JPH07239692A - Muffler - Google Patents

Abstract

PURPOSE:To provide a muffler with a high operation speed and preventing from being affected by disturbance and noise and remarkably reducing the device cost. CONSTITUTION:An output (x') of a digital filter is supplied to an absolute value arithmetic unit 42, and the absolute value output is supplied to a maximum value comparator 43, and the maximum value xp being the output of the comparator is squared by a squarer 46 to be added to a prescribed value (beta) by an adder 47, and a prescribed value C is divided by the addition result Pxp at an analog divider 48, and a (DELTAn) is obtained. The result multiplying the DELTAn by a muffle error signal (e) is multiplied by the (x') and edges of delay lines 241, 242,...24n-1, and a next filter coefficient of an adaptive filter 7 is operated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a silencer.

[0002]

2. Description of the Related Art FIG. 6 shows a so-called active noise control (Active Noise Co).
In order to prevent this, noise sources of low frequency are generated in a noise source, for example, a large vibrating drum shaker, and this system is used to prevent this.

In FIG. 6, noise from a noise source is guided to the noise duct 1 from the left side in the figure, and a noise detection microphone 2 is provided as a means for detecting this noise. An error detection microphone 3 is provided. The detection output of the noise detection microphone 2 is an analog output, which is amplified by the amplifier 4 and converted into a digital value by the A / D converter 5. This digital output NS is supplied to the subtractor 6. The adaptive filter 7 includes a plurality of multipliers, and this number is called the number of taps. This multiplication coefficient can be updated, and the calculation output of the coefficient calculator 14 that calculates this coefficient is supplied to the digital output. y
Is supplied to the D / A converter 9, converted into an analog value, amplified by the amplifier 10 and supplied to the speaker 11. Also, the digital output y of the adaptive filter 7
Is provided to prevent noise from being sent to the noise detecting microphone 2 by the noise-suppressed sound wave from the speaker 11 being supplied to the digital filter 8. Therefore, from the digital output y to the output terminal of the digital noise signal NS, It has a delay transfer function from the D / A converter 9, the amplifier 10 and the speaker 11 to the noise detection microphone 2, the amplifier 4 and the A / D converter 5, and the digital output is the subtractor described above. 6 is supplied to the adaptive noise source 6 and subtracted from the digital noise signal NS to be supplied to the adaptive filter 7 as a net noise signal x, thereby preventing howling from occurring.

The coefficient calculator 14 is provided with the mute error detection microphone 3
The analog output of the A / D converter 1
The mute error signal e is obtained by converting into a digital value in step 3, and this and the digital filter 15 (FIR filter-Finit
It has a function of receiving the output x ′ of the e Impulse Response filter) and calculating the coefficient of the adaptive filter 7. In addition, digital filter 1
5 is to equalize the silence error signals e and x in terms of time,
It has a transfer characteristic from the output terminal of the digital output y to the detection point S of the muffling error detection microphone 3. The coefficient calculator 14 repeats the calculation until the muffling error signal e becomes substantially zero, and updates the filter coefficient of the adaptive filter 7. As a result, the sound wave at the point s becomes zero, that is, the speaker 11 Although the sound wave from the sound source and the sound wave from the noise source have the same amplitude, the sound wave is such that the density is reversed to prevent the noise from leaking to the outside of the noise.

The coefficient calculator 14 is of a so-called NLMS type or LMS type. First, the NLMS type will be described with reference to FIG. In FIG. 7, x ′ is the output of the digital filter 15 in FIG. 6, which is sequentially supplied to the _N−1 delay devices 24 ₁ , 24 ₂ , ..., 24 _N−1 . That is, if the sampling period is T, the output of the delay device 24 ₁ is the noise signal NS before T from the present,
The output of the second delay device 24 ₂ is the noise signal N before the time 2T.
It is S. Thereafter, similarly, the output of the delay unit 24 _N-1 outputs the digital output x ′ before (N−1) × T time. That is, the delay devices 24 ₁ , 24 _2, ... 24 _N-1
Acts like a shift register,
First, the output x ′ of the digital filter 15 is directly supplied to the multiplier 25 ₀ , and hereinafter, the digital outputs x ′ before the times T, 2T, ... Are multiplied by the multipliers 25 ₁ , 25 ₂ ,.
Supplied to _N-1 .

