JPH06161468A - Noise eliminating device and on-vehicle acoustic device - Google Patents

Abstract

PURPOSE:To provide the noise eliminating device which accurately eliminates a noise even when control transfer characteristics have a zero point or deep dip and the on-vehicle acoustic device which uses the noise eliminating device. CONSTITUTION:The device consists of a noise detector 1, adaptive filters 2 which receive its detection signal, speaker groups 3 which consist of plural speakers cascaded to the adaptive filter 2 respectively, a microphone group 6 of plural microphones which detect differences between a sound from a noise source and sounds from both the speaker groups 3, adders 11 which add the output signals of the microphone group 6, computing elements 7 which perform convolutional arithmetic between the transfer characteristics from the speaker group 3 to the microphone group calculated from the output signal of the adder 11 and the noise detection signal, normalization units 8 which normalize the values of the output signals of the computing element 7, and multipliers 9 which multiply the normalization outputs by the error output of the microphone group 6 and output update signals for the coefficients of the adaptive filters 2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise canceller using active noise control in a noisy environment and a vehicle-mounted acoustic device using the same.

[0002]

2. Description of the Related Art In recent years, an active noise control method has been proposed in which a control sound is output from a speaker by using a digital signal processing technique to mute environmental noise at a listening position.

FIG. 4 shows a block diagram of a noise canceling apparatus to which the above-mentioned conventional noise canceling method is applied.

The detection signal of the noise source by the noise detector 1 is
The control transfer characteristics 4a ₁ , 4b ₁ , 4c ₁ , 4 are radiated by the speakers 3a-3d through the adaptive filters 2a-2d.
The noise signal that has passed through the noise transfer characteristic 5a on the one hand through d ₁ is canceled out, and the error is detected by the microphone 6a. The detection signal of the noise source is also input to the convolution calculator 7a in which the characteristic of the control transfer characteristic is set, passes through the normalizer 8a, and then is multiplied by the error detection signal 10a by the multiplier 9a, and this signal is used as an update signal. It is input to the adaptive filter 2a.

Similarly, the detection signal of the noise source from the noise detector 1 is passed through the adaptive filters 2a to 2d to the speaker 3a.
~ 3d, control transfer characteristics 4a ₂ , 4b ₂ , 4
The noise signal which has passed through c ₂ and 4d _{2 and} has passed through the noise transfer characteristic 5b is canceled and the error is detected by the microphone 6b. The noise source detection signal is also input to the convolution calculator 7b in which the control transfer characteristic is set, passes through the normalizer 8b, and then the multiplier 9b detects the error detection signal 10b.
And this signal is multiplied as an update signal by the adaptive filter 2
Input to b.

Similarly, the noise source detection signal from the noise detector 1 is radiated by the speakers 3a to 3d through the adaptive filters 2a to 2d, and the control transfer characteristics 4a ₃ and 4 are obtained.
The noise signal which has passed through b ₃ , 4c ₃ and 4d _{3 and} has passed through the noise transfer characteristic 5c is canceled and the error is detected by the microphone 6c. The noise source detection signal is also input to the convolution calculator 7c in which the characteristic of the control transfer characteristic is set,
After passing through the normalizer 8c, the multiplier 9c multiplies the error detection signal 10c, and this signal is input to the adaptive filter 2c as an update signal.

Further, the noise source detection signal from the noise detector 1 is passed through the adaptive filters 2a to 2d to the speakers 3a to 3d.
The control transfer characteristics 4a ₄ , 4b ₄ , 4c are radiated by 3d.
_The noise signal which has passed through ₄ and 4d _{4 and} has passed through the noise transfer characteristic 5d is canceled, and the error is detected by the microphone 6d. Further, the detection signal of the noise source is also input to the convolution calculator 7d in which the characteristic of the control transfer characteristic is set, and the normalizer 8
After passing d, the multiplier 9d multiplies the error detection signal 10d, and this signal is used as an update signal by the adaptive filter 2d.
Entered in.

[0008]

However, in order to obtain a sufficient muffling volume and muffling area in the passenger compartment or the like, the above-mentioned conventional method requires control by a multi-system control transfer system, and the adaptive filter There is a problem in that the cost increases because a large number of devices are required and the scale of the device is increased. Further, the conventional method requires a stereo reproduction speaker in addition to the muffling speaker, which causes a problem of high cost.

The present invention is intended to solve the above-mentioned problems. Noise can be accurately eliminated even when there is a zero point or a deep dip in the control transfer characteristic, and noise can be reduced at a plurality of positions where a wide silence region can be obtained. An object of the present invention is to provide a noise canceling device that can realize the same sound deadening effect in a small-scale device when canceling, and an in-vehicle acoustic device using the noise canceling device.

