JP2008260420A - Active type noise control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an active type noise control device with practicality suppressing offensive quantization noise in a low noise condition, even if an operation at low bit rates having no high calculation accuracy is used. <P>SOLUTION: In this active type noise control device, an adaptive type notch filter is used. When the sum or the sum of squares of each absolute value of each of a filter coefficient W1[n] of a first 1-tap digital filter and a filter coefficient W2[n] of a second 1-tap digital filter becomes lower than a prescribed value, a noise control signal is interrupted to prevent the quantization noise from being heard. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an active noise control apparatus that actively reduces vibration noise generated from rotating equipment such as a vehicle engine.

  In a conventional active noise reduction device, a method of performing adaptive control using an adaptive notch filter is known (see, for example, Patent Document 1). FIG. 7 shows a configuration equivalent to the configuration of the conventional active noise reduction device described in Patent Document 1. In FIG.

  In FIG. 7, the discrete calculation for realizing the active noise reduction device is executed in the discrete calculation processing unit 15. The engine speed detector 1 outputs a pulse train having a frequency proportional to the engine speed as an engine pulse p. For example, the engine pulse p is generated by taking out the output of the crank angle sensor. The frequency detector 2 calculates and outputs a noise frequency f based on the engine pulse p. The reference signal generation unit 18 has a sine wave table 3 that holds the value of each point obtained by equally dividing one cycle of the sine wave in a memory, and the data is selected from the sine wave table 3 by the selection means 19, and the frequency is A reference sine wave signal x1 [n] and a reference cosine wave signal x2 [n] equal to the noise frequency f are generated and output. The correction signal generation unit 20 represents a reference sine wave signal correction value table 21 that simulates a transfer characteristic value from the speaker 10 to the microphone 11 (a reference sine wave signal correction value at a frequency f [Hz] is represented as C1 [f]. ) And the reference cosine wave signal correction value table 22 (the reference cosine wave signal correction value at the frequency f [Hz] is expressed as C2 [f]), and the corrected sine wave signal r1 [n] and the corrected cosine wave. A signal r2 [n] is generated and output.

r1 [n] = x1 [n] × C1 [f] + x2 [n] × C2 [f] (1)
r2 [n] = x2 [n] × C2 [f] −x1 [n] × C1 [f] (2)
The first one-tap digital filter 7 filters x1 [n] with a filter coefficient W1 [n] held therein to generate a first control signal y1 [n]. The second 1-tap digital filter 8 filters the reference cosine wave signal x2 [n] with a filter coefficient W2 [n] held therein to generate a second control signal y2 [n]. The power amplifier 9 amplifies a signal obtained by adding the first control signal y1 [n] and the second control signal y2 [n]. The speaker 10 outputs the output signal from the power amplifier 9 as noise canceling sound. The microphone 11 detects sound generated as a result of interference between noise and noise canceling sound as an error signal ε [n]. Based on the corrected sine wave signal r1 [n] and the error signal ε [n], the first adaptive control algorithm calculation unit 12 uses, for example, a filter coefficient W1 based on an LMS (Least Mean Square) algorithm which is a kind of steepest descent method. [N] is updated sequentially. Similarly, the second adaptive control algorithm calculation unit 13 sequentially updates the filter coefficient W2 [n] based on the corrected cosine wave signal r2 [n] and the error signal ε [n].

The sequential updating formulas of the coefficients W1 and W2 are W1 [n + 1] = W1 [n] −μ × r1 [n] × ε [n] (3)
W2 [n + 1] = W2 [n] −μ × r2 [n] × ε [n] (4)
It becomes. Here, μ is a constant called a convergence coefficient, and is related to the time for which the coefficients W1 and W2 converge to the optimum value.