On the other hand, the mute error signal e is used as a single multiplier 23.
Which receives a digital output Δn described below,
These multiplication results are used as the second group of multipliers 22 ₀ , 22
₁ , 22 ₂ , ... 22 _N-1 is supplied. That is,
These multipliers 22 ₀ , 22 ₁ , 22 ₂ , ... 22 _N-1
Therefore, e × Δn and now, T time ago, 2T time ago, ...
·· (N-1) T before time noise output x 'and is multiplied the multiplication result subtracter 21 ₀ that is directly connected to the adaptive filter 7, 21 _1, 21 _2, .... 21
Is supplied to the _N-1, each filter coefficient h ₀ of the adaptive filter 7 at the present time, h _1, h _2, from ··· h _N-1 of these is subtracted, the subtraction result h ₀ new new, h ₁ new new, h ₂ ne
w, ... h _N-1 new are set as the next filter coefficients of the adaptive filter 7. The output x ′ of the digital filter 15 is, on the other hand, the first group of multipliers 25 ₀ , 25
₁ , 25 ₂ , ... 25 _N-1 , T, 2T, 3 respectively
The digital outputs of T, ... _N-1 T time ago are supplied, and the square calculation of these is performed to adders 26 ₀ , 26
_1, 26 _2, is supplied to the · · · 26 _N-1. The addition results of these adders are sequentially added to the adders 26 ₁ , 26 on the downstream side.
₂ , ... 26 _N-1 are added as shown by the edge, and the total output Px is supplied to the divider 27. The predetermined value β is supplied to the most upstream adder 26 ₀ and is included in the output Px of the final-stage adder 26 _N−1 . This is the predetermined value α in the divider 27. When dividing by Px, if each multiplier 25 ₀ , 25 ₁ , 25
₂ , ... 25 _{N- 1} If the sum of the outputs is zero, the divider α
/ Px becomes infinite. Β is given to prevent this. The division result of the divider 27 is Δn. As described above, in the single multiplier 23, the mute error signal e
Is multiplied by.

[0007] 'while indicating, the coefficient calculator 14' 8 coefficient calculator 14 of the LMS type in, as apparent in comparison with FIG. 7, the multipliers 25 _0, 25 _1, 25
₂ , ... 25 _N-1 and adders 26 ₀ , 26 ₁ , 26
₂ , ... 26 _N-1 and the divider 27 are omitted.
Therefore, the predetermined value μ is set instead of Δn as the calculation result. This is a so-called step size, and the mute error signal e is multiplied by this μ, and the result is multiplied by the multipliers 22 ₀ , 22 ₁ , 22 ₂ , ...
_It is supplied to _N-1 , and the multiplication as described above is performed to update the coefficient of each filter of the adaptive filter 7.

The conventional coefficient calculator NLMS type and LMS type are configured and operate as described above, but each has its advantages and disadvantages. Next, this will be described.

When the output x'of the digital filter 15 is a constant value, Δn calculated by the NLMS type coefficient calculator 14
Is the reciprocal of the sum of squares as described above, and is an amount inversely proportional to the square of x ′. Further, since the muffling error signal e is an amount substantially proportional to the output signal x ′ of the digital filter 15,
By the calculation of e × Δn, this result becomes an amount inversely proportional to the power of x ′. These are also the second group 22 ₀ ,
22 ₁ , ... 22 _{N− 1} is multiplied by x ′, so that all components proportional to the update amount x ′ are deleted.
This means that the update amount gain is determined only by the step size α. Therefore, the digital output x '
The convergence speed is constant irrespective of the size of, and stability is not impaired if α is determined so as to converge stably. However, as is clear from FIG. 7, the NLMS type requires at least 2N multipliers, which is determined by the number of taps of the adaptive filter 7.
This tap number usually requires several hundreds, which is a disadvantage that a very long calculation time is required.

Next, in the LMS type shown in FIG.
Δn of the coefficient calculator 14 in the form is calculated using the estimated value of the digital value x ′ to obtain the step size μ. Therefore, as in the coefficient of NLMS type calculator 14, that is the square of the variation of the 'because ^two components is not canceled and [Delta] n x' x included in the update amount, it affects as it becomes variation of the step size There are drawbacks.

