JPH0335295A - Silencer - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a muffler, and more particularly, to a muffler that automatically changes coefficients depending on the situation and muffles noise in an optimal state.

(Prior art) Conventionally, as a device for reducing noise at a certain location,
There is a known sound muffling device that artificially generates a sound wave of the same magnitude and opposite phase to the noise at that location, and actively cancels the noise by making the artificially created sound interfere with the noise. It is being

Figure 6 shows an example in which a silencer is used to cancel the noise at the doorway in order to prevent the noise generated by the equipment from leaking outside through the doorway of the room. 1 indicates a device that generates noise, 2 indicates a room in which device 1 is installed, 3 indicates an entrance to room 2, and 4 indicates a silencer.

The silencer 4 usually detects the noise generated by the device 1 by using a sensor,
For example, the detection signal is detected by a microphone 10 and sent to the signal processing device 1 via an amplifier 11 and an A/D converter 12.
3, and the signal processed by this signal processing device 13 is sent to the D/
The signal is supplied via the A converter 14 to a control sound generator, for example, a speaker 15 located near the entrance/exit 3.

A coefficient sequence (impulse response) for minimizing the sound at the doorway 3 using the sound detected by the microphone 10 is preset in the signal processing device 13 as a calculated value. The signal processing device 13 then outputs a signal obtained by multiplying the human signal by the coefficient sequence.

This output y (n) is, however, h (i) can be expressed as a coefficient sequence, and X (n-1) can be expressed as the i-th most recent sample data of the input signal (signal from the microphone 10).

By using such a silencer 4, the doorway 3
The noise generated by the device 1 can be actively canceled out, and the noise generated by the device 1 can be prevented from leaking to the outside through the doorway 3.

However, each acoustic transfer function, which is the basis for calculating the coefficient sequence, is not always constant. This acoustic transfer function is expressed as
It changes due to changes in the temperature of the transmission path, changes in the characteristics of the speaker 15, changes in the characteristics of the device 1, etc. as indicated by . For example, the seventh
As shown in Figure (a), when the temperature in the room 2 rises due to the heat generated by the device 1, the speed of sound changes. When the speed of sound changes in this way, the acoustic transfer function changes. Further, as shown in FIG. 7(b), if the speaker 15 continues to be energized, the winding temperature increases, the resistance value changes, and the output changes. This also changes the acoustic transfer function. Furthermore, as shown in FIG. 7(c), as the position of noise generation in the device 1 changes with the passage of operating time, the acoustic transfer function changes accordingly.

If the acoustic transfer function changes in this way, it becomes impossible to perform sound muffling effectively. Therefore, in order to achieve good noise reduction, it is necessary to change the values of the coefficient sequence itself in accordance with changes in the acoustic transfer function.

Therefore, in order to deal with this problem, recently, a so-called adaptive technology has been developed in which another microphone 16 is installed at the entrance/exit 3 and the coefficient sequence is automatically changed so that the output of this microphone 16 becomes zero. Silencers with active control capabilities have been proposed.

The muffling device equipped with this adaptive active control function changes the coefficient sequence at regular intervals according to equation (2).

h(i) m1ll -h(1) 11111 +Ke
X (n l)...(2) Here, h (1) , , is the i-th FIR after the update
Filter coefficient (impulse response) h(1). where eye is the i-th FIR filter coefficient (impulse response) before updating, K is a constant that determines the change rate of h, e is the error signal (signal from microphone 16), and X(n-i) is the human input signal (microphone 10). signal from ) the i-th most recent sample data, is.

