JPH04190396A - Noise eliminating device - Google Patents

Abstract

PURPOSE:To precisely suppress sound even though a zero point or a deep dip is present in a frequency characteristic as one control transmission characteristic, by transmitting control sounds for suppressing sound with the use of a plurality of control transmission paths. CONSTITUTION:A detection signal for a noise detector 1 is inputted to adaptive filters 2, 4 from which outputs are issued from speakers 3, 5. Then, they pass through control transmission characteristic C1, C2 and reaches, as control sounds, a microphone 13 so that noise from a noise source 12 is suppressed. Where the speakers 3, 5 are set so that the characteristics C1, C2 are different from each other. Errors between the control sounds and the noise are detected by a microphone 13 and are delivered to multipliers 8, 11. Meanwhile, a detection signal from a noise detector 1 is subjected to computation by convolution computers 8, 9 having the characteristics C1, C2, and is then multiplied with the error signals from the microphone by the multipliers 8, 11. Thus multiplied values are inputted as coefficient updating signals to the adaptive filters 2, 4. With this arrangement, it is possible to obtain a sound suppressing effect even though the frequency characteristic as one of the transmission characteristic has a zero point or a deep dip.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a noise canceling device using active noise control in a noisy environment.

2. Description of the Related Art In recent years, active noise control devices have been proposed that use digital signal processing technology to output control sound from a speaker to mute environmental noise at a listening position.

Conventionally, this type of noise canceling device generally had the configuration shown in FIG. 7. The configuration will be explained below with reference to FIG.

As shown in the figure, a detection signal of a noise source 12 by a noise detector 1 is emitted from a speaker 3 through an adaptive filter 2 to obtain a control transfer characteristic CI.

This control transfer characteristic CI is canceled out by the noise signal that has passed through the noise transfer path SI, and the error is detected by the microphone 13. Further, the noise detection signal is input to a convolution calculator 6 which sets the characteristics of the control transfer characteristic CI, passes through a normalizer 7, is multiplied by an error detection signal by a multiplier 8, and is input to the adaptive filter 2 as an update signal. Ru.

Problems to be Solved by the Invention However, in the conventional example described above, the adaptive filter and speaker are in one system, and the configuration allows only control using the control transfer system CR, so there is a zero point or deep dip in the frequency characteristic of the transfer characteristic of c■. If such a frequency exists, it is impossible to control that frequency, so there is a problem in that a silencing effect cannot be obtained at that frequency.

Also, since there is only one system for error detection, microphone 1
There was a second problem in that noise could only be canceled near the 0 position.

Furthermore, if a plurality of control circuits are provided as an extension of the conventional example in order to solve the above two problems, there is a third problem in that the scale of the device increases proportionally.

The present invention solves the above-mentioned problems, and can perform accurate noise reduction even when there is a zero point or deep dip in the control transfer characteristic, erase noise even at multiple positions, and provide a wide erasure area. It is an object of the present invention to provide a noise canceling device that can reduce the scale of the device even if a plurality of control circuits are provided.

Means for Solving the Problems In order to achieve the above object, the first invention provides a noise detector installed near a noise source, first and second adaptive filters receiving the noise detection signals, first and second speakers respectively connected in cascade to both of the adaptive filters;
a microphone that detects an error between the sound from the noise source and the sound from both of the speakers; a first computing unit that performs a convolution operation between the transfer characteristic from the first speaker to the microphone and the noise detection signal; a first normalizer that normalizes the magnitude of the output signal of the arithmetic unit; and a first normalizer that multiplies the normalized output by the error output of the microphone and outputs an update signal of the coefficients of the first adaptive filter. and a multiplier of
a second arithmetic unit that performs a convolution operation between the transfer characteristic from the second speaker to the microphone and the noise detection output; a second normalizer that normalizes the magnitude of the output signal of the arithmetic unit; It consists of a second multiplier that multiplies the normalized output and the error output of the microphone and outputs an update signal for the coefficients of the second adaptive filter.