If the digital value x'is approximately equal to the estimated value, no problem will occur, but if it is small, the convergence speed will deteriorate and
When it becomes large, it does not converge and diverges and becomes unstable.
However, the LMS type has a great advantage that the number of multipliers can be reduced to about 1/2 as compared with the NLMS type, and the step size μ is often set to a considerably small value with a stable margin.

On the other hand, in the case of giving this small step size μ and causing the fixed-point type microprocessor to calculate, if the digital value x'is small, the silencing effect is limited due to the cancellation of digits in the calculation, and it is sufficient. There is also a drawback that it does not have such a silencing effect. In addition, since the convergence speed is slow, it is not possible to mute promptly against external disturbance such as noise, and also large disturbance causes divergence and becomes unstable. Therefore, it is difficult to apply to a noise source whose digital value x ′ fluctuates greatly.

[0013]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and enables high-speed operation by shortening the long operation time, which is a drawback of the NLMS type, and reduces convergence speed and stability. It is an object of the present invention to provide a silencer that does not depend on an input signal.

[0014]

[Means for Solving the Problems] The above-mentioned objects are: noise detection means, silencing error detection means, sound wave generation means, and first A / D for converting the analog output of the noise detection means into a digital value. A converter, an adaptive filter that receives the first digital output of the first A / D converter, a coefficient calculator that calculates the filter coefficient of each tap of the adaptive filter, and a digital output of the adaptive filter to an analog value A D / A converter for converting the analog output of the muffling error detecting means into a digital value, a first digital output of the first A / D converter, and The first digital output is the second
A second digital output of the A / D converter and a digital filter having a transfer function for delaying the second digital output so as to be equivalent in time, the second digital output of the second A / D converter and the second digital output of the digital filter. And 3 digital outputs are supplied to the coefficient calculator, and the coefficient calculator repeatedly performs coefficient calculation so that the second digital output of the second A / D converter converges to zero, and the calculation result is the adaptive result. It is supplied to the filter to update the filter coefficient of each tap, and the D /
In the silencer adapted to supply the analog output of the A converter to the sound wave generating means, a maximum value detecting means for detecting the maximum value or the maximum absolute value of the third digital output of the digital filter, and a squared value. An arithmetic unit and a divider for dividing a predetermined value by the detected value of the maximum value detecting means, and a correction coefficient to be multiplied by the second digital output of the second A / D converter is calculated by these. It is achieved by a silencer characterized by having done.

[0015]

The maximum value of the output of the digital filter is selected by the maximum value detecting means, and the correction coefficient of the adaptive filter is calculated by squaring this value and dividing the predetermined value by this value. , The number of multipliers is L
Since it is the same as the MS type and the number of multipliers is not as large as that of the NLMS type, the calculation speed can be increased. On the other hand, the maximum amount of noise such as disturbance is taken, so that a sufficient silencing effect can be obtained without being affected by the noise.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A silencer according to an embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a circuit diagram of a silencer according to an embodiment of the present invention. The parts corresponding to those of the conventional example are designated by the same reference numerals, and the detailed description thereof is omitted.

That is, according to the present embodiment, the output x'of the digital filter 15 has the delay devices 24 ₁ , 24 ₂ , ...
It is the same as being sequentially shifted by 24 _N−1 , which is, on the other hand, supplied to the absolute value calculator 42, and its absolute value is taken and supplied to the maximum value comparator 43. This comparator has two input terminals and outputs the larger output x _p of these inputs. This output terminal is supplied to one input of the maximum value comparator 43 via the ζ multiplication means 44 and the delay device 45. ζ is 0 <ζ <1. Further, this output terminal is supplied to the multiplier 46, and the output x _p of the maximum value comparator 43 is squared and supplied to the adder 47. A predetermined value β is supplied to one of the input terminals, and the addition result Px
_p is supplied to the divider 48. Here, the predetermined value C is set to Px
An operation of dividing by _p is performed, and the operation result is [Δn].

[Δn] is supplied to the multiplier 49, which is also supplied with the muffling error signal e, calculates e × [Δn], and calculates the multipliers 22 ₀ , 22 ₁ , 22 ₂ , ... 22 _N-1
, And the output x ′ of the digital filter 15 is multiplied by the current value, T time before, 2T time before, ... (N−1) T time before, and subtractors 21 ₀ , 21 ₁ , 21 ₂ ,
... 21 _N-1 and output h ₀ new that determines the next filter coefficient by subtracting this value from the current filter coefficient outputs h ₀ , h ₁ , h ₂ , ... h _N-1 , H ₁ new, h ₂ n
ew, ... h _N-1 new is supplied to the adaptive filter 7.