[0010]

In order to achieve the above object, a noise canceling apparatus of the present invention includes a noise detector for detecting noise from a noise source, and first and first noise detector signals to which the noise detecting signal is input. Second adaptive filter, first and second speaker groups each including a plurality of speakers cascade-connected to the first and second adaptive filters, noise from a noise source, and the first and second First and second microphone groups composed of a plurality of microphones for respectively detecting an error from a sound from a speaker group, and first and second summing output signals of the first and second microphone groups, respectively. A second adder, the noise detection signal, and a first transfer characteristic from the first speaker group to the first microphone group calculated from the output signal of the first adder, the first addition From the output signal of the vessel Second transfer characteristics from the second speaker group to be calculated to the first microphone group, the second
Third transfer characteristic from the first speaker group to the second microphone group calculated from the output signal of the adder of the second speaker group and the second transfer characteristic from the second speaker group calculated from the output signal of the second adder First and second arithmetic units for performing a convolution operation with a fourth transfer characteristic up to the microphone group,
And the first and second normalizers for normalizing the magnitudes of the output signals of the second and third arithmetic units, respectively, and the normalized output and the error signal output from the first and second microphone groups. The first and second multipliers output the updated signals of the coefficients of the first and second adaptive filters, respectively.

Furthermore, the on-vehicle audio system of the present invention has a structure in which a sound reproducing device for inputting an audio signal to the speaker having the above structure is added, and the speaker and the microphone are arranged in the vehicle compartment.

[0012]

With the above structure, the noise canceller of the present invention has a plurality of microphones and a plurality of speakers arranged therein, uses a plurality of control transmission paths, and simultaneously transmits the muffling control sound while performing signal addition processing. Even if there is a zero point or a deep dip in the frequency characteristic of one control transfer characteristic, it can be compensated by another transfer characteristic, and the sound can be silenced accurately with a small-scale hardware configuration. Further, since the vehicle-mounted audio device of the present invention can mute sound while performing stereo reproduction with the speaker, it is possible to provide an inexpensive vehicle-mounted audio device.

[0013]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a noise canceller according to a first embodiment of the present invention. The configuration and operation will be described in detail below.

The signals detected by the noise detector 1 are input to the adaptive filters 2a and 2b, and the speakers 3a and 3b, which are the first speaker group, and the speakers, which are the second speaker group, are radiated as sound waves. The same control signal is input to each of 3c and 3d, and the signals output from these speakers are control transfer characteristics (hereinafter referred to as control sound).
Noise transmission characteristics (hereinafter referred to as noise) 5 reach the microphones 6a to 6d via 4a _{1 to} 4d ₄ as control sounds.
Erase the sounds from a-5d. Here, the speakers 3a to 3
d is placed at an arbitrary position in a desired space where the sound is muted. The error between the control sound and the noise is caused by the plurality of microphones 6a ...
6d, microphones 6a and 6b, 6
The microphones 6a are grouped into c and 6d.
The error detection signals 10a and 10b output from the output terminals 6b and 6b are added by the adder 11a to the multiplier 9a, and the error detection signals 10c and 10c output from the microphones 6c and 6d are also added.
10d is added by the adder 11b and input to the multiplier 9b.

On the other hand, the detection signal obtained by the noise detector 1 is input to the convolution calculators 7a and 7b to calculate the characteristics of the control sound, and the signals are normalized by the normalizers 8a and 8b, respectively. The added signals 12a, 12a,
12b is multiplied and is input to the adaptive filters 2a and 2b as a coefficient update signal. For example, a learning identification method or the like can be used for such an updating algorithm (see “Application of Digital Signal Processing” P219, Corona Company, Institute of Electrical Communication). The portion including the adaptive filter, the convolutional computing unit, the normalizer, the multiplier, and the adder surrounded by the broken line in FIG. 1 is hereinafter referred to as an ANC control unit.

With the above arrangement, even if the frequency characteristic of the transfer characteristic has a zero point or a deep dip, it is possible to control the frequency by setting the control transfer characteristic to an optimum value. Further, as apparent from a comparison of FIG. 1 with FIG. 4, if the configuration of this embodiment is adopted, an adaptive filter, a convolution operator, a normalizer,
Only two multipliers are required, and a good muffling effect can be obtained with a small-scale hardware configuration.

FIG. 2 is a block diagram of a noise canceller according to the second embodiment of the present invention. The configuration and operation will be described in detail below.