A second aspect of the present invention also provides a noise detector installed near a noise source, first and second adaptive filters that receive the noise detection signals, and a speaker that receives a summed signal of the outputs of both the adaptive filters. , first and second microphones that detect an error between the sound from the noise source and the sound from the speaker, and a first and second microphone that performs a convolution operation between the transfer characteristic from the speaker to the first microphone and the noise detection signal. 1 arithmetic unit, a first normalizer that normalizes the magnitude of the output signal of the arithmetic unit, and a coefficient of the first adaptive filter that multiplies the normalized output by the error output of the first microphone. a first multiplier that outputs an update signal; a second arithmetic unit that performs a convolution operation between the transfer characteristic from the speaker to the second microphone and the noise detection output; and a second multiplier that performs a convolution operation of the noise detection output; A second multiplier that multiplies the normalized output by the error output of the second microphone and outputs an update signal for the coefficients of the second adaptive filter. .

The third aspect of the present invention also provides a noise detector installed near a noise source, first and second adaptive filters that receive the noise detection signals, and first and second adaptive filters connected in cascade to both of the adaptive filters, respectively. a second speaker, a microphone that detects an error between the sound from the noise source and the sound from both of the speakers, a normalizer that normalizes the magnitude of the noise detection signal, and an error between the normalized output and the microphone. a multiplier that multiplies the output; a first arithmetic unit that performs a convolution operation between the multiplier output and a characteristic obtained by normalizing the transfer characteristic from the first speaker to the microphone by its energy; and a second speaker and the microphone. a second arithmetic unit that performs a convolution operation with the output of the multiplier and a characteristic obtained by normalizing the transfer characteristic up to the energy thereof, and updates the coefficients of the first adaptive filter with the output of the first arithmetic unit. The coefficients of the second adaptive filter are updated using the output of the second arithmetic unit.

Furthermore, a fourth aspect of the present invention includes a noise detector installed near a noise source, first and second adaptive filters that receive the noise detection signals, and a speaker that receives a summed signal of the outputs of both the adaptive filters. , first and second microphones that detect an error between the sound from the noise source and the sound from the speaker; a normalizer that normalizes the magnitude of the noise detection signal; a first multiplier that multiplies the error output of the microphone; and a first arithmetic unit that performs a convolution operation between the transmission characteristic from the speaker to the first microphone normalized by its energy and the output of the first multiplier. , and a second multiplier that multiplies the normalized output by the error output of the second microphone, and a characteristic obtained by normalizing the transfer characteristic from the speaker to the second microphone by its energy, and a second multiplier. and a second arithmetic unit that performs a convolution operation with the output of the first arithmetic unit.
The coefficients of the second adaptive filter are updated using the output of the second arithmetic unit.

Furthermore, the fifth aspect of the present invention includes a noise detector installed near the noise source, and a first, second, and second noise detectors that receive the noise detection signals. a third and fourth adaptive filter; a first speaker that inputs a summed signal of the output of the first adaptive filter and the second adaptive filter; and an output of the third adaptive filter and the fourth adaptive filter. a second speaker that inputs a sum signal from the noise source, first and second microphones that detect an error between the sound from the noise source and the sound from both speakers, and from the first speaker to the first microphone. a first arithmetic unit that performs a convolution operation between the transfer characteristic of the noise detection output and the noise detection output; a first normalizer that normalizes the output signal of the arithmetic unit; and an error between the normalized output and the first microphone. The first to multiply the output
a multiplier, a second computing unit that performs a convolution operation between the transfer characteristic from the first speaker to the second microphone and the noise detection output, and a second normalization unit that normalizes the output signal of the computing unit. a second multiplier that multiplies the normalized output by the error output of the second microphone; and a third multiplier that performs a convolution operation between the transfer characteristic from the second speaker to the first microphone and the noise detection output. a computing unit, a third normalizer that normalizes the output signal of the computing unit, a third multiplier that multiplies the normalized output by the error output of the first microphone, and a second speaker. a fourth arithmetic unit that performs a convolution operation between the transfer characteristic up to the second microphone and the noise detection output; a fourth normalizer that normalizes the output signal of the arithmetic unit; and a fourth normalizer that normalizes the output signal of the arithmetic unit; and a fourth multiplier for multiplying the error output of the second microphone by the error output of the second microphone. The outputs of the 2nd, 3rd and 4th multipliers cause the output of the 1st, 2nd, . The coefficients of the third and fourth adaptive filters are being updated.

