JP5728378B2 - Noise reduction device - Google Patents

Description

  The present invention relates to a noise reduction device, and more particularly to an active control type noise reduction device that reduces noise by active control.

  Conventionally, environmental standards have been established for road traffic noise and railway noise, and road construction that does not meet the standards and operation of high-speed vehicles that do not meet the standards are not allowed.

  From such a background, sound barriers are actively developed, and active noise control (ANC) technology for reducing diffracted sound from the sound barriers by active control is being developed.

  As ANC technology, a plurality of point sound sources are arranged in a straight line to form a finite-length control sound source, and a control sound is radiated from the control sound source in the front and diagonal directions. A noise reduction device for reducing noise is known (Patent Document 1).

JP 2006-138130 A

  However, in the technique described in Patent Document 1, since the range in which noise can be reduced is limited to the range of the width of the control sound source in the length direction, the noise diffused in a spherical shape from the noise source is reduced over a wide range. There is a problem that you can not.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a noise reduction device that can reduce noise diffused from a noise source over a wide range. It is another object of the present invention to provide a control sound source speaker (secondary sound source) for effectively reducing low-frequency noise, which can be effectively used by the active noise control ANC.

In order to achieve the above object, the noise reduction device according to the first aspect of the invention emits a control sound with respect to the noise to be reduced, and ½ of the wavelength corresponding to the frequency band of the noise to be reduced. A speaker unit including a plurality of speakers arranged at intervals equal to or less than the length of the microphone, a microphone that collects noise from a noise source and outputs an acoustic signal corresponding to the collected noise, and the acoustic signal And generating means for generating a reduction signal for reducing noise on the side opposite to the noise source from the speaker unit, and an envelope surface of the wave surface of the control sound emitted from the speaker as the wave surface of the noise. as the corresponding, the reduction signal, the delay time calculated in accordance with each determined distance between the plurality of loudspeakers for sound control sound based on the reduced signal delayed and the noise source each A processing unit that performs a delaying process and corrects the control sound emitted from the speaker to be smaller as the delay time is larger for each of the speakers, and a signal processed by the processing unit Is input to each of the plurality of speakers.

  According to the noise reduction device of the first invention, the microphone collects the noise from the noise source, outputs the acoustic signal corresponding to the collected noise, and the generating means generates the speaker based on the acoustic signal. A reduction signal is generated to reduce noise on the side opposite to the noise source from the unit.

  Then, the processing means emits the control signal based on the reduced signal delayed from the noise source so that the envelope of the wave front of the control sound emitted from the speaker corresponds to the wave front of the noise. A process of delaying by a delay time calculated according to each of the distances from the plurality of speakers is performed, and the loudness of the control sound emitted from the speaker is reduced as the delay time is increased for each speaker. to correct.

  Then, the signal processed by the processing means is input to each of the plurality of speakers by the input means, and the control sound is emitted by the plurality of speakers of the speaker unit. At this time, since the envelope of the wavefront of the sound wave obtained by synthesizing the emitted control sounds corresponds to the noise wavefront, the noise is reduced by the control sound.

  In this way, the reduction signal input to each speaker is delayed so that the envelope of the wavefront of the control sound emitted from the plurality of speakers based on the reduction signal corresponds to the wavefront of the noise. By emitting sound from the speaker, the noise diffused from the noise source can be reduced over a wide range.

The noise reduction device according to the second aspect of the invention emits a control sound with respect to noise to be reduced, and has an interval equal to or less than a length corresponding to ½ of the wavelength corresponding to the frequency band of the noise to be reduced. A speaker unit including a plurality of speakers arranged in a line, vibration detection means for detecting vibration of a noise source and outputting a vibration signal corresponding to the detected vibration, and based on the vibration signal, the speaker unit Generating means for generating a reduction signal for reducing noise on the side opposite to the noise source, and the reduction signal so that the envelope of the wave front of the control sound emitted from the speaker corresponds to the wave front of the noise. , it performs processing for each only delay time calculating delay in response to each of the calculated distances between the plurality of loudspeakers for sound control sound based on the reduced signal delayed and the noise source, Processing means for correcting the loudspeaker so that the control sound emitted from the loudspeaker decreases as the delay time increases, and the signals processed by the processing means are And an input means for inputting to.

  According to the noise reduction device of the second invention, the vibration detection unit detects the vibration of the noise source, outputs the vibration signal corresponding to the detected vibration, and the generation unit generates the speaker unit based on the vibration signal. A reduction signal for reducing noise on the side opposite to the noise source is generated.

  Then, the processing means emits the control signal based on the reduced signal delayed from the noise source so that the envelope of the wave front of the control sound emitted from the speaker corresponds to the wave front of the noise. A process of delaying by a delay time calculated according to each of the distances from the plurality of speakers is performed, and the loudness of the control sound emitted from the speaker is reduced as the delay time is increased for each speaker. to correct.

  Then, the signal processed by the processing means is input to each of the plurality of speakers by the input means, and the control sound is emitted by the plurality of speakers of the speaker unit. At this time, since the envelope of the wavefront of the sound wave obtained by synthesizing the emitted control sounds corresponds to the noise wavefront, the noise is reduced by the control sound.

  In this way, the reduction signal input to each speaker is delayed so that the envelope of the wavefront of the control sound emitted from the plurality of speakers based on the reduction signal corresponds to the wavefront of the noise. By emitting sound from the speaker, the noise diffused from the noise source can be reduced over a wide range.

  Further, a plurality of speakers of the above speaker unit can be provided on the wall.

  As described above, according to the noise reduction device of the present invention, the control sound is emitted from a plurality of speakers such that the envelope surface of the control sound wave emitted based on the reduction signal corresponds to the noise wave front. By making it sound, the effect that the noise which spreads from a noise source can be reduced in a wide range is acquired.

