JP3532584B2 - Active control noise reduction device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an active control noise reduction device, and more particularly to an active control noise reduction device capable of preventing the occurrence of an oscillation phenomenon that tends to occur during control.

[0002]

2. Description of the Related Art Recently, an active control noise reduction device to which an acoustic control technique is applied has been proposed in order to deal with noise problems of various devices.

The most basic type of such an active control noise reduction device is constructed as shown in FIG. In this figure, an example in which the sound emitted from the sound source 2 housed in the duct 1 is prevented from leaking to the outside from the opening 3 of the duct 1 by the active control noise reduction device 4 is shown. .

That is, the active control noise reduction device 4 detects the sound emitted from the sound source 2 with the microphone 5 and introduces the output signal of the microphone 5 into the signal processing device 6 having a filter having a predetermined filter coefficient. The secondary sound source, that is, the speaker 7, is operated by the signal obtained by passing through the filter, and the sound emitted from the speaker 7 actively cancels the sound source sound in the opening 3 that is the control target point. .

However, in the active control noise reduction device 4 configured as described above, oscillation (howling) occurs.
It has a fatal disadvantage that it is easy to occur. That is, the radiated sound from the speaker 7 goes up upstream and is detected again by the microphone 5. Therefore, an electro-acoustic feedback circuit is formed, and there is a problem that the system oscillates depending on the gain and a sufficient noise reduction effect cannot be obtained.

Therefore, in order to prevent such a problem, for example, "Nishimura / Arai, Active Control of Radiated Sound at Duct Exit, Technical Report, Vol. 88. No. 105 (19).
88) ”proposes the two-microphone method.

In this method, as shown in FIG. 14, the sound emitted from the sound source 2 is detected by the two microphones 5a and 5b, and the output signal of one microphone 5a and the transmission corresponding to the interval between the two microphones are transmitted. The detection signal is a signal obtained by synthesizing the output signal of the other microphone 5b passing through the delay function section 8 having characteristics with the signal adder 9 to eliminate the influence of the return sound from the speaker 7.

However, this so-called two-microphone method is effective for a long duct structure, but cannot be sufficiently effective for a duct having a relatively short length for the following reason. That is, the sound component P _M detected by the two microphones is expressed as the sum of the two components as follows. P _M = P _MO + P _MA (1)

Here, P _MO is a component from only the sound source 2, and P _MA is a component from only the secondary sound source 7. Therefore, in order to satisfy the condition of P _MA = 0, it is necessary to find the transfer function G ₁ to be set in one of the two microphones.

Now, if the distance between the two microphones is ΔL and the target sound wave is limited to the component of the period number range of the wavelength that is longer than the cross-sectional dimension of the duct, that is, the range that can be regarded as a plane wave, P _MA is It looks like this:

[0011]

[Equation 1]

Here, P _PA is a traveling wave component from the secondary sound source 7, and P _rA is a backward wave component. In addition, k is the wave number, and when the excitation frequency is f and the sound velocity is c, k = 2πf / c
Is the value indicated by. P _{MA in the} equation (2) is a howling component, and it is necessary to find G ₁ such that the equation (2) becomes zero.

However, since P _PA and P _rA cannot be directly identified, it is not possible to find G _{1 at} which P _MA = 0. For example, for a long duct,
Energy of the component from the secondary sound source 7 is absorbed by the wall surface or the like during propagation. Therefore, in such cases, P _PA
Can be approximately zero. in this case,

[0014]

[Equation 2] That is, P _MA can be made substantially zero by giving G ₁ a delay corresponding to the interval between the two microphones.

However, a sound absorption effect during propagation cannot be expected in a duct having a short length. Therefore, howling cannot be prevented even if the active control noise reduction technique of the two-microphone method is applied to a small product.