The silencer of the present invention is constructed as described above, and its operation will be described below.

The absolute value calculator 42 takes the absolute value of the output x'and supplies it to the maximum value comparator 43. The output of the maximum value comparator 43 (initial value) at time t = 0
Is usually 0, and the maximum value is gradually taken. Ζ is a damping coefficient between 0 and 1, but the absolute value of the output x'is decreasing. In some cases, this can be followed. The output of the divider 48 is [Δ
n], which is multiplied by the muffling error signal e by the multiplier 49 and multiplied by the multipliers 22 ₀ , 22 ₁ , 22 ₂ , ... 22 _N-1.
, And the filter coefficient of the adaptive filter 7 is updated. Other functions are the same as those of the conventional NLMS.
The operation is the same as that of the coefficient calculator of the form.

When the digital value x ′ is a constant value, the square component of x ′ is calculated by the square calculation of the maximum value x _p , and NM
Similar to the S-type coefficient calculator, [Δn] is an amount inversely proportional to the square of the digital value x ′. Therefore, the component of the update amount proportional to x'is canceled. Conventional NL
In the MS type, N number of multipliers were required to calculate Δn, whereas in the present invention, in addition to absolute value calculation and maximum value comparison, a small amount of calculation is required compared to the conventional case where only two multipliers are used. Only need a number. Therefore, the calculation time can be greatly shortened.

Further, the calculation time is almost the same as that of the LMS type, while having the excellent feature of the NLMS type that the convergence speed and stability do not depend on the input signal. Further, even when the calculation is performed by a fixed-point type microprocessor, [Δn] becomes large when x ′ is small, and therefore the precision of the arithmetic operation is unlikely to occur, and a sufficient noise reduction effect can be obtained even for a minute input. .

Further, while the amount of calculation for calculating the correction coefficient of the filter of the adaptive filter 7 has conventionally increased in proportion to the number of taps, according to the configuration of the embodiment of the present invention, it depends on the number of taps. No, it is always a small amount,
It is clear that the greater the number of taps, the greater the shortening effect.

Here, the Δn value of the conventional NLMS type will be compared with [Δn] obtained by the configuration of the embodiment of the present invention.
The results are shown in FIG. 2, where the swing level is scaled with a maximum value of 1.0 and expressed as a dimensionless amount. The vertical axis shows the amplitude of the correction coefficient.
The calculation conditions are α = 0.546 × 10 ⁻³ and C = 1.2.
2 × 10 ⁻⁴ , ζ = 0.94, and the noise emitted from the noise source is a single sine wave of 100 Hz given by three types of amplitude. In FIG. 2, dotted lines 1a, 2
a and 3a are outputs of Δn according to the conventional configuration, 1a is a sine wave amplitude of 7.6 × 10 ⁻³ , 2a is an amplitude of 1.5 × 10 ⁻² , and 3a is Swing width 3.1 x 1
This is the case with 0 ^-2 .

The thick solid lines 1b, 2b and 3b are outputs of [Δn] according to the embodiment of the present invention, and the conditions are 1a and 2 of the conventional example.
a and 3a are the same.

As is apparent from the graph of FIG. 2, it can be seen that [Δn] approximately equal to Δn is obtained at any swing level.

3 shows the case where random noise is used as the digital value x ', the dotted line 4a is the output of Δn according to the conventional configuration, and the solid line 4b is the output of [Δn] according to the configuration of the embodiment of the present invention. is there. As is clear from this, it is understood that the two are almost equal even in random noise.

As shown in FIGS. 2 and 3, Δn and [Δ
[n] has a certain error, but this system has the property of obtaining an optimum value by repeating many convergence calculations, and does not calculate it once. Therefore, Δn
Does not need to be exactly equal to [Δn], and it is clear that even with an error of the degree shown in FIG. 2 and FIG.

Although the embodiment of the present invention has been described above, the present invention is not limited to this, and various modifications can be made based on the technical idea of the present invention.

For example, FIG. 4 shows the first modification, but parts corresponding to those in the above embodiment are designated by the same reference numerals, and detailed description thereof will be omitted.