Further, the sixth aspect of the present invention includes a noise detector installed near a noise source, and a first and a second detector that receive the noise detection signal. a third and fourth adaptive filter; a first speaker that receives a sum signal of the outputs of the first adaptive filter and the second adaptive filter; and an output of the third adaptive filter and the fourth adaptive filter; a second speaker that inputs the summed signal of , first and second microphones that detect an error between the sound from the noise source and the sound from both speakers, and a normalizer that normalizes the magnitude of the noise detection signal. a first multiplier that multiplies the normalized output by the error output of the first microphone; and a first multiplier that multiplies the normalized output by the error output of the first microphone; a first arithmetic unit that performs a convolution operation with the output of the first multiplier, and a convolution operation between the first multiplier output and a characteristic obtained by normalizing the transfer characteristic from the second speaker to the first microphone by its energy; a second arithmetic unit that performs
a second multiplier that multiplies the error output of the microphone, and performs a convolution operation between the transmission characteristic from the first speaker to the second microphone normalized by its energy and the output of the second multiplier. a third arithmetic unit, and a second arithmetic unit;
a fourth arithmetic unit that performs a convolution operation of a characteristic obtained by normalizing the transfer characteristic from the first speaker to the second microphone by its energy and the output of the second multiplier; Second. The first . Second. The coefficients of the third and fourth adaptive filters are updated.

Effects of the Invention With the above-described configuration, firstly, the present invention transmits the silencing control sound using a plurality of control transmission paths, so even if there is a zero point or deep dip in the frequency characteristic of one control transmission characteristic, the other Since this is compensated for by the transmission characteristics of Second, by providing a plurality of error detection microphones and adaptive filters according to the silencing area, noise from a noise source can be muted at a plurality of microphone positions, and the silencing area can be widened. Thirdly, because the normalization process that requires a large amount of calculation in algorithms such as the learning identification method is moved to the stage before the convolution calculation process and is processed in common, when the number of adaptive filters and control transfer characteristics is two or more, The scale of the entire device can be reduced.

EXAMPLE Hereinafter, a first example of the present invention will be described with reference to FIG. The purpose of this embodiment is to perform accurate muffling even when a zero point or deep dip exists in the frequency characteristic of the control transfer characteristic from the speaker to the error detection microphone.

The detection signal of the noise detector 1 is input to the adaptive filters 2 and 4, and the output thereof is radiated from the respective speakers 3.5, passes through the control transfer characteristics C1 and C2, and reaches the microphone 13 as a control sound, and the noise source is Eliminate noise from 12. Here, the speaker 3.5 is set so that CI and C2 are different. Microphone 1 calculates the error between this control sound and the noise.
3 and input to the multiplier 8.11. On the other hand, the detection signal of the noise detector 1 is calculated by a convolution calculator 6.9 having characteristics of C1 and 02, respectively, and is multiplied by the error signal of the microphone 13 by a multiplier 8.11. 4 as those coefficient update signals. This updating algorithm can use, for example, a learning identification method ("Applications of Digital Signal Processing J p. 21
9 Coronasha, Institute of Electrical and Communication Engineers).

With this configuration, even if there is a zero point or a deep depth in the frequency characteristic of the C1 transfer characteristic, the frequency can be controlled by the C2 transfer characteristic, so that a good noise reduction effect can be obtained.

Note that this embodiment can be naturally extended and applied even when two or more sets of adaptive filters and speakers are provided.

Next, a second embodiment of the present invention will be described with reference to FIG. The purpose of this embodiment is to widen the area in which noise is eliminated.

The error detection microphones 13 and 14 are installed in the silencing area, the detection signal of the noise detector 1 is input to the adaptive filters 2 and 13, and the outputs of the adaptive filters 2 and 13 are added together and radiated from the speaker 3 to determine the control transfer characteristic. The control sound reaches the error detection microphone 13 through C11 and the error detection microphone 14 through the control transfer characteristic C12, thereby canceling the noise from the noise source 12, respectively. Errors between the noise from the noise source 12 and these control sounds are detected by microphones 13 and 14, the detection output of the microphone 13 is input to the multiplier 8, and the detection output of the microphone 14 is input to the multiplier 11. The control transfer characteristic C1l from the speaker 3 to the microphone 13 is set in the convolution calculator 6, and after performing a convolution operation with the noise detection signal, it is normalized by the normalizer 7, and the multiplier 8 converts it into the error detection signal of the microphone 13. The multiplied result is input to the adaptive filter 2 as a coefficient update signal. Similarly, the control transfer characteristic C12 from the speaker 3 to the microphone 14 is set in the convolution calculator 9, and after performing a convolution operation with the noise detection signal, it is normalized by the normalizer 10, and the error of the microphone 14 is detected by the multiplier 11. The signal is multiplied by the signal and inputted to the adaptive filter 13 as a coefficient update signal. The update algorithm is the same as in the first embodiment. In this embodiment, with this configuration, the noise from the noise source 12 can be muted at both the microphone 13 and microphone 14 positions by providing a plurality of error detection microphones and adaptive filters depending on the silencing area.