The signals detected by the noise detector 1 are input to the adaptive filters 2a and 2b, and the speakers 3a and 3b which are the first speaker group and the speakers which are the second speaker group so as to radiate sound waves. The same control signal is input to each of 3c and 3d, and signals output from these speakers are control transfer characteristics (hereinafter referred to as control sound) 4a.
₁ is reached as the control sound microphone 6a~6d via ～4D _4, the noise transfer characteristics (hereinafter, referred to as noise). 5a-
Erase the sound from 5d. Here, the speakers 3a to 3d are arranged at arbitrary positions in a desired space where the sound is silenced. The error between the control sound and the noise is caused by a plurality of microphones 6a to 6d.
The microphones 6a and 6b and 6c and 6d are grouped into the microphones 6a and 6d.
The error detection signals 10a and 10b output from b are added by the adder 11a to the multiplier 9a, and the error detection signals 10c and 10c output from the microphones 6c and 6d are also added.
d is added by the adder 11b and input to the multiplier 9b.

On the other hand, the detection signal obtained by the noise detector 1 is input to the convolution calculators 7a and 7b to calculate the characteristics of the control sound, and the signals are normalized by the normalizers 8a and 8b, respectively. The added signals 12a, 12a,
12b is multiplied and is input to the adaptive filters 2a and 2b as a coefficient update signal. Such an update algorithm can also use the learning identification method described above. Also,
The stereo signals of the L channel and the R channel output from the sound reproducing devices 14a and 14b are respectively supplied to the speaker 3
It is configured to input to a, 3b and 3c, 3d.

With the above arrangement, even if the frequency characteristic of the transfer characteristic has a zero point or a deep dip, it is possible to control the frequency by setting the control transfer characteristic to an optimum value. Further, as is clear from comparison of FIG. 2 with FIG. 4, if the configuration of this embodiment is adopted, an adaptive filter, a convolution operator, a normalizer,
The number of multipliers required is two each, and it is possible to obtain a good muffling effect and a stereo effect with a small-scale hardware configuration while using both the muffling speaker and the music reproducing speaker.

Further, in this embodiment, the number of microphones used is four and two microphones are added, but the number of microphones and the number of microphones to be added may be arbitrary, and a larger number produces a better muffling effect. Obtainable.

FIG. 3 is a block diagram of a vehicle-mounted acoustic device in which the noise canceller according to the second embodiment of the present invention is mounted on an automobile. The configuration and operation will be described in detail below.

By the noise detector 1 arranged in the passenger compartment,
A noise signal such as an engine sound is input to the adaptive filter 2a, and the same control signal is output to the speakers 3a and 3b, which are the first speaker group, and the speakers 3c and 3d, which are the second speaker group, so that sound waves are emitted. Is entered. Then, the sound waves emitted from the speakers reach the microphones 6a to 6d as control sounds (not shown), and the sounds from the noises (not shown) are erased. Here, the speakers 3a and 3b are provided on the headrest portion of the driver's seat 15a,
The speakers 3c and 3d are arranged in the headrest portion of the passenger seat 15b. The error between the control sound and the noise is detected as an error signal by the microphones 6a to 6d arranged in the headrest part of the driver's seat and the headrest part of the passenger's seat. The error signals output from 6a and 6b and 6c,
AN so that the error signal output from 6d becomes zero.
The signal processed by the C control units 13a and 13b is the speaker 3
a, 3b and speakers 3c, 3d are driven. This signal is normally exchanged via the ANC control cable, but the effect is not changed not only by the cable but also by radio waves or the like. Further, the R channel signal of the stereo signal output from the sound reproducing device 14a is sent to the speaker 3a located in the driver's seat, and the L channel signal is sent to the speaker 3a.
Similarly, the R channel signal of the stereo signal output from the sound reproduction device 14b is input to the speaker 3c located in the passenger seat, and the L channel signal is input to the speaker 3b.

With the above configuration, even if there is a zero point or a deep dip in the frequency characteristic of the transfer characteristic in the vehicle interior, the control transfer characteristic can be set to an optimum value to control the frequency. it can. Also,
As is clear from comparison of FIG. 2 with FIG. 4, if the configuration of this embodiment is adopted, only two adaptive filters, convolutional arithmetic units, normalizers and multipliers are required, and a muffling speaker and music reproduction are provided. It is possible to obtain a good muffling effect and a stereo effect with a small-scale hardware configuration while also using as a speaker for audio.

Further, in this embodiment, two microphones are arranged in each of the driver's seat and the passenger's seat, but one microphone may be arranged in each of the driver's seat, the passenger's seat, the rear right seat and the rear left seat. Well, if done in this way, it is possible to mute the entire passenger compartment. Further, the number of microphones and the number to be added may be arbitrary, and a larger number can provide a better muffling effect.