However, even with the above-mentioned muffling device having the adaptive active control function, there are the following problems. This muffler usually changes the coefficient sequence at regular intervals and operates while keeping the value of K in equation (2) constant. The reason why the value of K is kept constant is that the standard for changing the value of K is not clear. However, if the value of K is kept constant, the following problem may occur. That is, when the time constant of the change in the physical system that determines the acoustic transfer function and the time constant of the change in h due to K are comparable, oscillation occurs. Furthermore, if the time constant of the change in h due to K is larger than the time constant of the change in the physical system, control that follows the change in the physical system will not be possible. Furthermore, if K is increased to make the time constant of change in h extremely small, the stability of control will deteriorate. In addition, if it is known in advance that the change in the physical system is gradual, it is necessary to reduce the value of K, but if the value of K is extremely small, it may be difficult to process the signal in terms of bit loss etc. A problem arises. Therefore, it is extremely difficult to select the value of K to be set. In addition, in conventional devices, the coefficient sequence is changed at regular intervals, but if the above period is shortened and changes are made too frequently, the stability of the control system deteriorates.
Furthermore, if the above-mentioned period is made longer and the change frequency is made smaller, there is also a problem that it becomes impossible to follow changes in the acoustic transfer function.

Furthermore, in the conventional device, if an external sound enters after the coefficient identification process is completed and converged, the device malfunctions and changes the coefficients, reducing the silencing effect.

(Problems to be Solved by the Invention) As mentioned above, in the conventional noise suppressor equipped with an adaptive active control function, the change rate of the coefficient sequence, that is, the convergence rate of the control system is always constant, so・Depending on the mode, there was a problem that the stability and convergence of the control system deteriorated.

SUMMARY OF THE INVENTION An object of the present invention is to provide a noise reduction device that can efficiently and appropriately change the convergence rate of a control system in response to operating conditions and modes, and that can always perform stable noise reduction control. Another object of the present invention is to provide a muffling device that performs stable adaptive active muffling control without changing the coefficients due to malfunction due to external noise or the like after the coefficients have been identified.

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the silencer according to the present invention has the following features:
The convergence rate of control is adjusted by switching between adaptive active control in which coefficients are adaptively changed according to predetermined conditions and active control in which coefficients are fixed.

Here, the predetermined condition refers to, for example, the frequency of coefficient updates determined by the operating conditions/mode, or the frequency of coefficient updates determined by the elapsed time from the start of operation of the equipment that is the source of noise. It means.

(Function) In normal adaptive active silencing control, the coefficient is updated every fixed period according to the above-mentioned formula (2). In the silencing device according to the present invention, the frequency of coefficient updating is changed depending on the driving conditions and motors. This is equivalent to essentially changing the convergence rate K when updating the coefficients. Therefore, the optimal convergence rate is automatically selected in accordance with the operating conditions and mode, and stable Adaptive active silencing control that satisfies both performance and convergence is realized.Furthermore, after coefficient identification,
Stable adaptive active noise reduction control is achieved without changing coefficients due to malfunctions caused by external noise.

(Example) Examples will be described below with reference to the drawings.

FIG. 1 shows a schematic configuration of an example in which a silencer 21 according to an embodiment of the present invention is used to prevent noise generated in a device 1 from leaking outside through an entrance 3 of a room 2. ing.

The muffling device 21 is configured as follows. That is,
The noise or vibration generated by the device 1 is converted into an electrical signal by a sensor 22, such as an acceleration pickup or a microphone, placed near the device 1, and this signal is sent to an amplifier 23 and an A/D converter. Signal processing device 2 via 24
It has been introduced in 5. On the other hand, doorway 3, which is the point to be muted,
A microphone 26 for detecting the error in the control result is placed at
The signal is introduced into the signal processing device 25 via a converter 28.

The signal processing device 25 uses a coefficient sequence h set internally.
(i) as F I R (Finite IrApuls
e1? response) is used as a filter to create an output signal.

At this time, the coefficient sequence h(i) is appropriately updated (changed) so that the output of the A/D converter 28, that is, the sound pressure at the doorway 3, which is the point to be muted, becomes zero. This update was mentioned above (2
) according to the formula. Specifically, the previous update value hm
and the previous updated value hm-1, and (h--
When the value obtained by multiplying the absolute value of hm-+ ) by the count value N of the update frequency counter becomes larger than a predetermined constant value ε, that is, hm(l h −-h #-11) N > ε
...When (3) is reached, the coefficient sequence is updated by executing the formula (2).The update frequency counter counts clock pulses of a constant period, and the coefficient sequence is updated at the time when the coefficient sequence is updated. The count value is cleared by .