Note that this embodiment can be naturally extended and applied even when two or more sets of adaptive filters and microphones are provided.

Next, a third embodiment of the present invention will be described with reference to FIG. The purpose of this embodiment is to perform accurate muffling even when there is a zero point or deep tip in the frequency characteristic of the control transfer characteristic from the speaker to the error detection microphone, and to reduce the scale of the noise control device. .

The detection signal of the noise detector 1 is input to the adaptive filter 2.4, and is radiated from the speakers 3 and 5, respectively, to obtain the control transfer characteristic C.
The control sound reaches the error detection microphone 13 through I and C2, and eliminates noise from the noise source 12. Here, the speaker 3.5 is set so that C1 and C2 are different. Microphone 1 calculates the error between this control sound and the noise.
3 and input to the multiplier 7. The noise detection signal from the noise detector 1 is normalized by a normalizer 7, multiplied by an error signal from an error detection microphone 13 by a multiplier 8, and input to convolution calculators 6 and 9. The convolution calculator 6 has a characteristic CI/IC1r'' which is the normalized control transfer function C1, and the convolution operator 9 has a characteristic C which has normalized the control transfer function C2.
2/rc212 is set in advance, and the output from the multiplier 8 and the convolution operation are applied to the adaptive filters 2 and 4.
are input as those coefficient update signals.

With the above configuration, this embodiment can accurately silence even when there is a zero point or a deep dip in the frequency characteristic of the control transfer characteristic from the speaker to the error detection microphone. The normalization process, which requires a large amount of calculations, and the multiplier process with the error signal are moved to the front stage of the convolution calculation process, making them common processes. Therefore, when there are two or more adaptive filters and control transfer characteristics, the overall device can be reduced in scale.

Note that this embodiment can be naturally extended and applied even when two or more sets of adaptive filters and speakers are provided.

Next, a fourth embodiment of the present invention will be described with reference to FIG. The purpose of this embodiment is to widen the area in which noise is eliminated and to reduce the scale of the noise control device.

Error detection microphones 13.14 are installed in the erasure area, and the noise signal detected by the noise detector 1 is normalized in signal magnitude by the normalizer 7, and then multiplier 15.16 calculates the error of the microphone 13.14. It is multiplied by a signal and input to a convolution calculator 8.9. The convolution operator has a characteristic C1l/ICllI2 obtained by normalizing the control transfer function C1l,
A characteristic C12/I C1112 obtained by normalizing the control transfer function C12 is set in advance in the convolution calculator 9, and the output from the multiplier and the convolution operation are respectively applied and input to the adaptive filters 2 and 4 as coefficient update signals. do.

In this embodiment, with the above configuration, by providing a plurality of error detection microphones and adaptive filters according to the silencing area, the noise from the noise source 12 can be muted at both the positions of the microphone 13 and the microphone 14. , the normalization process that requires a large amount of calculation in algorithms such as the learning identification method is moved to the stage before the convolution calculation process and is made into a common process, so when there are two or more adaptive filters and control transfer characteristics, the overall system The scale can be reduced.

Note that this embodiment can be naturally extended and applied even when two or more sets of adaptive filters and microphones are provided.

Next, a fifth embodiment of the present invention will be described with reference to FIG. The purpose of this embodiment is to perform accurate muffling even when there is a zero point or deep dip in the frequency characteristic of the control transfer characteristic from the speaker to the error detection microphone, and to widen the area in which noise is canceled. There is.

