JP2004037801A - Silencer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a silencer for efficiently silencing low-frequency stationary sounds over a wide range. <P>SOLUTION: Speakers 8a to 8d are arranged in proximity to a noise source 2 generating the noises having thtree-dimentional directivity in a contactless state. The noises from the noise source 2 are collected and microphones 10a to 10d collect noise signals. A controller 12 receives input of at least the noise signals and forms and supplies the silencing signals of the antiphase from the phase of the noise signals and of equal amplitude to the respective speakers independently from each other with each of the speakers. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a noise reduction device for eliminating noise, and more particularly to an active noise reduction device.
[0002]
[Prior art]
In the active silencer, a speaker is provided as an interference sound source in a place where the noise from the noise source propagates, and noise is collected by noise detection means provided between the noise source and the speaker, for example, a reference microphone. In some cases, the collected noise is converted into a digital signal and input to a digital adaptive filter to generate an interference signal, and the interference signal is amplified from a speaker to muffle the noise. An interference sound, which is a sound in which the interference sound based on the interference signal interferes with the noise, such as an error sound, is collected by an error microphone, digitized, and an adaptive filter is adjusted based on the error sound to mute the sound. Is performed optimally.
[0003]
In such a noise reduction device, when performing noise reduction over a wide range, the closer the noise source and the speaker are, the more their wavefronts match in a wide range. Therefore, they are arranged as close as possible. However, since the digital processing is performed, a time t for the digital processing is required, and the minimum distance between the noise source and the microphone is limited to the sound speed * t (m). Actually, the distance between the reference microphone and the speaker is a problem. However, considering that the reference microphone is located between the noise source and the speaker, the maximum distance between the reference microphone and the speaker is determined by the noise of the reference microphone. It will be the case where it is located in the immediate vicinity of the source, which is considered as described above.
[0004]
Here, if the noise to be silenced is a low-frequency steady sound, the noise source and the speaker can be brought closer to each other beyond the above-mentioned limit. Normally, it is necessary to generate the interference sound at the same timing as the original noise that generated the interference sound, but if it is a low-frequency steady sound, both the past noise wave and the current noise wave are the same. This is because it can be considered that the interference sound generated based on the past noise wave can be adapted to the current noise.
[0005]
In order to mitigate sound over a wide area by another method, a technique called multi-active muffling is adopted. This is disclosed in, for example, Japanese Patent Publication No. 1-501344, Japanese Unexamined Patent Application Publication No. 3-274897, and the like. In the signal processing calculation of a multi-system active silencer, not only the error path of its own system but also the error path of its own system The filter coefficients calculated using the error paths in other systems are used for the adaptive filter. With this method, it is also possible to mute the sound in a range that cannot be covered by only one interference sound source.
[0006]
[Problems to be solved by the invention]
However, in the low-frequency steady sound, the noise reduction can be expanded by bringing the noise source and the speaker closer to each other to muffle the sound. However, there is a limit to the approach between the noise source and the speaker. Since the muffling range is limited to the vicinity of the reference microphone, it is not possible to efficiently muffle the three-dimensional space simply by bringing the noise source and the speaker close to each other.
[0007]
In addition, in the case of multi-active silencing, it is necessary to increase the number of systems in order to achieve silencing in a three-dimensional space. If the number of systems is increased, it is necessary to calculate an error path between the systems. Is required, the apparatus becomes complicated, and the calculation takes time, which is not practical.
[0008]
SUMMARY OF THE INVENTION An object of the present invention is to provide a silencer that silences low-frequency stationary sound efficiently over a wide range.
