 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a system for attenuating noise
electronically and, in particular, to an electronic noise attenuation
system which is capable of attenuating non-steady noise occurring in
propagation passages such as duct lines or the like by exercising an
adaptive control using a computer system including a digital filter
therein.
2. Description of the Prior Art
Conventionally, there has been widely put into practical use a passive
noise attenuation apparatus which attenuates noise occurring within ducts
by use of the interference due to the duct structure or the noise
absorption due to a porous material attached to the duct. However, this
type of noise attenuation apparatus is found disadvantageous in that it is
too big in size, it involves too much loss of pressure, and so on.
On the other hand, there is also available an active noise attenuation
apparatus which has been long proposed and employs another method of the
reduction of unwanted sounds within the duct. That is, recently, special
interest has been given to an electronic noise attenuation system of such
active type in which noise propagated from a source of noise is sensed, a
cancellation sound having the same sound pressure and an opposite phase
with respect to the sensed noise is generated against the noise to thereby
provide sound wave interference between the noise and the cancellation
sound, and thus the noise can be cancelled forcibly by the sound wave
interference. With the rapid progress of an electronic device, signal
processing technique and the like, there have been recently published
various kinds of study results on such active electronic noise attenuation
method and apparatus.
However, there are still left many problems to be solved and thus such
electronic noise attenuation method or apparatus has not yet come into a
stage of seriously practical application.
A technical problem in putting into practice such electronic noise
attenuation system consists in the construction of a model which can be
used as a basis for design of a control system of the electronic noise
attenuation system. The model must be able to cope with the following
points. At first, there is necessary a filter which is capable of
cancelling noise of continuous spectra. That is, if a cancellation sound
can be generated with respect to the noise of continuous spectra such as
automotive noise, air current noise and the like as well as the noise of
discrete spectra such as transformer noise, compressor noise and the like,
the applications of the electronic noise attenuation system can then be
expanded further. To realize this, a filter is required which is able to
provide arbitrary amplitude characteristics and phase characteristics.
Secondly, it is necessary to prevent the feedback of the cancellation sound
with respect to a sensing microphone. That is, in the electronic noise
attenuation system, there is interposed the sensing microphone between a
source of noise and a source of cancellation sounds within a propagation
passage through which sound waves are propagated, and it is necessary to
create an electric signal to drive the cancellation sound source which
generates sound waves to cancel the propagated sound waves from the noise
source, in accordance with the sounds sensed by the sensing microphone and
by some proper signal generation means. In this case, the sound waves
generated from the cancellation sound source is also caught by the sensing
microphone and, as result of this, there is produced an acoustic feedback
system between the cancellation sound source and the sensing microphone.
For this reason, it is essential to take a countermeasure to cope with
this situation. Especially in order to make compact the electronic noise
attenuation system and to allow it to be mounted at an arbitrary position
in a pipe line such as a duct line, the sensing microphone and the
cancellation sound source must be located adjacent to each other.
Therefore, the above-mentioned acoustic feedback has a great influence on
the electronic noise attenuation system and thus the countermeasure to
cope with this problem is very important.
Thirdly, it is necessary to make it possible to correct the characteristics
of electro-acoustic transducers such as a microphone, speaker and the like
used in the electronic noise attenuation system. That is, in order to
stabilize the control function of the electronic noise attenuation system,
it is essential that the control system of the electronic noise
attenuation system is provided with a function to correct the minute
amount of deterioration of the characteristics of the electro-acoustic
transducers. This is another problem to be solved.
In view of this, we have already found and proposed models for an
electronic noise attenuation system which can cope with the
above-mentioned problems (Japanese Patent Application No.60-139293,
No.60-139294, No.61-7115, No.62-148254.)
According to the electronic noise attenuation system that we have proposed,
the above-mentioned third problem can be solved properly: that is, by
properly controlling the characteristics of a digital filter for creating
an electric signal to be given to a cancellation sound source, the system
can cope with the variations of the propagation characteristics of a sound
wave propagation passage (e.g., a duct) as well as the variations of the
characteristics of a control system (which includes a speaker as a
cancellation sound source, a microphone as a sensor and the like).
