonargram display).
Alternatively, (step S9), or reverse frequency conversion may be performed in a predetermined time window that is sequentially shifted at regular time intervals to return to the time domain, and the time waveform may be output (steps S10 and S11).

[0057]

FIG. 13 shows an embodiment of a sound source search system according to the present invention [1]. In this embodiment, a measurement surface K is provided with a plurality of microphones and 40 microphones are arranged. Scanning microphones M _{11 to} M ₅₈ are arranged in a matrix.

These scanning microphones M₁₁~ M₅₈Is the sound source S
When receiving the sound pressure from
A₁₁~ A₅₈To each amplifier A₁₁
~ A ₅₈A / D converter C provided corresponding to₁₁~ C
₅₈Are converted from analog signals to digital signals at
It

The respective output signals of the A / D converters C _{11 to} C ₅₈ are sent to the holographic operation unit HC composed of a personal computer, a workstation or the like, and the operation according to the above acoustic holographic operation formulas (3) to (5) is performed. Done.

Then, the calculation result in the holography calculation unit HC is sent to the display D such as a CRT or a plotter and the sound source distribution map is displayed.

FIG. 14 shows an embodiment of the sound source search system according to the present invention [2]. Instead of using the scanning microphones arranged at multiple points as in the embodiment shown in FIG. 13, the measurement plane K is used. One scanning microphone M1 and a fixed reference microphone M2
The scanning microphone M1 is attached to the support rods B1 and B2 so that the scanning microphone M1 is horizontally moved by the microphone traverse device TVS on the measurement surface K with respect to the sound source S.

Since the support rod B1 can move on the support rod B2, the scanning microphone M1 is eventually used.
Will be scanned in the horizontal and vertical directions on the measurement surface K, respectively.

The sound pressure outputs of the scanning microphone M1 and the reference microphone M2 are amplified by amplifiers A1 and A2, respectively, and then A
After being converted into digital signals by the / D converters C1 and C2, the holographic calculation unit HC in a computer PC such as a personal computer or a workstation performs the acoustic holographic calculation as described above, and displays it on the display D such as a CRT or plotter. The sound source distribution map will be displayed.

Microphone traverse device TVS
The controller CNT for controlling the
The moving speed of the microphone traverse device TVS controlled by the control software unit CS provided in the C is adjusted beforehand with the speed at which the output of the microphone M1 is sampled.

[0065]

As described above, according to the sound source search method of the present invention [1], the sound pressure data measured by the microphones respectively arranged at the multiple measurement points with respect to the sound source are input and stored, and each measurement is performed. A reproduction formula of the sound pressure data obtained by calculating the respective back propagation distances of the sound waves from the sound pressure data of the points to the plurality of reproduction points set on the reproduction surface, and advancing and converting the back propagation time obtained from the back propagation distances. Since it is configured to search the position of the sound source by obtaining for each playback point, it is possible to perform a simple process of moving the measured sound pressure on the time axis for the back propagation time in the playback calculation, and no frequency conversion is required. Therefore, the processing speed becomes faster. It can also be applied to the case where the sound source emits a transient sound wave, and the waveform of the sound wave can be reproduced at the same time as the position of the sound source.

Further, in the present invention [2], at least one scanning microphone is moved with respect to a sound source having periodicity, and sound pressure data and reference sound pressure data measured at each measurement point by the scanning microphone and the reference microphone. Then, the respective back propagation times are calculated from the above, and the measured sound pressure data and the reference sound pressure data that are corrected by advancing only the back propagation time are frequency-converted in a predetermined time window that is sequentially shifted at regular time intervals, Phase of the measured sound pressure data is standardized using the reference sound pressure data, and the sound source position is searched by obtaining and integrating the phase standardized sound pressure data for each reproduction point by a reproduction formula. You may comprise, and in this case, there exists an effect as shown in FIG.

That is, for example, in the embodiment of FIG. 14, a measurement surface of 1.2 m in the horizontal direction × 0.9 m in the vertical direction is installed at a position 0.5 m from the sound source, and a horizontal direction 7 is placed on this surface.
Points, seven measurement points in the vertical direction are arranged, and two speakers as sound sources are arranged at a distance of about 0.3 m, and one side emits a sound wave (created by an electronic musical instrument or the like) searched by the present invention, and the other side. In the case of a specific example in which a sound wave that becomes noise is emitted from the speaker of No. 2, since these two sound waves have the same frequency component and are different only in the temporal change, they cannot be distinguished by the conventional acoustic holography method. .

