JP6669011B2 - Active noise control device, active noise control program, and active noise control method - Google Patents

Description

  The present invention relates to an active noise control device, an active noise control program, and an active noise control method, and can be applied to, for example, a noise control device, a noise control program, and a noise control method used for noise leaking from a room. .

  In a duct such as a housing of an air conditioner or an electric device, a driving sound of the device or a fan becomes a noise source and generates noise. The generated noise propagates through the space inside the duct and is emitted from the outlet of the duct, which causes a noise problem. Many researches and developments have been made on methods for reducing the noise caused by such ducts.

  One technique for reducing duct noise is to use an active noise control (hereinafter referred to as "ANC") device. The ANC device is a method of generating control sound having the opposite phase and the same amplitude of noise to actively reduce (mute) noise.

  ANC devices include a feedforward type and a feedback type.

  A feed-forward type ANC apparatus is based on a reference microphone input signal, an error microphone input signal, a measured secondary path impulse response and a measured feedback path impulse response, and a FIR (Finite Impulse Response) adaptive filter. Is successively updated by a Filtered-X LMS (Least Mean Square) algorithm to generate a control signal for reducing noise at the installation point (erasing point or control point) of the error microphone, and the control signal is output from the control speaker. Is output, the noise at the installation point of the error microphone is reduced.

  The feedback-type ANC apparatus sequentially updates the filter coefficient of the FIR-type adaptive filter by the Filtered-X LMS algorithm based on the input signal of the error microphone and the measured impulse response of the secondary path, thereby setting the error microphone. A control signal for reducing noise at the point is generated, and a control signal is output from the control speaker to reduce noise at the installation point of the error microphone.

  By using these ANC devices, by suppressing the secondary vibration at the stage where the space is divided and a structure (for example, a wall of a room) becomes a secondary vibrator in the process of spreading vibration noise in the space, A method for reducing noise is proposed by Patent Document 1.

  The ANC device disclosed in Patent Document 1 detects original vibration of a noise source, and vibration of a boundary surface between a space in which primary noise and a space in which secondary noise propagates due to the vibration of the noise source. Then, the FIR filter is updated by the LMS algorithm using the detected vibration signal of the original vibration and the vibration signal of the boundary surface, a cancellation vibration signal is generated, and the cancellation vibration signal is generated by the vibration generation means installed at the boundary surface. And reduce the noise at the interface.

JP 2010-286684 A

  However, in the active noise control device described in Patent Literature 1, a sound generated from a position where vibration of a noise source does not easily propagate to a boundary surface (for example, a sound generated from a human mouth, output from a speaker installed on a stand). The original signal does not directly propagate to the boundary, and the input signal of the vibration sensor installed on the noise source (hereinafter referred to as vibration sensor) and the input signal of the vibration sensor installed on the boundary are The noise cannot be reduced because they are different.

  Even if the vibration sensor is changed to a microphone in order to record sound generated in space, since the frequency characteristics of the input signal of the microphone and the frequency characteristics of the input signal of the vibration sensor at the boundary surface are different, an adaptive filter must be used. Even if the update is performed by the LMS algorithm, the update of the filter coefficient is not performed correctly, the vibration of the boundary surface cannot be reduced, and the performance of the ANC device cannot be exhibited.

  Therefore, even if the boundary surface is vibrated by sound, the active noise control device and the active noise control program can correctly update the filter coefficient of the adaptive filter, reduce the vibration of the boundary surface, and reduce the vibration sound due to the vibration of the boundary surface. , And an active noise control method is desired.

  According to a first aspect of the present invention, there is provided an active noise control apparatus for calculating a control vibration signal for reducing vibration output from a vibration speaker using an input signal of a microphone and an input signal of a vibration sensor. A microphone input signal correction unit for correcting the frequency characteristics of the vibration sensor to the frequency characteristics of the vibration sensor, (2) a microphone input signal corrected by the microphone input signal correction unit, a vibration sensor input signal, and a measured vibration speaker. An adaptive filter unit that calculates the control vibration signal using an impulse response of a secondary path to the vibration sensor and updates a filter coefficient of the adaptive filter.