Error detection microphones 13 and 14 are installed in the silencing area, and the detection signal of the noise detector 1 is passed through adaptive filters 25 and 26,
27.28, add the output of the adaptive filter 25.26, and radiate it from the speaker 3 to obtain the control transfer characteristic C1l.
The outputs of the adaptive filters 27 and 28 are similarly added to the error detection microphone 13 through the control transfer characteristic C12, and the outputs of the adaptive filters 27 and 28 are added and radiated from the speaker 5. The control sound reaches the detection microphone 13 and the error detection microphone 14 through the control transfer characteristic C22, and cancels the noise from each noise source 12. The error between the noise from the noise source 12 and these control sounds is detected by the microphones 13 and 14, respectively.
The detection output of the microphone 13 is sent to the multiplier 2.
9.31, and the detection output of the microphone 14 is input to a multiplier 30.32. On the other hand, the detection signals of the noise detector 1 are detected by C11, CI2, . The coefficients of C21 and C22 are input to convolution calculators 17, 18 and 19, 20, and after being calculated by these calculators, they are each input to a normalizer 21.
．． 22. Perform normalization processing in 23.24, and multiplier 2
9, 30, 31° 32 are multiplied by the error signal of the microphone 13, 14, and the adaptive filters 25, 26, 26, respectively
27 and 28 as those coefficient update signals.

With this configuration, since the silencing control sound is transmitted using multiple control transmission paths, the frequency characteristics of one control transmission characteristic,
Even if there is a zero point or a deep dip, it can be compensated for with other transfer characteristics, making it possible to perform accurate muffling. In addition, by providing multiple error detection microphones and adaptive filters according to the muffling area, the noise from the noise source 12 can be suppressed. Noise can be muffled at both the microphone 13 and microphone 14 positions.

Also, this embodiment can be naturally expanded and applied even when two or more sets of adaptive filters, speakers, and microphones are provided.

Next, a sixth embodiment of the present invention will be described with reference to FIG. This embodiment aims at accurately muting even when there is a zero point or a deep dip in the frequency characteristic of the control transfer characteristic from the speaker to the error detection microphone, widening the noise erasing area, and noise control. The purpose is to reduce the size of the device.

The error detection microphones 13 and 14 are installed in the silencing area, and the detection signal of the noise detector 1 is passed through the adaptive filters 25, 26,
27 and 28, add the outputs of the adaptive filters 25 and 26, radiate it from the speaker 3, pass through the control transfer characteristic C1l, and send it to the error detection microphone 13.
Similarly, the outputs of the adaptive filters 27 and 28 are added and radiated from the speaker 5 through the error detection microphone 14 through the control transfer characteristic C21, and the error through the control transfer characteristic C22. It reaches the detection microphone 14 as a control sound, and each noise source 1
Eliminate the noise from 2.

Errors between the noise from the noise source 12 and these control sounds are detected by the microphones 13 and 14, respectively.
The detection output of the microphone 14 is input to the multiplier 29, and the detection output of the microphone 14 is input to the multiplier 31.

On the other hand, the detection signal of the noise detector is normalized by the normalizer 21, and then the microphone 13 is normalized by the multiplier 29.31.
．． 14 error signals, and each convolution operator 17.
18 and a convolution calculator 19 and 20. The convolution calculator 17 has the characteristic C1l/IC11r2 which is the normalized control transfer function C1l, and the convolution generator 18 has the characteristic C21/IC12I2 which has the normalized control transfer function C21.
The convolution operator 19 receives the characteristic C12/IC12I2 obtained by normalizing the control transfer function C12, and the convolution operator 2
0 has the characteristic C22/I which is the normalized control transfer function C22.
Set C2212 in advance and add multiplier 29.3
The outputs from 1 are subjected to convolution calculations and input to adaptive filters 25, 26, 27, and 28 as their coefficient update signals.

With this configuration, the silencing control sound is transmitted using a plurality of control transmission paths, so even if there is a zero point or deep dip in the frequency characteristic of one control transmission characteristic,
Accurate muffling can be achieved by supplementing with other transfer characteristics.
Further, by providing a plurality of error detection microphones and adaptive filters depending on the silencing area, noise from the noise source 12 can be muted at both the microphone 13 and microphone 14 positions. In addition, the normalization process, which requires a large amount of calculation in algorithms such as the learning identification method, has been moved to the previous stage of the convolution calculation process, making it a common process.
When the number of adaptive filters and control transfer characteristics is two or more, the scale of the entire device can be reduced.