Referring now to FIG. 1, there is shown a basic structure of a monopole
sound source type of adaptive electronic noise attenuation system
including two sensor microphones M1, M2.
In this structure, the output of the sensor microphone M2, which is located
on the down stream side of the figure, is as an error signal. The basic
operation of the structure is to update the transfer function of a digital
filter 2 from the input X of the digital filter 2 and the output E of the
sensor microphone M2 so that the energy of the output E can be a minimum
value under some evaluation standard or other.
Now, if an actual electronic noise attenuation system is modeled according
to FIG. 1, then a model shown in FIG. 2 can be obtained. The model shown
in FIG. 2 is constructed on the assumption that a sound wave to be fed
back from a cancellation sound speaker (an additional sound source) S to
the sensor microphone M1 is cancelled electrically at a point of addition
20 and thus it is not input to the digital filter 2.
What is important here is the existence of a transfer function D with a
time delay representing the transfer characteristics of speaker, duct and
the like from the output of the digital filter 2 to the addition point of
the error signal.
By the way, in order to be able to apply a well-known adaptive control
algorithm such as VS-LMS (Variable Step-Least Means Square) or the like,
not only the input X of an adaptive digital filter must be defined clearly
but also it is necessary to clarify the connection of the output Y of the
digital filter with an error signal E. In the case of a system in which
after the output of the digital filter 2 is determined the error signal E
can be observed in an instant or a system in which the error signal E has
already been decided at latest by the time of updating of the next
coefficient of the digital filter, basically there arises no problem and
thus the well-known algorithm can be applied. An echo canceller filter is
a good example to deal with an acoustic signal and in this filter the
output Y of the filter is reflected, as it is, in the error signal E. In
contrast to this, in the electronic noise attenuation system shown in FIG.
1, the film output is not connected, as it is, with the error signal E but
the error signal E can be obtained only by means of the electro-acoustic
conversion characteristics of speaker, transfer characteristics from
speaker to microphone, process of super-position (interference) of
acoustic signals in space, and the acoustic-electric conversion
characteristics of microphone. That is, if the above-mentioned transfer
function D is not taken into consideration, then a sound cancellation
effect cannot be obtained at all.
Further, in our previous application for patent (Japanese Patent
Application No. 62-148254), as shown in FIG. 8, the restriction of an
acoustic feedback is effective only when the transfer function from the
speaker S to the microphone M1 is practically equal to that from the
speaker S to the microphone M2. Most of linear duct equipment can satisfy
this requirement.
However, when a sound cancelling device is constructed by mounting speaker
to the bent portion of a duct, the above-mentioned structure is not able
to perform its function to the full. For this reason, the present
invention is proposed. Since the restriction of the acoustic feedback is
performed by means of identification of the transfer function of a
feedback system, the invention can be applied to any duct whatever shape
it has. Also, the invention can apply even to an active sound cancellation
system in a three-dimensional sound field (outdoor or indoor).
SUMMARY OF THE INVENTION
The present invention aims at eliminating the drawbacks found in the
above-mentioned prior art systems.
Accordingly, it is an object of the invention to provide an electronic
noise attenuation system which is capable of performing an adaptive
control in consideration of the transfer function of a transmission system
from a sound source for cancellation to a microphone for evaluation and is
also capable of restriction of an acoustic feedback in an arbitrary duct
shape.