  An active noise control program according to a second aspect of the present invention is installed in an active noise control device that calculates a control vibration signal for reducing vibration output from a vibration speaker using an input signal of a microphone and an input signal of a vibration sensor. (1) a microphone input signal correction unit that corrects a frequency characteristic of an input signal of a microphone to a frequency characteristic of a vibration sensor; (2) a microphone input signal corrected by the microphone input signal correction unit; The control vibration signal is calculated using the input signal of the above and the measured impulse response of the secondary path from the vibration speaker to the vibration sensor, and functions as an adaptive filter unit for updating the filter coefficient of the adaptive filter. And

  According to a third aspect of the present invention, there is provided a microphone input signal correction unit for correcting an active noise control method for calculating a control vibration signal for reducing vibration output from a vibration speaker using an input signal of a microphone and an input signal of a vibration sensor. (1) the microphone input signal correction unit corrects the frequency characteristics of the microphone input signal to the frequency characteristics of the vibration sensor, and (2) the microphone input signal correction unit corrects the microphone input signal. The control vibration signal is calculated using the input signal of the microphone corrected by the unit, the input signal of the vibration sensor, and the measured impulse response of the secondary path from the vibration speaker to the vibration sensor. It is characterized in that the coefficient is updated.

  According to the present invention, even at the interface vibration generated by sound propagating in the space, the sound is received by the microphone, and the frequency characteristic of the microphone input signal is changed to the frequency characteristic of the vibration sensor input signal installed at the interface. By performing the correction, the filter coefficient of the adaptive filter is updated appropriately, the vibration of the boundary surface can be reduced, and the vibration sound due to the vibration of the boundary surface can be reduced.

It is a block diagram showing the composition of the ANC device of an embodiment. It is a block diagram showing the composition of the ANC device at the time of the impulse response measurement of the secondary route of an embodiment. FIG. 3 is a block diagram illustrating a configuration of an ANC device when calculating a correction filter according to the embodiment.

(A) Main Embodiment Hereinafter, embodiments of an active noise control device, an active noise control program, and an active noise control method of the present invention will be described in detail with reference to the drawings.

  This embodiment exemplifies a case in which the above-described ANC apparatus, ANC program, and ANC method of the present invention are applied to a room noise control system, for example.

(A-1) Configuration of Embodiment FIG. 1 is a block diagram illustrating a configuration of an ANC apparatus 100 according to an embodiment of the present invention.

  The ANC device 100 according to the embodiment of the present invention may be configured as, for example, a dedicated board, or may be realized by writing an ANC program into a DSP (digital signal processor). And may be realized by software (ANC program) executed by the CPU, but can be functionally represented by FIG.

  1, an ANC apparatus 100 according to this embodiment includes a reference microphone 101, a microphone amplifier 102, an AD converter 103 (103a, 103b), an ANC unit 104, a vibration sensor 105, a vibration sensor amplifier 106, a DA converter 107, It has a vibration speaker amplifier 108 and a vibration speaker 109.

  The reference microphone 101 is a microphone that receives a sound or the like emitted from the sound source S in the room R1. Note that the sound emitted from the sound source S is, for example, a human voice (a television, an audio device, or the like) that outputs a human voice or sound. Note that a wall W as a boundary surface exists between the room R1 and the room R2.

  The microphone amplifier 102 amplifies the input signal received by the reference microphone 101.

  The AD converter 103 (103a, 103b) converts an input signal amplified by the microphone amplifier 102 and a later-described vibration sensor amplifier 106 from an analog signal to a digital signal. Hereinafter, the input signal converted by the AD converter 103a is referred to as a “reference microphone input signal” and the input signal converted by the AD converter 103b is referred to as a “vibration sensor input signal”.

  The ANC unit 104 generates a control vibration signal for reducing the vibration of the wall W generated by the sound source S from the reference microphone input signal and the vibration sensor input signal. The ANC unit 104 measures the impulse response of the secondary path P in advance when the sound source S is not generated. The ANC unit 104 stores the measured impulse response of the secondary path P in a secondary path filter unit 115 described later. Further, when the sound source S is generated, the ANC unit 104 calculates a correction filter for correcting the frequency characteristic of the reference microphone input signal to the frequency characteristic of the vibration sensor 105 described later. Further, the ANC unit 104 holds the calculated correction filter in the microphone signal correction unit 113 described later.

  The vibration sensor 105 is a sensor that receives a vibration sound of the wall W. The vibration sensor 105 of the present embodiment uses a contact-type sensor that is fixed to a wall W for measuring vibration sound and measures the vibration sound by vibrating the sensor itself together with the vibrating body. A non-contact type sensor that measures the vibration sound without contacting the wall W may be used. The place where the vibration sensor 105 is installed is not limited, but it is preferable that the vibration sensor 105 is installed near the vibration speaker 109.