It is the schematic which shows the structure of the noise reduction system which concerns on the 1st Embodiment of this invention. It is an image figure which shows the case where a center speaker is determined. It is an image figure which shows the case where a delay time is determined. It is a figure which shows the center frequency of the frequency band of the control sound emitted from each of a some speaker. It is a table | surface which shows a center frequency and a frequency band. It is a figure for demonstrating a Fresnel ring zone. It is the schematic which shows the structure of the noise reduction system which concerns on the 2nd Embodiment of this invention. It is the schematic which shows the structure of the noise reduction system which concerns on the 3rd Embodiment of this invention. It is the schematic which shows the structure of the noise reduction system which concerns on the 4th Embodiment of this invention. It is the schematic which shows the sound-proof wall and speaker of the noise reduction system which concern on the 6th Embodiment of this invention. It is the schematic which shows the sound barrier and speaker of the noise reduction system which concerns on the 7th Embodiment of this invention. It is the schematic which shows the sound-proof wall and speaker of the noise reduction system which concern on the 8th Embodiment of this invention. It is the schematic which shows arrangement | positioning of the speaker unit of the noise reduction system which concerns on the 9th Embodiment of this invention. It is the schematic which shows arrangement | positioning of the speaker unit which has arrange | positioned the speaker at equal intervals. It is a graph which shows the frequency characteristic of the speaker unit of the noise reduction system which concerns on the 9th Embodiment of this invention, and the speaker unit by an equal interval speaker. It is an image figure for demonstrating the directivity by two point sound sources. It is the schematic which shows the structure of the noise reduction system which concerns on the 10th Embodiment of this invention.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In the present embodiment, a case where the present invention is applied to a noise reduction system in a free sound field without a soundproof wall will be described.

  As shown in FIG. 1, the noise reduction system 10 according to the first embodiment controls a speaker unit 14 in which a plurality of speakers 12 are arranged in a row and a control sound emitted from the speakers 12. Control block 16, a plurality of power amplifiers 36 that amplify the signal output from the control block 16 and input to each speaker 12, and are installed near the noise source and collect noise from the noise source The sensor microphone 18 that outputs an acoustic signal corresponding to the collected noise, the microphone amplifier 20 that amplifies the acoustic signal, and the signal output from the microphone amplifier 20 are subjected to inverse filtering to generate a reduced signal. And an inverse filter DSP 22 serving as a generating means for outputting to the control block 16 and an error filter installed at the control point and fed back to the inverse filter DSP 22. And Maiku 24, and a signal processing unit 40 for performing predetermined signal processing. In the explicit method control, the error microphone 24 is necessary when measuring the impulse response in order to determine the inverse filter, but it is clearly stated that it is not necessary during the control. As for the control point, since the sound wave from the noise source propagates in a spherical wave, it may be anywhere from the speaker unit 14 as long as it is opposite to the noise source. The power amplifier 36 corresponds to input means of the present invention.

  In the speaker unit 14, a plurality of speakers 12 are arranged in a straight line, and the speaker 12 is controlled by the control block 16 so as to emit sound in a frequency band equal to or lower than the maximum frequency of the noise frequency band to be reduced. Is done.

In addition, the plurality of speakers 12 are arranged at equal intervals, and the interval d is equal to or less than ½ of the wavelength λ of the sound having the maximum center frequency Fmax in the frequency band of noise to be reduced as shown by the following equation. .
d ≦ λ / 2
For example, in the case of Fmax = 500 kHz, when the sound speed c is 340 m / s, the distance d is 0.34 m.

  The distance d between the speakers 12 may be equal to or less than λ / 2, and the maximum center frequency Fmax of the frequency band of noise to be reduced changes to a higher frequency Fmaxh, and the distance d is λmaxh / 2 (λmaxh). = C / Fmaxh), a speaker moving means is provided by laying rails (not shown) or the like along the arrangement position of the speakers 12 so that each of the speakers can be moved manually or mechanically. Each speaker 12 may be moved by an automatic method (for example, individual speaker movement by an electric motor or the like) so that the interval d with respect to the frequency Fmaxh is λmaxh / 2 or less.

  The inverse filter DSP 22 includes an A / D converter 26 that converts an analog signal output from the microphone amplifier 20 into a digital signal, and an inverse filter 28 that does not change when a filter coefficient is set in advance. The filter coefficient of the inverse filter 28 is determined by either the explicit method or the LMS method, and outputs a signal having the same amplitude and opposite phase with respect to the signal output from the A / D converter 26.

  In addition, the control block 16 is provided with a filter group 33 provided with a band-pass filter 32 that performs filtering so as to pass a signal in a predetermined frequency band for each of the n speakers 12, and n Each of the speakers 12 is provided with a plurality of delay circuits 30 that delay signals in accordance with the arrangement order of the speakers 12 and a plurality of D / A converters 34 that convert digital signals into analog signals.

  The delay circuit 30 delays the signal input to the speaker 12 so that the virtual sound source of the speaker 12 is arranged in an arc shape centered on the noise source. The delay time of the signal input to each speaker 12 is determined by the signal processing unit 40 as follows.

  First, as shown in FIG. 2, the position of the noise source is specified by visual inspection and confirmation by a drawing, and the radius R of the arc-shaped array speaker is determined, and from the noise source to the center speaker of the arc-shaped array speaker group. A radius R which is a distance of is determined.

Then, as shown in FIG. 3, the delay time Δ with respect to the central speaker is calculated by the following equation.
Δ = {(R ² + D _{± i} ² ) ^1/2 −R} / c (1)
Based on the above formula, the delay time Δ is calculated for each speaker 12, and the delay time Δ is set in the delay circuit 30 corresponding to each speaker 12. Thereby, each of the delay circuits 30 delays the signal by the calculated delay time Δ.