In order to achieve the above object, according to the invention, there is
provided an electronic noise attenuation system which achieves attenuation
of a sound wave propagated from a source of noise in a propagation passage
of a sound wave by generating another sound wave 180.degree. out of phase
and having the same sound pressure with the propagated sound wave to
produce sound wave interference between the two sound waves at a given
position in said propagation passage, said system comprising: first
mechano-electric transducer means disposed at a position closer to the
noise source from the above-mentioned given position in the propagation
passage to sense the propagated sound wave from the noise source and
convert it into an electric signal; electro-mechanical transducer means
interposed between the position of the first mechano-electric transducer
means and the given position in the propagation passage to generate a
sound wave for cancelling the propagated sound wave from the source of
noise at the given position; second mechano-electric transducer means
interposed between the position of the electro-mechanical transducer means
and the given position or disposed at the given position to sense the
propagated sound waves from the electro-mechanical transducer means as
well as from the source of noise and convert them into electric signals;
operation means for inputting therein the output signal of the first
mechano-electric transducer means and a drive signal to be given to the
electro-mechanical transducer means to find a difference between them;
drive signal generating means for inputting therein the output signal of
the operation means to generate on the basis of a given transfer function
a drive signal to be given to the electro-mechanical transducer means so
that the amount of sound cancellation of the electronic noise attenuation
system can be maximized; and, control means for determining a transfer
function to be given to the drive signal generating means, setting up in
the drive signal generating means a control parameter to specify the
transfer function, and correcting the control parameter according to the
variations of the propagation characteristics of the propagation passage
as well as to the variations of the characteristics of the control system
of the electronic noise attenuation system, characterized in that the
control means outputs a pseudo-signal to the electro-mechanical transducer
means to generate a sound wave in the sound wave propagation passage,
specifies in accordance with the output signal of the second
mechano-electric transducer means a transfer function with a time delay
representing the transfer characteristics of a transfer system including a
sound wave propagation passage ranging from the output terminal of the
drive signal generating means to the second mechano-electric transducer
means and an electric signal transmission path so that the output signal
of the second mechano-electric transducer means can be minimized, and
determines a transfer function to be given to the drive signal generating
means in accordance with a given adaptive algorithm in consideration of
the specified transfer function with a time delay.
In the electronic noise attenuation system according to the present
invention, a sound wave based on an artificial signal is generated in a
sound wave propagation passage from electro-mechanical transducer means
which serves as a source of an additional sound, and, for this sound wave,
a transfer function with a time delay representing the transfer
characteristics of a transfer system, which includes a sound wave
propagation passage ranging from the output terminal of drive signal
generating means to second mechano-electrical transducer means and an
electric sound transmission path, is specified by control means so that
the output signal (error signal) of the second mechano-electric transducer
means for evaluation of sound cancellation effects can be minimized.
In addition, the control means is able to determine a transfer function to
be given to the above-mentioned drive signal generating means in
accordance with a given adaptive algorithm in consideration of the
transfer function with a time delay specified in the above-mentioned
manner.
Thanks to the above-mentioned construction, an electronic noise attenuation
system can be realized which enjoys a high effect on noise cancellation.
BRIEF DESCRIPTION OF THE DRAWINGS
The exact nature of this invention, as well as other objects and advantages
thereof, will be readily apparent from consideration of the following
specification relating to the accompanying drawings, in which like
reference characters designate the same or similar parts throughout the
figures thereof and wherein:
FIG. 1 is a view to show on principle the basic structure of an electronic
noise attenuation system according to the present invention;
FIG. 2 is an explanatory view of a modeled version of the electronic noise
attenuation system shown in FIG. 1;
FIG. 3 is an explanatory view of an embodied model of the electronic noise
attenuation system including a controller in consideration of a transfer
function D with a time delay;
FIG. 4 is a block diagram of an embodied structure of the electronic noise
attenuation system to which the model shown in FIG. 3 is applied;
FIG. 5 is an explanatory view of a blocked embodiment of the operation of
the control part of the electronic noise attenuation system shown in FIG.
1;
FIGS. 6 and 7 are respectively explanatory views of the modifications of
the control part of the above-mentioned electronic noise attenuation
system; and,
FIG. 8 is a view of the structure of a conventional electronic noise
attenuation system.
DETAILED DESCRIPTION OF THE INVENTION
Detailed description will hereunder be given of the preferred embodiment of
an electronic noise attenuation system according to the present invention
with reference to the accompanying drawings.
Referring now to FIG. 1, there is shown a basic structure of an electronic
noise attenuation system according to the present invention. Although
FIGS. 1 and 2 were already discussed simply for convenience' sake in the
chapter of (Description of the Related Art), they will be described here
again in detail because the above discussion is not sufficient for
understanding of the present invention.
