JPH05108082A - Electronic silencing system - Google Patents

Abstract

(57) [Abstract] [Purpose] To improve the calculation accuracy when measuring the transfer function between the speaker and microphone, and enable higher-performance silencing. [Structure] In an electronic muffling system that recursively updates the filter coefficient of an adaptive digital filter by the so-called Filtered-x LSM algorithm, when setting the filter coefficient of the digital filter that generates the reference signal before the muffling operation starts, Between the noise signal generation circuit 41 and the identification digital filter 51, a digital filter 81 which passes only a frequency similar to the frequency reproducible by the speaker 13 is arranged, whereby a highly accurate transfer function (that is, filter) is provided. Coefficient). The transfer function measured in this way is added as a filter coefficient of a digital filter that generates a reference signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic silencing system, and more particularly to an electronic silencing system which generates acoustic waves of opposite phase and the same sound pressure with respect to acoustic waves propagating from a noise source and actively muffs sound waves. Regarding muffling system.

[0002]

2. Description of the Related Art An electronic noise suppression system called active noise control is used in which noise propagating in a ventilation path is silenced by adding sound waves of opposite phase and same amplitude. FIG. 4 is a block diagram showing an example of a conventional electronic silencing system. The electronic silencing system includes a sensor microphone 11 that detects noise information in the air duct 10, a speaker 13 that cancels sound waves propagated in the air duct 10, and an error microphone 12 that detects an error after canceling the sound waves propagated. The controller 14 is configured to generate a drive signal to be given to the speaker 13.

The controller 14 includes an A / D converter 61,
61, a D / A converter 62, an adder 30, digital filters 31, 32 and 33, and an arithmetic circuit 91. A noise signal indicating the noise detected by the sensor microphone 11 is added to the adder 30 via the A / D converter 61. The output signal from the digital filter 31 is added to the other input of the adder 30.
0 adds these input signals, and the added signal x
(n) is output to the digital filters 32 and 33.
The digital filter 31 is for canceling the acoustic feedback given to the sensor microphone 11 by the sound wave from the speaker 13.

By the way, the digital filter 33 can be realized by an FIR filter having variable tap weights (filter coefficients) and an adaptive algorithm for controlling the FIR filter. The adaptive algorithm uses the above input x (n) and the error microphone 12. From the information of the error signal e (n) input through the A / D converter 61, the filter coefficient of the adaptive digital filter is set so that the energy of the error signal e (n) is minimized under some evaluation criteria. Adjust.

Next, a method of setting the filter coefficient of the digital filter 33 to the optimum value will be described. The output y (n) of the digital filter 33 is given by the convolution of the input x (n) and the filter coefficient w _i . The error output e (n) can be expressed as Can be expressed as In the equation (2), r (n) is a reference signal filtered by the digital filter 32 showing the transmission characteristic from the speaker 13 to the error sensor 12, Is.

[0006] The following vector representation for the sake of simplicity, R = [r (n) , r (n-1), ... r (n-I + 1)] T W = [w O, w 1, ... w I- ₁ ] ^T , the above equation (2) can be expressed by the following equation: e (n) = d (n) + ^RT · W (4)

Here, the mean square error (MSE: mean-squ)
are error) E [e (n) ² ], J = E [e (n) ² ] = E [d (n) ² ] + 2W ^T E [R ^T d (n) ] + W ^T E [R ^T R] W (5), and MSE becomes a quadratic function of the filter coefficient. The second derivative is the first, and if the derivative is set to 0, the minimum value J
A solution with _min is obtained.

Now, it is a steepest descent algorithm.
In the Filtered-xLSM algorithm, the instantaneous squared error e (n) ² itself is used as the estimator of MSE J, and the estimator ∇ _n of the derivative (gradient ∇) of J is given by Then, the filter coefficient of the adaptive digital filter is recursively updated by the following equation using the above ∇ _n . _{W n + 1 = W n +} μ (-∇ n) = W n -2μR n T e (n) ... (7) where the mu controls the magnitude of the correction amount in each of the repeated positive scalar parameters Is. The above equation (7) means that the filter coefficient is sequentially updated in the direction opposite to the gradient vector (∇ _n ) (in the direction of the steepest descent of the error surface). It reaches _min , and the filter coefficient has an optimum value.

In FIG. 4, the arithmetic circuit 91 has a reference signal r (n) added from the digital filter 32.
And the error signal e (n) applied from the error microphone 12 through the A / D converter 61, the filter coefficient W (z) is calculated based on the above equation (7), and the filter coefficient W (z) is calculated.
At (z), the filter coefficient W (z) of the digital filter 33 is updated.

