JPH0511771A - Noise control device - Google Patents

Abstract

PURPOSE:To enable formation of a noise eliminating zone constantly in the vicinity of ears a human even with his head moving, when noise elimination is desired in the vicinity of his ears. CONSTITUTION:A noise elimination operation is started after the filter coefficients w(1) through w(m) for the respective predetermined proposed control pints P1 through Pm are memorized by a coefficient memory part 6. When control point to be noise-eliminated moves e.g. from the position of the proposed control point P1 to the position of the proposed control point P2, a control point selective part 7 selects the proposed control point P2, and transmits the information to a signal processing switching part 8. The signal processing switching part 8 reads the filter coefficient w(2) corresponding to the proposed control point P2 from the coefficient memory part 6, and set the value as the filter coefficient of a signal processing part 2 and switches the signal processing. This arrangement allows the movement of the noise-controlled point from P1 to P2 to be followed and rapid noise elimination to be realized in the vicinity of the position P2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention produces a sound as a secondary sound that cancels noise.
The present invention relates to a noise control device that attenuates noise at a certain point.

[0002]

2. Description of the Related Art Conventionally, for example, the literature "Realization of Active Noise Control Chair"
The study group report EA90-2 ”discloses a noise control device that controls a noise by generating a sound that cancels the noise as a secondary sound. FIG. 11 is a diagram showing a configuration example of a conventional noise control device of this type. In the noise control device of FIG. 11, a noise observation unit 51 such as a microphone that observes a causal signal of noise at a point of a noise generation source 50 and converts it into an electric signal x (n), and an electric noise obtained by the noise observation unit 51. A signal processing unit 52 that performs signal processing on the signal x (n) by an appropriate digital filter, a speaker that generates a secondary sound according to the signal s (n) resulting from the signal processing by the signal processing unit 52, and the like. Of the secondary sound source section 53, the noise y (n) propagated from the noise source 50 and the secondary sound as the control sound propagated from the secondary sound source section 53, which are arranged at a point where noise is to be removed, are left together. A residual noise observation unit 54 such as a microphone that inputs noise, converts the sound pressure thereof into an electric signal, and outputs it as an output signal e (n), and a residual noise observation unit 54.
And a coefficient updating unit 55 for updating the filter coefficient of the digital filter in the signal processing unit 52 based on the output signal e (n) from the residual noise observing unit 54. By updating the filter coefficient of the processing unit 52 and performing appropriate signal processing so that the output signal e (n) becomes zero, noise near the point where the residual noise observation unit 54 is arranged is removed. It is like this. In each of the above signals, n represents time.

[0003]

By the way, in the above-mentioned conventional noise control device, although the output signal e (n) is controlled so as to be zero, the coefficient updating unit 55 has the LM.
The S (Least Mean square) algorithm, that is, the coefficient updating algorithm by the least square error method is often used. That is, in the system configuration of FIG.
The output signal e (n) from the residual noise observation unit 54 in
It is expressed by the following equation.

[0004]

[Equation 1]

Here, C _j is a transfer function between the secondary sound source section 53 and the residual noise observing section 54, and the convolution of s (n−j) and the transfer function C _j is the residual noise observing section 54. It becomes the secondary sound signal input to. Further, w _i (n) is a filter coefficient of the digital filter in the signal processing unit 52, and in the LMS algorithm, adaptive signal processing for updating w _i (n) for each sample is performed.

That is, the LMS in the coefficient updating unit 55
In the algorithm, the squared error σ of the output signal e (n)
The filter coefficient w _i is updated for each sample so that (n), that is, {e (n)} ² decreases with time n. More specifically, in Equation 1, when e (n) is squared, a quadratic expression regarding w _i is obtained. Therefore, in the LMS algorithm, the squared error σ (n) is calculated as w
When viewed as a quadratic expression for _{i, the} quadratic surface Z of the following expression
The filter coefficient w _i is updated for each sample.

[0007]

[Equation 2]

In this case, the filter coefficient w _i (n + 1) at time (n + 1) is given by the following equation, where the convergence coefficient is α.

[0009]

[Equation 3]

As can be seen from equation 3, in order to update w _i (n), a transfer function C _j between the secondary sound source section 53 and the residual noise observing section 54 is required in calculation. This transfer function C _j
As the above, what has been measured in advance by the system identification method is generally used. In order to obtain the predetermined silencing effect by applying the above LMS algorithm, the transfer function C _j once measured and set is set. It is necessary to keep it unchanged during system operation. That is, in the conventional noise control device, it is necessary to fix the positions of the secondary sound source section 53 and the residual noise observation section 54 so that the transfer function C _j once set does not change. However, the point where noise is desired to be removed is, for example, near the human ear that is performing a predetermined work, and when the residual noise observation unit 54 such as a microphone is used near the human ear and the head of the human moves, As a result, the residual noise observation unit 54 also moves. At this time, the transfer function C _j changes and the LMS algorithm does not work effectively, and as a result, w _i (n) is not always updated in the direction in which σ (n) decreases, and the muffling effect is obtained. There were times when I couldn't.

As described above, the conventional noise control device can effectively muffle sound only at a fixed position, and when the human head moves, a mute zone is not formed in the vicinity of the human ear, and the muffling sound is eliminated. There is a problem that the effect may not be obtained.

It is an object of the present invention to provide a noise control device capable of always forming a muffling zone near the human ear even when the head moves, for example, when the sound is to be silenced near the human ear. There is.

[0013]

In order to achieve the above object, the invention according to claim 1 is a noise observing means for observing noise in a noise generating source, and a signal processing for processing a signal obtained by the noise observing means. Means, a secondary sound source means for generating a secondary sound for noise control in accordance with a signal resulting from the signal processing by the signal processing means, and noise control is performed from a plurality of predetermined control point candidates. The control point selection means for selecting a control point candidate corresponding to the desired position as a control point, and the signal processing switching means for switching the signal processing of the signal processing means according to the selected control point are provided. I am trying.

Further, according to the present invention, the position estimation means for estimating the position where the noise control should be performed, and the control point determination for selecting the control point corresponding to the estimated position from a plurality of control point candidates. And means.

According to the third aspect of the present invention, the noise observing means for observing the noise at the noise generating source, the signal processing means for processing the signal obtained by the noise observing means, and the signal processing means for signal processing. Secondary sound source means for generating a secondary sound for noise control, and control point selecting means for selecting a control point to be controlled at the current time from a plurality of predetermined control points according to the resulting signal And signal processing switching means for switching the signal processing contents of the signal processing means in accordance with the selected control point.

The invention according to claim 4 has a noise observing means for observing the noise at the noise source, a signal processing means for processing the signal obtained by the noise observing means, and a plurality of control speakers. And a secondary sound source means for generating a secondary sound for noise control according to a signal resulting from signal processing by the signal processing means, and at least one used for control at a current time from a plurality of speakers of the secondary sound source means. The present invention is characterized by comprising control speaker selection means for selecting one control speaker and signal processing switching means for switching the signal processing content of the signal processing means according to the selected control speaker.