FIG. 15 shows the result of an experiment conducted on a duct having a length as short as 1.5 m. This is when a white noise sound is generated from the speaker which is the secondary sound source. The upper waveform X is the output of one of the two microphones, and the lower waveform Y is G ₁ . This is the combined output of the two microphones when set as in equation (3). The output Y is not reduced at all. From this, it is understood that the howling prevention effect cannot be obtained in the case of the duct having a short length. This is because P _PA ≠ 0 due to reflection due to impedance at the end opposite to the opening of the duct. As described above, the conventional active control noise reduction device employing the two-microphone method has a problem that howling cannot be prevented when the length of the duct is short.

As another means for preventing howling, when a signal having a correlation with a target noise, for example, a sound source is a compressor, the vibration of the compressor is detected and this is actively controlled to reduce noise. It can be considered as an input of the device. However, in this method, for example, in noise generated by a fan, noise generated by whirling of a motor or blades,
That is, the electromagnetic noise and the rotating noise that have a correlation can be reduced, but it is difficult to reduce the so-called flow noise generated by friction with the wall surface of the duct. It is more effective to use a microphone to reduce this flow noise. However, using a microphone causes howling for the above reasons.

[0018]

As described above, in the conventional active control noise reduction device adopting the two-microphone method, in particular, the noise is canceled at the opening of the duct having a short length. However, there is a problem that howling occurs and a good noise reduction effect cannot be obtained.

Therefore, the present invention adopts the two-microphone method,
Moreover, it is an object of the present invention to provide an active control noise reduction device that can exert a good noise reduction effect even in the case of a structure such as a duct having a short length.

[0020]

SUMMARY OF THE INVENTION The present invention provides an active control noise reduction device for controlling a primary sound emitted from a noise source,
A secondary sound source that generates a secondary sound that cancels the primary sound emitted from the noise source, and a secondary sound source that is disposed at a predetermined distance in the direction from the noise source to the secondary sound source The first microphone on the side closer to the sound source and the second microphone on the side farther than the secondary sound source, and the transfer function from the secondary sound source to the second microphone for the primary transfer function which is the transfer function from the secondary sound source to the first microphone. The output signal of the secondary microphone is multiplied by a coefficient corresponding to the ratio of a certain secondary transfer function, the difference between the multiplication result signal and the output signal of the first microphone is obtained, and the signal corresponding to this difference is subjected to predetermined processing. And a signal processing means for generating a secondary sound signal obtained as described above and supplying the secondary sound signal to a secondary sound source.

The present invention relates to an active control noise reduction device for actively controlling a primary sound emitted from a noise source arranged in a duct, and a secondary sound attached to the duct to cancel the primary sound sent from the noise source to the duct. And a secondary sound source for generating the noise, which is provided in the duct and is arranged between the noise source and the secondary sound source at a predetermined distance in the direction from the noise source to the secondary sound source.
And a second microphone on the side closer to the secondary sound source for outputting the second output signal and a second microphone on the side farther from the secondary sound source, respectively.
A coefficient connected to the first and second microphones and corresponding to a ratio of a secondary transfer function which is a transfer function from the secondary sound source to the second microphone with respect to a primary transfer function which is a transfer function from the secondary sound source to the first microphone. And a multiplication means for multiplying the output signal of the second microphone by the coefficient and outputting a multiplication result signal, the multiplication result signal and the first
Difference calculating means for obtaining a difference from the output signal of the microphone and outputting the difference signal, and processing the difference signal and supplying a secondary sound signal obtained by multiplying the difference signal by a predetermined coefficient to the secondary sound source. Provided is an active control noise reduction device characterized by being constituted by a signal processing means.

[0022]

According to the configuration of the present invention, the noise generated from the noise source can be effectively canceled by the two-microphone method without causing howling.

According to the configuration of the present invention, a good noise reduction function can be achieved by the two-microphone method without generating howling even in a short duct in which many reflected sound components are generated.

[0024]

Embodiments will be described below with reference to the drawings.

FIG. 1 shows an active-controlled noise reduction device according to an embodiment of the present invention, in which sound emitted from a sound source 12 housed in a duct 11 leaks out of an opening 13 of the duct 11. An example in which the noise is prevented by the active control noise reduction device 14 is shown.