In FIG. 1, in a propagation passage 1 for sound waves, two sensor
microphones M1, M2, which are respectively used to detect respectively
sound waves propagated from a source of noise, are disposed on the
upstream and downstream sides of a speaker S serving as a source of
additional sounds with the speaker S as the reference position thereof. To
a point of addition 20 are input the output signal of the sensor
microphone M1 and the output signal of a digital filter 22 for restriction
of acoustic feedback such that the output signal of the digital filter 22
is added to the output signal of the sensor microphone M1 while the former
is opposite to the latter in phase.
Also, the output signal of the point of addition 20 is input to an adaptive
digital filter 2 and a controller part 10. To the controller part 10 there
is input the output of the sensor microphone M2 as an error signal E.
In the above-mentioned structure, the propagated sound waves from the
source of noise are detected by the sensor microphones M1 and M2, and the
output signal of the sensor microphone M2 is input to the controller part
10 as the error signal E.
At the point of addition 20 the outputs of the sensor microphone M1 and the
digital filter 22 for restriction of acoustic feedback are added to each
other in mutually opposing phases and the addition output thereof is input
to the digital filter 2 and the controller part 10.
The controller part 10 performs such addition and output that the error
signal E can be a minimum value. In other words, the controller part 10 is
a device of an adaptive type which, in accordance with the input X of the
digital filter and the error signal E, determines a transfer function to
be given to the digital filter 2, and also supplies the digital filter 2 a
filter coefficient which is a control parameter for specifying the thus
determined transfer function. In the digital filter 2, the input signal X
is processed or converted to a signal having given a given amplitude and
phase characteristic in accordance with the filter coefficient given
thereto. The output signal of the digital filter 2 is converted from
digital to analog and is then output to the speaker S, namely the source
of additional or cancelling sounds, which is adapted to generate
cancelling sound waves for cancelling the propagated waves from the source
of noise at the position of the sensor microphone M2. In this manner, the
propagated sound waves from the source of noise can be cancelled at the
position of the sensor microphone M2.
The above-mentioned cancelling sound waves from the speaker S can be
detected or sensed by the sensor microphone M1 and, the detected
components of the sensor microphone M1, that is, the sensed cancelling,
sound waves can be cancelled by adding the output signal of the digital
filter 22 representing the transfer characteristics from the sound
cancelling digital filter 2 to the point of addition 20 with the phase
thereof reversed, to the output signal of the sensor microphone M1 in the
point of addition 20, so that the acoustic feedback from the speaker S to
the sensor microphone M1 can be restricted. That is, the digital filter 22
acts as a digital filter for restricting the acoustic feedback.
In FIG. 2 which shows a modeled version of the electronic noise attenuation
system shown in FIG. 1, reference character G designates a transfer
function representing the propagation characteristics of sound waves
within the propagation passage 1 between the sensor microphones M1 and M2
and the conversion characteristics of the sensor microphone M1 and M2.
And, D, as described before, designates a transfer function representing
transfer characteristics which include sound wave propagation
characteristics of the propagation passages existing from the output
terminal of the digital filter 2 to the point of addition for the error
signal, that is, passages from the output terminal of the digital filter 2
to the speaker S and from the speaker S to the microphone M2 as well as
the conversion characteristics of electro-acoustic transducers themselves
such as the speaker S and the sensor microphone M2.
Next, in FIG. 3, there is shown a model obtained by embodying the
electronic noise attenuation system including a controller in
consideration of the above-mentioned transfer function D. In this model,
the VS-LMS algorithm is employed in the controller part 10 as an adaptive
control algorithm and the multiplication of the output signal X at the
point of addition 20 by the transfer function D is considered as the input
signal of the digital filter 2, whereby the coefficient of the digital
filter 2 can be updated. Therefore, by replacing the input signal X by
X.multidot.D as the input of the operation according to the VS-LMS
algorithm, the updating of the filter coefficient according to the VS-LMS
algorithm is possible.
The transfer function D, as will be discussed later, can be obtained by the
controller part 10 prior to the operation of the system, thereby
determining a filter coefficient which specifies the transfer function D.
While the system is in operation, the filter coefficient is fixed and the
digital filter 2 is controlled adaptively according to the VS-LMS
algorithm.