The invention according to claim 5 has a noise observing means for observing the noise at the noise source, a signal processing means for processing the signal obtained by the noise observing means, and a plurality of control speakers. The secondary sound source means for generating a secondary sound for noise control according to the signal resulting from the signal processing by the signal processing means, and the control target at the current time from a plurality of predetermined control points Control point selecting means for selecting a control point, control speaker selecting means for selecting at least one control speaker used for control at the current time from the plurality of speakers of the secondary sound source means, and the selected control point and / or Alternatively, there is provided a signal processing switching means for switching the signal processing content of the signal processing means according to the control speaker.

Further, in the invention according to claim 6, in the noise control device according to claim 3, 4 or 5, when the space to be silenced is divided into regions of a predetermined size, the control speaker and / or At least one control point is arranged for each divided area so as to control the area. In this case, when an area to be controlled at the current time is selected from the areas. Further, the control point selecting means is adapted to select a control point in charge of the selected area, and the control speaker selecting means is adapted to select a speaker in charge of the selected area. It is characterized by being.

According to the invention described in claim 7, claim 6
In the noise control device according to the description, position detection means for selecting a region to be controlled at the current time from the region is further provided, and the position detection means detects the position of a person who uses the device, The feature is that an area including the human head is selected as an area to be silenced.

According to the invention of claim 8, claim 6
In the noise control device described above, one control speaker is assigned to a plurality of adjacent regions, and these regions are shared by the one control speaker.

According to the invention described in claim 9, claim 3
Alternatively, in the noise control device described in 5, the control point is divided into an identification control point and an estimation control point, and with respect to the identification control point, the impulse response from the control speaker of the secondary sound source unit is identified and the estimation control is performed. Regarding points, the impulse response from the control speaker of the secondary sound source unit is estimated from the impulse response between the control speaker of the secondary sound source unit and the identification control point, and the processing in the signal processing means is performed based on those impulse responses. It is characterized by being adapted to determine.

According to a tenth aspect of the present invention, in the noise control apparatus according to the ninth aspect, when estimating an impulse response from a certain control speaker to a certain estimated control point, the estimated control point is set. The amplitude and time delay of the corresponding reflected wave are detected between the impulse responses reaching multiple identification control points around the, and the reflected waves are determined according to the positional relationship between the estimated control point and those identification control points. It is characterized by estimating the impulse response with amplitude and time delay.

[0023]

In the noise control device according to the first aspect, a control point candidate corresponding to a position where noise control is to be performed is selected as a control point from a plurality of predetermined control point candidates,
The signal processing of the signal processing means is switched according to the selected control point. As a result, even if the position where the noise control should be performed changes, the control point corresponding to that position is selected and the signal processing corresponding to this is immediately performed. Can be formed.

Further, as described in claim 2, when selecting the control point, the position where noise control is to be performed is estimated,
If a control point corresponding to the estimated position is selected from the control point candidates, it is possible to form the sound deadening zone by following the free movement of the position where the noise control should be performed.

Further, in the inventions according to claims 3, 4, 5, 6, and 7, it is possible to control only a range required at that time, and a large number of control point microphones and control speakers are provided. However, only some of the control point microphones and control speakers are actually used.

In the invention according to claim 8, one control speaker is assigned to a plurality of adjacent areas, and these areas are shared by the one control speaker.

In the invention according to claims 9 and 10, the control point is divided into an identification control point and an estimation control point, and the identification control point is an impulse response from the control speaker of the secondary sound source section. For the estimated control points, the impulse response from the control speaker of the secondary sound source section is estimated from the impulse response between the control speaker of the secondary sound source section and the identification control point, and the impulse response is estimated based on those impulse responses. First, the processing in the signal processing means is determined.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a configuration diagram of an embodiment of a noise control device according to the present invention. Referring to FIG. 1, in the noise control device of this embodiment, a noise observation unit 1 such as a microphone for observing a causal signal of noise at a point of a noise generation source 50 and converting it into an electric signal x (n), and a noise observation A signal processing unit 2 that performs signal processing on the electric signal x (n) obtained by the unit 1 by, for example, an appropriate digital filter to process the phase and amplitude of the electric signal x (n), and a signal processing unit 2 Is arranged in any of a plurality of predetermined control point candidates P1 to Pm, and a secondary sound source unit 3 such as a speaker that generates a secondary sound according to the signal s (n) resulting from the signal processing by Noise y propagated from the noise source 50 at the position of the control point candidate
Residual noise observation of a microphone or the like that converts sound pressure as residual noise, which is a combination of (n) and the secondary sound as the control sound propagated from the secondary sound source unit 3, into an electric signal and outputs it as an output signal e (n) Output signal e from section 4 and residual noise observation section 4
For each control point candidate P1 to Pm when the coefficient updating unit 5 that updates the filter coefficient of the filter in the signal processing unit 2 based on (n) and the residual noise observing unit 4 are positioned at each control point candidate P1 to Pm, respectively. The coefficient storage unit 6 in which the filter coefficients whose coefficients have been updated are stored in association with the control point candidates P1 to Pm, and the control point candidates P1 during noise control processing.
To Pm, a control point selection unit 7 that selects, as a control point, a control point candidate corresponding to a position where noise control should be performed.
And a signal processing switching unit 8 that switches the signal processing of the signal processing unit 2 according to the selected control point.

As the noise observing section 1, a sensor for detecting a sound wave generated from a machine or the like and converting the sound wave into an electric signal is used, or the noise from the noise generating source 50 is accompanied by the rotation of the motor. In the case where the noise is observed, the noise observation unit 1 may be, for example, one that converts the rotation speed of the motor into an electrical signal x (n) that has the electrical frequency as it is, or vibration of the machine. When noise is generated by the noise observing unit 1, a vibration pickup or the like that converts vibration into an electric signal x (n) is used.

The residual noise observing unit 4 and the coefficient updating unit 5 are
The coefficient updating unit 5 is used before the noise control process is actually performed, and the coefficient updating unit 5 uses, for example, the above-mentioned LMS algorithm.
The filter coefficient of the signal processing unit 2 is updated in the direction in which the output signal e (n) from the residual noise observing unit 4 decreases, and this is stored in the coefficient storage unit 6 in advance. Further, specifically, in this embodiment, the signal processing switching unit 8 reads the filter coefficient corresponding to the control point selected by the control point selecting unit 7 from the coefficient storage unit 6, and the read filter coefficient is the signal processing unit. The signal processing is switched by setting it as the filter coefficient of the second filter.

Next, the operation of the noise control device having such a configuration will be described. A digital filter is used as the filter in the signal processing unit 2. In this case, in this embodiment, the filter coefficient of the digital filter for each control point candidate P1 to Pm is determined in advance by the coefficient updating unit 5 before the noise control process is actually performed.

That is, the coefficient updating unit 5 squares the output signal e (n) output from the residual noise observing unit 4 by using, for example, the above-described LMS algorithm (coefficient updating algorithm by the least square error method). The filter coefficient is updated so that the error becomes smaller with time, and the filter coefficient for each of the control point candidates P1 to Pm is determined in advance. For example, the filter coefficient at the position of the control point candidate P2 is the output signal e (n) from the residual noise observation unit 4 set at this position when the residual noise observation unit 4 is installed at the position of the control point candidate P2. Is determined to be “0” by using the above-described Equation 3.