[0009]
[Means for Solving the Problems]
In the noise suppressor according to the present invention, a plurality of interference sound sources are arranged in a predetermined area in close proximity to a noise source that generates standing wave noise three-dimensionally. As an interference sound source, for example, a speaker can be used. Noise detection means is provided between the noise source and the interference sound source. The noise detection means generates a noise detection signal that has detected the noise. The noise detection means can be, for example, closely attached to the noise source. An adaptive filter is provided in the predetermined area corresponding to each interference sound source. These adaptive filters filter the noise detection signal and supply an interference signal to the interfering sound source such that the interfering sound source causes the interfering sound to be emitted from the interfering sound source. Corresponding to each interference sound source and each adaptive filter, an interference state sound detection means is provided in the predetermined area. The interference state sound detection means generates an interference state sound signal that detects an interference state sound in a state where the interference sound interferes with the noise. The control unit updates the coefficient of the adaptive filter based on the interference state sound signal from the interference state sound detection unit and the noise detection signal. In this update, the interference state sound detection unit and the corresponding This is performed in consideration of a transfer function between the interfering sound source and a transfer function between the interference state sound detecting means and the interfering sound source that does not correspond thereto. That is, when viewed from the interfering sound source, so-called multi-active silencing is performed in consideration of not only the transfer function with the corresponding interfering state sound detector but also the transfer function with the non-corresponding interfering state sound detector. Has been done.
[0010]
In the silencer configured as described above, the so-called multi-active silence is performed after the interference sound source is arranged close to the noise source. Since the interfering sound source is arranged close to the noise source, it functions so as to eliminate the generation of the noise itself from the noise source, and the noise reduction range can be expanded. Further, since the interference sound from the closely arranged interference sound source is generated by the multi-active muffling, it can be muffled over a wider range.
[0011]
The distance between each of the interference sound sources and the noise detection means can be determined, for example, as follows. After the noise signal detection means generates the noise detection signal, the noise is moved from the position of the noise detection means to the interference sound source in a time shorter than a time required for generating an interference signal corresponding to the noise detection signal. The distance between each of the interference sound sources and the noise detecting means is selected so as to reach the position. This distance also indicates the distance between the noise source and each interfering sound source when the noise detection means is arranged in contact with the noise source.
[0012]
BEST MODE FOR CARRYING OUT THE INVENTION
The active silencer according to the first embodiment of the present invention silences noise from a noise source 2 as shown in FIG. As shown in FIG. 1B, for example, the noise source 2 is disposed in the middle of a building 4 having a height of 6 m and a building 6 having a height of 3 m, which are located at an interval of 6 m. As the noise source, for example, a speaker that generates a standing wave, for example, a low-frequency standing wave, specifically, a sine wave of 100 Hz is used.
[0013]
Around the noise source 2, a plurality of, for example, four interference sound sources, for example, speakers 8a to 8d are provided. These speakers 8a to 8d are arranged at 90-degree intervals around the noise source 2 and face the opposite side of the noise source 2, respectively. These speakers 8a to 8d are located close to the noise source 2, for example, approximately 30 cm from the center of the noise source 2, respectively.
[0014]
Outside the speakers 8a to 8d, interference state sound detecting means, for example, error microphones 10a to 10d are arranged so as to be located on the same straight line as the speakers 8a to 8d. These microphones 10a to 10d are located at a position of, for example, about 60 cm from the center of the noise source 2.
[0015]
The silencer is provided with a control device 12 to which the same sine wave signal as that supplied to the noise source 2 is supplied as a noise detection signal. Note that a noise detection unit that actually collects noise from the noise source 2, for example, a reference microphone, may be provided, and a noise detection signal representing the noise collected by the reference microphone may be supplied to the control device 12. The control device 12 includes, for example, a DSP, a CPU, and a memory, and configures a total of four adaptive digital filters corresponding to the speakers 8a to 8d. A noise detection signal is supplied to these digital filters, and an interference signal having the same phase and an opposite amplitude as the noise signal is generated and supplied to the corresponding speakers 8a to 8d to louder the interference sound. Each of the error microphones 10a to 10d detects a sound in a state where the interference sound from the speakers 8a to 8d actually interferes with the noise, that is, an interference state sound, and generates an interference state sound detection signal, for example, an error signal. This error signal is supplied to the control device 12, and each coefficient of each adaptive digital filter is updated in the control device 12 based on the error signal.