[0027]

As is apparent from the above description, the noise canceling apparatus of the present invention has a plurality of microphones and a plurality of speakers arranged therein and uses a plurality of control transmission paths to simultaneously perform signal addition processing. Since the mute control sound is transmitted while performing, even if there is a zero point or a deep dip in the frequency characteristic of one control transfer characteristic, it is compensated for by another transfer characteristic, and the muffling is accurately performed with a small-scale hardware configuration. be able to. Further, since the vehicle-mounted noise canceller of the present invention can be used both as an audio reproducing speaker and a noise canceller, it is possible to provide an inexpensive vehicle silencer.

[Brief description of drawings]

FIG. 1 is a block diagram of a noise canceller according to a first embodiment of the present invention.

FIG. 2 is a block diagram of a noise canceller according to a second embodiment of the present invention.

FIG. 3 is a configuration diagram of a vehicle-mounted noise canceller in which a noise canceller according to a second embodiment of the present invention is mounted on an automobile.

FIG. 4 is a block diagram of a noise canceller in a conventional example.

[Explanation of symbols]

1 the noise detector 2a~2d adaptive filter 3a~3d speaker 4a ₁ ~4d ₄ control transfer characteristic 5a~5d noise transfer characteristics 6a~6d microphone 7a~7d convolver 8a~8b normalizer 9a~9b multipliers 10a~ 10d Error detection signal 11a, 11b Adder 12a, 12b Addition signal 13a, 13b ANC control part 14a, 14b Sound reproduction device 15a Driver's seat 15b Passenger's seat 15c Rear seat 16a, 16b ANC control signal 17a, 17b Sound reproduction signal

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroyuki Hashimoto 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (4)

[Claims] 1. A noise detector for detecting noise from a noise source, first and second adaptive filters to which the noise detection signal is input, and cascade connection to the first and second adaptive filters, respectively. First and second consisting of a plurality of speakers
Speaker group, noise from a noise source, and the first and second
A first and a second microphone group composed of a plurality of microphones for respectively detecting an error from a sound from a speaker group of the first microphone, and a first addition circuit for adding output signals of the first and second microphone groups, respectively. And a second adder, the noise detection signal, and a first transfer characteristic from the first speaker group to the first microphone group calculated from the output signal of the first adder, the first transfer characteristic. A second transfer characteristic from the second speaker group to the first microphone group calculated from the output signal of the adder, and a second transfer characteristic from the first speaker group calculated from the output signal of the second adder. A first convolution operation with a third transfer characteristic up to the microphone group and a fourth transfer characteristic from the second speaker group up to the second microphone group calculated from the output signal of the second adder And the second And adder, the first and second normalizing each magnitude of said first and second operator of the output signal
A normalizer for multiplying the normalized output by the error signals output from the first and second microphone groups, respectively, and outputting updated signals of the coefficients of the first and second adaptive filters. A noise canceller comprising a first multiplier and a second multiplier. 2. A noise detector for detecting noise from a noise source, first and second adaptive filters to which the noise detection signal is input, and cascade connection to the first and second adaptive filters, respectively. First and second consisting of a plurality of speakers
And a sound reproducing device for inputting an audio signal to the first and second speaker groups, and an error between a noise from a noise source and a sound from the first and second speaker groups, respectively, is detected. A first and second microphone group composed of a plurality of microphones, first and second adders for adding output signals of the first and second microphone groups, respectively, and the noise detection signal. A first transfer characteristic from a first speaker group to a first microphone group calculated from the output signal of the first adder, the first
Second transfer characteristic from the second speaker group to the first microphone group calculated from the output signal of the adder of the second adder, and from the first speaker group to the second microphone calculated from the output signal of the second adder. Third transfer characteristic up to the microphone group of
First and second arithmetic units for performing a convolution operation with a fourth transfer characteristic from the second speaker group to the second microphone group calculated from the output signal of the second adder; First and second normalizers that normalize the magnitudes of the output signals of the first and second arithmetic units, respectively, and the normalized output and the error signals output from the first and second microphone groups. And a first and a second multiplier for outputting the update signals of the coefficients of the first and second adaptive filters, respectively. 3. A vehicle-mounted acoustic device in which the noise canceller according to claim 2 is mounted on an automobile, and a speaker and a microphone constituting the noise canceller are arranged in a vehicle interior space. 4. The vehicle-mounted acoustic device according to claim 3, wherein at least two speakers are arranged in each headrest portion of a driver's seat and a passenger's seat.