And it is possible to reduce noise by repeating such a process described above at a predetermined cycle (hereinafter referred to as a sampling cycle).
JP 2004-361721 A

  In such a conventional active noise reduction device, the discrete arithmetic processing unit 15 performs discrete numerical calculation to generate a noise control signal having the same amplitude and opposite phase as the noise. The discrete calculation processing unit 15 is generally implemented by digital calculation using a microcontroller. In general, in a digital calculation using a microcontroller, the processing time becomes longer as the number of calculation digits is larger and the product operation is used more frequently. Even in the active noise control apparatus having the above-described conventional configuration, as shown in the equations (1), (2), (3), and (4), product calculation is frequently used in the calculation, and the calculation load is large. Therefore, depending on the performance of the microcontroller, the number of calculation digits must be reduced in order to complete predetermined calculation processing within the sampling period, and the number of calculation digits such as an operation with 8-bit accuracy is small. There are many cases where the method must be adopted. When such 8-bit computation is performed, the quantization noise becomes very large. Since this quantization noise is always constant without depending on the magnitude of the output, it does not matter at all when the output signal is large. However, when the output signal is small, only the noise is heard, which is very disturbing. In other words, when such a conventional active noise reduction device is used to reduce a large noise, the output from the active noise reduction device is also large, and the effect of reducing the noise can be enjoyed without worrying about the quantization noise. When the noise is low, the output from the active noise reduction device is small, and the quantization noise is noticeable rather than the effect of reducing the noise, and there is a problem that only very annoying noise is noticeable.

  It is an object of the present invention to provide a practical active noise control apparatus that suppresses annoying quantization noise in a low noise state even when low-bit calculation with low calculation accuracy is used.

  The active noise control apparatus of the present invention includes a control target noise frequency detection unit that detects a frequency of noise to be controlled due to a noise source, and a frequency that is the same as the noise frequency detected by the control target noise frequency detection unit. A sine wave generating means for generating a reference sine wave, a cosine wave generating means for generating a reference cosine wave, a first one-tap digital filter to which a reference sine wave signal from the sine wave generating means is input, and the cosine wave Noise obtained by adding the second one-tap digital filter to which the reference cosine wave signal from the generation unit is input, the output from the first one-tap digital filter, and the output from the second one-tap digital filter. Analog amplifying means for amplifying the control signal and interference for interfering with the noise to be controlled caused by the noise source when the signal from the analog amplifying means is inputted. An interference signal generating means for outputting a signal; and an error signal detecting means for detecting an error signal resulting from interference between the interference signal output from the interference signal generating means and the noise to be controlled due to the noise source; The first coefficient updating means for updating the filter coefficient of the first one-tap digital filter and the second coefficient updating means for updating the filter coefficient of the second one-tap digital filter, the first coefficient The updating means and the second coefficient updating means receive the error signal from the error signal detecting means, the reference sine wave signal from the sine wave generating means, and the reference cosine wave signal from the cosine wave generating means from the error detecting means. The noise in the error signal detecting means is reduced by the corrected sine wave signal and the corrected cosine wave signal corrected by the transfer characteristic to the interference signal generating means. In the active noise reduction apparatus, when the sum of the absolute value of the filter coefficient of the first one-tap digital filter and the absolute value of the filter coefficient of the second one-tap digital filter is less than a predetermined value, the value of the noise control signal Is forcibly set to 0.

  In another embodiment, when the sum of squares of the filter coefficient of the first one-tap digital filter and the filter coefficient of the second one-tap digital filter is less than a predetermined value, the value of the noise control signal is forcibly set to 0. It is characterized by.

  In another embodiment, when the sum of the absolute values of the filter coefficient of the first one-tap digital filter and the filter coefficient of the second one-tap digital filter is less than or equal to a predetermined value, the amplification degree of the analog amplification means is lowered. It is characterized by that.

  In another embodiment, when the sum of squares of the filter coefficient of the first one-tap digital filter and the filter coefficient of the second one-tap digital filter is less than a predetermined value, the amplification degree of the analog amplifying means is lowered. Features.