The signal processing device 25 also has the following functions. That is, when an operation stop command is introduced from the operation control system 31 that controls the start and stop of operation of the device 1,
The values of hm and hm- at the time when this command is introduced are fixed and stored, and a storage completion signal is given to the operation control system 31 when the storage is completed. Note that the operation control system 31 issues an operation stop command to the device 1 after receiving the storage completion signal. Further, when the signal processing device 25 receives an operation restart command from the operation control system 31, it starts the silencing control using the previously stored hm and hm as initial values. In addition, when the operation stop command information of the equipment that is the sound source is input manually, the coefficient is held fixed, and the next time the operation restart command information of the equipment is input, the held coefficient is supplied as the initial value. This concept can be applied to other than ordinary active muffling devices and the adaptive active muffling device of the present invention.

The output signal of the signal processing device 25 is given to a speaker 30 arranged near the entrance/exit 3 via a D/A converter 29, and is used for muffling the sound at the entrance/exit 3.

Next, the operation of the silencer 21 configured as described above will be explained.

This silencer 21 is installed under beam 1, hs-1 under unused condition.
The value corresponding to is not set. Therefore, at the start of use, values corresponding to hm and hm-1 are artificially set.

FIG. 2 is a diagram showing the flow of processing when the noise at the entrance/exit 3 is continuously muffled using this muffling device 21.

First, the sensor 22 acquires the control signal X(ni) (S,). The coefficient sequence hm(i) that was last updated for this signal
is used as a FIR filter to generate control output signal y (n
) is generated (82), and this signal is used to generate sound from the speaker 30 (S,). Due to the sound produced by the speaker 30, the sound coming from the device 1 and the sound coming from the speaker 30 are synthesized at the doorway 3, which is the control target point. If the control result is good, the sound pressure at the entrance/exit 3 will be completely zero. However, normally the sound pressure rarely becomes completely zero. Obtain the error signal e as a control result using the microphone 26 (S4)
The amount of change in h is hm−h.

is calculated (S5). Next, it is determined whether or not to update the coefficients based on the above equation (3) (S6). Then, when it is determined not to update, the process moves to S2. Therefore, in this case, silencing control is continued using the same coefficient sequence hm(i). Furthermore, when it is determined that the update is to be performed, KeX (ni) shown in equation (2) above is output (S
7).

When KeX (n-i) is output, a coefficient sequence update command is issued (Sa), whereby the coefficient sequence used at 82 is updated to hm++(t) according to equation (2). In S2, y (n) is calculated using hmm+1(i). This process is repeated thereafter.

With such a control method, when the update amount is large, updates are performed frequently, and when the update amount is small, updates are performed less frequently. Therefore, the optimal convergence rate is automatically selected in accordance with the operating conditions and modes, thereby realizing adaptive active noise control that satisfies both hm stability and convergence. In addition, after the coefficient identification, even if external sound is incident, basically hm-hm of equation (3),
Since the value of + is small, stable adaptive active control is achieved without changing the coefficients due to malfunction.

FIG. 3 is a diagram showing the relationship between the start and stop of operation of the device 1 and the operation of the silencer 21. This muffling device 21 is configured in consideration of the case where the equipment 1 is intermittently stopped (for example, a compressor, etc.) as shown in FIG. 4.

That is, when an operation stop command is given from the operation control system 31 during the adaptive active silencing control (S++) explained using FIG. 2, the coefficient /II h used at that time
@, h a-r are fixed and stored in the storage device (S+□). When the storage is completed, a storage completion signal is given to the operation control system 31. Operation control system 31 is device 1
The operation of the controller is stopped (S13). Then, when an operation restart command is given from the operation control system 31, the silencing control operation shown in FIG. Start.

With such an operation method, the filter coefficient string is not updated while the device 1 is stopped, so that noise is not set in the filter coefficient string while the device 1 is stopped. In addition, since adaptive control is started using the retained coefficient sequence as the initial value when restarting operation, convergence can be accelerated compared to the case where the initial value is a coefficient sequence that has been cleared to zero or has noise set. It is possible to quickly shift to an adaptive active silencing state that suits the situation.