The digital filter 33 receives the input signal x (n)
And a given filter coefficient W (z) are convoluted to output a driving signal y (n) of the speaker 13, and the driving signal y (n) is sent to the speaker 13 via the D / A converter 62. Added. By the way, in this electronic muffling system, because of control by feedforward, the transfer function (acoustic feedback) from the speaker 13 to the sensor microphone 11 and the transfer function from the speaker 13 to the error microphone 12, that is, the digital filters 31 and 32. The filter coefficient must be set before the start of muffling work.

Next, a method of setting these two filter coefficients will be described with reference to FIG. The noise signal generation circuit 41 is
For example, an M-sequence doubt random signal is generated as a noise signal. This noise signal is converted into an analog signal by the D / A converter 61, and then radiated into the air duct 10 through the speaker 13, and its sound wave is detected by the error microphone 12 and A / A is detected.
It is input to the adder 71 via the D converter 62.

At the same time, the noise signal is also input to the digital filter 51, and the output thereof is the adder 71.
Other inputs have been added. The adder 71 subtracts both signals and feeds back the calculation result, that is, the error, to the digital filter 51. When this operation is repeatedly executed and adaptive control is performed so as to reduce the output error of the adder 71, the transfer function of the system from the speaker 13 to the error microphone 12 is replaced as the filter coefficient of the digital filter 51. Then, the filter coefficient thus identified is the digital filter 3 of FIG.
It is set as a filter coefficient of 2.

Similarly, the filter coefficient indicating the transfer function from the speaker 13 to the sensor microphone 11 is identified,
It is set as a filter coefficient of the digital filter 31 of FIG.

[0014]

However, the noise signal generated from the noise signal generating circuit 41 is DC to
Although the signal includes a frequency up to 20 kHz, the speaker 13 actually emits a sound wave of 50 Hz or less due to a mechanical limit. That is, while the noise signal of the entire frequency band is input to the digital filter 51,
A phenomenon occurs in which the error microphone 12 cannot detect an acoustic signal of 50 Hz or less.

Therefore, the transfer function in the frequency band below 50 Hz cannot be identified, and the error subtracted by the adder 71 is not reduced. Due to this error, the measurement accuracy of the transfer function in the entire frequency band deteriorates, and there is a drawback in that the sound volume is adversely affected. The present invention has been made in view of such circumstances, and an object thereof is to improve the calculation accuracy when measuring the transfer function between the speaker and the microphone, and to provide a higher-performance electronic silencing system.

[0016]

In order to achieve the above object, the present invention provides a noise information detecting means for detecting a sound wave propagating from a noise source, a speaker for emitting a sound wave for canceling the sound wave propagating, An error microphone for detecting a sound wave propagating from a speaker and the noise source; and an output signal of the noise information detecting means, and a driving signal to be given to the speaker by convolution with a given first filter coefficient. A second digital filter which takes in the output signal of the noise information detecting means and performs a convolution operation with a second filter coefficient indicating the transmission characteristic from the speaker to the error microphone; The reference signal from the second digital filter and the error signal from the error microphone Based on the calculation means, the filter coefficient given to the first digital filter is updated so that the input signal from the error microphone is minimized, and a noise signal is generated before the silencing operation is started, and the speaker is generated based on the noise signal. Noise signal generating means for driving, a third digital filter for taking in the noise signal and performing a convolution operation with a given third filter coefficient, and a noise from a speaker driven based on the noise signal. The third filter coefficient is identified so that the error signal from the error microphone to be detected matches the output signal of the third digital filter, and the identified third filter coefficient is used as the second digital filter. And a means for giving it as the second filter coefficient of Between the digital filter, the speaker is characterized in that it is provided a filter for passing only the same frequency as renewable.

[0017]

The present invention confirms by experiment that the frequency characteristic of the noise signal used for measuring the transfer function between the speaker and the microphone before the start of the muffling operation affects the mute volume, and the speaker can be reproduced. Focusing on the frequencies and frequencies that can be muted, a filter that passes only frequencies similar to those reproducible by the speaker is arranged between the noise signal generating means and the identification digital filter, thereby achieving high accuracy. It enables the measurement of transfer functions.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of an electronic silencing system according to the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 is a block diagram showing an embodiment of a transfer function measuring system in an electronic sound deadening system according to the present invention, and particularly shows a block diagram for measuring a filter coefficient given to a digital filter 32 shown in FIG. .. The same parts as those in FIG. 4 are designated by the same reference numerals, and detailed description thereof will be omitted.