The signal processing unit 40 calculates an amplitude correction coefficient Ki for correcting the amplitude of the control sound emitted from the speaker 12 and sets it in the delay circuit 30 as follows. When the speaker 12 is linearly arranged, the actual speaker 12 exists ahead by a distance corresponding to the delay time Δ with respect to the arc of the virtual arrangement, so that a louder sound than when the speaker 12 exists at the location of the virtual arrangement is generated. Release to control point. Therefore, a correction coefficient Ki for correcting the magnitude of the control sound to be small is calculated by the following equation.
Ki = R / (Δi × c + R) (i = 1, 2,...) (2)
Each band-pass filter 32 is a filter that allows signals of all frequencies to pass from an input signal, and a plurality of center frequencies of the signal frequency band are increased one by one in descending order to increase the frequency band of the center frequency. Each of the plurality of combined groups is configured by any one of filters that allow signals in the frequency bands included in each of the groups to pass therethrough. For example, a filter that passes signals of all frequencies from an input signal, and a first group of frequencies that includes only a frequency band of the maximum center frequency 2 ^n-1 f (for example, 4f = 250 Hz) from the input signal. A filter that passes a signal in a band, a second band that includes a maximum center frequency 2 ⁿ⁻¹ f (250 Hz) frequency band and a second largest center frequency 2 ⁿ⁻² f (for example, 2f = 125 Hz). A filter that passes signals in the frequency band of the group, a frequency band of the maximum center frequency 2 ^n-1 f (250 Hz), a frequency band of the second largest center frequency 2 ^n-2 f (125 Hz), and a third large central frequency 2 ^{n-3 f (e.g.,} f = 63 Hz) third passing a signal of a third group of frequency band containing the frequency band of the It is composed of either filter.

  Further, as shown in FIG. 4, the filter group 33 includes a filter that passes signals of all frequencies as the corresponding bandpass filter 32 so as to emit control sounds of all frequency bands from the central speaker 12. Be placed.

  Moreover, about the other speaker 12, the speaker which exists in the position used as the space | interval below the length equivalent to 1/2 of the wavelength corresponding to the minimum center frequency contained in the frequency band in any one of a 1st group-a 3rd group 12, any one of the first filter to the third filter is disposed as the corresponding bandpass filter 32 so that the control sound in any frequency band of the first group to the third group is emitted. The speaker 12 receives a signal in a frequency band included in any one of the first group to the third group.

  For example, the first group of frequency bands having a minimum center frequency of 250 Hz as the bandpass filter 32 corresponding to the speaker 12 located at a position 0.66 m away from the center speaker 12 that is ½ of the wavelength of 250 Hz. The filter which passes the signal of this is arrange | positioned. Further, as a band-pass filter 32 corresponding to the speaker 12 located at a position separated by 1.32 m which is ½ of the wavelength of 125 Hz from the central speaker 12, the second group of frequency bands having a minimum center frequency of 125 Hz. The filter which passes the signal of this is arrange | positioned. Further, as a band-pass filter 32 corresponding to the speaker 12 that is located at a distance of 2.64 m, which is ½ of the wavelength of 63 Hz, from the center speaker 12, a third group of frequency bands having a minimum center frequency of 63 Hz. The filter which passes the signal of this is arrange | positioned.

Here, the center frequency will be described. As a frequency analyzer, there is a constant ratio frequency analyzer, and a frequency analyzer with a constant ratio (Δf / f) includes a narrow band 1/3 octave, a 1/1 octave analyzer, etc. depending on the width of Δf. A 1/3 octave or 1/1 octave analyzer is used for analysis of noise, vibration, etc., and an international standard is determined (IEC Publication 225, Octave, half-octave and third octave filters, 1996). ). The center frequency is determined as shown in FIG. 5. When the minimum center frequency is f, all the center frequencies are 2 ⁿ⁻¹ f (n is a natural number).

In the 1/1 octave analyzer, when the upper limit frequency of the frequency band is f1, the lower limit frequency is f2, and the center frequency is f0, it is expressed by the following equation.
f0 = (f1 × f2) ^1/2
f1 = ^{2−1 / 2} × f0
f2 = 2 ^1/2 × f0
The bandwidth is expressed as follows.
f2 / f1 = (2 ^1/2 × f0) / (2 ^−1/2 × f0) = 2 ^1/1
In the case of a 1/3 octave analyzer, it is expressed as follows.
f0 = (f1 × f2) ^1/2
f1 = 2 ⁻¹ / ⁶ × f0
f2 = 2 ^1/6 xf0
f2 / f1 = (2 ^1/6 × f0) / (2 ⁻¹ / ⁶ × f0) = 2 ^1/3
In the case of a 1 / n octave analyzer, it is expressed as follows.
f0 = (f1 × f2) ^1/2
f1 = ^{2−1 / 2n} × f0
f2 = 2 ^{1 / 2n} × f0
f2 / f1 = (2 ^{1 / 2n} × f0) / ( ^{2−1 / 2n} × f0) = 2 ^{1 / n}
In addition, the signal processing unit 40 determines the filter coefficient W (t) of the inverse filter 28 by, for example, an explicit method as follows.

  First, the cross-correlation function hs (t) between the noise x (t) from the noise source entering the sensor microphone 18 installed in the vicinity of the noise source and the noise ys (t) reaching the control point from the noise source is measured. , The cross-correlation function hc () between the noise x (t) from the noise source entering the sensor microphone 18 and the noise yc (t) emitted from the speaker 12 with the inverse filter set to W (t) = δ (t). t) is measured (measurement of impulse h (t) by cross-correlation method).

  Then, assuming that the transfer functions of hs (t) and hc (t) are Hs (ω) and Hc (ω), the inverse filter W (t) is determined by the following equation.

ここで、Ｆ^−１は逆フーリエ変換を示す。

Here, F ⁻¹ indicates an inverse Fourier transform.

  Next, the Fresnel zone will be described. As shown in FIG. 6, all points on the primary waveform are considered as a continuous radiation source of a spherical secondary wavelet, and an inclination factor (inclination factor) is used to describe the directionality of the secondary radiation. ) When K (θ) is introduced, K (θ) is given by the following equation.

ここで、θは、音源Ｓから観測点Ｐへのベクトルと１次波面に垂直なベクトルｋとのなす角度である。また、音源ＳからＰへの直接波は、以下の式で表される。

Here, θ is an angle formed by a vector from the sound source S to the observation point P and a vector k perpendicular to the primary wavefront. The direct wave from the sound source S to P is expressed by the following equation.

一方、２次小波から合成される波動は、各フレネル輪帯からの寄与の合計であるため、以下の式で表される。

On the other hand, the wave synthesized from the secondary wavelet is the sum of contributions from each Fresnel zone, and is expressed by the following equation.