Referring now to FIG. 4, there is shown the concrete structure of an
electronic noise attenuation system to which the model shown in FIG. 3 is
applied. In FIG. 4, within the propagation passage 1 there are provided
the sensor microphones M1, M2 such that they are disposed with the speaker
S, the source of cancellation sound, between them.
Numerals 30, 32 respectively designate microphone amplifiers for amplifying
the output signals of the microphones M1, M2, respectively, and 34 stands
for a power amplifier which amplifies a drive signal to be output to the
speaker S up to a given level.
Also, 50, 52 respectively designate A/D converters, 54 a D/A converter, and
1000 a control part.
The control part 1000 comprises a control processor 100 for generally
controlling the whole system, digital signal processors 102, 104 which
respectively serve as a noise generator for measuring an adaptive digital
filter to be discussed later, a digital filter of a fixed coefficient type
and the above-mentioned transfer function D, and serial/parallel interface
adapters 106, 108 converting a serial signal to a parallel signal or a
parallel signal to a serial signal, all of which are connected to one
another by means of bus lines 200.
Now, description will be given of the operation of the electronic noise
attenuation systems shown in FIG. 1 with reference to FIG. 5. FIG. 5 is a
block diagram of the operation of the control part 1000. In FIG. 5, before
the system is put into operation, a switch 208 is changed over to a point
of contact and a pseudo-random noise is output from a noise generator 206
to the D/A converter 54.
On the other hand, the digital signal processor 104 is used to provide an
adaptive digital filter 210. The adaptive digital filter 210 identifies
the transfer function D of the digital filter 202 in accordance with an
input signal (pseudorandom noise) from the noise generator 206 and the
output signal (error signal) of the A/D converter 52 that is the output
signal from the sensor microphone M2.
Also, similarly, in accordance with an input signal from the noise
generator 206 and the output signal of the A/D converter 50 that is the
output from the sensor microphone M1, an adaptive digital filter 410
identifies the transfer function F of the digital filter 22 for
restriction of acoustic feedback.
Next, the switch 208 is changed over to a point of contact b to thereby
make the electronic noise attenuation system ready for operation. Then,
the filter coefficient representing the transfer function D identified by
the digital filter 210 is set in the digital filter 202 and, similarly,
the filter coefficient representing the transfer function F identified by
the digital filter 410 is set in the digital filter 22. The digital
filters 202 and 22 are shared by the digital signal processor 102 in the
functions thereof, and the adaptive digital filter 204 and the adaptive
digital filter coefficient updating algorithm realizing circuit 220 are
shared by the digital signal processor 104 in the functions thereof. The
adaptive digital filter 204 corresponds to the digital filter 2 in the
model shown in FIG. 3.
In this state, to the point of addition 20 there are input electric signals
respectively through the A/D converter 50 and digital filter 22 and, in
the point of addition 20, the output signal of the A/D converter 50 and
the inverted version of the output signal of the digital filter 22 are
added together. In addition, in the digital filter 202, the output signal
X of the point of addition 20 is multiplied by the transfer function D
that is set in the digital filter 202.
The adaptive digital filter coefficient updating algorithm realizing
circuit 220 takes therein the output signal of the A/D converter 52 as the
error signal and, in accordance with this signal and the output
X.multidot.D of the digital filter 202, updates the filter coefficient of
the adaptive digital filter 204. The adaptive digital filter 204 performs
a given operation on the output signal X of the point of addition 20 and,
by means of the switch 208, outputs the resultant to the D/A converter 54
as the drive signal for the speaker S to cancel the propagated sound waves
from the source of noise at the position where the sensor microphone M2 is
set. The operation of the point of addition 20 in FIG. 5 is performed by
the control processor 100 and, besides this, the control processor 100
transmits and receives signals to and from the electronic noise
attenuation system and other systems (not shown) to which the electronic
noise attenuation system is applied, such as air conditioning system and
the like. Further, the control processor 100 monitors, the operation of
the electronic noise attenuation system and, if anything wrong occurs in
the system, performs processings to cope with it. In addition, the control
processor 100 is able to check the noise cancelling digital filter 204 for
its on/off operation on updating of the filter coefficient, so that the
operation of the digital filter 204 can be controlled adaptively and thus
the digital filter 204 is able to cope with unstable situations.