The main part of the active control noise reduction device 14 is a predetermined interval ΔL in the duct 11 in the extending direction of the duct 11.
The microphones 15a and 15b, which are arranged apart from each other and detect the sound emitted from the sound source 12, and the microphone which has passed through the delay function unit 16 having a transfer function G ₁ corresponding to the output signal of the microphone 15a and the distance between both microphones. A signal adder 17 for synthesizing the output signal of 15b, a signal processing device 18 incorporating a filter having a predetermined filter coefficient for processing the output signal of the signal adder 17, and a signal obtained through the filter. And a speaker 19 as a secondary sound source that operates in accordance with.

That is, this main part is constructed in the same manner as the conventional device adopting the two-microphone method, and the sound emitted from the speaker 19 is used to actively cancel the sound source sound in the opening 13 which is the control target point. I have to.

In this embodiment, as an additional element of the active control noise reduction device 14, the left side of the sound source 12 in the figure, that is, the upstream side of the sound source 12 when the direction from the sound source 12 to the speaker 19 is used as a reference. Speaker 19
A sound absorbing material 20 is arranged as an antireflection means for preventing reflection of sound coming from the side.

With such a configuration, the signal adder 17
The output P _M of the above is a composite of the component P _MO from only the sound source 12 and the component P _MA from only the speaker 19, as in the above-mentioned equation (1). And the component P _MA is further (2)
As shown in the equation, it is divided into a P _PA component and a P _rA component. Among them, the P _PA component is a reflection component from the left end of the duct 11, but as shown in FIG. 2, among the sounds emitted from the speaker 19, the sound that progresses toward the left end of the duct 11 is
Since the sound absorbing material 20 having a high sound absorbing effect absorbs the energy, the energy of P _PA eventually becomes almost zero.

Therefore, by setting the transfer function of the delay function section 16 to G ₁ as in the equation (3), the sound component emitted from the speaker 19 and directly entering the microphones 15a, 15b and reflected by the opening 13 are reflected. Are microphones 15a and 15b
The sound component that enters can be removed, and the component P _MO from the sound source 12 only
Will be able to get. Therefore, even in the case of a duct having a short length, a good noise reduction function can be exerted without causing howling.

In FIG. 3, under the condition that the sound absorbing material 20 is installed at the left end of the duct 11, when the white noise sound is generated from the speaker 19, the output of one of the two microphones and the white nozzle signal are shown. The impulse response function obtained from the transfer function of is shown. In this output, the first wave F is the direct sound from the speaker 19, and the second wave S is the reflected sound from the opening 13. From this figure, the sound absorbing effect of the sound absorbing material 20 provided at the left end can be confirmed.

FIG. 4 shows an impulse response obtained with two microphones under the condition that the sound absorbing material 20 is installed at the left end of the duct 11, and FIG. 5 shows a time waveform of sound pressure. . In the impulse response shown in FIG. 4, almost no component from the speaker 19 appears. Further, in the time waveform of FIG. 5, the waveform C is one microphone output and the waveform D is 2
This is the microphone output. The size of the two-microphone structure is almost zero.

As described above, by arranging the sound absorbing material 20 on the upstream side of the sound source 12, the effect of the two-microphone method can be sufficiently brought out and the noise reduction performance can be improved.

The above-described embodiment is an example in which the sound source is housed in a dead end duct and noise is prevented from leaking outside from the opening of the duct. Although the configuration is effective when the sound source is a compressor, the present invention is not limited to such a configuration. FIG. 6 shows an example of reducing the noise of the fan used for cooling the board of the computer by using the active control noise reduction device 14 according to the present invention.

That is, reference numeral 21 in the drawing denotes a substrate on which an IC chip or the like is mounted, and this substrate is arranged in the room 22. Ventilation openings 23a and 23b are formed on the walls of the room 22 at diagonal positions, and a fan 24 is installed in the ventilation openings 23a.