FIG. 15A shows a sound wave emitted from one sound source in a three-dimensional display as shown in FIG. 8, and FIG. 15B shows a search result of the sound wave (step S in FIG. 6).
9). Further, FIG. 7C shows a sound wave simply measured at a position 0.5 m in front of the sound source, and comparing FIGS. 11A and 11B, in FIG. Although the influence of the other sound source remains slightly in the 0.3 KHz band), it can be seen that the intensity change and the decay time of each frequency component in the other frequency bands are reproduced almost accurately.

Further, in the figure (c), the sound waves of two sound sources appear at the same time, and the sound waves are completely different from those in the figure (a), and the search degree of the sound wave is far more than the result of the figure (b). I know that is bad.

[Brief description of drawings]

FIG. 1 is a diagram showing propagation of a sound wave on a time axis for explaining a working principle of a sound source search method according to the present invention.

FIG. 2 is a waveform diagram for explaining the operation principle of the sound source search system according to the present invention corresponding to FIG.

FIG. 3 is a diagram for explaining an acoustic holography method on a time axis in order to explain the operation principle of the sound source search method according to the present invention.

FIG. 4 is a waveform diagram for explaining the operation principle of the sound source search system according to the present invention corresponding to FIG.

FIG. 5 is a diagram showing a reproduced waveform by a sound source search method according to the present invention.

FIG. 6 is a flowchart showing a search processing process by the sound source search method according to the present invention [2].

FIG. 7 is a diagram showing three-dimensionally a sound pressure spectrum at a sample time in the sound source search method according to the present invention [2].

FIG. 8 is a sonargram diagram in which sound pressure data is three-dimensionally displayed in time / frequency / amplitude in the sound source search method according to the present invention [2].

FIG. 9 is a waveform diagram for explaining a sound wave sampling method when performing time-frequency conversion in the sound source search method according to the present invention [2].

FIG. 10 is a sonargram diagram obtained by performing time-frequency conversion in the sound source search method according to the present invention [2].

FIG. 11 is a complex display diagram for explaining phase correction (standardization) by a reference microphone in the sound source search method according to the present invention [2].

FIG. 12 is a diagram showing integrated vectors obtained in the sound source search system according to the present invention [2].

FIG. 13 is a diagram showing an embodiment (multipoint microphone system) of a sound source search system according to the present invention.

FIG. 14 is a diagram showing an embodiment (microphone moving method) of a sound source searching method according to the present invention.

FIG. 15 is a graph diagram for explaining the effect of the sound source search method according to the present invention [2].

FIG. 16 is a diagram for explaining sound pressure reproduction by a conventional acoustic holography method.

FIG. 17 is a diagram showing a state in which a sound source is reproduced by a conventional acoustic holography method.

FIG. 18 is a diagram showing a composite vector on the reproduction surface obtained by the acoustic holography method.

[Explanation of symbols]

During M1 scanning microphone M ₁₁ ~M ₅₈ Microphone M2 reference microphone S source K measurement plane TVS microphone traverse device HC holography calculation section view, the same reference numerals denote the same or corresponding parts.

Claims (4)

[Claims] 1. A plurality of microphones arranged at a measurement point for a sound source and sound pressure data measured by each microphone at each measurement point are input and stored, and are set arbitrarily on a predetermined reproduction surface from each measurement point. And a calculation means for searching the position of the sound source by obtaining a reproduction formula of the sound pressure data converted by advancing the back propagation time obtained from each distance to the plurality of reproduction points, for each reproduction point. Characteristic sound source search method. 2. At least one scanning microphone arranged with respect to a sound source having a periodicity, a means for moving the scanning microphone, a reference microphone, and each measurement by the moving scanning microphone and the reference microphone. Input and store sound pressure data and reference sound pressure data measured at each point, and calculate each back propagation time from each distance from each measurement point to multiple playback points arbitrarily set on the specified playback surface. Then, the measured sound pressure data and the reference sound pressure data corrected by advancing by the back propagation time are frequency-converted in a predetermined time window that is sequentially shifted at regular intervals, and then the reference is made to the measured sound pressure data. The sound pressure data is used to standardize the phase, and the sound pressure data obtained by the phase standardization is calculated for each reproduction point by a reproduction formula, and is integrated to calculate the position of the sound source. Sound source search system to. 3. A sound source search system according to claim 2, wherein said calculation means outputs the result of said integration as a three-dimensional display of amplitude-time-frequency. 4. The sound source according to claim 2, wherein the calculation means inversely frequency-converts the result of the integration in a predetermined time window that is sequentially shifted at regular time intervals and outputs it as a time waveform. Search method.