ここで、上記（５）式と（６）式とは完全に等価でなければならない。また、輪帯Ｋ_１に対してθ＝０となるため、Ｋ_１＝１となる。従って、以下の式が与えられる。

Here, the above equations (5) and (6) must be completely equivalent. Further, since θ = 0 with respect to the annular zone K ₁ , K ₁ = 1. Thus, the following equation is given:

ここで、ε_Ａは、半径ρの１次波面上における単位面積あたりの２次小波の強度であり、また、ε_０／ρは１次波面上における単位面積あたりの１次波の強度である。従って、２次小波の発生源の強度は、１次波の強度の１／λとなる。

Here, ε _A is the intensity of the secondary wave per unit area on the primary wavefront of radius ρ, and ε ₀ / ρ is the intensity of the primary wave per unit area on the primary wavefront. . Therefore, the intensity of the secondary wave generation source is 1 / λ of the intensity of the primary wave.

  Next, the operation of the noise reduction system 10 according to the first embodiment will be described. First, the noise radiated from the noise source is collected by the sensor microphone 18, the acoustic signal is amplified by the microphone amplifier 20, and input to the inverse filter DSP 22. Then, the inverse filter 28 of the inverse filter DSP 22 performs digital filtering processing using the digital signal input from the A / D converter 26 and the set filter coefficient.

  At this time, by passing through the inverse filter 28, the waveform of the signal is shaped and time-shifted, and a reduced signal with the input signal in reverse phase is generated. Therefore, sound is emitted from the speaker 12 based on the reduced signal. A sound reduction effect can be obtained by the control sound.

  The reduced signal filtered by the inverse filter 28 is input to each delay circuit 30, and each delay circuit 30 delays the signal by a delay time set according to the arrangement order of the corresponding speakers 12. The delayed signal is input to the band-pass filter 32, and only a signal in a predetermined frequency band passes through. The passed signal is input to the D / A converter 34, and is D / A converted. To the speaker unit 14. Then, the sound wave corresponding to the filtered signal, that is, the noise anti-phase wave and the sound wave of a predetermined frequency band is emitted as a control sound from each speaker 12 of the speaker unit 14 toward the control point. Is done.

  At this time, since the control sound is emitted from the virtual sound source arranged on the arc centered on the noise source, the control sound is emitted in an arc shape corresponding to the noise diffusing in a circle from the noise source, The control sound emitted from each speaker 12 is synthesized, and the synthesized sound wave becomes a spherical sound wave. The envelope of this wave front corresponds to the wave front of the noise from the noise source, and the control sound is in the opposite phase of the noise.

  Further, in the present embodiment, there are three speakers 12 that emit 63 Hz control sound, five speakers 12 that emit 125 Hz control sound, and nine speakers 12 that emit 250 Hz control sound. Yes, the strength of each virtual sound source arranged on the circular arc is approximately 1 / λ, so that the theory of secondary sound sources in the Fresnel zone can be applied.

  Therefore, the noise from the noise source is canceled out by the control sound from the speaker 12 in a wide range including the control point.

  As described above, according to the noise reduction system according to the first embodiment, the envelope surface of the wave surface of the control sound emitted from the plurality of speakers based on the signal subjected to the inverse filtering process becomes the wave surface of the noise. Correspondingly, the delayed filtered signal to be input to each speaker is delayed and the control sound of the frequency band included in each of the groups combining the frequency bands of the center frequency is emitted from each speaker. As a result, the noise diffused from the noise source can be reduced over a wide range.

  In addition, by applying the principle of the Fresnel zone, the control sound is emitted as a spherical sound wave from a plurality of speakers in response to the sound wave propagating from the noise source, so that the sound in the near field (Near field) Field control can be performed to reduce noise.

  Further, active control can be performed even for a free sound field, that is, when there is no sound barrier, and noise can be reduced.

  In addition, according to the present embodiment, the area in which the effect of reducing noise by 5 dB or more can be greatly increased as compared with the front direction control and the diagonal direction from the linear sound source of the conventional linear arrangement.

  Further, since the control sound in the frequency band of the predetermined center frequency group may be emitted from the speakers arranged at predetermined intervals, the number of speakers can be reduced.

  In the above embodiment, the case where the speakers are arranged at equal intervals has been described as an example. However, the present invention is not limited to this, and 2n + 1 speakers are arranged at intervals proportional to λ that differ for each frequency. Also good. In this case, the number of speakers can be greatly reduced as compared with the number of speakers of a linear sound source having a conventional linearly spaced linear arrangement.

  In the control block, the signal delayed by the delay circuit is input to the bandpass filter. However, the signal filtered by the bandpass filter may be input to the delay circuit.

  Moreover, although the case where the plurality of speakers are arranged in a straight line has been described as an example, the present invention is not limited to this and may be arranged in an arc shape. Also in this case, for each speaker, the delay time is determined and set in the delay circuit based on the distance between the noise source and the speaker so that the virtual sound source of the speaker is arranged on the arc with respect to the noise source. do it.

  Next, a noise reduction system according to the second embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The second embodiment is different from the first embodiment in that the position of the noise source is specified based on the sound wave information.

  As shown in FIG. 7, in the noise reduction system 210 according to the second embodiment, a wire 240 is provided along the arrangement direction of the plurality of speakers 12 of the speaker unit 14, and the speaker proximity microphone 238 is suspended from the wire. Thus, the speaker proximity microphone can be moved.

  When specifying the position of the noise source, the noise from the noise source is collected while moving the speaker proximity microphone 238 to a predetermined position close to each of the plurality of speakers 12, and the sound corresponding to the collected noise is collected. The signal is output to the signal processing unit 40.

  Next, processing for specifying the position of the noise source in the signal processing unit 40 will be described. First, based on the sound wave information, the two sound waves x (t) and y (t) from the noise source incident on the sensor microphone 18 and the speaker proximity microphone 238 at a position close to each speaker 12 (referred to as Mi). A cross-correlation function φxy (τ) that is an inverse Fourier transform of the cross spectrum Φxy (ω) is obtained.