  The active noise control apparatus of the present invention estimates the noise level by calculating a filter coefficient of the first one-tap digital filter and a filter coefficient of the second one-tap digital filter, and the calculation is performed. When the value is less than or equal to the predetermined value, the noise control signal is stopped or the amplification degree of the analog amplification means for amplifying the noise control signal is lowered. This prevents the quantization noise from being heard when the quantization noise is conspicuous, and the noise reduction operation is performed as usual when the noise is loud. The effect that an effect can be exhibited is obtained.

(Embodiment)
Hereinafter, an active noise control apparatus according to Embodiment 1 of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram of an active noise control apparatus according to Embodiment 1 of the present invention.

  In FIG. 1, an engine speed detector 1 outputs a pulse train having a frequency proportional to the engine speed as a noise source mounted on a vehicle as an engine pulse p. The frequency detection unit 2 as the control target noise frequency detection means calculates and outputs the control target noise frequency f [Hz] from the engine pulse p. The sine wave table 3 serving as discretized sine wave data holds a sine value at each point obtained by dividing one cycle of the sine wave into N equal parts. The sine wave generating means 5 reads out data at a predetermined interval based on the control target noise frequency f from the sine wave table for each sampling period, and generates a reference sine wave signal x1 [n]. Similarly, the cosine wave generating means 6 reads out data from the sine wave table 3 at a predetermined interval based on the control target noise frequency f at every sampling period, but at the same time point, the point preceding the sine wave generating means by N / 4. Is generated as a reference cosine wave signal x2 [n]. When the read point exceeds N, a point obtained by subtracting N from the read point must be set as a new read point. The characteristic table 4 holds, for each frequency, a phase characteristic conversion value P [f] obtained by converting the phase characteristic from the speaker 10 to the microphone 11 into a relative point movement amount of the number N of points in the sine wave table 3. Based on the control target noise frequency f, the reference signal generation unit 14 reads the phase characteristic conversion value P [f] at the control target noise frequency f from the characteristic table 4, and based on them, the reference sine wave signal r1 [n], the reference cosine wave A signal r2 [n] is generated.

  Next, the first one-tap digital filter 7 holds the first filter coefficient W1 [n] inside, and based on the reference sine wave signal x1 [n] and the first filter coefficient W1 [n]. 1 control signal y1 [n] is output. The second one-tap digital filter 8 holds the second filter coefficient W2 [n] inside, and performs the second control based on the reference cosine wave signal x2 [n] and the second filter coefficient W2 [n]. The signal y2 [n] is output. The power amplifier 9 as an analog amplifying unit amplifies a signal obtained by DA-converting the noise control signal obtained by adding the first control signal y1 [n] and the second control signal y2 [n]. The speaker 10 as the interference signal generating means outputs the output signal from the power amplifier 9 as noise canceling sound. The microphone 11 serving as the error signal detection means detects a sound generated as a result of interference between the control target noise generated due to engine vibration and the noise canceling sound as an error signal ε [n]. The first adaptive control algorithm calculation unit 12 as the first coefficient updating unit 12 uses the filter coefficient W1 [of the first one-tap digital filter 7 based on the reference sine wave signal r1 [n] and the error signal ε [n]. n] are updated sequentially. The second adaptive control algorithm computing unit 13 as the second coefficient updating means 13 uses the filter coefficient W2 [of the second one-tap digital filter 8 based on the reference cosine wave signal r2 [n] and the error signal ε [n]. n] are updated sequentially. As described above, the discrete arithmetic processing unit 15 is configured by software.

  16 calculates a first filter coefficient W1 [n] in the first one-tap digital filter 7 and a second filter coefficient W2 [n] in the second one-tap digital filter 8, and the values are equal to or larger than a predetermined value. In this case, a coefficient calculating means for generating a control signal is provided, and 17 is a switch for performing an opening / closing operation by a control signal from the coefficient calculating means 16, and when the calculated value in the coefficient calculating means 16 becomes a predetermined value or more. Is opened and the noise control signal is cut off.

  Next, a specific operation of this apparatus will be described.

