JPH08196772A - Noise controller for sewing machine - Google Patents

Abstract

PURPOSE: To simultaneously and actively reduce, without lowering the mechanical strength of a sewing machine, the machine noise generated by the driving of the sewing machine at plural points of measurement. CONSTITUTION: The oscillation signal is detected from the oscillation-detecting sensor 202 by driving the sewing machine 203. Simultaneously, the machine noise at plural points of measurement is detected by each of both microphones 205a and 205b. Based on the oscillation signal and plural noise signals, the noise-controlling part 201 simultaneously controls the plural speakers 204a and 204b. The control sound generated from each of both speakers 204a and 204b then works to weaken the noise at plural points of measurement. Accordingly, the noise can be actively reduced at plural points of measurement.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise control device for a sewing machine, which actively reduces the noise of the sewing machine generated by the rotation of the upper shaft of the sewing machine and the vertical reciprocating motion of a needle bar.

[0002]

2. Description of the Related Art In a conventional sewing machine, in order to reduce noise generated by driving the sewing machine, vibration and the like are reduced by reducing the weight of motion machine mechanical parts of the sewing machine and balancing the rotary machine parts. I was trying to reduce noise.

Further, there has been proposed a simple noise suppressing device which detects and accumulates noise of a sewing machine in advance, generates a waveform having an opposite phase to that, and suppresses the noise of the sewing machine by using the waveform having the opposite phase.

[0004]

However, reducing the weight of the motion system of the sewing machine lowers the mechanical strength of the motion system as a whole, and this method has a limit, which causes noise. There is a limit to noise reduction.

Further, in the rotating system of the sewing machine, since the respective mechanical parts make complicated mutual movements, it is very difficult to balance them so as to reduce noise.

Further, according to the simple noise suppressing device,
It is very difficult to accurately generate a waveform with the same amplitude and opposite phase with respect to the noise generated by the high-speed rotation of the sewing machine. Further, in practice, the noise of the sewing machine is affected by disturbances and the time delay of the noise transmission system, so it is considered difficult to suppress the noise while following the noise.

The present invention has been made in order to solve the above-mentioned problems, and the noise of the sewing machine generated by driving the sewing machine can be simultaneously and simultaneously observed at a plurality of observation positions without lowering the mechanical strength of the sewing machine. An object of the present invention is to provide a noise control device for a sewing machine that is actively reduced.

[0008]

In order to achieve the above object, a noise control device for a sewing machine according to the present invention detects a vibration generated by driving the sewing machine and a vibration detecting means for detecting a vibration generated by driving the sewing machine. A plurality of noise detection means for detecting at a plurality of observation positions, a plurality of control sound generation means capable of generating a control sound for canceling the noise generated by driving the sewing machine, a signal detected by the vibration detection means, and each of the above Based on the signal detected by the noise detecting means, there is provided control means for controlling the control sound generating means so that the noise detected by each of the noise detecting means is minimized at the plurality of observation positions.

The vibration detecting means may be provided in the sewing machine main body.

Further, the control means is provided with a transfer function setting means for presetting a transfer function indicating a characteristic or a mutual interference characteristic of a mechanical system part of the sewing machine, an electric system part for control, or the like.
You may comprise so that the said control sound generation means may be controlled based on the said transfer function with each detection signal of the said vibration detection means and the said noise detection means.

Further, a storage means for storing the transfer function set by the transfer function setting means is provided, and when the control means is operated, the transfer function stored in the storage means is read and applied to the control means. It may be configured to do so.

[0012]

According to the noise control device for a sewing machine of the present invention having the above-mentioned structure, the vibration detecting means detects the vibration signal correlated with the noise of the sewing machine, and the noise detecting means detects the vibration signals at a plurality of observation positions. Detect noise. Then, based on the signal detected by the vibration detection means and the signal detected by each noise detection means, the control means controls the control sound generation means so that noises at the plurality of observation positions are simultaneously minimized. Then, a control sound for canceling the noise is generated from the control sound generating means.

By providing the vibration detecting means in the sewing machine main body, the vibration of the sewing machine main body can be directly detected. In this case, since the directly detected signal is applied to the control means, the control of the control sound generation means can be made more reliable.

Further, since the transfer function set by the transfer function setting means is taken into consideration in the control by the control means, the control of each control sound generating means can be performed more finely and highly accurately. The noise of the sewing machine can be further reduced.

Further, a storage means for storing the transfer function set by the transfer function setting means is provided, and when the control means is operated, the transfer function stored in the storage means is read and applied to the control means. Therefore, it is not necessary to set the transfer function each time the control means is operated, and noise can be efficiently reduced.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be described in detail below with reference to the drawings.