The fan 24 is used in a so-called push system. Therefore, the ventilation port 23a is on the suction side, and the ventilation port 23b is on the discharge side. The ventilation port 23a forming the suction side communicates with a suction duct 25 extending so as to change the direction by 90 degrees with respect to the central axis of the ventilation port 23a. In addition, the ventilation port 23b forming the discharge side
Communicates with the discharge duct 26 extending so as to change the direction by 90 degrees with respect to the central axis of the ventilation port 23b. The sound absorbing material 27, 28, 29, 30, which has a high sound absorbing effect, is formed on the inner surface of the suction duct 25 and the inner surface of the discharge duct 26.
31 and 32 are attached.

In this example, in order to prevent the noise and the flow noise of the fan 24 from leaking from the opening 33 of the discharge duct 26 to the outside, the active control noise reduction configured as in the example shown in FIG. A device 14 is provided. Therefore, in this example, the sound absorbing material 30 provided on the inner surface of the lower portion in the drawing on the inner surface of the discharge duct 26 has a function of preventing reflection of the sound coming from the speaker 19 side. With this configuration, noise that tends to leak from the outlet of the cooling channel can be effectively canceled by the two-microphone method.

In FIG. 7, active control noise reduction devices 14a and 14b are provided on both the discharge duct 26 side and the suction duct 25 side.
Are provided respectively from the opening 33 of the discharge duct 26 and the opening 34 of the suction duct 25 to the fans 24a, 24b.
An example is shown in which the noise and the flow noise of the vehicle are prevented from leaking to the outside.

In this example, the sound absorbing material 28 on the suction duct 25 side prevents reflection of the sound coming from the speaker 19 side, and the sound absorbing material 30 on the discharge duct 26 side is the sound absorbing material 30.
It plays the role of preventing the reflection of the sound coming from the 9 side. With this configuration, noise that tends to leak from the inlet and the outlet of the cooling channel can be effectively canceled by the two-microphone method.

In the example shown in FIG. 1, the sound absorbing material 20 is arranged on the upstream side of the sound source 12 to absorb the sound coming from the speaker 19 side. However, as shown in FIG. Sound source 1 upstream of sound source 12
A horn-shaped opening 41 that widens as it moves away from the opening 2 may be provided.

It is well known that the horn is used for the purpose of reducing reflection due to radiation impedance at its mouth and improving acoustic radiation efficiency. Therefore,
By providing the horn-shaped opening 41, the sound coming from the speaker 19 side is reflected and the microphones 15a and 15b are reflected.
Therefore, it is possible to suppress the entry and to obtain the same effect as that of the above embodiment. The effect can be further improved by mounting the sound absorbing material 42 on the inner surface of the opening 41. FIG. 9 shows a schematic configuration of an active control noise reduction device 14d according to another embodiment of the present invention.

In each of the above-described embodiments, the sound coming from the speaker 19 side is reflected on the upstream side of the sound source 12, and the reflected sound enters the microphones 15a and 15b is suppressed by a mechanical method such as a sound absorbing material or a horn. However, in this embodiment, it is suppressed by a purely electric signal processing method.

That is, in the active-control noise reduction device 14d according to this embodiment, the output of one of the two microphones 15a and 15b, which is located on the side of the speaker 19 which is the secondary sound source, is set to the coefficient unit 51. Through one input end of the signal adder 52 via the
The output of 5a is introduced to the other input terminal of the signal adder 52,
A signal processing device 1 incorporating a filter having a predetermined filter coefficient for the difference signal between the two outputs obtained by the signal adder 52.
Introduced in 8.

As the coefficient unit 51, an adaptive controller under the condition that the adaptive operation is stopped is actually used as will be described later. Then, the coefficient unit 51 includes a speaker 1
The output of the microphone 15a obtained when allowed to radiate white noise sound from 9 and P _1, P output of the microphone 15b
_{When 2} , the coefficient is H = −P ₁ / P ₂ , that is, the ratio of the transfer function from the microphone 15b to the microphone 15a is set as the coefficient.