That is, according to this modification, the output of the absolute value calculator 42 is supplied to the adder 52, the predetermined value β is supplied to the other input terminal, and the addition result is the maximum value comparator 43. Is supplied to. Further, this output x _p is delayed by the delay device 51 via the ζ multiplication means 44, and the maximum value comparator 43
This delay device delays the time by 2T of 2 cycles. this is,
This configuration is suitable when more gradual damping is required, and the delay period may be further increased. That is, 3T, 4T
Etc. may be any delay sample. Further, the output x _p of the maximum value comparator 43 is supplied to the divider 53, where the predetermined value C is divided by this, and the result is the multiplier 54.
Is supplied to the output terminal and is squared to output [Δn]. Even in such a modification, it is clear that the same operation and effect as those of the above-described embodiment can be obtained.

FIG. 5 shows a second modification. Similarly, parts corresponding to the above-mentioned embodiment are designated by the same reference numerals, and detailed description thereof will be omitted, but the output x'of the digital filter is omitted. Is supplied to the multiplier 61 to be squared, and is supplied to the maximum value comparator 43 to be compared with the signal delayed by one cycle after being reduced by ζ times as in the above-described embodiment, and the maximum value x _p is output, This is supplied to the adder 47, and the predetermined value β is supplied to the other input terminal. This addition result Px
_p is used in the divider 48 to divide large, resulting in [Δn]. It is obvious that this modification also has the same operation and effect as those of the above embodiment.

Further, various modifications are conceivable in which the order of division, the addition position of β, and the like are exchanged in accordance with basic arithmetic rules such as the exchange rule. Further, the predetermined value β is for the purpose of preventing division by zero, and there are various addition positions that can achieve this purpose, but the same purpose can be obtained by adding the maximum value x _p before squaring in addition to the modification. Is achieved. Of course, this calculation becomes unnecessary when x _p = 0 does not hold.

[0035]

As described above, according to the muffler of the present invention, it is possible to shorten the calculation time and prevent it from being affected by disturbance or noise.

[Brief description of drawings]

FIG. 1 is a wiring diagram of a silencer according to an embodiment of the present invention.

FIG. 2 is a graph showing a comparison between [Δn] in the example and Δn in the conventional example.

FIG. 3 is a comparative graph diagram similar to FIG. 2 when a noise source generates random noise.

FIG. 4 is a wiring diagram showing a main part of a first modified example.

FIG. 5 is a wiring diagram showing a main part of a second modified example.

FIG. 6 is a wiring diagram of a conventional active noise control system.

FIG. 7 is a wiring diagram showing an NLMS type of a coefficient calculator in the conventional example.

FIG. 8 is a wiring diagram showing the same LMS type.

[Explanation of symbols]

15 Digital Filter 22 ₀ Multiplier 22 ₁ Multiplier 22 ₂ Multiplier 22 ₂ Multiplier 22 _N-1 Multiplier 42 Absolute Value Calculator 43 Maximum Value Comparator 46 Squarer 47 Adder 48 Divider 49 Divider

Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location H03H 21/00 8842-5E

Claims (1)

[Claims] 1. A noise detecting means and a muffling error detecting means,
Sound wave generating means, a first A / D converter for converting an analog output of the noise detecting means into a digital value, and the first A / D converter
An adaptive filter that receives the first digital output of the A / D converter, a coefficient calculator that calculates the filter coefficient of each tap of the adaptive filter, and a D for converting the digital output of the adaptive filter into an analog value.
/ A converter, a second A / D converter for converting the analog output of the muffling error detection means into a digital value, and a first digital output of the first A / D converter, and the first digital output And a digital filter having a transfer function for delaying so as to be temporally equivalent to the second digital output of the second A / D converter, the second digital output of the second A / D converter and the digital filter. The third digital output of the filter is supplied to the coefficient calculator, and the coefficient calculator repeatedly performs coefficient calculation so that the second digital output of the second A / D converter converges to zero, and the calculation result is obtained. Are supplied to the adaptive filter to update the filter coefficient of each tap, and the analog output of the D / A converter is supplied to the sound wave generating means. From the maximum value detecting means for detecting the maximum value of the third digital output or the maximum value of the absolute value, a square calculator, and a divider for dividing a predetermined value by the detected value of the maximum value detecting means. And the second A / D
A silencer characterized in that a correction coefficient to be multiplied by the second digital output of the converter is calculated.