Similarly, the filter coefficient at the position of another control point candidate, such as the control point candidate Pm, is also installed at this position when the residual noise observing unit 4 is installed at the position of this control point candidate Pm. The determined output signal e (n) from the residual noise observation unit 4 is determined by using Equation 3 so as to be "0".

When determining the filter coefficient for each of the control point candidates P1 to Pm using Equation 3, the transfer function C of the sound from the secondary sound source unit 3 to the position of each of the control point candidates P1 to Pm is determined. _{Although j} is required, this transfer function C _j may be measured in advance for each control point candidate P1 to Pm by using a known system identification method such as LMS using white noise.

In this way, prior to the actual noise control processing, the residual noise observing section 4 is sequentially installed at the positions of the control point candidates P1 to Pm, and the filter coefficient for each control point candidate P1 to Pm is set as a coefficient. When determined by the updating unit 5, the determined filter coefficients w (1) to w (m) are stored in the coefficient storage unit 6 for each of the control point candidates P1 to Pm as shown in FIG.

After the filter coefficients w (1) to w (m) for each control point candidate P1 to Pm are stored in the coefficient storage unit 6 in advance, the silencing operation is actually performed. In this case, the control point selection unit 7 selects the control point at which the sound should be muted.
For example, when the control point candidate P1 is selected from the control point candidates P1 to Pm as the control points to be muted,
The control point selection unit 7 gives the signal processing switching unit 8 that the control point candidate P1 has been selected.

The signal processing switching unit 8 switches the signal processing of the signal processing unit 2 according to the selected control point. Specifically, the optimum filter coefficient w corresponding to the control point candidate P1
(1) is read from the coefficient storage unit 6 and set as the filter coefficient of the signal processing unit 2, that is, the digital filter. As a result, the secondary sound source unit 3 immediately outputs a secondary sound that effectively removes the noise y (n), that is, a control sound at the position of the control point candidate P1.
You can quickly mute near the position. Next, when the control point selecting unit 7 selects, for example, the control point candidate Pm as the control point and the control point to be silenced moves from P1 to Pm, for example, the optimum filter coefficient w (corresponding to the control point candidate Pm m) is immediately set as the filter coefficient of the signal processing unit 2, whereby the secondary sound source unit 3 immediately outputs the secondary sound that effectively removes the noise y (n) at the position of the control point candidate Pm. It is output, and it is possible to quickly mute the vicinity of the position of the control point candidate Pm.

As described above, in this embodiment, among the filter coefficients measured and stored in advance at various positions, the optimum one for the current position is selected according to the position change and is selected as the filter coefficient of the signal processing unit 2. Since the setting is made, even if the position to be muted varies, it is possible to always follow the position variation and form the muffling zone at the position to mute.

Further, in this embodiment, it is not necessary to repeatedly update the filter coefficient by the LMS algorithm or the like during the noise control processing, and therefore, even when the position to be muted suddenly changes, this position fluctuation is caused. Since the contents of the signal processing of the signal processing unit 2 can be immediately switched to and the optimum filter coefficient can be immediately set in the signal processing unit 2, a high silencing effect can always be obtained.

By the way, the control point selection unit 7 can have various configurations according to the application to which the noise control device is applied. For example, when it is intended to apply to muffling in a vehicle in which the position of the seat is switched step by step, the control point selection unit 7 can be configured as a manual switch that switches the position of the seat stepwise. At this time, in the state where a person is seated on the seat, each position of the human head corresponding to the stepwise switching position of the seat is predetermined as a control point candidate, and the optimum filter coefficient at each position is predetermined and stored. I will let you. As a result, even if a person moves the seat position by a manual switch in the automobile and the human head moves accordingly, the control point corresponding to the manual switch, that is, the head position is automatically set. Selected,
A muffling zone can always be formed near this head position.

Alternatively, the control point selection unit 7 may be configured to automatically estimate the position of the user's head and select the control point. FIG. 2 is a diagram showing a configuration example of this type of control point selection unit 7, and in this example, the control point selection unit 7 includes a position estimation unit 40 that estimates the position of the user's head and a position estimation unit. The control point determining unit 44 determines an optimum control point based on the position of the user's head estimated by 40. Specifically, the position estimation unit 40 includes an ultrasonic speaker 42 mounted near the user's head, and a signal generation unit 4 that generates an ultrasonic output signal to the ultrasonic speaker 42.
1 and three ultrasonic microphones 43, which are installed in advance at predetermined positions and detect ultrasonic waves output from the ultrasonic speaker 42. It should be noted that FIG. 2 particularly shows only the elements necessary for the actual noise control process, and the residual noise observation unit 4 and the coefficient updating unit 5 that are not generally used during the actual noise control process are shown in FIG. Not shown.

When the control point selector 7 is constructed as shown in FIG. 2, the ultrasonic speaker 42 is mounted near the human head, for example near the ear, and the signal generator 41 outputs an impulse as a signal for a predetermined time. When it is generated every time, an ultrasonic impulse is output from the ultrasonic speaker 42 mounted near the human ear every predetermined time. The three ultrasonic microphones 43 installed at predetermined positions (for example, spatially above the human head) detect ultrasonic impulses output from the ultrasonic speaker 42 and propagated in the space. At this time, the position estimation unit 40 measures the propagation time until the ultrasonic impulse output from the ultrasonic speaker 42 reaches the ultrasonic microphone 43, and based on this propagation time, the position estimation unit 40 communicates with the ultrasonic speaker 42. To calculate the distance. Since the ultrasonic speaker 42 is located below the three ultrasonic microphones 43, the ultrasonic speaker 42, that is, the human speaker The spatial position near the ear can be uniquely determined.

When the position estimating unit 40 estimates the position near the human ear, the control point determining unit 44 determines the optimum position among the preset control point candidates P1 to Pm based on the estimated position. The appropriate control point. For example, the control point candidate closest to the estimated position is determined and selected as the optimum control point.

In this way, when the optimum control point is selected, the filter coefficient corresponding to this control point is read out from the coefficient storage unit 6 and the signal processing unit 2 is operated in the same manner as described above.
And the secondary sound from the secondary sound source unit 3 based on this is able to form a sound deadening zone near the human ear. In particular, in the example of FIG. 2, the control point selection unit 7 automatically estimates the position of the human head, for example, near the ear, and selects the control point based on the estimated position. The muffling zone is formed following the free movement of the parts, and it is possible to obtain the same effect as that in a practically large area where the sound can be silenced.

In the above description, the ultrasonic speaker 4
Although the ultrasonic waves output from 2 are impulses,
Instead of impulse, burst wave, pink noise, white noise, etc. may be used. Further, instead of the generation of ultrasonic waves, position estimation may be performed using radio waves. Alternatively, the position can be estimated by mounting the infrared generator on the head and using an infrared camera.

Further, in the above description, one control point candidate is selected as the control point from the plurality of control point candidates P1 to Pm. However, for example, it is selected at the estimated position depending on the processing convenience. Two or more control point candidates close to each other may be selected as control points.