  The cross-correlation function φxy (τ) is expressed as follows. The autocorrelation function for time functions x (t) and y (t) is

と表され、相互相関関数は

And the cross-correlation function is

と表される。ここで、φｘｘ（τ）、φｘｙ（τ）をフーリエ変換すると、パワースペクトルは、

It is expressed. Here, when φxx (τ) and φxy (τ) are Fourier transformed, the power spectrum is

と表され、クロススペクトルは、

And the cross spectrum is

と表される。また、系のインパルス応答をｈ（ｔ）とし、その伝達関数をＨ（ω）とすると、

It is expressed. If the impulse response of the system is h (t) and its transfer function is H (ω),

となる。また、周波数領域では以下のように表される。

It becomes. In the frequency domain, it is expressed as follows.

上記の相互相関関数φｘｙ（τ）は、センサーマイク１８を時間軸ゼロ（基点）として、騒音がセンサーマイク１８からスピーカ近接マイク３８（Ｍｉとする）に到達するまでに要する時間Ｔｉである。ここで、スピーカ近接マイク３８からセンサーマイク１８までの間の距離をｄｉとし、ｃを音速とすると、以下の式で表される。
Ｔｉ＝ｄｉ／ｃ
従って、センサーマイク１８からスピーカ近接マイク３８までの間の距離ｄｉは、ｄｉ＝Ｔｉ×ｃとなるため、スピーカ近接マイク３８の各々を円の中心として、半径ｄｉの円弧を描いて、円弧が交わる点を騒音源の位置として特定する。

The cross-correlation function φxy (τ) is the time Ti required for noise to reach the speaker proximity microphone 38 (Mi) from the sensor microphone 18 with the sensor microphone 18 as the time axis zero (base point). Here, when the distance from the speaker proximity microphone 38 to the sensor microphone 18 is di and c is the speed of sound, it is expressed by the following equation.
Ti = di / c
Accordingly, since the distance di from the sensor microphone 18 to the speaker proximity microphone 38 is di = Ti × c, each of the speaker proximity microphones 38 has an arc of radius di with each circle as the center, and the arcs intersect. The point is specified as the position of the noise source.

When the position of the speaker proximity microphone 38 when the minimum time Tmin of Ti is the closest position to the noise source and the distance from the sensor microphone 18 to the closest speaker proximity microphone 38 is R, the distance R is: It is expressed by the following formula.
R = Tmin · c
Then, similarly to the first embodiment, the delay time with the central speaker as a reference may be calculated and set in the delay circuit 30.

  In the above embodiment, a case where only one speaker proximity microphone is provided has been described as an example, but a plurality of speaker proximity microphones may be provided. In this case, corresponding to each of the plurality of speakers, a speaker proximity microphone is fixedly installed at a position close to the corresponding speaker, and noise from a noise source is collected by each speaker proximity microphone, An acoustic signal corresponding to the generated noise may be output to the signal processing unit.

  Next, a noise reduction system according to the third embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  In the third embodiment, in the control block, the delay circuit and the band pass filter are provided not for each speaker but for each combination of the delay time and the type of frequency band to be passed. Is different.

  As shown in FIG. 8, in the noise reduction system 310 according to the third embodiment, the noise source is located near the bisector of the speaker unit 14, and the speaker 12 at the center of the speaker unit 14 is connected to the noise source. It is located in the nearest place.

  In the delay circuit 30 of the control block 316, each of the speakers 12 positioned on one side when the speaker unit 14 is divided by the speaker 12 at the center is at the center as in the method described in the first embodiment. Based on the speakers 12, the delay time Δ is calculated, and the delay time Δ is set in the delay circuit 30 corresponding to each speaker 12.

  Each bandpass filter 32 of the filter group 333 passes a signal of all frequencies from an input signal, and passes a signal of the first group frequency band including only the maximum center frequency 4f frequency band. A filter, a filter that passes a signal in a second group of frequency bands including a frequency band of the maximum center frequency 4f and a frequency band of the second largest center frequency 2f, and a frequency band of the maximum center frequency 4f, second It is composed of any one of a third filter that passes signals in a third group of frequency bands including a large center frequency 2f frequency band and a third largest center frequency f frequency band.

  The band-pass filter 32 for the center speaker 12 is composed of a filter that passes signals of all frequencies, and the band-pass filters 32 for the other speakers 12 are connected to the two speakers 12 symmetrically with respect to the center speaker 12. These band-pass filters 32 are arranged so that the frequency band signals that each filter passes are not longer than the length corresponding to ½ of the wavelength corresponding to the minimum center frequency included in the frequency band. The first filter to the third filter are arranged so as to be input to each of the speakers 12 existing at the positions.

  For example, the band-pass filter 32 corresponding to the speaker 12 adjacent to the center speaker 12 is formed of a first filter that passes a signal in the frequency band of the maximum center frequency 4f, and further corresponds to the band-pass filter corresponding to the adjacent speaker 12. The filter 32 is composed of a second filter that passes signals in the frequency band of the center frequency 4f and the frequency band of the center frequency 2f, and the bandpass filter 32 corresponding to the adjacent speaker 12 is a frequency band of the center frequency 4f. The band-pass filter 32 corresponding to the speaker 12 is composed of a third filter that passes signals in the frequency bands of the center frequencies 4f, 2f, and f. Yes.

  Note that the operation of the noise reduction system 310 is the same as in the first embodiment, and a description thereof will be omitted.

  As described above, by connecting the delay circuit and the band pass filter in correspondence with the speaker so as to be symmetrical for each combination of the delay time and the frequency band group of the band pass filter, the delay circuit and the band pass filter are connected. Since the number can be reduced, the cost can be reduced.

  Next, a noise reduction system according to the fourth embodiment will be described. In addition, about the part which has the same structure as 1st Embodiment and 3rd Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The fourth embodiment differs from the third embodiment in that the speakers of the speaker unit are arranged not in a straight line but in an arc shape and in that the reduction signal input to the speaker is not delayed. ing.