  Generation of the reference sine wave signal x1 [n], generation of the reference cosine wave signal x2 [n], generation of the reference sine wave signal r1 [n], generation of the reference cosine wave signal r2 [n], first Generation of the control signal y1 [n], generation of the second control signal y2 [n], detection of the error signal ε [n], update of the filter coefficient W1 [n], and filter coefficient W2 [n] And the second sine wave signal x3 [n] are all executed in the same cycle. Hereinafter, this period is described as T [seconds].

  For example, the frequency detector 2 generates an interrupt at each rising edge of the engine pulse p, measures the time between the rising edges, and calculates the frequency f of the control target noise based on the measurement result.

  The sine wave table 3 equally divides one cycle of the sine wave into N, and holds discrete data of sine values at each point on the memory. When an array storing sine values from the 0th point to the (N-1) th point is represented by z [m] (0 ≦ m <N), the relational expression (5) is established.

z [m] = sin (360 ° × m / N) (5)
For example, FIG. 3 and Table 1 show a graph and a table of z [m] when N = 3000 with 8-bit accuracy, respectively.

  The characteristic table 4 is a phase characteristic obtained by converting the amplitude characteristic array G [f] representing the amplitude characteristic of the transmission characteristic from the speaker 10 to the microphone 11 and the phase characteristic into a relative point movement amount of the number N of points of the sine wave table 3. The converted value array P [f] is held in the memory (f is the frequency [Hz]).

  When the amplitude characteristic at f [Hz] is β [f] (dB) and the phase characteristic is θ [f] (degrees), the relational expression (6) is established.

P [f] = N × θ [f] / 360 (6)
For example, FIG. 5 shows an example of the phase characteristic θ [f] when N = 3000 and the range of the noise frequency to be controlled is 30 Hz to 100 Hz, and Table 2 shows the corresponding phase characteristic array P [f].

  The sine wave generating means 5 stores the current read position i [n] of the sine wave table 3 in the memory, and moves the current read position every cycle based on the control target noise frequency f by the equation (7). Let

i [n + 1] = i [n] + N × f × T (7)
However, when the calculation result of the right side of Expression (7) is N or more, the result of subtracting N from the calculation result of the right side of Expression (7) is i [n + 1].

  At the same time, the sine wave generating means 5 generates the reference sine wave signal x1 [n] having the same frequency as the control target noise frequency f by the equations (8) and (9).

ix1 = i [n] (8)
x1 [n] = z [ix1] (9)
However, when the calculation result of the right side of Expression (8) is N or more, ix1 is obtained by subtracting N from the calculation result of the right side of Expression (8).

  Further, the cosine wave generating means 6 generates a reference cosine wave signal x2 [n] having the same frequency as the control target noise frequency f and advanced by a quarter of a period from the reference sine wave signal x1 [n] using the equation (10). And the equation (11).

ix2 = i [n] + N / 4 (10)
x2 [n] = z [ix2] (11)
However, when the calculation result of the right side of Expression (10) is N or more, ix2 is obtained by subtracting N from the calculation result of the right side of Expression (10).

  At the same time, the reference signal generation unit 14 converts the amplitude characteristic value and the phase characteristic of the transfer characteristic from the speaker 10 to the microphone 11 at the control target noise frequency f into the relative point movement amount of the number N of points in the sine wave table 3. The phase characteristic conversion value is extracted as P [f] from the characteristic table 4, and the reference sine wave signal r1 [n] and the reference cosine wave signal r2 [n] are created by the following method.

