JP3273051B2 - Active noise control device and adaptive noise control method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an active noise control device for reducing noise in an acoustic space by causing noise and a control sound having the opposite phase to interfere with each other. The present invention relates to an improvement of an active noise control device suitable for reducing noise in the vehicle, for example, "muffled noise" which is an engine noise, and an adaptive noise control method executed there.

[0002]

2. Description of the Related Art In general, as a method of reducing noise transmitted from a noise source into an acoustic space, a control sound having a phase inverted from that of the noise is generated in the acoustic space, and the noise is caused to interfere with each other. Are already known or adopted.

[0003] As an example, for example, see the paper "Multiple Error LMS Algorithm and Its Application to Active Control of Sound and Vibration (A Multiple Error LMSAlg
olithm and Its Application to Active Control of So
und and Vibration) "
On Acoustic, Speech and Signal Processing, ASSP-35, No. 10, 1987
(IEEE TRANS.ON ACOUSTICS, SPEECH, AND SIGNAL P
ROCESSING, VOL.ASSP-35, NO.10, October 1987)) or JP-A-1-501344 (UK Patent Application 86
No. 24053) introduces an active vibration control device and method for reducing engine noise in a passenger car as an acoustic space using an LMS (LMS) algorithm.

In the conventional active noise control, a reference signal having the same frequency characteristic as a sound wave to be controlled is generally generated based on an ignition signal or the like. On the other hand, one or a plurality of noise detection means (specifically, for example, a microphone or the like) detects noise in a vehicle interior which is an acoustic space, and cancels the noise detected by the noise detection means. The amplitude and phase of the reference signal are controlled by a digital active filter, and a control sound is generated from an additional sound generation means (specifically, a speaker or the like) based on the output signal. As a result, the control sound from the additional sound generating means interferes with the noise transmitted from the noise source, thereby reducing the noise in the vehicle interior. In the above-mentioned papers and the like, various applications and control algorithms using a microprocessor are introduced as a detailed analysis of the characteristics and behavior of the vehicle interior noise and a method of controlling the amplitude and phase of the reference signal. (In particular, page 1430,
V. in the right column. Algorithm change (V. MODIFICATION OF
THE ALGORITHM)).

[0005]

In the above-mentioned active noise control apparatus according to the prior art, the characteristics of the noise detection means and the additional sound generation means, and the sound in the vehicle interior, which is an acoustic space for performing silencing control, are provided. If the propagation characteristics, etc., differ from the initial value or the design value due to aging or usage conditions, not only the desired noise reduction characteristics cannot be obtained, but rather, sound waves having the same phase as the noise will be generated, Control may fall into a divergent state. In such a case, it is necessary to stop the control so that the user or the occupant does not feel uncomfortable.

[0006] In some cases, if a control on / off switch is provided so that a user or an occupant feels the effect of the control, the control may be stopped similarly.
However, stopping the control means that the additional output sound is suddenly turned off, which often generates a short, unpleasant, pulse-like noise called a pop sound, which is harmful to the user or the occupant. And so on.

Accordingly, the present invention has been made in view of the above-mentioned problems in the prior art. Specifically, the present invention eliminates the generation of pop noise when the control is stopped, and furthermore, It is an object of the present invention to provide an active noise control device and method which can be realized without adding any hardware such as an electronic circuit.

[0008]

SUMMARY OF THE INVENTION In order to reduce noise transmitted from a noise source in a predetermined acoustic space, the present invention provides a reference signal generating means for generating a reference signal correlated with the noise, Noise detection means for detecting noise in a space; and control means for taking in the reference signal and the detection output value of the noise detection means and controlling the amplitude and phase of the reference signal so as to acoustically interfere with the noise. And an additional sound generating means for taking in the reference signal controlled by the amplitude and the phase and generating an additional sound for noise reduction, wherein the control means adaptively updates a filter coefficient, and Including adaptive digital filter means for changing the amplitude and phase, the output means comprises a detection output value of the detection means having a first contribution coefficient and an output value of the filter means having a second contribution coefficient. Evaluation function An active noise control apparatus for executing a coefficient update algorithm for adaptively updating a filter coefficient of a filter means so as to minimize the filter coefficient, wherein the filter coefficient is an update of the immediately preceding filter coefficient. A first term multiplied by a detected output value having a contribution coefficient and a value of a reference signal filtered by sound propagation characteristics; and an output value and a reference signal having the second contribution coefficient. The control means sets the first contribution coefficient to zero when the control is off, and sets the second contribution coefficient to a normal term when the control is on when the control is off. An active noise control device is disclosed, wherein the active noise control device is set larger than that at the time of control, and executes a coefficient update algorithm so as to minimize the evaluation function in the set state.