  The vibration sensor amplifier 106 amplifies the input signal detected by the vibration sensor 105.

  The DA converter 107 converts a control vibration signal generated in the ANC unit 104 from a digital signal to an analog signal.

  The vibration speaker amplifier 108 amplifies the control vibration signal converted into an analog signal.

  The vibration speaker 109 outputs an amplified control vibration signal.

  Next, a detailed configuration of the ANC unit 104 will be described.

  The ANC unit 104 includes a reference microphone input terminal 110, a vibration sensor input terminal 111, a control vibration signal output terminal 112, a microphone signal correction unit 113, an adaptive filter unit 114, a secondary path filter unit 115, and an adaptive filter coefficient update unit 116. Have.

  The reference microphone input terminal 110 is an interface for inputting a reference microphone input signal to the ANC unit 104.

  The vibration sensor input terminal 111 is an interface for inputting a vibration sensor input signal to the ANC unit 104.

  The control vibration signal output terminal 112 is an interface that outputs the control vibration signal generated by the ANC unit 104 to the outside.

  The microphone signal correction unit 113 corrects the reference microphone input signal using the held correction filter and outputs the corrected signal.

  The adaptive filter unit 114 calculates and outputs a control vibration signal using the corrected reference microphone input signal and the adaptive filter.

  The secondary path filter unit 115 calculates and outputs a filtered reference signal from the corrected reference microphone input signal and the measured impulse response of the secondary path P.

  The adaptive filter coefficient updating unit 116 updates the filter coefficient of the adaptive filter using an adaptive algorithm (for example, an LMS algorithm) using the filtered reference signal and the vibration sensor input signal.

(A-2) Operation of Embodiment The operation of the ANC apparatus 100 according to the embodiment of the present invention at the time of noise control will be described in detail.

  First, when the sound source S is not generated, the ANC apparatus 100 measures the impulse responses c to (k) of the secondary path P. Various methods can be widely applied to the method of measuring the impulse response c to (k) of the secondary path P. For example, as shown in FIG. 2, a white noise generation unit that causes the ANC unit 104 to generate white noise 201, the white noise generation unit 201 generates white noise, the vibration speaker 109 outputs the vibration of the white noise, the vibration sensor 105 receives the sound, and the secondary path adaptive filter coefficient update unit 203 executes the adaptive algorithm. (For example, using an LMS algorithm) to update the filter coefficient of the secondary path adaptive filter unit 204, and when the error calculated by the secondary path error calculation unit 202 converges, the secondary path adaptive filter unit There is a method of setting the filter coefficient of 204 as the impulse response c 応 答 (k) of the secondary path P. Further, the method of measuring the impulse responses c to (k) of the secondary path P is such that an impulse signal is output from the vibration speaker 109 a plurality of times, and the average of the impulse responses received by the vibration sensor 105 is averaged. A measurement method that sets the response c to (k) may be used.

  When the measurement of the impulse responses c to (k) of the secondary path P is completed, the ANC apparatus 100 stores the measured impulse responses c to (k) of the secondary path P in the secondary path filter unit 115.

  When the sound source S starts to be generated in the room R1, vibration of the wall W starts to be generated, and noise starts to be generated in the room R2. The ANC device 100 also starts operating when the sound source S starts to be generated.

  The sound of the sound source S in the room R1 propagates through the space in the room R1, reaches the position of the reference microphone 101, and is input to the reference microphone 101.

  The input signal input to the reference microphone 101 is amplified by the microphone amplifier 102, converted from an analog signal to a digital signal by the AD converter 103 a, and is input to the reference microphone input terminal 110 of the ANC unit 104. Is entered as

  On the other hand, when the sound of the sound source S in the room R1 propagates through the space in the room R1 and propagates to the wall W, the sound collides with the wall W, causing the wall W to vibrate, thereby generating vibration. Vibration sound is generated by the vibration of W.

  The vibration sensor 105 detects a vibration sound generated when the wall W vibrates.