Reference sine wave signal r1 [n]
ix3 = i [n] + P [f] (12)
r1 [n] = z [ix3] (13)
Reference cosine wave signal r2 [n]
ix4 = i [n] + N / 4 + P [f] (14)
r1 [n] = z [ix3] (15)
Next, the first one-tap digital filter 7 holds the first filter coefficient W1 [n] inside, and based on the reference sine wave signal x1 [n] and the first filter coefficient W1 [n]. 1 control signal y1 [n] is output. The second one-tap digital filter 8 holds the second filter coefficient W2 [n] inside, and performs the second control based on the reference cosine wave signal x2 [n] and the second filter coefficient W2 [n]. The signal y2 [n] is output. The power amplifier 9 as an analog amplifying unit amplifies a signal obtained by DA-converting the noise control signal obtained by adding the first control signal y1 [n] and the second control signal y2 [n]. The speaker 10 as the interference signal generating means outputs the output signal from the power amplifier 9 as noise canceling sound. The microphone 11 serving as the error signal detection means detects a sound generated as a result of interference between the control target noise generated due to engine vibration and the noise canceling sound as an error signal ε [n]. The first adaptive control algorithm calculation unit 12 as the first coefficient updating unit 12 uses the error signal ε [n] and the reference sine wave signal r1 [n] to filter coefficient W1 [n of the first one-tap digital filter 7. ] Are updated sequentially. The second adaptive control algorithm computing unit 13 as the second coefficient updating means uses the coefficient updating error signal ε [n] and the reference cosine wave signal r2 [n] to filter coefficients of the second one-tap digital filter 8. W2 [n] is updated sequentially. As described above, the discrete arithmetic processing unit 15 is configured by software.

W1 [n + 1] = W1 [n] −μ × r1 [n] × ε [n] (16)
W2 [n + 1] = W2 [n] −μ × r2 [n] × ε [n] (17)
By executing these processes for each sampling period, the frequency f component in the error signal ε [n] is reduced.

  On the other hand, the coefficient calculation means 16 calculates the absolute values of the first filter coefficient W1 [n] in the first 1-tap digital filter 7 and the second filter coefficient W2 [n] in the second 1-tap digital filter 8. The sum is calculated, and when the calculated value is equal to or less than the predetermined value, the switch 17 is opened so that the noise control signal is not input to the power amplifier 9. The sum of the absolute values of the first filter coefficient W1 [n] in the first one-tap digital filter 7 and the second filter coefficient W2 [n] in the second one-tap digital filter 8 is indirectly noise. When the value is small, quantization noise is relatively conspicuous. Therefore, the predetermined value is determined by comprehensively considering the effect of noise reduction and the range where the quantization noise can be heard. do it. The sum of the absolute values of the first filter coefficient W1 [n] in the first 1-tap digital filter 7 and the second filter coefficient W2 [n] in the second 1-tap digital filter 8 has a predetermined value. When it exceeds or falls, the switch 17 is opened and closed. However, the value of the noise control signal at that time cannot be specified, and a pop sound may be generated. In such a case, it is also preferable to apply control such that the opening and closing of the switch 17 is slightly delayed and the switch 17 is opened and closed at the moment when the noise control signal becomes zero.

  In the second embodiment of the present invention, the coefficient calculating means 16 uses the first filter coefficient W1 [n] in the first 1-tap digital filter 7 and the second filter coefficient W2 [ n] is calculated, and the switch 17 is opened when the calculated value is equal to or less than a predetermined value. This has an effect that the quantization noise can be reduced while the noise magnitude can be estimated more accurately than in the first embodiment.

  Needless to say, it is also preferable to delay the opening / closing timing of the switch 17 as in the first embodiment and to perform control to open / close the switch 17 at the moment when the noise control signal becomes zero.

  FIG. 2 is a block diagram showing the configuration of the third embodiment of the present invention. Although most of the configurations are the same as those of the first embodiment, the power amplifier 9 as the analog amplifying means is configured so that the amplification degree can be varied by the control signal of the coefficient calculating means 16. The sum or square sum of the absolute values of the first filter coefficient W1 [n] in the first 1-tap digital filter 7 and the second filter coefficient W2 [n] in the second 1-tap digital filter 8 is When the value is lower than the predetermined value, the coefficient calculation means 16 outputs a control signal and lowers the amplification degree of the power amplifier 9 so that the quantization noise is not noticeable.

  At this time, the sum or square sum of the absolute values of the first filter coefficient W1 [n] in the first one-tap digital filter 7 and the second filter coefficient W2 [n] in the second one-tap digital filter 8 are used. When the value falls below or exceeds the predetermined value, the amplification factor of the power amplifier 9 is changed. However, if the amplification factor is changed suddenly, there is a concern about pop noise. It is preferable to control to change the degree.