As shown in FIG. 1, the noise signal generating circuit 41
Digital filter 8 that cuts low frequencies in the latter stage
1 is added. Here, the digital filter 81
The frequency characteristic of is determined so as to cut off frequencies of 50 Hz or less, judging from the reproducible frequency of the speaker 13. Therefore, the frequency band of 50 Hz or less that is not measured by the error microphone 12 is not input to the digital filter 51, and the filter coefficient indicating the transfer function from the speaker 13 to the error microphone 12 including the conversion characteristics of the speaker 13 and the error microphone 12 is not input. Can be accurately measured.

FIG. 2 is a graph showing the silencing effect of the electronic silencing system according to the present invention. The solid line shows the silencing volume according to the present invention when the filter coefficient is measured by providing the digital filter 81 for cutting the low frequency band. In the figure, the broken line shows the conventional volume of sound when the filter coefficient is measured without providing the digital filter 81 that cuts the low frequency.

As shown in the figure, it was confirmed by experiments that the noise reduction level is improved especially for noise near 125 Hz. FIG. 3 is a block diagram showing another embodiment of the transfer function measuring system in the electronic silencing system according to the present invention.
The same parts as those in FIG. 1 are designated by the same reference numerals,
Detailed description thereof will be omitted.

As shown in FIG. 3, the speaker 13 inputs a noise signal from the noise signal generating circuit 41 via the D / A converter 61 without passing through the digital filter 81.
The effect in this case is similar to that of the above-described embodiment.
In this embodiment, the case where the filter coefficient indicating the transfer function from the speaker 13 to the error microphone 12 is accurately measured has been described, but the digital filter 31 as shown in FIG. 4 is used to cancel the acoustic feedback. In the electronic sound deadening system having the configuration, the filter coefficient of the digital filter 31 can be set in the same manner as above. Further, in the present embodiment, the electronic silencing system for silencing noise propagating in the air duct has been described.
Not limited to this, it can be applied to an electronic silencing system for silencing noise propagating in a three-dimensional space.Furthermore, noise information detection means is not limited to a sensor microphone, and noise is generated due to mechanical vibration. If so, a vibration pickup sensor may be used.

[0023]

As described above, according to the electronic muffling system of the present invention, the noise signal used when measuring the transfer function between the speaker and the microphone passes only the frequency similar to the frequency reproducible by the speaker. Since the output is performed through the filter, the measurement accuracy of the transfer function is improved, and higher-performance silencing control can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a transfer function measuring system in an electronic silencing system according to the present invention.

FIG. 2 is a graph showing the silencing effect of the electronic silencing system according to the present invention.

FIG. 3 is a block diagram showing another embodiment of the transfer function measuring system in the electronic silencing system according to the present invention.

FIG. 4 is a block diagram showing an example of an electronic silencing system during silencing.

FIG. 5 is a block diagram showing a transfer function measurement system in a conventional electronic silencing system.

[Explanation of symbols]

 11 ... Sensor microphone 12 ... Error microphone 13 ... Speaker 14 ... Controller 30, 71 ... Adder 31, 32, 33, 51, 81 ... Digital filter 41 ... Noise signal generation circuit 91 ... Arithmetic circuit

Claims (1)

[Claims] 1. A noise information detecting means for detecting a sound wave propagating from a noise source, a speaker for radiating a sound wave for canceling the sound wave, and an error microphone for detecting a sound wave propagating from the speaker and the noise source. A first digital filter that takes in an output signal of the noise information detecting means and creates a drive signal to be given to the speaker by a convolution operation with a given first filter coefficient; and an output signal of the noise information detecting means. From a second digital filter that takes in and convolves a second filter coefficient indicating a transmission characteristic from the speaker to the error microphone to create a reference signal, and a reference signal from the second digital filter and the error microphone Based on the error signal and the error input signal from the microphone is minimized A calculation means for updating the filter coefficient given to the first digital filter, a noise signal generation means for generating a noise signal before the start of the muffling operation, and driving the speaker based on the noise signal; A third digital filter that performs a convolution operation with the third filter coefficient that is taken in and given, an error signal from the error microphone that detects noise from a speaker driven based on the noise signal, and the third digital filter. Means for identifying the third filter coefficient so that the output signal of the digital filter of the second digital filter coincides with the output signal of the second digital filter, and providing the identified third filter coefficient as the second filter coefficient of the second digital filter. In the electronic silencing system, the speaker reproduces between the noise signal generating means and the third digital filter. An electronic silencing system, characterized in that a filter is arranged to pass only frequencies similar to possible frequencies.