Further, in the above-mentioned embodiment, in order to simplify the explanation, the noise source is one, and the secondary sound source section 3 and the residual noise observing section 4 are each one. The present invention can be similarly applied to a case where a plurality of secondary sound source sections or residual noise observation sections are provided. Further, the coefficient updating unit 5, the signal processing switching unit 7 and the like are connected to the signal processing unit 1.
The coefficient updating unit and the like may or may not be included in the signal processing unit 1 and is not essential to the present invention.

FIG. 3 is a conceptual diagram of a noise control device in which a large number of control speakers are provided in the secondary sound source section, and FIG. 4 is a diagram showing a concrete configuration example of the noise control device of FIG. That is,
In FIGS. 3 and 4, a large number of control points are further prepared,
It is intended to provide a large number of control speakers in the secondary sound source section in proportion to the number of control points to form a muffling zone in a wide range. Note that, in FIGS. 3 and 4, a portion corresponding to the coefficient updating unit 5 and the like in FIG. 1 is included in the signal processing unit 10.

Referring to FIG. 3, in this noise control device, there is one noise source, and L control points P _l (1
The sound pressure in ≦ l ≦ L) is controlled by using M control speakers SP _m (1 ≦ m ≦ M) of the secondary sound source section.

Further, referring to FIG. 4, this noise control device has a noise observation section 9, a signal processing section 10, a secondary sound source section 11, and a plurality of control point microphones MIC _l (1 ≦ l ≦
L) and. The noise observing unit 9 detects noise with the noise source microphone 201, filters this electric signal with the low-pass filter 202, and converts it into a digital signal with the A / D converter 203, as in the above-described embodiment. It is supposed to do. In addition, the signal processing unit 1
0 is a digital FIR filter of length I W (m) = {w
_i (m)} (i = 0 to I−1, 1 ≦ m ≦ M), the control filter storage area 12 that holds the coefficient, and the noise source signal from the noise observation unit 9, that is, the electric signal x (n). The noise source signal storage area 13 having a length I for holding I time and each filter W holding the noise source signal x (n) stored in the noise source signal storage area 13 in the control filter storage area 12
(M) is filtered, and the control signal s
_A control signal calculation circuit 14 for generating _m (n) and a modeling circuit 15 for calculating the transfer function C by the LMS algorithm.
When, using modeling filter C _'lm of length J modeling filter storage area 16 for L × M pieces hold, and the content of the content and the modeling filter storage area 16 of the noise source signal storage region 13, the filtering results r _lm (n)
A reference signal calculation circuit 17, a reference symbol storage area 18 that holds L × M filtering results r _lm (n) for a length I, the contents of the reference symbol storage area 18, and each control point P _l . A control filter updating circuit 1 for updating the contents of the filter coefficient by using the output e _l (n) from the arranged control point microphone MIC _l and the contents of the control filter storage area 12.
9 and 9.

In addition, the secondary sound source section 11 includes a signal processing section 10
D / A converter 20 for converting the control signal s (n) from the control signal calculating circuit 14 into an analog signal, a low-pass filter 21 for filtering the analog signal from the D / A converter 20, and a low-pass filter. The power amplifier 22 amplifies the signal from the filter 21 and supplies it to the control speaker SP _m (1 ≦ m ≦ M).

Next, the operation of the noise control device having such a configuration will be described. The signal processing unit 10 stores a control filter storage area 12 that holds the coefficient of the digital FIR filter W (m) having a length I, and a noise source signal storage that holds the noise source signal x (n) for the past I time. The area 13 is provided in, for example, a memory, and the control signal calculation circuit 14 of the signal processing unit 10 converts the noise source signal x (n) stored in the noise source signal storage area 13 into the control filter storage area 12 The target control signal s _m (n) is generated by performing the filtering process with each filter W (m) stored in the. When this control signal s _m (n) is given to the secondary sound source unit 11 as the output of the signal processing unit 10, the secondary sound source unit 11 outputs D /
A converter 20, low-pass filter 21, power amplifier 2
After treatment with 2, it applied to the control speaker SP _m, and outputs to the outside world by the control speaker SP _m.

By the way, the filter coefficient in the signal processing unit 10 is obtained, for example, by the following sequential update rule. Now, at each control point P _l , the microphone MIC _l
Is installed, and the noise propagating through the room and reaching the control point P _l at time n is output from the control speaker SP _m from the output point of d _l (n), W (m) to the control point P _l . The impulse response of the output path to the installed microphone MIC _l is set as C _lm = {C _j (lm)}. At this time,
The output e _l (n) of the microphone MIC _l is given by the following equation.

[0054]

[Equation 4]

In order to find an optimum filter that minimizes the sound pressure at the control point, the sum E of the squared errors of the output given by
Determine (m).

[0056]

[Equation 5]

In the above equation, E [•] represents the average. From Eqs. 4 and 5, E can be regarded as a quadratic form of w _i (m), and the matrix that determines the quadratic coefficient is a semidefinite value.
In addition, since it is a positive definite value in most cases in practice, E (n)
There is only one w _i (m) that minimizes Filtered-XL to search w _i (m) that takes the minimum value sequentially
In the MS algorithm, −∂ {e
The value of w _i (m) is updated in the direction of (n) ² } / ∂w _i (m). That, Δw _i (m, n) When the update amount of w _i (m) a at time n, w _i at time (n + 1) (m) is given by the following equation.

[0058]

[Equation 6]

In the above equation, C (lm) is an acoustic impulse response, and in order to update the filter coefficient by the equation 6, C (lm) must be identified beforehand by some means. There is. Modeling circuit 15 of the signal processing unit 10, modeling filter storage area 16
It is provided in order to achieve the identification of C (lm), and the input noise signal covering the frequency band is intended to control the speaker SP _m in the secondary tone generator 11, originating from the control speaker SP _m Using the signal e _l (n) output from the microphone MIC _l installed at the control point P _l through the transfer function C (lm) in the room, the modeling circuit 15 measures the measured value of C (lm) by the LMS method. And obtain the modeling filter C ′ (lm) as
The data is stored in the modeling filter storage area 16.

The length J is stored in the modeling filter storage area 16.
FIR filter, ie modeling filter C ′
If (lm) is held and the transfer function C _lm is identified with sufficient accuracy by the FIR filter C ′ _lm = {C ′ _j (lm)} (j = 1 to J−1) of length J, then In order to calculate 6, the reference signal calculation circuit 17 first calculates the contents of the noise source signal storage area 13 and the modeling filter storage area 16
Using the contents of and, r _lm (n) is calculated by the product-sum calculation of the following equation and written in the reference symbol storage area 18.

[0061]

[Equation 7]

Next, the control filter updating circuit 19 determines the contents of the reference symbol storage area 18 and the microphone MIC.
_The following equation is calculated using the output e _l (n) from _l and the contents of the control filter storage area 12, and the contents of the filter coefficient are updated accordingly.

[0063]

[Equation 8]

When the environment is not changed so much and the filter coefficient is fixed and the predetermined processing is performed during the normal operation, the above-described filter coefficient is updated before the operation, and this calculation is performed when the convergence is sufficiently achieved. Just stop. In this case, the signal processing unit 10 causes the control signal calculation circuit 14 to perform only the filtering process of s _m (n) of the equation 4. At this time, the microphone installed at the control point may be removed.