  The vibration sound detected by the vibration sensor 105 is amplified by the vibration sensor amplifier 106, converted from an analog signal to a digital signal by the AD converter 103 b, and input to the vibration sensor input terminal 111 of the ANC unit 104 by the vibration sensor input signal e (n). Is entered as

  When the operation of the ANC apparatus 100 starts and signals start to be input to the reference microphone input terminal 110 and the vibration sensor input terminal 111 of the ANC unit 104, first, the ANC unit 104 calculates a correction filter h (k). Various methods can be widely applied to the calculation method of the correction filter h (k). For example, as shown in FIG. 3, an adaptive filter coefficient for a correction filter is calculated using a reference microphone input signal and a vibration sensor input signal. The update unit 302 updates the filter coefficient of the adaptive filter unit 303 for the correction filter using an adaptive algorithm (for example, an LMS algorithm), and the correction filter when the error calculated by the error calculation unit 301 for the correction filter converges. There is a method in which the filter coefficient of the adaptive filter unit 303 is used as a correction filter h (k). Further, the calculation method of the correction filter h (k) is that the reference microphone and the vibration sensor are installed close to each other before the sound source S is generated, the speaker outputs white noise, and the reference microphone input signal and the vibration sensor input signal are output. A calculation method may be used in which the filter coefficient of the adaptive filter is updated by an adaptive algorithm (for example, an LMS algorithm), and when the error converges, the filter coefficient of the adaptive filter is used as the correction filter h (k).

  When the calculation of the correction filter h (k) is completed, the ANC apparatus 100 stores the calculated correction filter h (k) in the microphone signal correction unit 113.

  When the calculation of the correction filter h (k) is completed, the ANC apparatus 100 starts generating a control vibration signal for reducing the vibration of the wall W.

The microphone signal correction unit 113 converts the corrected reference microphone signal x ′ (n) from the reference microphone input signal x (n) and the correction filter h (k) held in the microphone signal correction unit 113 using Expression (1). calculate.

FLEN _{h in} equation (1) is the filter length of the correction filter h (k). Equation (1) is obtained by convolving the reference microphone input signal x (n) with the correction filter h (k), thereby correcting the frequency characteristic of the reference microphone input signal to the frequency characteristic of the vibration sensor input signal. This is an equation for calculating (n). The microphone signal correction unit 113 outputs the corrected reference microphone signal x '(n) to the adaptive filter unit 114 and the secondary path filter unit 115.

The adaptive filter unit 114 calculates the control vibration signal y (n) from the corrected reference microphone signal x ′ (n) and the filter coefficient AF (n, k) of the adaptive filter using the equation (2).

FLEN _AF in the equation (2) is the filter length of the adaptive filter. Equation (2) is an equation that calculates the control vibration signal y (n) by convolving the corrected reference microphone signal x ′ (n) and the filter coefficient AF (n, k) of the adaptive filter. The adaptive filter unit 114 outputs the calculated control vibration signal y (n) to the control vibration signal output terminal 112 of the ANC unit 104.

  The control vibration signal y (n) output from the control vibration signal output terminal 112 of the ANC unit 104 is converted from a digital signal to an analog signal by the DA converter 107, amplified by the vibration speaker amplifier 108, and then transmitted from the vibration speaker 109. Is output.

  At the start of generation of the sound source S, the control vibration signal y (n) is not a signal for reducing the vibration, but cannot reduce the vibration sound due to the vibration of the wall W because the filter coefficient of the adaptive filter unit 114 is not converged. When the filter coefficient of the adaptive filter converges, a signal for reducing the vibration of the wall W is generated, and the vibration sound due to the vibration of the wall W that can be heard in the room R2 is reduced.

On the other hand, the secondary path filter unit 115 converts the filtered reference signal r ′ (n) into the equation (3) from the corrected reference microphone signal x ′ (n) and the measured impulse response c '(k) of the secondary path P. Calculate using

FLEN _{c ~ in the} expression (3) is the measured filter length of the impulse response c ~ (k) of the secondary path P. Equation (3) is an equation for calculating a filtered reference signal r ′ (n) by convolving the corrected reference microphone signal x ′ (n) with the measured impulse response c ~ (k) of the secondary path P. Secondary path filter section 115 outputs filtered reference signal r ′ (n) to adaptive filter coefficient update section 116.

The adaptive filter coefficient updating unit 116 updates the filter coefficient AF (n, k) by using an adaptive filter updating algorithm using the filtered reference signal r ′ (n) and the vibration sensor input signal e (n). In this embodiment, as shown in the following equation (4), the adaptive filter updating algorithm updates the filter coefficient AF (n, k) of the adaptive filter using the LMS algorithm.