Further, according to the present invention, a reference signal correlated with noise from a noise source is detected, and an additional sound for noise reduction is generated in a predetermined acoustic space based on the reference signal. Is a noise control method for reducing noise by acoustically interfering in the acoustic space, further detecting the noise in the acoustic space, a noise detection value having a first contribution coefficient A coefficient update for adaptively updating a filter coefficient so as to minimize an evaluation function constituted by an output value of a digital filter for changing an amplitude and a phase of the reference signal having a second contribution coefficient. In the adaptive noise control method for executing an algorithm, the filter coefficient is filtered by a detection output value having the first contribution coefficient and a sound propagation characteristic as an update to the immediately preceding filter coefficient. First the value of the reference signal is multiplied by
And a second term obtained by multiplying the output value having the second contribution coefficient and the reference signal, and when the control is off, the first contribution coefficient is set to zero. The second contribution coefficient is set to be larger than that at the time of normal control when the control is on, and a coefficient updating algorithm is executed under the setting state so as to minimize the evaluation function. An adaptive noise control method is disclosed.

[0010]

According to the present invention, when the control is turned off, only the output value of the filter for changing the amplitude and phase of the reference signal of the evaluation function is made to contribute, and the contribution coefficient is made larger than that in the normal control. A coefficient update algorithm so as to minimize the evaluation function.

More specifically, according to a control algorithm and a filtered X-LMS algorithm for minimizing the evaluation function J according to an embodiment of the present invention, the control algorithm for controlling the amplitude and phase of the reference signal is used. The output ym (n) of the output adaptive digital filter is represented by the following equation [Equation 1].

(Equation 1) Here, the coefficient Wmi (i; filter order) of the output adaptive digital filter is adaptively updated.
The updating method is based on the following equation [Equation 2].

(Equation 2) Where: α; convergence coefficient of the steepest descent method n; discretization time (sampling time) Wmi (n); value of the coefficient of the i-th tap of the m-th output digital filter at time n x (n) The reference signal Rlm (n) at time n; the value at time n of the reference signal x (n) filtered by the acoustic transfer characteristics.

When the control is stopped, the contribution ratio coefficients pl = 0 and qm = (predetermined values) for all l and m are set to the same values as those in the normal control with respect to the above equation [Equation 2]. The output adaptive digital filter is updated as described above. That is, in the above equation (Equation 2), the filter coefficient Wmi
Since the update amount, that is, the change, is calculated, the weighting coefficient is set as described above, and the above equation (Equation 2) becomes the following equation (Equation 3).

(Equation 3) When the filter coefficients are updated by repeatedly calculating based on this equation [Equation 3], the output ym (n) decreases exponentially, and thus a gradual change in the output is reduced by an extra electronic circuit or the like. It can be easily obtained without any additional hardware. In this way, it is possible to suppress the generation of the output pop sound.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the accompanying drawings. First, FIG. 2 shows a configuration of an active noise control system to which the present invention is applied. This embodiment is an example in which the present invention is applied to an automobile, particularly a passenger car. Vibration noise accompanying rotation of an engine is transmitted to a vehicle interior, which is an acoustic space, resulting in so-called "muffled noise" noise. And this active noise control system reduces this "muffled sound" noise.

In FIG. 2, as the noise detecting means, for example, a plurality of (in this embodiment, L) microphones 2,
2 is mounted in a vehicle interior, which is an acoustic space, more specifically, on a vehicle interior roof. The outputs of the microphones 2 and 2 are respectively filtered for high-frequency components through low-pass filters 3 and 3 such as an anti-aliasing filter. The signals from which these high frequency components have been filtered are input to a plurality of A / D converters 4, where they are converted into digital values and output as residual noise data el.