  As shown in FIG. 9, the noise reduction system 410 according to the fourth embodiment controls a speaker unit 414 in which a plurality of speakers 12 are arranged in an arc shape and a control sound emitted from the speakers 12. A control block 416, a plurality of power amplifiers 36, a sensor microphone 18, a microphone amplifier 20, and an inverse filter DSP 22 are provided.

The plurality of speakers 12 of the speaker unit 414 are arranged at equal intervals with an interval corresponding to ½ of the wavelength corresponding to the maximum center frequency 2 ⁿ⁻¹ f, and a circle whose center of curvature is the position of the noise source Arranged in an arc. Note that the noise source is not limited to the position of the center of curvature of the speaker unit 414, but may be located in a region within a predetermined range near the center of curvature.

  In the control block 416, a plurality of D / A converters 34 are provided for each of the n speakers 12, and the signal output from the inverse filter 28 passes a signal in a predetermined frequency band. A filter group 433 including a plurality of bandpass filters 32 that perform filtering is provided.

  The filter group 433 is provided with a filter that passes signals of all frequencies from an input signal as a band-pass filter 32 corresponding to the center speaker 12, and the center speaker 12 is provided as the other band-pass filter 32. Are provided with a plurality of filters connected to the two speakers 12 symmetrically.

  The bandpass filter 32 connected to the two speakers 12 symmetrically is a filter that passes a signal in the first group frequency band including only the frequency band of the maximum center frequency 4f, and the frequency band of the maximum center frequency 4f. And a filter that passes signals in the second group of frequency bands including the frequency band of the second largest center frequency 2f, and the frequency band of the maximum center frequency 4f, the frequency band of the second largest center frequency 2f, and 3 It is composed of any third filter that passes signals in the third group of frequency bands including the frequency band of the second largest center frequency f. Further, as the band-pass filter 32 connected to the speaker 12 symmetrically, the frequency band signal that each filter passes is not longer than a length corresponding to ½ of the wavelength corresponding to the minimum center frequency included in the frequency band. The first filter to the third filter are arranged so as to be input to each of the speakers 12 existing at the positions of the intervals.

  Next, the operation of the noise reduction system 410 according to the fourth embodiment will be described. Noise radiated from the noise source is collected by the sensor microphone 18, amplified by the microphone amplifier 20, and input to the inverse filter DSP 22. The inverse filter 28 of the inverse filter DSP 22 performs digital filtering using the digital signal input from the A / D converter 26 and the set filter coefficient.

  Then, the signal filtered by the inverse filter 28 is input to each band pass filter 32 of the filter group 433, and only a signal of a predetermined frequency band passes, and each of the passed signals is connected to one connected band pass filter 32. Alternatively, the signals are input to the two D / A converters 34, D / A converted, and output to the speaker unit 314 through the power amplifier 36. Then, a sound wave corresponding to the filtered signal, that is, an anti-phase wave of noise is emitted as a control sound from each speaker 12 of the speaker unit 314 toward the control point.

  At this time, since the control sound is emitted from the speaker 12 arranged in an arc shape with the noise source as the center of curvature, the control sound is emitted in an arc shape corresponding to the noise spreading in a circle from the noise source. The control sound emitted from each speaker 12 is synthesized, and the synthesized sound wave becomes a spherical sound wave. The envelope of this wave front corresponds to the wave front of the noise from the noise source, and the control sound is in the opposite phase of the noise.

  Further, in the present embodiment, there are three speakers 12 that emit 63 Hz control sound, five speakers 12 that emit 125 Hz control sound, and nine speakers 12 that emit 250 Hz control sound. In addition, since the strength of each speaker 12 arranged in an arc shape is approximately 1 / λ, the theory of secondary sound sources in the Fresnel zone can be applied.

Therefore, the noise from the noise source is canceled out by the control sound from the speaker 12 of the speaker unit 414 in a wide range including the control point. As described above, according to the noise reduction system according to the fourth embodiment. The control sound of the frequency band included in each of the groups obtained by combining the frequency bands of the center frequency based on the inverse filtered signal is emitted to each speaker arranged in an arc shape centering on the noise source. Thus, since the envelope surface of the wave surface of the control sound emitted based on the signal subjected to the inverse filtering processing corresponds to the wave surface of the noise, the noise diffused from the noise source can be reduced in a wide range.

  Next, a noise reduction system according to a fifth embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The fifth embodiment is different from the first embodiment in that there are a plurality of noise sources.

  In the noise reduction system according to the fifth embodiment, a sensor microphone 18, a microphone amplifier 20, an inverse filter DSP 22, a delay circuit 30, and a filter group 33 are provided for each noise source.

  Further, outputs from each of the bandpass filters 32 of the filter group 33 for each of the plurality of noise sources are added by a digital adder (not shown) and output to each D / A converter 34.

  A control sound is emitted from the speaker 12 of the speaker unit 14 via each D / A converter 34 and each power amplifier 36. At this time, since the virtual sound source of the speaker 12 is virtually arranged in an arc shape for each noise source, the control sound emitted from each speaker 12 is synthesized, and the synthesized sound wave is a spherical sound wave for each noise source. It becomes. The envelope of this wave front corresponds to the wave front of the noise from each noise source, and the control sound is in the opposite phase of each noise, so that it is from a plurality of noise sources in a wide range including the control point. Noise is canceled by the control sound.

  As described above, even when there are a plurality of noise sources, a signal that has been subjected to inverse filtering is delayed for each noise source, and a signal of a predetermined frequency band is processed by applying a bandpass filter. Noise from a plurality of noise sources can be reduced by the control sound emitted from the speaker.

  Next, a noise reduction system according to the sixth embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  As shown in FIG. 10, the noise reduction system according to the sixth embodiment is provided with a soundproof wall 650, and a plurality of speakers 12 of the speaker unit 14 are installed on the soundproof wall 650.

  As described above, even when the soundproof wall is provided, the speaker is arranged on the soundproof wall, the reduction signal input to the speaker is delayed, and the control sound in a predetermined frequency band is emitted to be diffused from the noise source. Noise can be reduced over a wide range.