Note that this embodiment can be naturally extended and applied even when two or more sets of adaptive filters, speakers, and microphones are provided.

Effects of the Invention As is clear from the above explanation, according to the present invention, the silencing control sound is transmitted using a plurality of control transmission paths.
Even if there is a zero point or a deep dip in the frequency characteristics of one control transfer characteristic, it is compensated for by the other transfer characteristics, so that accurate muffling can be achieved. Further, by providing a plurality of error detection microphones and adaptive filters according to the silencing area, noise from a noise source can be muted at a plurality of microphone positions, and the silencing area can be widened. In addition, because the normalization process, which requires a large amount of calculation in algorithms such as the learning identification method, is moved to the stage before the convolution calculation process and made into a common process, the overall device size increases when the number of adaptive filters and control transfer characteristics is two or more. can be made smaller.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a first embodiment of the present invention, FIG. 2 is a block diagram of a second embodiment of the present invention, FIG. 3 is a block diagram of a third embodiment of the present invention, and FIG. The figure is a block diagram of the fourth embodiment of the present invention, FIG. 5 is a block diagram of the fifth embodiment of the present invention, FIG. 6 is a block diagram of the sixth embodiment of the present invention, and FIG. FIG. 1 is a configuration diagram of a conventional noise canceling device. 1...Noise detector, 2, 4, 25, 26, 27
゜28.35, 36, 37.38...Adaptive filter, 3,5...Speaker, 13.14...
...Error detection microphone, 6, 9, 17, 18,
19゜20... Convolution operator, 7, 10.2
1. 22.23.24... Normalizer, 8.1
1,15,16゜29.30,31.32...
Multiplier. Name of agent: Patent attorney Haruaki Ogata and two others Kual Ku' Kuqfu

Claims (6)