Further, in order to perform noise control, a noise correlation signal correlated with noise from a noise source, that is, a so-called reference signal is required. For example, in the case of controlling the “muffled sound” in the cabin of an automobile, the reference signal may be a rotational vibration of the engine, that is, an engine rotation signal or a sine wave signal having a frequency that is an integral multiple of the ignition signal. Good. In this embodiment, the ignition signal (more specifically,
For example, the generation timing of a high-voltage ignition signal generated from an engine control device or the like is taken in from a terminal 21 and shaped by a shaping circuit 22 so that the processor can easily process the signal.

On the other hand, the control device for controlling the noise is constituted by, for example, a microprocessor 5, in which an adaptive filter 51 is included as a function control block. The adaptive filter 51 is a so-called output adaptive digital filter. The microprocessor 5 performs the so-called active noise control by taking in the above-mentioned input signal. As an algorithm (control flow), for example, a filtered X-LMS algorithm is generally used.

The output calculated by the adaptive filter 51 of the microprocessor 5 is then sent to a plurality of D / A converters 6 and converted into analog quantities. These analog signals are respectively smoothed through output smoothing filters 7 and 7 and then amplified by a plurality of amplifiers 8 and 8, respectively. In the example, the additional sound is output from M speakers 9, 9, and is controlled, that is, reduced by acoustically interfering with the engine noise in the vehicle interior.

The active noise control system of the present invention is further provided with a switch 23 for switching on / off the control, whereby a user or an occupant can arbitrarily switch the noise control. You can experience the effects and select them as needed.

FIG. 1 shows a control flow according to an embodiment of the present invention.
Then, the output adaptive digital filter 51 (hereinafter, W
Filter) output. This output is the output of the microprocessor 5. The output ym (n) of the W filter is represented by the following equation (Equation 4).

(Equation 4) Here, Wmi (n); the value x (n) of the i-th tap coefficient at time n for the m-th filter W; the reference signal at time n. Regarding a system for performing silencing control, consider a predetermined evaluation function J of the following [Equation 5] configured so that the filter output value ym reduces the detection output value el.

(Equation 5) Here, E ｛｝; expected value el; residual noise detected by the l-th noise detection means ym; output of the m-th W filter pl, qm; first and second contributions which are weighting factors of respective values. Is the rate coefficient.

To reduce the noise in the acoustic space, the so-called steepest descent method is used to minimize the evaluation function J, and the coefficient of the W filter 51 used for controlling the amplitude and phase of the reference signal is used. Wmi (i; filter order) is adaptively updated. The updating method is based on a filtered X-LMS algorithm, which is a coefficient updating algorithm, shown in the following equation [Equation 6].

(Equation 6) Here, α: convergence coefficient n of the steepest descent method; discretization time (sampling time) Wmi (n); for the m-th W filter, the value of the i-th tap coefficient at time n x (n); Reference signal Rlm (n): value at time n of reference signal x (n) filtered by sound propagation characteristics.

According to the present invention, for example, the divergence state of control (for example, the characteristics of a microphone as noise detection means and the characteristics of a speaker as additional sound generation means, and the sound propagation characteristics in a vehicle interior as an acoustic space, etc.) However, if it differs from the initial value or the design value due to aging, usage conditions, etc., not only will the desired silencing characteristics not be obtained,
Rather, when a sound wave having the same phase as the noise is generated, or when it is determined that the switch 23 has been turned off, the output stop is determined in step 102. Is performed. As a result, if it is determined that output stop is necessary, the flow proceeds to step 103. Otherwise (ie, when it is determined that the output stop is unnecessary), the process proceeds to step 104,
The aforementioned W filter is updated.

On the other hand, in step 103, in order to stop the output, the contribution coefficient pl = 0, qm = predetermined value (compared to the normal control) for all l and m with respect to the above equation [Equation 6]. To a larger value). When the contribution coefficient is set as described above, the above equation [Equation 6] is changed to the following [Equation 7].

(Equation 7)

Here, the above equation [Equation 7] indicates the amount of change of the filter W. As is clear from the above equation [Equation 4], the output ym (n) is Wmi (n). Since the above equation [Equation 7] is proportional to the following equation [Equation 8]

(Equation 8) Is considered to be similar in shape. This indicates a function that decreases exponentially. When the filter coefficient is updated by repeatedly calculating based on the above equation (Equation 7), the output ym (n)
Decreases exponentially, and a slow output change can be easily obtained. That is, in this way, it is possible to suppress the generation of the pop sound generated when the control is turned off only by a simple software change.