  Next, a noise reduction system according to a seventh embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  As shown in FIG. 11, the noise reduction system according to the seventh embodiment is provided with a soundproof wall 750 that is bent at an angle θ, and the plurality of speakers 12 of the speaker unit 14 are placed on the soundproof wall 750. Is installed.

  Thus, even when a speaker is arranged on a soundproof wall that is bent, the noise diffused from the noise source can be reduced over a wide range.

  Next, a noise reduction system according to the eighth embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  As shown in FIG. 12, the noise reduction system according to the eighth embodiment is provided with a soundproof wall 850 extending in a straight line of two stages, and the plurality of speakers 12 of the speaker unit 14 are It is installed on the soundproof wall 850.

  Thus, even when a speaker is arranged on a soundproof wall that is a two-step straight line, the noise diffused from the noise source can be reduced over a wide range.

  The soundproof wall in which the speaker is arranged on the wall may have a curved shape centering on the side opposite to the noise source. Also in this case, the noise diffused from the noise source can be reduced over a wide range by delaying the reduction signal.

  Moreover, although the case where a soundproof wall is provided has been described as an example, even in a free sound field having no soundproof wall, a reduction signal is sent to a plurality of speakers that are not arranged linearly as described above. By delaying, the noise diffused from the noise source can be reduced over a wide range. For example, when the site boundary is not linear, the noise diffused from the noise source can be reduced over a wide range even when speakers are arranged along the site boundary.

In the above description, the example of controlling the noise by dividing the frequency band of the noise to be reduced into three frequency bands has been described, but the frequency band of the noise to be reduced is generally expressed as a minimum center. frequency as f, ^{2 n-1} times f (n is a natural number) can be divided ^f, 2f, 4f, ···, into n frequency bands having a center frequency of ^{2 n-1} f of.

  In this case, a group in which the number of center frequencies is increased one by one in the descending order and the frequency bands of the center frequencies are combined is as follows.

Center frequency of frequency band included in first group: 2 ^n-1 f
Center frequencies of frequency bands included in the second group: 2 ^n-1 f, 2 ^n-2 f
Center frequency of frequency band included in third group: 2 ^n-1 f, 2 ^n-2 f, 2 ^n-3 f
Center frequencies of frequency bands included in the fourth group: 2 ^n-1 f, 2 ^n-2 f, ..., 2 ^n-4 f
...
Center frequency of the frequency band included in the nth group: 2 ^n-1 f, 2 ^n-2 f,... F
In addition, when the bandpass filter is arranged corresponding to the speaker unit, the bandpass filter is arranged from the bandpass filter corresponding to the left end of the speaker unit or the bandpass filter (end portion) corresponding to the right end to the center of the speaker unit. On the other hand, it can arrange so that a pass band may become the following arrangements.
"2 ^n-1 f, 2 ^n-2 f, ..., f" (end), "2 ^n-1 f", "2 ^n-1 f, 2 ^n-2 f, ..., 2f ", 2 ^n-1 f", "2 ^n-1 f, 2 ^n-2 f, ..., 4f", "2 ^n-1 f", "2 ^n-1 f, 2 ^n-2 f" , ..., 8f "," 2 ^n-1 f ", ...," 2 ^n-1 f "," 2 ^n-1 f, 2 ^n-2 f, ..., f "(central part ). In addition, you may use the band-pass filter which allows the signal of a full frequency band to pass through in the center part.

  Next, a ninth embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The ninth embodiment differs from the first embodiment in that the control sound in the frequency band of one center frequency is emitted from each speaker and that the speakers are arranged at unequal intervals. ing.

  The wavelength λ of the low-frequency sound wave is long. For example, λ at 100 Hz is 3.4 m. When the maximum frequency of the frequency band to be controlled is 500 Hz, the speakers that emit the control sound are arranged at equal intervals with a half of the wavelength λ of 500 Hz, that is, with an interval of d = λ / 2 = 0.34 m or less. It was the prior art to do. Since the wavelength λ of the sound wave of 63 Hz is 5.4 m, in order to reduce 100 Hz, the length of the speaker emitting the control sound is at least twice as long as λ, that is, at least 17 speakers as shown in FIG. Needed to be installed.

  In the ninth embodiment, a speaker unit that efficiently reduces low-frequency sound waves with a minimum number of speakers is presented. That is, as shown in FIG. 13, based on the Huygens-Fresnel theory, the wavelength λ (λ = c) corresponding to the center frequency f of the frequency band (center frequency 63, 125, 250, 500 Hz, etc.) of the noise to be controlled. The speakers 12 are arranged at unequal intervals at intervals equal to or less than the half length d (d = λ / 2) of (sound speed) / f).

  By arranging in this way, it is possible to reduce noise with a frequency of 63 Hz with nine speakers 12. This is realized by the accompanying effects described below.

  FIG. 15 shows the frequency characteristics of the speaker unit 712 according to the ninth embodiment and the speaker unit in which 17 speakers as shown in FIG. 14 are arranged at equal intervals.

The thin line in FIG. 15 represents the frequency characteristics of one speaker constituting the speaker unit in FIGS. 13 and 14. Further, the frequency characteristics of the speaker unit by the 17 equally spaced speakers shown in FIG. The sound pressure level increases by the number of speakers (17) ( ₁₀ log ₁₀ (17) = 12.3 dB), but the frequency characteristics are the same as in the case of one.

However, in the case of the speaker unit 712 using the unequally spaced speakers according to the present embodiment in FIG. 13, the frequency characteristics are as shown by the chain line in FIG. There are two speakers 12 for each band of the center frequencies 63, 125, 250, and 500 Hz, and the power increases by 3 dB (= ₁₀ log ₁₀ (2)). Compared to, it is flatter.

  The reason for this is that the lower frequency is arranged at intervals of 1/2 of the wavelength λ, and the larger the interval, the larger the radiation area is secured at the lower frequency when the speaker unit shown in FIG. 13 is regarded as one speaker. This is because the mutual radiation impedance in the low sound range is increased, and the efficiency is improved in the low sound.