On the other hand, when the acoustic environment change is predicted during the operation, the filter coefficient needs to be updated during the operation as well, so that the control signal calculation circuit 14 executes S of the equation (4).
_In parallel with the calculation of _m (n), the reference signal calculation circuit 17 calculates the equation 7 and the control filter update circuit 1 in parallel.
The calculation of Equation 8 by 9 will be performed.

As described above, in the noise control device of FIG. 4, a large number of control points and control loudspeakers are used, and the filtering operation by the digital filter is performed, whereby the muffling zone can be formed in a wide range and the wide range can be formed. Even if the position where the noise control is to be performed fluctuates slightly due to the formed muffling zone, the muffling effect can be obtained.

However, when a wide range of muffling zones is formed by the apparatus of FIG. 4, the amount of filtering calculation required for control increases in the order of the product of the number of control points and the number of control speakers, and The filtering operation by the digital filter requires many multiply-add operations.
Therefore, if a large number of control points and control speakers are used, the amount of calculation becomes enormous and the hardware scale becomes very large.

More specifically, the amount of calculation in the apparatus of FIG. 4 requires M pieces of filtering of length I to calculate the control signal s _m (n) in the control signal calculation circuit 14 (Equation 4). , And when updating the filter coefficient, r _l (n)
Requires L × M filterings of length J to calculate
In addition, a product-sum operation of length I is performed to update W (m) in Equation 8 by L ×
M pieces are required, and the sum of products of length I is M ×
(L + 1) times, and the product sum of length J is required L × M times. Therefore, when the number L of control points and the number M of control speakers increase, the calculation amount increases in the order of the product. In particular, when I = J, it is necessary to execute the product-sum operation of length I (2L + 1) · M times. Since these calculations need to be performed in real time, if M and L are very large, the hardware cost becomes enormous and a realistic control device cannot be constructed. In other words, in order to obtain the muffling effect, it is necessary to use digital signal processing because it is necessary to control the phase and amplitude of the control sound with high accuracy while adapting to the sound field. Requires a large number of sum-of-products operations, so the hardware cost becomes enormous when the number of control points and control speakers increases. Also,
If the number of control points and control loudspeakers is large, it is necessary to install microphones at a large number of control points at the time of identifying C ′ in the initial stage or at each stage, which is very troublesome. This also causes a problem.

The inventor of the present application improved the device shown in FIG. 4 and further devised a noise control device having a practical hardware scale and capable of silencing a wide range without troublesome identification.

FIG. 5 is a diagram showing a first structural example of such a noise control device. The noise control device of FIG. 5 is provided with a plurality of control points and a plurality of control speakers similarly to the noise control device of FIG. 4, but the noise control device of FIG. 4 should be muted. Position detection unit 2 for detecting a region
6 and the control target at the current time from the plurality of control points 1
Control point selection unit 24 for selecting one or more control points
And a control speaker selection unit 25 for selecting one or a plurality of control speakers to be used at the present time from the plurality of control speakers.

More specifically, the space to be silenced is divided into regions of appropriate size, and one or a plurality of control speakers are in charge of controlling each of the divided regions. One or a plurality of control points are responsible for controlling the area. In this case, the control point selection unit 24 is configured to select an area to be controlled at the current time from the above areas, and select a control point in charge of the selected area. , Is configured to select the control speaker in charge of the selected area. That is, as shown in FIG. 6A, the control point selection unit 24 is provided with an area condition holding unit 27 that holds the coordinate condition that forms the boundary of each area, and when the area number is given. A control point number conversion table or a control point number conversion formula calculation circuit 28 that returns the number of the control point responsible for controlling the area to the signal processing unit 10 is provided. The control speaker selection unit 25 also has the same area condition holding unit 2 as the control point selection unit 24, as shown in FIG. 6B.
9 and a control speaker number conversion table or control speaker number conversion formula calculation circuit 30 which returns the number of the control speaker responsible for controlling the area to the signal processing unit 10 when the area number is given. .

Next, the operation of the noise control device configured as shown in FIG. 5 will be described with reference to FIG. FIG. 7 is a diagram showing an example of a space in which sound is desired to be silenced. Here, as a space to be silenced, consider a two-dimensional plane of a certain height in the room.
A virtual XY coordinate system is set in the dimensional space. This is cut into meshes at regular intervals and divided into square areas. The control points for controlling each divided area are the four vertices of the area. A control speaker is installed in the upper portion UL of each apex of a square in which these 9 regions are collected in 3 × 3 9 units, and the control speaker is in charge of controlling these 9 regions RG. The above-mentioned area division method and the setting method of the control point and the control speaker are merely examples, and the method described below can be applied almost as it is even when other setting methods are adopted.

Hereinafter, the number of control points in charge of controlling a certain area will be M ', and the number of control speakers will be L'. In this case, M '= L' = 4.

The total number of set areas is T, the total number of control points is L, and the total number of control speakers is M. The serial number of the area is t (1≤t≤T) and the control points are P _l ( 1 ≦ l ≦
L), the control speaker is represented by SP _m (1 ≦ m ≦ M), and particularly when one area is focused, the control point in charge of controlling the area t is P (t, μ) (1 ≤μ≤
L '), SP the control speaker in charge of the control of the area t
It is represented by the notation (t, ν) (1 ≤ ν ≤ M '). Further, the signal processing unit 10 includes one input terminal NS for receiving a noise source signal and L ′ input terminals μ (1 ≦ μ ≦ for receiving an output of a microphone installed at a control point according to circumstances.
L ′) and M ′ input terminals ν (1 ≦ ν ≦ M ′) connected to any of the control speakers.

Since the control filtering performed by the signal processing unit 10 needs to perform a unique process for each region t, the filter coefficient is (t, ν) (1≤t≤T, 1≤).
It is necessary to hold every ν ≦ M ′). Therefore, the filter coefficient that creates the signal output from the speaker SP (t, ν) that controls the area t is W (t, ν) = {w _i (t,
ν)} (i = 0 to I-1). W (t,
ν) (1 ≤ t ≤ T, 1 ≤ ν ≤ M ') is the control filter storage area 1 accessible by the index of (t, ν)
Held at 2.

The control point selection unit 24 and the control speaker selection unit 25 hold the conditions of the X and Y coordinates forming the boundary of each area in the area condition holding units 27 and 29. For example, in the present example, the region t is a _t ≤X≤
If b _t and c _t ≦ Y ≦ d _t are set, the data {(a _t , b _t ), (c _t , d _t )} is stored in the area condition holding units 27 and 29. Keep it.

Further, the control point number conversion table or the control point number conversion formula calculation circuit 28 of the control point selection unit 24, when given the serial number t of the area, controls the control point P (t, μ) (1 ≤ μ ≤ L ') serial number l
(1 ≦ l ≦ L) is returned, and the control speaker number conversion table of the control speaker selection section or the control speaker number conversion formula calculation circuit 30 controls the area t when the serial number t of the area is given. Control speaker S in charge of
Serial number m of P (t, ν) (1 ≤ ν ≤ M ') (1 ≤ m ≤
M) shall be returned.