[0024]

According to the present invention, by changing the coefficients of the adaptive digital filter for output and the updating thereof based on a certain rule, it is easy to add hardware such as extra electronic circuits without adding any hardware. In addition, it is possible to avoid pop noise that is unpleasant for a user or the like that occurs when noise control is stopped.

[Brief description of the drawings]

FIG. 1 is a diagram showing a control flow in an active noise control device applied to an automobile according to an embodiment of the present invention.

FIG. 2 is a diagram showing a system configuration of an active noise control device applied to an automobile according to an embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 2 microphone 22 waveform shaping circuit 23 on / off switch 3 low-pass filter 4 A / D converter 5 microprocessor 51 adaptive digital filter for output 6 D / A converter 7 smoothing filter 8 amplifier 9 additional sound output means

Continuation of the front page (56) References JP-A-5-61487 (JP, A) JP-A-5-61483 (JP, A) JP-A-3-266516 (JP, A) JP-A-3-178846 (JP) JP-A-3-25295 (JP, A) JP-A-2-285799 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G10K 11/178 B60R 11/02 F01N 1/00 F01N 1/06 H03H 17/02 601 H03H 21/00

Claims (9)

(57) [Claims] 1. A reference signal generating means for generating a reference signal having a correlation with the noise in order to reduce noise transmitted from a noise source in a predetermined acoustic space, and detecting noise in the acoustic space. Noise detection means, control means for taking in the reference signal and the detection output value of the noise detection means, and controlling the amplitude and phase of the reference signal so as to acoustically interfere with the noise; An additional sound generating means for receiving a controlled reference signal and generating an additional sound for noise reduction, wherein the control means adaptively updates a filter coefficient to change an amplitude and a phase of the reference signal. An evaluation function including digital filter means and comprising an output value of a detection means having a first contribution coefficient and an output value of a filter means having a second contribution coefficient is minimized. To Phil An active noise control device for executing a coefficient updating algorithm for adaptively updating a filter coefficient of the filter means, wherein the filter coefficient has the first contribution coefficient as an update to the immediately preceding filter coefficient. The first term in which the detected output value is multiplied by the value of the reference signal filtered by the sound propagation characteristic, and the second term in which the output value having the second contribution coefficient is multiplied by the reference signal The control means sets the first contribution coefficient to zero when the control is off, and sets the second contribution coefficient larger than that in the normal control when the control is on. An active noise control device characterized by executing a coefficient updating algorithm so as to minimize the evaluation function in the set state. 2. The active noise control device according to claim 1, wherein a switch for switching an on / off state of the control is provided in the acoustic space. 3. The active noise control device according to claim 1, wherein said control means comprises a microprocessor. 4. The active noise control device according to claim 1, wherein the coefficient update algorithm that minimizes the evaluation function executed by the control device is a filtered X-LMS algorithm. 5. The active noise control device according to claim 1, wherein said additional sound generating means comprises one or more speakers, and said noise detecting means comprises one or more microphones. . 6. The active noise control device according to claim 1, wherein said reference signal generating means includes an anti-aliasing filter. 7. A reference signal correlated with noise from a noise source is detected, and an additional sound for noise reduction is generated in a predetermined acoustic space based on the reference signal. A noise control method for reducing noise by acoustically interfering in a space, further detecting noise in the acoustic space, and applying a first contribution coefficient to the noise detection value. A coefficient updating algorithm for adaptively updating the filter coefficients so as to minimize the evaluation function formed by the output value of the digital filter for changing the amplitude and the phase of the reference signal having the contribution coefficient of 2. In the adaptive noise control method to be executed, the filter coefficient is, as an update to the immediately preceding filter coefficient, a detection value having the first contribution coefficient and a reference signal filtered by a sound propagation characteristic. And a second term obtained by multiplying an output value having the second contribution coefficient and a reference signal by a reference value.
The first contribution coefficient is set to zero, the second contribution coefficient is set to be larger than that in the normal control when the control is on, and the evaluation function is minimized under this setting state. An adaptive noise control method characterized by executing a coefficient updating algorithm. 8. The adaptive noise control method according to claim 7, wherein said coefficient updating algorithm is a filtered X-LMS algorithm. 9. The adaptive noise control method according to claim 7, wherein a change in an output value of said filter is changed exponentially.