  The active noise control device according to the present invention is useful as a highly practical active noise control device that can reduce quantization noise that is noticeable when the noise is low and can cope with low noise to high noise.

Block diagram for explaining an active noise control apparatus according to Embodiment 1 of the present invention Block diagram for explaining an active noise control apparatus according to Embodiment 3 of the present invention Characteristic diagram showing an example of a sine wave table in the active noise control device Characteristic diagram showing an example of transfer characteristics from the speaker to the microphone in the active noise control device Block diagram showing the configuration of a conventional active noise reduction device

Explanation of symbols

1 Engine speed detector 2 Frequency detector (Controlled noise frequency detection means)
DESCRIPTION OF SYMBOLS 3 Sine wave table 4 Characteristic table 5 Sine wave production | generation means 6 Cosine wave production | generation means 7 1st 1 tap digital filter 8 2nd 1 tap digital filter 9 Power amplifier (analog amplification means)
10 Speaker (Interference signal generation means)
11 Microphone (error signal detection means)
12 1st adaptive control algorithm calculating part (1st coefficient update means)
13 2nd adaptive control algorithm calculating part (2nd coefficient update means)
DESCRIPTION OF SYMBOLS 14 Reference signal production | generation part 15 Discrete operation processing part 16 Coefficient calculation means 17 Switch 18 Reference signal generation part 19 Selection means 20 Correction signal generation part 21 Reference sine wave signal correction value table 22 Reference cosine wave signal correction value table

Claims (4)

Control target noise frequency detection means for detecting the frequency of the noise to be controlled due to the noise source, and sine wave generation for generating a reference sine wave having the same frequency as the noise frequency detected by the control target noise frequency detection means Means, a cosine wave generating means for generating a reference cosine wave, a first one-tap digital filter to which a reference sine wave signal from the sine wave generating means is input, and a reference cosine wave signal from the cosine wave generating means are input. A second one-tap digital filter, and analog amplifying means for analog-amplifying a noise control signal obtained by adding the output from the first one-tap digital filter and the output from the second one-tap digital filter; An interference signal generator that receives the signal from the analog amplification means and outputs an interference signal for causing interference with noise to be controlled caused by the noise source. Error signal detecting means for detecting an error signal resulting from interference between the interference signal output from the interference signal generating means and the noise to be controlled caused by the noise source, and the first one-tap digital filter The first coefficient updating means for updating the filter coefficient and the second coefficient updating means for updating the filter coefficient of the second one-tap digital filter, and the first coefficient updating means and the second coefficient updating. The means transmits the error signal from the error signal detection means, the reference sine wave signal from the sine wave generation means, and the reference cosine wave signal from the cosine wave generation means to the transfer characteristic from the error detection means to the interference signal generation means. In the active noise reduction device configured to reduce the noise in the error signal detection means by the corrected sine wave signal and the corrected cosine wave signal When the sum of the absolute value of the filter coefficient of the first 1-tap digital filter and the absolute value of the filter coefficient of the second 1-tap digital filter is less than or equal to a predetermined value, the value of the noise control signal is forcibly set to 0. An active noise control device. The value of the noise control signal is forcibly set to 0 when the sum of squares of the filter coefficient of the first one-tap digital filter and the filter coefficient of the second one-tap digital filter is equal to or less than a predetermined value. The active noise control device according to claim 1. When the sum of the absolute value of the filter coefficient of the first one-tap digital filter and the absolute value of the filter coefficient of the second one-tap digital filter is less than or equal to a predetermined value, the amplification degree of the analog amplifying means is lowered. The active noise control device according to claim 1. 2. The degree of amplification of the analog amplifying means is lowered when the sum of squares of the filter coefficient of the first one-tap digital filter and the filter coefficient of the second one-tap digital filter is a predetermined value or less. The active noise control apparatus as described.

Images