Although in the adaptive digital filters 204, 210, 410 shown in FIG. 2
there is used the VS-LMS algorithm, this is not limitative, but other
adaptive algorithm such as the BLMS (Block Least Mean Square) or the FLMS
(Fast Least Mean square) or the like may be employed. Also, in the
above-mentioned embodiment the point of addition 20 is set at a position
where the digital operation can be performed, but the point of addition 20
may be set, together with the digital filter 22, externally of the
controller and the addition thereof may be executed at the stage of an
analog signal.
Further, in the system construction shown in FIG. 4, there are used two
digital signal processors and one control processor, but a microprocessor
having a high function can be used in place of them to perform their
functions. Moreover, the digital signal processors 102 and 104 can be
replaced with high-speed multiplying/adding devices, respectively.
Now, description will be given in more detail of the application of the
invention by use of expressions in a block diagram according to FIG. 5.
Here, the parts that are used in common with FIG. 5 are given the same
designations and the description thereof is omitted here.
In a case when a special noise is to be cancelled, that is, in a case where
electro-mechanical transducer means for generating an additional or
cancelling sound is weakly connected to first mechano-electric transducer
means for detecting a propagated signal from a source of noise to convert
it into an electric signal, an acoustic feedback group need not be taken
into consideration. For example, when the first mechano-electric
transducer means such as a vibration pickup or the like is used to detect
the vibration speed components of a source of noise not a sound pressure,
or when, in structure, the first mechano-electric transducer means is
weakly connected to the electro-mechanic transducer means for generating
the additional sound because the former is disposed remotely from the
latter, the input and error signals shown in FIG. 5 can be realized in a
further more simplified construction. In the most simplifid case, as shown
in FIG. 6, the noise detect signal can be directly used as the input
signal of the adaptive digital filter 204. However, even in this case, due
to the fact that there is essentially present the transfer function with a
time delay between the electro-mechanic transducer means for generating
the additional sound and the mechano-electric transducer means for
detecting the error signal, it is necessary to secure a highly applicable
adaptive digital filter system according to the invention as shown in FIG.
1, which provides an excellent noise cancelling effect.
Also, in FIG. 1, the digital filter 22 for restriction of acoustic feedback
is formed of a digital filter of a fixed coefficient type, but, however,
it is well known that a wider range of application can be provided if the
digital filter 22 is composed of an adaptive digital filter.
In FIG. 7, there is shown a concrete structure of the above-mentioned
adaptive digital filter, in which E designates an error signal of the
digital filter and X an input signal thereof. The adaptive digital filter
may be used in combination with a digital filter 2 for adapter
controlling/noise cancelling or may be used indepently.
As can be understood from the foregoing description, the present invention
not only can apply to an electronic noise attenuation system but also can
apply to all adaptive control systems including a transfer function with a
time delay.
As has been described hereinbefore, in the electronic noise attenuation
system according to the present invention, the electro-mechanic transducer
means as the source of additional sound, prior to operation of the system,
generates a sound wave in the propagation passage of sound waves according
to a pseudo-signal, the control means, responsive to the sound wave
generated by the electro-mechanic transducer means, specifies a transfer
function with a time delay representing the propagation characteristics of
propagation passages of sound waves existing from the output terminal of
the drive signal generating means for generating a drive signal for the
electro-mechanic transducer means to the second mechano-electric
transducer means and the transfer characteristics of the transfer systems
including the transfer paths of electric signals so that the output signal
(error signal) of the second mechano-electric transducer means for
evaluation of the noise cancelling effects of the generated sound wave can
be a minimum value, and the control means, in consideration of the
specified transfer function with a time delay, determines a transfer
function to be given to the drive signal generating means in accordance
with a given adaptive algorithm. Therefore, according to the invention, an
electronic noise attenuation system which can enjoy a high noise
cancelling effect can be realized.
It should be understood, however, that there is no intention to limit the
invention to the specific forms, but on the contrary the invention is to
cover all modifications, alternate constructions and equivalents falling
with in the spirit and scope of the invention as expressed in the appended
claims.
* * * * *
|
|
 |
|
 |
Description |
|