First, the signal processing section 10 performs the following processing in the initialization stage. That is, the signal processing unit 10
A signal including a frequency band of noise to be controlled such as band-limited white noise is generated by a signal generation circuit (not shown) of its own, and a certain control speaker SP _m
To output from. A microphone MIC ₁ is installed at the control point P _l (1 ≦ l ≦ L) to capture the acoustic signal output from the control speaker SP _m and transmitted to the outside world. The modeling circuit 15 receives the signal output from the control speaker SP _m and the output signal of MIC _l , and outputs the LMS described above.
The transfer function C ′ _lm is calculated by the method described above. When the transfer function C ′ _lm is calculated for each of l and m, the results are written in the modeling filter storage area 16 of the signal processing unit 10 in an accessible state by the index of (l, m). Retained.

Next, during the control, the region t is selected as follows. That is, the position detector 26
For example, when the position of the head of a person who uses the device is detected, the position detection information is passed to the control point selection unit 24 and the control speaker selection unit 25. In the case of the present example, since the space to be silenced is a two-dimensional space, the XY coordinates set in this two-dimensional space are passed to the control point selection unit 24 and the control speaker selection unit 25. It should be noted that position detection can be performed using ultrasonic waves as described with reference to FIG. 2, or a red light source can be attached to the vicinity of the ears or the top of the head and turned on to turn on the entire space to be muted. The RGB output is acquired by the TV camera installed so that it can be projected.
The red (R) output is subjected to A / D conversion to obtain a multi-valued digital image, and by detecting a pixel whose value is a certain threshold value or more, the position of the red light source, that is, the position of the human head is detected. It may be adapted to detect.

The control point selection unit 24 includes an area condition holding unit 27.
Boundary conditions (a _t , b _t ), (c _t ,
reads d _t), the coordinate values by the position detection unit 26 has determined to _{(X, Y), a t} ≦ X <b t, performs inequality evaluation of c _t ≦ Y <d _t, if this is true, This t is obtained as the number of the area to be controlled. The control speaker selection unit 25 also
The number t of the area to be controlled is obtained by the same method as that of the control point selection unit 24.

In this way, when the area t to be controlled is selected by the control point selecting section 24 and the control speaker selecting section 25, this is passed to the signal processing section 10. In addition, the control point selection unit 24 uses the control point number conversion table to determine the index (t, μ) (1 ≦ μ ≦ L ′) of the control point in charge of controlling the area t as a serial number (l (1), l). (2), l
(3), l (4)), and passes this to the signal processing unit 10.
Similarly, the control speaker selection unit 25 controls the index (t, ν) (1 ≦ ν of the control speaker in charge of controlling the area t).
≦ M ′) according to the control speaker number conversion table
The serial number (m (1), m (2), m (3), m (4)) is converted, and the control speaker SP _{m (ν)} is connected to the output terminal ν of the signal processing unit 10. In addition, the serial number _{m (ν)} is _assigned to the signal processing unit 10.
Pass to.

The signal processing unit 10 has a noise source signal storage area 10 of length I, where x (n) and x (n-
The data 1), ..., X (n-I + 1) are stored. The control signal calculation circuit 14 filters this data by W (t, ν) read from the control filter storage area 12 in accordance with the given area number t as in the following expression, and outputs the control speaker S through each output terminal ν.
Output from P (t, ν) to the outside world.

[0083]

[Equation 9]

At this time, the filter coefficient W (t,
As ν), it is necessary to design one that can obtain an appropriate control effect, and the design method will be described below.

First, in the initialization stage, the same processing as the above-described control signal generating method is performed to generate a control sound from the speaker SP (t, ν) (1≤ν≤M ').
In addition to the processing described above, the control point selection unit 24 connects the output of the microphone MIC _{l (ν)} to the input terminal μ of the signal processing unit 10. The signal processing unit 10 supplies the serial numbers (l (1), l (2), l (3), l) of the control points given from the control point selection unit 24.
(4)) and the control speaker serial numbers (m (1), m (2), m (3), m (4)) given from the control speaker selection unit 25, from the modeling filter storage area 16 Using the read contents of C ′ _{l (μ) m (ν) and} the contents of the noise source signal storage area 13, r _μν (n), w as in the case of the _equations 7 and 8.
Calculate _i (t, ν; n + 1). That is, the reference signal calculation circuit 17 calculates r _μν (n) by the following equation.

[0086]

[Equation 10]

Further, the MIC obtained through the input terminal μ
_Using the output e _{l (μ)} (n) of _{l (μ)} , the control filter updating circuit 19 calculates w _i (t, ν; n + 1) by the following equation.

[0088]

[Equation 11]

This result is written in the control filter storage area 12, and W (t, ν) (1≤ν≤M ') is thereby written.
Can be gradually optimized.

In the case of fixed control, the above-mentioned processing is performed at the initialization stage of the apparatus to sufficiently converge it, and the contents of the control filter storage area 12 are fixed during control. Also,
When performing adaptive control, the above processing is continuously performed even during the control state. In the adaptive case, a microphone is installed at the control point during control. In the above process, one area t is fixed. This t goes through 1 ≦ t ≦ T in the initialization stage in order, and is the index of the selected area to be controlled in the control state in the adaptive control.

As described above, in the noise control device of FIG. 5, since the position detecting section 26 detects the region (for example, the position of a person) to be muted and controls only the range required at that time, a large number of Even when a microphone and a speaker are provided, control can be performed with a small number of microphones and speakers, and the amount of calculation and the amount of hardware can be greatly reduced despite the fact that a wide range of noise can be suppressed.

In the above example, the selection is based on the area, but this method is not always necessary. For example, when it is desired to obtain a high silencing effect only in the vicinity of a certain point, it is possible to select a control point immediately on that side. On the contrary, when the sound deadening effect does not have to be as high as the difference, but it is desired to mute sound in a relatively wide range, it is possible to select control points sparsely arranged in that range.

In the above example, both the control point selecting section 24 and the control speaker selecting section 25 are provided.
A configuration in which only one of these is provided is also possible.
That is, in the above example, it is assumed that a plurality of control points and control speakers are provided. However, the present invention is not applicable when a plurality of control points or control speakers are provided. However, the control speaker can be configured with only one control point, or the control point can be selected with only one control speaker. .

When a control speaker is prepared for each of the divided areas, if a separate speaker is used for each area, T × M ′ speakers are required, and when T is large. Increases the number of speakers,
There are physical and cost issues in placement.
On the other hand, as shown in FIG. 6, if the control speakers are arranged at the vertices of the squares, the control speakers are shared between the adjoining squares, and one control speaker also serves to control a plurality of areas. It is possible to reduce the number of control speakers, and also to reduce labor and the like when installing the control speakers.

In addition, in an environment in which the variable position of a person is limited to only a few places, such as when moving a fixed seat, the area that the person should control with a manual switch each time the person moves is changed. It is also possible to select a region, and in this case, the region can be selected without using the position detecting unit 26 of FIG. 4 or even if it is not provided.