[0016]
The control device 12 is configured so as to be capable of multi-active silencing. When updating the coefficients of the digital filters, only the transfer function between each digital filter and the corresponding error microphone is considered. Instead, it also considers the transfer function between other error microphones. For example, when a digital filter corresponding to the error microphone 10a is considered, this digital filter naturally considers the transfer function between the error microphone 10a and the transfer function between the error microphone 10b and the error microphone 10b. The coefficient is updated in consideration of the transfer function with the error microphone 10d and the transfer function with the error microphone 10d. The same applies to other digital filters. By performing multi-active silencing in this manner, silencing can be performed over a wide range.
[0017]
The speakers 8a to 8d are arranged close to the noise source 2 and 30 cm, and the sine wave signal to the noise source 2 is supplied to the control device 12 as it is. This is equivalent to the fact that the virtual reference microphone and each of the speakers 8a to 8d are arranged at an interval of 30 cm. The distance is the time required for the reference microphone to generate a noise detection signal and to generate an interference signal corresponding to the noise detection signal (the noise detection signal is digitally processed by the control device 12 and released from the speaker). In this case, the noise arrives from the installation position of the reference microphone to the installation positions of the speakers 8a to 8d in a time shorter than the time required for sounding.
[0018]
FIGS. 2A and 2B conceptually show a state in which the range in which sound can be silenced changes depending on the distance between the noise source 2a and the interference sound source 11. The solid line shows the maximum value of the sound wave, and the broken line shows the maximum value. What is shown is the minimum value of the sound wave. In FIG. 2A, the noise source 2a and the interference sound source 11 are relatively separated from each other, and both are arranged so that the maximum value of the sound wave from the noise source 2a and the minimum value of the sound wave from the interference sound source 11 coincide. In this case, the range where the maximum value of the sound wave from the noise source 2a and the minimum value of the sound wave from the interference sound source 11 substantially overlap each other is relatively narrow as shown by a thin line. On the other hand, as shown in FIG. 2B, the noise source 2a and the interference sound source 11 are arranged close to each other, and the maximum value of the sound wave from the noise source 2a and the minimum value of the sound wave from the interference sound source 11 match. In such a case, as shown by a thin line, the range in which the two substantially overlap each other is widened. In general noise, when an attempt is made to perform active noise reduction, when noise at the position of the reference microphone existing between the noise source 2a and the interference sound source 11 reaches the position of the interference sound source 11, the noise is used as a source. It is necessary that the generated interference sound is generated. Therefore, a certain distance is required between the reference microphone and the interference sound source 11. Since a reference microphone needs to be installed between the noise source 2a and the interference sound source 11, if a certain distance is required between the reference microphone and the interference sound source 11, the interference microphone and the noise source 2a A further distance is required between and. When the reference microphone is brought into contact with the noise source 2a, the distance between the noise source 2a and the interference sound source 11 can be reduced most. Even in such a case, the distance between the noise source 2a and the interference sound source 11 is small. Some distance is required.
[0019]
However, in the case of standing wave noise, since its waveform is constant, the waveform of the past noise and the waveform of the current noise are the same, so that an interference sound is generated based on the noise collected by the reference microphone in the past. Even so, silencing becomes possible. Therefore, the distance between the interference sound source 11 and the noise source 2a is set to be shorter than the above-mentioned certain distance (when the noise at the position of the reference microphone reaches the interference sound source, the interference based on the noise is still present). The reference microphone and the interference sound source can be installed at a distance where no sound is generated, that is, at a state where the interference sound is delayed with respect to the noise.) In the case of active noise reduction, the noise can be reduced over a wide range. However, even if the noise source 2a and the interference sound source 11 are brought close to each other, there is a limit, and the standing wave noise cannot be efficiently silenced in the three-dimensional space.