[Claims] (1) a noise detector installed near a noise source; first and second adaptive filters that receive the noise detection signals; first and second speakers respectively connected in cascade to both of the adaptive filters; a microphone that detects an error between the sound from the noise source and the sound from both of the speakers; and a first computing unit that performs a convolution operation between the transfer characteristic from the first speaker to the microphone and the noise detection signal. ,
a first normalizer that normalizes the magnitude of the output signal of the arithmetic unit; and a first normalizer that multiplies the normalized output by the error output of the microphone and outputs an update signal of the coefficients of the first adaptive filter. a first multiplier, a second arithmetic unit that performs a convolution operation between the transfer characteristic from the second speaker to the microphone and the noise detection output, and normalizes the magnitude of the output signal of the arithmetic unit. A noise canceling device comprising a second normalizer and a second multiplier that multiplies its normalized output by an error output of the microphone and outputs an update signal for coefficients of the second adaptive filter. (2) a noise detector installed near the noise source, first and second adaptive filters that receive the noise detection signals, and a speaker that receives the summed signal of the outputs of both the adaptive filters;
first and second microphones that detect an error between the sound from the noise source and the sound from the speaker; and performing a convolution operation between the transfer characteristic from the speaker to the first microphone and the noise detection signal. a first arithmetic unit;
A first normalizer normalizes the magnitude of the output signal of the arithmetic unit, and multiplies the normalized output by the error output of the first microphone to obtain an update signal of the coefficients of the first adaptive filter. a first multiplier that outputs a signal; a second arithmetic unit that performs a convolution operation between the transfer characteristic from the speaker to the second microphone and the noise detection output; and a second arithmetic unit that normalizes the magnitude of the output signal of the arithmetic unit. a second normalizer for
A noise canceling device comprising a second multiplier that multiplies the normalized output by the error output of the second microphone and outputs an update signal for the coefficients of the second adaptive filter. (3) a noise detector installed near the noise source, first and second adaptive filters that receive the noise detection signals, and first and second adaptive filters that are cascade-connected to both of the adaptive filters, respectively.
a speaker, a microphone that detects an error between the sound from the noise source and the sound from both the speakers, a normalizer that normalizes the magnitude of the noise detection signal, and an error between the normalized output and the microphone. a first arithmetic unit that performs a convolution operation on the multiplier output and a characteristic obtained by normalizing the transfer characteristic from the first speaker to the microphone by its energy;
a second arithmetic unit that performs a convolution operation between a characteristic obtained by normalizing the transmission characteristic up to the speaker and the microphone by its energy and the output of the multiplier; Update the adaptive filter coefficients,
A noise canceling device characterized in that coefficients of the second adaptive filter are updated using the output of the second arithmetic unit. (4) a noise detector installed near the noise source, first and second adaptive filters that receive the noise detection signals, and a speaker that receives the summed signal of the outputs of both the adaptive filters;
first and second microphones that detect an error between the sound from the noise source and the sound from the speaker; a normalizer that normalizes the magnitude of the noise detection signal; a first multiplier that multiplies an error output of a microphone; and a first multiplier that performs a convolution operation between a characteristic obtained by normalizing a transfer characteristic from the speaker to the first microphone by its energy and the output of the first multiplier. a second multiplier that multiplies the normalized output by the error output of the second microphone; and a characteristic in which a transfer characteristic from the speaker to the second microphone is normalized by its energy. a second arithmetic unit that performs a convolution operation with the output of the second multiplier, updates the coefficients of the first adaptive filter, and uses the output of the second multiplier to update the coefficients of the second adaptive filter. A noise canceling device characterized by updating filter coefficients. (5) a noise detector installed near a noise source; first, second, third, and fourth adaptive filters that receive the noise detection signals; the first adaptive filter; and the second adaptive filter. a first speaker that inputs a signal added to the output of the filter, a second speaker inputted a signal added to the output of the third adaptive filter and the output of the fourth adaptive filter, and a sound from the noise source. and a first and second microphone that detects an error between the sound from the first speaker and the sound from both the speakers, and a first microphone that performs a convolution operation between the transfer characteristic from the first speaker to the first microphone and the noise detection output. a first normalizer that normalizes the output signal of the calculator; a first multiplier that multiplies the normalized output by the error output of the first microphone; a second arithmetic unit that performs a convolution operation between the transfer characteristic from the speaker to the second microphone and the noise detection output; a second normalizer that normalizes the output signal of the arithmetic unit; and a second normalizer that normalizes the output signal of the arithmetic unit. a second multiplier that multiplies the output by the error output of the second microphone;
a third arithmetic unit that performs a convolution operation between the transfer characteristic up to the microphone and the noise output; a third normalizer that normalizes the output signal of the arithmetic unit; The third multiplier is the error output of the microphone.
a multiplier, a fourth arithmetic unit that performs a convolution operation between the transfer characteristic from the second speaker to the second microphone and the noise detection output, and a fourth arithmetic unit that normalizes the output signal of the arithmetic unit. a normalizer, and a fourth multiplier that multiplies the normalized output by the error output of the second microphone, and the output of the first, second, third, and fourth multipliers A noise canceling device that updates coefficients of first, second, third, and fourth adaptive filters, respectively. (6) a noise detector installed near a noise source; first, second, third, and fourth adaptive filters that receive the noise detection signals; the first adaptive filter; and the second adaptive filter. a first speaker that inputs a sum signal of the outputs of the third adaptive filter and the fourth adaptive filter; a second speaker that inputs the sum signal of the outputs of the third adaptive filter and the fourth adaptive filter; first and second microphones that detect an error between the sound from both speakers; a normalizer that normalizes the magnitude of the noise detection signal; and a normalized output thereof and the first microphone.
a first multiplier that multiplies the error output of the microphone, a characteristic obtained by normalizing the transfer characteristic from the first speaker to the first multiplier, and the first speaker to the first microphone by energy. a first arithmetic unit that performs a convolution operation between and the output of the first multiplier; and a characteristic obtained by normalizing the transfer characteristic from the first speaker to the second microphone by its energy, and the first multiplication. a second multiplier that multiplies the normalized output and the error output of the second microphone; and a second multiplier that performs a convolution operation with the output of the second microphone; a third arithmetic unit that performs a convolution operation between the transfer characteristic normalized by the energy and the output of the second multiplier, and the transfer characteristic from the second speaker to the second microphone is calculated by the energy. a fourth arithmetic unit that performs a convolution operation between the normalized characteristic and the output of the second multiplier; A noise canceling device characterized by updating coefficients of second, third, and fourth adaptive filters.