Note that the present invention is not limited to the embodiments described above. That is, in the embodiment described above, when h is changed chm-
hm is calculated, and the update timing of h is determined based on this calculated value and the count value N of the update frequency counter. However, as shown in FIG. 7, changes in the elements that affect the acoustic transfer function depending on the system may always show a constant tendency from the start of operation. When targeting such a system, the frequency of adaptive control may be changed based on the elapsed time t from the start of operation. Fifth
The figure shows the processing flow of a silencer that employs this method.

In this example, S reads out the relationship between the elapsed time t and the required convergence amount from a database whose characteristics are stored in advance, and determines in S86 whether or not coefficient string update is instructed at the elapsed time t. That's what I do.

With such a control system, although the adaptive characteristics are slightly worse than those of the embodiments described above, the device configuration and processing routine can be significantly simplified.

[Effects of the Invention] As described above, according to the present invention, the convergence rate and stability of adaptive control according to operating conditions can be obtained relatively easily, and the optimum Adaptive active silencing control can be realized.

[Brief explanation of drawings]

Fig. 1 is a schematic configuration diagram of a silencing device according to an embodiment of the present invention, Fig. 2 is a diagram showing the flow of processing when the silencing device continuously performs silencing operation, and Fig. 3 is a diagram showing the same silencing device. Flow diagram showing the operational relationship between the equipment and the shutdown of equipment that is a noise source, No. 4
5 is a diagram showing an example of a mode in which equipment is stopped; FIG. 5 is a diagram showing a process flow when a silencing device according to another embodiment of the present invention performs a continuous silencing operation; FIG. 7 is a schematic configuration diagram of a conventional silencing device, and FIG. 7 is a diagram for explaining several reasons why the acoustic transfer function changes. DESCRIPTION OF SYMBOLS 1... Equipment which is a noise source, 2... Room, 3... Doorway, 21... Silencer, 22... Sensor, 26
...Microphone, 23.27...Amplifier, 24.
28... A/D converter, 25... Signal processing device, 2
9...D/A converter, 30...Speaker, 31...
・Operation control system.

Claims (8)

[Claims] (1) A control signal is created by multiplying the signal received by the sensor by a predetermined coefficient, and this control signal controls the sound generator to cancel the sound at the point to be muted, and the control at the point to be muted is performed. In a silencer equipped with an adaptive active silencing control function that adaptively changes the coefficient according to the result, there is an active silencing control state in which silencing control is performed with the coefficient fixed according to predetermined conditions, and an adaptive active silencing control state. A silencing device characterized by comprising a switching means for switching to a silencing control state. (2) The silencing device according to claim 1, wherein the switching means changes the frequency of switching to the adaptive active silencing control state in accordance with operating conditions and modes. (3) The silencing device according to claim 1, wherein the switching means changes the frequency of switching to the adaptive active silencing control state in accordance with the elapsed time from the start of control. (4) When the operation stop command information for the device that is the sound source is input, the coefficient is fixed and held, and the retained coefficient is set as the initial value when the operation restart command information for the device is input next. 2. The muffling device according to claim 1, further comprising supplying means. (5) A control signal is created by multiplying the signal received by the sensor by a predetermined coefficient, and this control signal controls the sound generator to cancel the sound at the point to be muted, and the control at the point to be muted is performed. A muffling device equipped with an adaptive active muffling control function that adaptively changes the coefficient in accordance with a result, characterized in that the update frequency of the coefficient is made variable according to predetermined conditions. (6) The muffling device according to claim 5, wherein the coefficient is updated when the following conditions are satisfied. (|h_m−h_m_−_1|)N>ε However, h_m is the previous update value h_m_−_1 is the previous update value N is the count value of the update frequency counter ε is a predetermined constant value (7) The muffling device according to claim 5, wherein the updating of the coefficient is variably set according to a predetermined elapsed time from the start of the control. (8) A control signal is created by multiplying the signal received by the sensor by a predetermined coefficient, and this control signal controls the sound generator to cancel the sound at the point to be muted, and the device that is the sound source The apparatus is characterized by comprising means for fixing and holding the coefficient when operation stop command information is input, and supplying the held coefficient as an initial value when equipment operation restart command information is input next. Silencer.