[0020]
On the other hand, active silence includes multi-active silence. In order to achieve noise reduction in a three-dimensional space, if the number of error microphones, adaptive digital filters, and interfering sound sources is one, the number of systems must be increased. Increases, and an enormous amount of calculation is required, which cannot be realized.
[0021]
Therefore, in this embodiment, the distance between the speakers 8a to 8d, which are the interference sound sources, and the noise source 2 is reduced, and then the multi-active silence with a certain number of systems is performed to efficiently perform the noise reduction. Standing wave noise is muted.
[0022]
FIG. 3A shows the sound pressure distribution of the noise generated by the noise source 2 in a state where the muffler is not operated. The sound pressure level was measured at each point.
[0023]
FIG. 2B shows the vertical and horizontal directions of the noise source 2 within a range of 6 m square when the speaker 8 a, the microphone 10 a, and the corresponding digital filter on the building 6 side are operated. The sound pressure level was measured at 49 points in total. That is, it shows the sound pressure distribution when single active silencing is performed. As is clear from the comparison of FIGS. 9A and 9B, the area of the sound pressure of 90 dB and 85 dB is narrow, while the area of 80 dB is narrow near the driven speaker 10 a. It has a considerable width in the direction of the other speakers 10b to 10c. Therefore, the single active silencing cannot perform silencing in a wide range.
[0024]
FIG. 3C shows the sound pressure distribution when all the speakers 8a to 8d, the microphones 10a to 10d, and all the digital filters are operated to execute multi-active silencing. Are the same as those shown in FIGS. At this time, the maximum sound pressure is 89 dB, and the sound pressure distribution area of 85 dB, 80 dB, and 75 dB is only a very small area near the center of the noise source 2, and the sound is widely silenced.
[0025]
FIG. 4 (a) shows measured values of the muffling volume when the speaker 8a, the microphone 10a and the corresponding digital filter on the building 6 side are operated, and the speaker 8a, the microphone 10a and 10 dB is performed in the installation direction of the speaker 8a, but only about 5 dB is obtained in the direction orthogonal to the speaker 8a and the microphone 10a, and almost large in the opposite direction to the installation direction of the speaker 8a and the microphone 10a. No silencing has been obtained. From now on, it is clear that single active silencing does not provide extensive silencing.
[0026]
FIG. 5 (b) shows measured values of the sound deadening volume when the multi-active sound deadening is performed by operating all the speakers 8a to 8d, the microphones 10a to 10d, and all the digital filters. A silencing volume of 15 dB is obtained in almost the entire region, and silencing is performed in a wide range.
[0027]
As described above, even for the standing wave noise, the speakers 8a to 8d are arranged close to the noise source 2 and the multi-active silence is performed, so that the noise can be effectively silenced over a wide range.
[0028]
FIG. 5 shows a silencer according to a second embodiment. In this silencer, for example, a transformer for extra high voltage is used as the noise source 2a. The transformer 2a is installed between the buildings 4a and 6a, each of which stands at an interval of about 8m and has a height of about 10m, at a position slightly biased toward the building 4a from the center thereof. The transformer 2a has a height, width, and depth of, for example, 2 m, and generates noise of various frequencies. Among them, a low-frequency standing wave of 120 Hz, which is a second harmonic of a power supply frequency of 60 Hz, is used. The main purpose is to mute the sound.
[0029]
A noise detecting means, for example, a vibration pickup 10 is arranged on the building 6a side of the transformer 2a. It is mounted directly on the face of the transformer 2a. The transformer 2a is arranged outdoors, and the noise detecting means is affected by wind and rain. However, since the vibration pickup 10 is used as the vibration detecting means, the influence from the surroundings can be minimized. . In consideration of the influence on the transformer 2a, the vibration pickup is not attached by welding or the like, but a magnet sheet is bonded to the vibration pickup 10 and attached to the transformer 2a by the magnet sheet. A relatively large vibrating surface is selected at the mounting position of the vibration pickup 10. The vibration signal picked up by the vibration pickup is supplied to the control device 12.