With such a structure, howling can be reduced without being influenced by the length of the duct 11 and the like, and the characteristics of the two-microphone method can be exhibited. Further, there is also an advantage that the above-mentioned sound absorption processing is unnecessary. The reason will be described below.

As described above, the two microphones 15
The component P _{MA of the} secondary sound source contained in the component P _M of the sound detected by a and 15b is expressed by the equation (2). Normally, a sound pressure signal measured by a sound level meter or the like has a traveling wave component P _PA and a backward wave component P _rA.
Since it is observed as the sum of the
It is not possible to identify _PA and _PrA . However, when P _PA and P _rA in the equation (2) are known and G ₁ with the equation (2) set to 0 is expanded, the following equation is obtained.

[0047]

[Equation 2] Where P _PA is the traveling wave component, P _rA is the backward wave component, and Δ L is the
1, the distance between the second microphone, L is between the secondary sound source and the second microphone
Interval, k is 2 π f / c (f is excitation frequency, c is sound velocity)

As shown in FIG. 10, the denominator in the equation (4) represents the sound pressure component P ₂ in the microphone 15b, and the numerator represents the sound pressure component P ₁ in the microphone 15a. Therefore, in G ₁ represented by the equation (4), the transfer function ratio when the output of the secondary sound source is A, that is, G ₁ = − (P ₁ / A) / (P ₂ / A) = -P becomes ₁ / P ₂ ... (5) that may be set.

To obtain the equation (5), an adaptive controller 51 'with a built-in digital filter may be used as shown in FIG. The algorithm of this adaptive control is
For example, an LMS algorithm or the like that minimizes the sum of squares of the error signal E is used. Then, the output P _{2 of the} microphone 15b located closer to the speaker 19 is introduced as the input signal of the adaptive control device 51 ', and the output P _{1 of the} microphone 15a located farther is introduced as the reference signal. To do.
When the speaker 19 is operated with the M-sequence signal and P ₁ and P ₂ are introduced, the error signal E is represented by the following equation. E = P ₁ -P ₂ H (6) In the adaptive operation, E in the equation (6) becomes 0 in the optimum state, so that the coefficient of the adaptive controller 51 ′ finally becomes H = P ₁ / P ₂ … Identified in (7).

This is equivalent to the equation (5) without any correction. Therefore, under the condition that the adaptive control device 51 'is identified, the component P of the secondary sound source detected by the two microphones is detected.
_MA becomes 0.

Therefore, in this embodiment, the adaptive controller 5 is set under the condition that the convergence coefficient is set to 0 and E in the equation (6) becomes 0.
The adaptive controller 51 'holding the coefficient H identified by the equation (7) is used as the coefficient unit 51 by stopping the adaptive operation at 1'. That is, when actually performing the noise reduction operation, the output of the signal adder 52, which has been used as the error signal E until now, is introduced into the signal processing device 18.

Therefore, the signal introduced into the signal introducing device 18 can be only the component from the sound source 12, and the duct 11 can be formed by the two-microphone method without causing howling.
The sound source sound from the opening 13 can be effectively canceled.

In FIG. 12, the time output of each part when the sound radiated from the secondary sound source is detected by the microphones 15a and 15b under the condition that the coefficient H of the coefficient unit 51 is set to the relationship shown in the equation (7). It is shown. That is, the microphone 15 is shown in (a).
The time output of b is shown, and the time output of the signal adder 52 is shown in (b). As can be seen from these figures, the time output of the signal adder 42 can be made sufficiently small, that is, the howling component can be sufficiently reduced. In this case, since the adaptive control device 51 'itself can be used as the coefficient unit 51, there is an advantage that the system can be simplified. Further, the configuration of the two-microphone method is equivalent to the high-pass filter processing applied to the sound source signal. Therefore, it is possible to reduce the sensitivity in the very low frequency region where the radiation efficiency of the speaker is poor. This has an advantageous effect when the active control device is constructed with adaptive signal processing. That is, in such an ultra-low frequency region, the emission efficiency of the speaker is poor, and therefore the adaptive control device should not identify the components in the above region as much as possible. However, actually, the component of the proper level is input. A high-pass filter is most effective for eliminating this, but such a filter causes a long duct length for active control. However, with a two-microphone configuration, an acoustic high-pass filter can be configured,
No electric circuit is needed. Then, the control band can be adjusted by the interval between the two microphones. Therefore, it is not necessary to consider the duct length.