Further, when the position detecting section 26 is provided, it is possible to further provide a mechanism for detecting the absence of a person in the space where the sound is desired to be silenced. For example, when the position detection unit 26 uses a TV camera as described above, this detection mechanism holds the level in the image of the red light source in advance, and a pixel whose difference from this level is within a certain threshold value. Is not present in the image, or when the detected human position is outside the target space, it is determined that no human is present, and a signal to that effect is provided to the signal processing unit 10. You can Also, when using ultrasonic waves, whether the ultrasonic receiver installed in the space does not detect ultrasonic waves of a specified frequency,
When the detected position of the human is outside the target space, it can be determined that the human does not exist and a signal to that effect can be given to the signal processing unit 10. Further, when a signal indicating that there is no human being is sent from the position detecting unit 26, the signal processing unit 1 outputs the signal “0” to the speaker without performing the filtering calculation at that time.
0 can be configured. As a result, it is possible to prevent useless generation of sound waves and reduce the power consumption cost of devices such as a power amplifier that drives a speaker.

FIG. 8 is a diagram showing a second configuration example of the noise control device which is an improvement of the device of FIG. 4 and is capable of silencing a wide range with practical hardware scale and without trouble. In addition,
Portions corresponding to those in FIG. 4 are designated by the same reference numerals. The noise control device of FIG. 8 is provided with a plurality of control points and a plurality of control speakers as in the noise control device of FIG. 4, but in the noise control device of FIG. This is different from the signal processing unit 10 of the noise control device of FIG.
That is, the signal processing unit 101 of the noise control device of FIG.
Control filter storage area 12 and noise source signal storage area 13
A control signal calculation circuit 14, a modeling circuit 15,
Modeling filter storage area 16, reference signal calculation circuit 17, reference symbol storage area 18, control filter update circuit 19, modeling filter estimation circuit 102, noise path calculation circuit 103, pseudo noise calculation circuit 104, noise It has a route storage area 105.

Further, in the noise control device of FIG. 8, the control points are arranged as shown in FIG. 9, and the identification control point BP indicated by a black circle and the estimated control point C indicated by a white circle.
It is divided into P and. In this example, for the sake of simplicity of explanation, it is assumed that the estimated control point CP exists at the midpoint between the two identified control points BP that oppose it. An estimated control point is P _0, and four identification control points adjacent to this are P ₁ ,
Let P ₂ , P ₃ , and P ₄ . The method of arranging the control points shown here is merely an example and is not necessary for the present invention.

The feature of the noise control device of FIG. 8 is the way of identifying the modeling filter C'which is performed in the initialization stage. That is, in the noise control device of FIG. 8, first, only the identification control point P _l (1 ≦ l ≦ 4) is provided with the microphone MIC _l.
Set up. The signal processing unit 101 includes a modeling circuit 15
, The microphone M installed at the identification control point P _l
Impulse response C ′ _lm = {C ′ _j (lm)} (j = 1 to J− between IC _l (1 ≦ l ≦ 4) and the speaker SP _m.
1) is identified in exactly the same manner as described above. Then, the modeling filter estimation circuit 102 estimates and determines the impulse response regarding the estimated control point CP.

It is now assumed that the impulse response _C'lm identified in the modeling circuit 15 is as shown in FIG. 10, for example. Impulse response C '
_{Since lm} represents a sound wave transmitted to the control point P _l when an impulse is emitted from the control speaker SP _m , the waveform in FIG. 10 corresponds to a reflected wave that has arrived at each control point P _l. . Therefore, the modeling filter estimation circuit 102 uses {C ′ _j1 (1m)} and {C ′ _j1 (3m)}.
In order to compare and, and extract the components of the reflected waves corresponding to both, {j1} and {j3} are used by using the evaluation function E of the following equation.
Correlate with.

[0101]

[Equation 12] E = [j1 · C ′ _j1 (1 m) −j3 · C ′ _j3 (3 m)] ²

A method such as DP matching may be used for this association. It should be noted that the reason that each value is multiplied by j in Equation 12 is that the sound wave attenuation is normalized by a model that is inversely proportional to the distance. The associated combinations are stored in the table as a set of (j0, j3). Impulse response _C'j0 (0m) corresponding to P ₀
For some j0,

[0103]

## EQU13 ## j0 = (j1 + j3) / 2

A set (j1, j3) satisfying the above conditions is obtained from the above table. This subscript is taken so that the time delay is exactly in the middle of both because P ₀ is at the midpoint of P ₁ and P ₃ . When there are a plurality of such sets in the table, one of them is selected by some method such as selecting the one with the smallest value of j1. By the above operation, only one set of (j1, j3) is determined for j0. The same operation is performed for {C ' _j2 (2m)} and {C' _j4 (4m)}, and for j0, (j2,
Only one set of j4) is defined. Using the set (j1, j3), (j2, j4) determined in this way, {C '
The value of _j0 (0m)} is estimated from the corresponding reflected wave according to the following formula.

[0105]

[Number 14] _{C 'j0 (0m) = [} (j0 / j1) · C' j1 (1m) + (j0 / j2) · C 'j2 (2m) + (j0 / j3) · C' j3 (3m) + (J0 / j4) ・ C ' _j4 (4m)] / 4

The above equation is an average of the values estimated from the respective identification control points BP. The modeling filter estimation circuit 102 writes the value of the transfer function obtained as described above in the modeling filter storage area 16, and uses this value when performing the calculation of Expression 7 when updating the control filter. .

Further, the noise path calculation circuit 103 uses the noise source signal x (n) and the output of the microphone MIC _l installed at the identification control point BP in the initialization stage to set LM.
Impulse response G _l = {g _i (l)} (i = 0 to I from the noise source to each control point P _l by a method such as the S method)
-1) is identified and written in the noise path storage area 105.
Then, the noise path calculation circuit 103 performs {g} for the estimated control point in exactly the same manner as the above-described estimation of C ′.
_i (l)} is estimated and written in the noise path storage area 105.

The pseudo noise calculation circuit 104 writes {g in the noise path storage area 105 as described above.
_{Based on i} (l)}, the microphone output e ′ _l (n) when determining the control filter coefficient is calculated as follows. That is, when P _l is the identification control point, the output e _l (n) of the MIC _l from the microphone is directly used as the microphone output e ′ _l (n) as in the following equation.

[0109]

[Equation 15] e ′ _l (n) = e _l (n)

On the other hand, when P _l is the estimated control point, the virtual microphone output e ′ _l (n) at the estimated control point is calculated by the following equation.

[0111]

[Equation 16]

In the noise control device of FIG. 8, the control filter update circuit 19 calculates e ′ calculated as above.
_{Using l} (n), the filter coefficient W _m
Is calculated sequentially.

[0113]

[Equation 17]

When this value is sufficiently converged, these processes are stopped, the contents of the control filter storage area 12 are fixed, and the control stage is entered.

As described above, in the noise control device of FIG. 8, it is possible to save the labor of identifying the impulse response by installing the microphone, which is required in the initialization stage of the device. That is, at the estimated control point, the estimation is performed without actually installing the microphone and actually identifying the impulse response, so that the labor for identifying C ′ can be reduced, and as a result, the apparatus can be practically used. It is possible to improve the simplicity of the aspect.

In the example of FIG. 8 as well, as in the example of FIG. 4, it is assumed that a plurality of control points and control speakers are provided. However, as described above, control is performed with only one control point. If multiple speakers are provided, or
The present invention can be applied to the case where a plurality of control points are provided but only one control speaker is provided.