[0030]
A plurality of, for example, four interference sound sources, for example, speakers 8a to 8d, are arranged at positions closer to the transformer 6a than the vibration pickup 10 so as to intersect a straight line connecting the buildings 4a and 6a. Further, they are arranged in an arc shape so as to surround the transformer 2a. A mute signal from the control device 12 is supplied to these speakers 8a to 8d.
[0031]
The control device 12 includes a total of four adaptive digital filters corresponding to the respective speakers 8a to 8d. These digital filter coefficients are provided with interference state sound detecting means, for example, error microphones 14a to 14d, corresponding to the respective speakers 8a to 8d, at a position where noise is to be most suppressed, for example, at a position close to the building 6a. These error microphones 14a to 14d generate an error signal indicating how noise and interference sound actually interfere with each other at the position where they are installed. These error signals are supplied to the control device 12, and the coefficients of each digital filter are updated so that multi-active silencing can be performed as in the first embodiment.
[0032]
Further, since the speakers 8a to 8d are provided near the transformer 2 so as to satisfy the conditions described in relation to the first embodiment, the speakers 8a to 8d can be silenced in a wide range in combination with the multi-active silencer. Is possible. In addition, since the sound is muted by the speakers 8a to 8d, a high voltage is supplied, and it is not necessary for a person to touch the transformer 2a, which is preferably not touched by a person. high.
[0033]
In the above-described embodiments, four speakers are used. However, the number of speakers is not limited to four, and a smaller number or a larger number of speakers may be used. Further, in the second embodiment, a reference microphone can be used as noise detection means instead of the vibration pickup.
[0034]
【The invention's effect】
As described above, according to the present invention, since the interfering sound source and the noise source are arranged close to each other and multi-active silencing is performed, silencing can be performed over a wide range.
[Brief description of the drawings]
FIG. 1 is a plan view and a side view of a muffler according to a first embodiment of the present invention.
FIG. 2 is an explanatory diagram of a state in which a muffling range changes depending on positions of a noise source and an interference sound source.
FIG. 3 is a sound pressure distribution diagram in the silencer of FIG.
FIG. 4 is a distribution diagram of a sound deadening volume in the sound deadening device of FIG. 1;
FIG. 5 is a plan view of a muffler according to a second embodiment of the present invention.
[Explanation of symbols]
2 2a Noise sources 8a to 8d Speaker 10 Noise pickup 12 Control device (control means)
14a to 14d error microphone (interference state sound detection means)

Claims (2)

A plurality of interference sound sources arranged in close proximity to a noise source that generates standing wave noise three-dimensionally in a predetermined area,
A noise detection unit that is provided between the noise source and the interference sound source and generates a noise detection signal that detects the noise;
An interference signal is provided in the predetermined area corresponding to each of the interfering sound sources, filters the noise detection signal, and outputs an interfering signal to the interfering sound source such that the interfering sound source emits an interfering sound that interferes with the noise. A plurality of adaptive filters to supply;
An interference state sound that is provided in the predetermined area corresponding to each of the interference sound sources and the adaptive filters and generates an interference state sound signal that detects an interference state sound in which the interference sound interferes with the noise. Detecting means;
Considering the transfer function between the interference state sound detecting means and the corresponding interference sound source, and the transfer function between the interference state sound detection means and the non-corresponding interference sound source, Based on the interference state sound signal and the noise detection signal from the interference state sound detection means, based on the control means for updating the coefficient of the corresponding adaptive filter,
Equipped silencer. 2. The noise reduction device according to claim 1, wherein a distance between each of said interference sound sources and said noise signal detection means is determined by said noise signal detection means generating a noise detection signal, and then an interference signal corresponding to said noise detection signal is generated. A noise reduction device that selects a distance between each of the interference sound sources and the noise detection means so that the noise reaches the position of the interference sound source from the position of the noise detection means in a time shorter than the time required for the noise detection.

Images