  Moreover, the directivity by two point sound sources is demonstrated using FIG. When the distance d between the two point sound sources is ½ of the wavelength λ, the sound wave is strongly emitted in the forward direction and has directivity as shown in FIG. Therefore, by configuring the speaker unit 712 as shown in FIG. 13 above, sound waves are efficiently emitted in the control point direction in each of the central frequency bands of 63, 125, 250, and 500 Hz.

  Note that the directivity in the rear direction is considerably small in the actual speaker BOX because it is behind the speaker BOX. Further, the directivity of the speaker included in the actual speaker BOX may be considered to be only the control point direction.

  Next, a tenth embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The tenth embodiment is different from the first embodiment in that a control sound in all frequency bands is emitted from each speaker without providing a bandpass filter.

  As shown in FIG. 17, in the noise reduction system 910 according to the tenth embodiment, the control block 316 includes a delay circuit 30 and a D / A converter 34, and signals of all frequencies are sent to the power amplifier 36. Output to the speaker 12.

  Since the other configuration of the noise reduction system 910 is the same as that of the first embodiment, description thereof is omitted.

  Next, the operation of the noise reduction system 910 according to the tenth embodiment will be described. First, the noise radiated from the noise source is collected by the sensor microphone 18, the acoustic signal is amplified by the microphone amplifier 20, and input to the inverse filter DSP 22. Then, the inverse filter 28 of the inverse filter DSP 22 performs digital filtering processing using the digital signal input from the A / D converter 26 and the set filter coefficient.

  The reduced signal filtered by the inverse filter 28 is input to each delay circuit 30, and each delay circuit 30 outputs a signal for a delay time set in accordance with each distance between the corresponding speaker 12 and the noise source. Delay. The delayed signal is input to the D / A converter 34, is D / A converted, and is output to the speaker unit 14 via the power amplifier 36. Then, a sound wave corresponding to the signal subjected to the inverse filtering process, that is, a sound wave in the entire frequency band, which is a reverse phase wave of noise, is emitted as a control sound from each speaker 12 of the speaker unit 14 toward the control point.

  At this time, since the control sound is emitted from the virtual sound source arranged on the arc centered on the noise source, the control sound is emitted in an arc shape corresponding to the noise diffusing in a circle from the noise source, The control sound emitted from each speaker 12 is synthesized, and the synthesized sound wave becomes a spherical sound wave. The envelope of this wave front corresponds to the wave front of the noise from the noise source, and the control sound is in the opposite phase of the noise.

  Therefore, the noise from the noise source is canceled out by the control sound from the speaker 12 in a wide range including the control point.

  As described above, according to the noise reduction system according to the tenth embodiment, the envelope surface of the wavefront of the control sound emitted from the plurality of speakers based on the signal subjected to the inverse filtering process is the noise wavefront. Correspondingly, the noise that is diffused from the noise source can be reduced over a wide range by delaying the signal that has been subjected to inverse filtering processing to be input to each speaker and emitting the control sound from each speaker. .

  In the first to tenth embodiments, the sensor microphone is used to collect noise from the noise source, and the acoustic signal corresponding to the collected noise is transmitted via the microphone amplifier. However, a vibration detector that detects the vibration of the noise source may be used. In this case, the vibration detector is used to detect the vibration of the noise source, and a vibration signal corresponding to the detected vibration is output to the inverse filter DSP via the microphone amplifier. A reduced signal may be generated by applying inverse filtering.

10, 210, 310, 410, 910 Noise reduction system 12 Speaker 14, 414, 712 Speaker unit 16, 316, 416, 916 Control block 18 Sensor microphone 20 Microphone amplifier 22 Inverse filter DSP
26 A / D converter 28 Inverse filter 30 Delay circuit 32 Band pass filter 33, 333, 433 Filter group 34 D / A converter 36 Power amplifier 40, 240 Signal processing unit 650, 750, 850 Sound barrier

Claims (3)

A speaker having a plurality of speakers arranged to emit a control sound with respect to noise to be reduced and arranged at intervals equal to or less than a length corresponding to ½ of the wavelength corresponding to the frequency band of the noise to be reduced Unit,
A microphone that collects noise from a noise source and outputs an acoustic signal corresponding to the collected noise;
Generating means for generating a reduction signal for reducing noise on the opposite side of the noise source from the speaker unit based on the acoustic signal;
The plurality of sound emitting control sounds based on the reduced signal obtained by delaying the reduced signal and the noise source so that an envelope surface of the wave surface of the controlled sound emitted from the speaker corresponds to the wave front of the noise. A process of delaying by a delay time calculated according to each of the obtained distances from the speaker, and the magnitude of the control sound emitted from the speaker as the delay time increases for each of the speakers Processing means for correcting so as to decrease,
Input means for inputting the signal processed by the processing means to each of the plurality of speakers;
A noise reduction device comprising :
A noise reduction device in which a delay time for the speaker is a value obtained by dividing a difference between a distance between the noise source and the speaker and a distance between the noise source and a central speaker among the plurality of speakers by a sound speed. . A speaker having a plurality of speakers arranged to emit a control sound with respect to noise to be reduced and arranged at intervals equal to or less than a length corresponding to ½ of the wavelength corresponding to the frequency band of the noise to be reduced Unit,
Vibration detecting means for detecting vibration of a noise source and outputting a vibration signal corresponding to the detected vibration;
Generating means for generating a reduction signal for reducing noise on the side opposite to the noise source from the speaker unit based on the vibration signal;
The plurality of sound emitting control sounds based on the reduced signal obtained by delaying the reduced signal and the noise source so that an envelope surface of the wave surface of the controlled sound emitted from the speaker corresponds to the wave front of the noise. A process of delaying by a delay time calculated according to each of the obtained distances from the speaker, and the magnitude of the control sound emitted from the speaker as the delay time increases for each of the speakers Processing means for correcting so as to decrease,
Input means for inputting the signal processed by the processing means to each of the plurality of speakers;
A noise reduction device comprising :
A noise reduction device in which a delay time for the speaker is a value obtained by dividing a difference between a distance between the noise source and the speaker and a distance between the noise source and a central speaker among the plurality of speakers by a sound speed. .   The noise reduction device according to claim 1, wherein a plurality of speakers of the speaker unit are provided on a wall.

Images