Further, the noise control device of the above-mentioned embodiment is
When used in an environment where mechanical devices generate noise in factories or living rooms, it is possible to reduce noise in a wide space such as a human working range. It can also be applied to partitions with noise control devices.

[0118]

As described above, according to the first aspect of the present invention, the control point candidate corresponding to the position where the noise control should be performed is selected from the plurality of predetermined control point candidates. Since the signal processing of the signal processing means is switched according to the selected control point, even if the position where noise control should be performed changes, the control point corresponding to that position is selected. The signal processing according to is immediately performed, and it is possible to immediately follow the position variation and form the sound deadening zone.

According to the second aspect of the invention, when selecting the control point, the position where the noise control is to be performed is estimated, and the control point corresponding to the estimated position is selected from the control point candidates. Therefore, the sound deadening zone can be formed by following the free movement of the position where the noise control should be performed.

According to the third, fourth, fifth, sixth and seventh aspects of the invention, it is possible to control only the range required at that time, and a large number of control point microphones and control speakers are provided. In that case, only some of the control point microphones and control speakers are actually used.
The amount of calculation can be suppressed and the hardware scale can be suppressed.

According to the invention described in claim 8, one control speaker is assigned to a plurality of adjacent areas, and these areas are shared by the one control speaker. Therefore, it is possible to reduce the number of control speakers and further reduce problems in installing the control speakers.

According to the ninth and tenth aspects of the present invention, the control point is divided into the identification control point and the estimation control point, and the identification control point is output from the control speaker of the secondary sound source section. The impulse response is identified, and regarding the estimated control points, the impulse response from the control speaker of the secondary sound source section is estimated from the impulse response between the control speaker of the secondary sound source section and the identification control point, and those impulse responses are Since the processing in the signal processing means is determined based on the original, it reduces the effort of identifying the impulse response by installing the control point microphone required at the initialization stage of the device, and in the practical aspect of the device. The convenience can be improved.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an embodiment of a noise control system according to the present invention.

FIG. 2 is a diagram illustrating a configuration example of a control point selection unit.

FIG. 3 is a schematic diagram of a noise control device in which a large number of control speakers are provided in a secondary sound source section.

FIG. 4 is a diagram showing a specific configuration example of the noise control device of FIG.

5 is a diagram showing a first configuration example of a noise control device obtained by further improving the noise control device of FIG.

6A and 6B are diagrams showing configuration examples of a control point selection unit and a control speaker selection unit, respectively.

FIG. 7 is a diagram for explaining the operation of the noise control device of FIG.

8 is a diagram showing a second configuration example of a noise control device obtained by further improving the noise control device of FIG.

9 is a diagram for explaining the operation of the noise control device of FIG.

FIG. 10 is a diagram showing an example of an impulse response identified by a modeling circuit.

FIG. 11 is a diagram showing a configuration example of a conventional noise control system.

[Explanation of symbols]

1 Noise Observation Department 2 Signal processing unit 3 Secondary sound source section 4 Residual noise observation section 5 Coefficient updating section 6 coefficient storage 7 Control point selection section 8 Signal processing switching section 40 Position estimation unit 41 Signal generator 42 ultrasonic speaker 43 Ultrasonic Microphone 44 Control point determination unit P1 to Pm control point candidates

Claims (10)

[Claims] 1. A noise observing means for observing noise in a noise generating source, a signal processing means for processing a signal obtained by the noise observing means, and a noise control according to a signal resulting from signal processing by the signal processing means. Secondary sound source means for generating a secondary sound for control, and control point selection means for selecting, as a control point, a control point candidate corresponding to a position where noise control should be performed from a plurality of predetermined control point candidates. And a signal processing switching means for switching the signal processing content of the signal processing means in accordance with the selected control point. 2. The noise control device according to claim 1, wherein
The control point selecting means includes a position estimating means for estimating a position where noise control should be performed, and a control point determining means for selecting a control point corresponding to the estimated position from a plurality of control point candidates. A noise control device characterized in that 3. Noise control means for observing noise at a noise source, signal processing means for processing a signal obtained by the noise observing means, and noise control according to a signal resulting from signal processing by the signal processing means. Secondary sound source means for generating a secondary sound for control, a control point selecting means for selecting a control point to be controlled at a current time from a plurality of predetermined control points, and a selected control point. And a signal processing switching means for switching the signal processing contents of the signal processing means according to the noise control apparatus. 4. A noise observing means for observing noise in a noise generating source, a signal processing means for processing a signal obtained by the noise observing means, and a plurality of control speakers, wherein the signal processing means performs signal processing. Control for selecting a secondary sound source means for generating a secondary sound for noise control according to the signal of the generated result and at least one control speaker used for control at the current time from a plurality of speakers of the secondary sound source means. A noise control device comprising: speaker selection means; and signal processing switching means for switching the signal processing content of the signal processing means according to the selected control speaker. 5. A noise observing means for observing noise in a noise generating source, a signal processing means for processing a signal obtained by the noise observing means, and a plurality of control speakers, the signal processing means performing the signal processing. Secondary sound source means for generating a secondary sound for noise control according to the resulting signal, and control point selection for selecting a control point to be controlled at the current time from a plurality of predetermined control points Means, a control speaker selection means for selecting at least one control speaker used for control from the plurality of speakers of the secondary sound source means at the current time, and the signal according to the selected control point and / or control speaker. A noise control device comprising: signal processing switching means for switching the signal processing content of the processing means. 6. The noise control device according to claim 3, 4, or 5, when the space to be silenced is divided into areas of a predetermined size, the control speaker and / or the control point are divided. For each area, at least one is arranged to be in charge of controlling the area, and in this case, when the area to be controlled at the current time is selected from the areas, the control point selecting means, The noise control is characterized in that a control point in charge of the selected area is selected, and the control speaker selection means is configured to select a speaker in charge of the selected area. apparatus. 7. The noise control device according to claim 6,
Position detection means for selecting an area to be controlled at the current time from the areas is further provided, and the position detection means detects a position of a person who uses the device, and includes an area including a human head. A noise control device characterized in that is selected as a region to be silenced. 8. The noise control device according to claim 6,
A noise control device characterized in that one control speaker is assigned to these areas for a plurality of adjacent areas, and these areas are shared by the one control speaker. 9. The noise control device according to claim 3 or 5, wherein the control point is divided into an identification control point and an estimation control point, and the identification control point is output from a control speaker of a secondary sound source section. The impulse response is identified, and regarding the estimated control points, the impulse response from the control speaker of the secondary sound source section is estimated from the impulse response between the control speaker of the secondary sound source section and the identification control point, and those impulse responses are A noise control device characterized in that the processing in the signal processing means is determined based on the original. 10. The noise control apparatus according to claim 9, wherein when estimating an impulse response from a certain control speaker to a certain estimated control point, a plurality of identification controls around the estimated control point are provided. The amplitude and time delay of the corresponding reflected waves are detected between the impulse responses reaching the points, and the amplitude and time delay corresponding to each reflected wave are obtained according to the positional relationship between the estimated control points and their identification control points. A noise control device characterized by being adapted to estimate an impulse response.