[0054]

As described above, according to the present invention,
The features of the two-microphone method can be maximized without being affected by the length of the duct or the like.

[Brief description of drawings]

FIG. 1 is a schematic diagram of an active control noise reduction device according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining that a component of a secondary sound source is not reflected in the same device.

FIG. 3 is a diagram showing an impulse sound detected by one microphone in the same device.

FIG. 4 is a diagram showing an impulse response of a combined output of two microphones in the same device.

FIG. 5 is a diagram showing a time output of one microphone and a combined time output of two microphones in the same device.

FIG. 6 is a schematic diagram of an example of reducing the fan noise of a computer with the same device.

FIG. 7 is a schematic diagram of another example of reducing the fan noise of a computer with the same device.

FIG. 8 is a schematic diagram showing an active control noise reduction device according to another embodiment of the present invention.

FIG. 9 is a schematic diagram showing an active control noise reduction device according to still another embodiment of the present invention.

FIG. 10 is a diagram for explaining coefficients set in a coefficient unit incorporated in the apparatus.

FIG. 11 is a diagram for explaining a work of setting a coefficient in a coefficient unit incorporated in the apparatus.

FIG. 12 is a diagram showing a time output of one microphone and a combined time output of two microphones obtained when only a secondary sound source is operated in the same device.

FIG. 13 is a schematic diagram showing a conventional active control noise reduction device adopting the one-microphone method.

FIG. 14 is a schematic diagram showing a conventional active control noise reduction device adopting the two-microphone method.

FIG. 15 is a diagram showing time output of one microphone and time output of two microphones in a conventional active control noise reduction device adopting the two-microphone method.

[Explanation of symbols]

11, 11a ... Duct 12 ... Sound source 13 ... Openings 15a, 15b
... microphones 14, 14a, 14b, 14c, 14d ... active control noise reduction device 16 ... delay function section 17, 52 ... signal adder 18 ... signal processing device 19 ... speakers 20, 28, 30, 42 as secondary sound sources ... Sound absorbing material 41 ... Horn-shaped opening 51 ... Coefficient unit 51 '... Adaptive control device

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G10K 11/178 G10K 11/16

Claims (2)

(57) [Claims] 1. An active control noise reduction device for controlling a primary sound emitted from a noise source, comprising: a secondary sound source that generates a secondary sound that cancels the primary sound emitted from the noise source; A first microphone that is disposed between the noise source and the secondary sound source at a predetermined distance in the direction toward the next sound source, and is closer to the secondary sound source; and a second microphone that is farther from the secondary sound source; A coefficient (G ₁ ) corresponding to a ratio of a secondary transfer function which is a transfer function from the secondary sound source to the second microphone with respect to a primary transfer function which is a transfer function from the secondary sound source to the first microphone.
Is calculated based on the following equation, and the calculated coefficient (G ₁ ) is calculated as
Coefficient means for multiplying the output signal of the microphone; Where P _PA is the traveling wave component, P _rA is the backward wave component, and ΔL is the
1, the distance between the second microphone, L is between the secondary sound source and the second microphone
Interval, k is 2πf / c (f is excitation frequency, c is sound velocity) Calculation means for obtaining a difference between the multiplication result signal of the coefficient means and the output signal of the first microphone, and a difference signal corresponding to the difference An active control noise reduction device comprising: a signal processing unit that generates a secondary sound signal obtained by performing a predetermined process and supplies the secondary sound signal to the secondary sound source. 2. The active control noise reduction apparatus according to claim 1, wherein the signal processing means includes means for supplying a signal obtained by multiplying the difference signal by a predetermined coefficient to the secondary sound source.