Μ in the equation (4) is a step size for updating the adaptive filter coefficient. Various methods can be widely applied to the algorithm for updating the filter coefficient AF (n, k) of the adaptive filter. For example, the filter of the adaptive filter can be applied by a learning identification method (Normalized LMS) of the equation (5). The coefficient AF (n, k) may be updated.

(A-3) Effects of the Embodiment As described above, according to the embodiment, the ANC apparatus 100 receives the sound with the microphone even if the boundary surface vibration is generated by the sound, and the frequency characteristic of the microphone input signal. Is corrected to the frequency characteristic of the vibration sensor input signal provided on the boundary surface, the filter coefficient of the adaptive filter is appropriately updated, and the vibration on the boundary surface can be reduced.

(B) Other Embodiments In the above-described embodiments, various modified embodiments have been described. However, the present invention can be applied to the following modified embodiments.

  (B-1) The ANC device 100 described in the above embodiment uses a vibration sensor. However, the device may not be a vibration sensor as long as it can detect vibration.

  (B-2) Although the ANC device 100 described in each of the above embodiments uses a vibration speaker, the device may not be a vibration speaker as long as it can generate vibration.

100: ANC device, 101: Reference microphone, 102: Microphone amplifier, 103 (103a, 103b): AD converter, 104: ANC unit, 105: Vibration sensor, 106: Vibration sensor amplifier, 107: DA converter, 108: Vibration speaker amplifier 109 109 Vibration speaker 110 Reference microphone input terminal 111 Vibration sensor input terminal 112 Control vibration signal output terminal 113 Microphone signal correction unit 114 Adaptive filter unit 115 Secondary path filter Unit, 116: adaptive filter coefficient update unit, 201: white noise generation unit, 202: error calculation unit for secondary path, 203: adaptive filter coefficient update unit for secondary path, 204: adaptive filter unit for secondary path, 301 ... Error calculator for correction filter 302, adaptive filter coefficient updater for correction filter 303, for correction filter Adaptive filter, R1 ... room 1, R2 ... room 2, W ... wall, P ... 2 primary route, S ... sound source.

Claims (4)

In an active noise control device that calculates a control vibration signal that reduces vibration output from a vibration speaker using an input signal of a microphone and an input signal of a vibration sensor,
A microphone input signal correction unit for correcting the frequency characteristics of the input signal of the microphone to the frequency characteristics of the vibration sensor,
The control vibration signal is calculated using the microphone input signal corrected by the microphone input signal correction unit, the input signal of the vibration sensor, and the measured impulse response of the secondary path from the vibration speaker to the vibration sensor. And an adaptive filter unit for updating a filter coefficient of the adaptive filter. The active noise control device is installed on the space where the sound propagates, of the space where the sound propagates, the boundary surface that divides the space, and the space adjacent via the boundary surface,
A microphone installed on the space side where the sound propagates and receiving the sound;
A vibration sensor installed on the boundary surface on the space side where the sound propagates and detecting vibration of the boundary surface caused by the sound,
The active noise control device according to claim 1, further comprising: a vibration speaker that outputs the control vibration signal for reducing the vibration of the boundary surface. Using the input signal of the microphone and the input signal of the vibration sensor, a computer mounted on the active noise control device that calculates a control vibration signal that reduces vibration output from the vibration speaker,
A microphone input signal correction unit for correcting the frequency characteristics of the input signal of the microphone to the frequency characteristics of the vibration sensor,
The control vibration signal is calculated using the microphone input signal corrected by the microphone input signal correction unit, the input signal of the vibration sensor, and the measured impulse response of the secondary path from the vibration speaker to the vibration sensor. An active noise control program characterized by functioning as an adaptive filter section for updating a filter coefficient of an adaptive filter. In an active noise control method for calculating a control vibration signal for reducing vibration output from a vibration speaker using an input signal of a microphone and an input signal of a vibration sensor,
It has a microphone input signal correction unit and an adaptive filter unit to correct,
The microphone input signal correction unit corrects the frequency characteristics of the microphone input signal to the frequency characteristics of the vibration sensor,
The adaptive filter unit uses the microphone input signal corrected by the microphone input signal correction unit, the input signal of the vibration sensor, and the measured impulse response of the secondary path from the vibration speaker to the vibration sensor, An active noise control method comprising calculating a control vibration signal and updating a filter coefficient of an adaptive filter.

Images