As shown in FIG. 2, the sewing machine main body 203
A vibration detecting sensor (vibration detecting means) 202 for detecting a vibration having a correlation with noise is provided on the bed under the needle bar. Further, at a plurality of observation positions (two positions in this embodiment) near the main body of the sewing machine, a pair of microphones 205a and 205b as noise detecting means for detecting noise generated by driving the sewing machine, and those microphones. 205
A pair of speakers 204a and 204b as control sound generating means for generating a control sound for canceling the noise of the sewing machine are arranged in the vicinity of a and 205b, respectively. The microphones 205a and 205b and the speaker 2 of them
The sewing machine main body 2 is denoted by 04a and 204b for convenience in FIG.
Although shown in the left space of 03, the sewing machine main body 2 is actually
03, a portion where particularly large noise is generated, for example, in a space near the needle bar or the shuttle mechanism, is fixedly installed on a support member (not shown) fixed to the sewing machine main body 203. There is.

The vibration detection sensor 202, the pair of microphones 205a and 205b, and the pair of speakers 204a and 20a.
4b is connected to the noise control unit 201 as a control means.

As shown in FIG. 3, the noise control unit 201 mainly includes an amplifier AMP 304, an A / D converter 306, a D / A converter 308, a low pass filter LPF 305 for removing high frequency components, and a low frequency filter. High-pass filter HPF307 for removing frequency components, adaptive filter 30
2, a digital filter 303 and an LMS adaptive algorithm 301. Each microphone 205
a, 205b based on the detection signals e ₁ and e ₂ , the vibration signal x of the vibration detection sensor 202 correlated with the noise of the sewing machine, and the transfer function previously obtained, the microphones 205
a, 205b so that the detection signals e ₁ and e _{2 of the} a and 205b are simultaneously minimized, the noise control unit 201 uses the speakers 204a and 2b.
The drive control signal is output to 04b. The LMS adaptive algorithm 301 calculates a coefficient to be given to the adaptive filter 302, and functions as the transfer function setting means of the present invention.

That is, when the sewing machine rotates, the vibration signal x detected by the vibration detection sensor 202 is sent to the amplifier AMP304,
High pass filter HPF307, low pass filter LP
It is input to the adaptive filter 302 and the digital filter 303 via the F305 and the A / D converter 306, respectively. The digital filter 303 generates a reference signal r _lm in consideration of mutual interference and the like by a transfer function obtained in advance, and inputs them to the LMS adaptive algorithm 301.

On the other hand, the noises at the respective observation positions are detected by the respective microphones 205a and 205b for detecting the respective noises, and the detection signals e ₁ and e ₂ are detected by the amplifier AMP 304,
Low pass filter LPF 305 and A / D converter 306
Is input to the LMS adaptive algorithm 301 via.
The LMS adaptive algorithm 301 adjusts the coefficients of the adaptive filter 302 in consideration of the transfer function of the acoustic space, and transmits these to the adaptive filter 302. Adaptive filter 3
The outputs y ₁ and y _{2 of} 02 drive each speaker 204. That is, each sensor 20 that detects an interference state between the noise of the sewing machine and the control sound generated from each speaker 204a, 204b.
The coefficients to be given to the adaptive filter 302 are determined so that the signals 5a and 205b are simultaneously minimized, and the coefficients are set in the adaptive filter 302.

The LMS adaptive algorithm 301
Always modifies the above-mentioned control parameters in response to changes in the characteristics of the mechanical parts of the sewing machine, electrical parts for control, etc. and changes in the mutual interference characteristics. As a result, the vibration detection sensor 20
The vibration signal detected by 2 is an amplifier 304, a high pass filter HPF307, and a low pass filter LPF30.
5 and the adaptive filter 302 via the A / D converter 306.
To the digital signal having a predetermined amplitude characteristic and phase characteristic based on the coefficient given from the LMS adaptive algorithm 301 by the adaptive filter 302. The digital signal is the D / A converter 3
08, low-pass filter LPF305 and amplifier 304
D / A converted and amplified by each speaker 204a, 2
04b is applied as a drive signal to each speaker 204.
From a and 204b, each microphone 205 at each observation position
A control sound is generated to minimize the residual noise of a and 205b. As a result, the noise of the sewing machine is attenuated at the positions of the microphones 205a and 205b.

Further, each speaker 2 is selected from the vibration signal of the sewing machine.
The adaptive filter 302 shown in FIG. 3 is composed of a finite impulse response filter (also referred to as FIR filter) for the reason that it is easy to give an arbitrary characteristic in order to synthesize the control sounds for driving the 04a and 204b. There is.

Here, the LMS adaptive algorithm 30
The operation of No. 1 will be described.

Now, the residual noise signal detected by the l-th microphone 205 is e _{l (n)} , and the noise signal detected by the l-th microphone 205 when no control sound is generated from the speakers 204a and 204b is d _l (n). ), The j-th term of the transfer function between the m-th speaker 204 and the l-th microphone 205 is c
_lmj , when the vibration signal x (n) is input and the i-th coefficient of each filter in each output channel in the adaptive filter 302 that drives the m-th speaker 204 is w _mi ,

[0026]

[Equation 1]

Is satisfied. Here, the term with (n) is a sample value at the sampling time, L is the number of microphones 205 (two in this embodiment), and M is the speaker 204.
(2 in this embodiment), I is the number of taps of the transfer function c _lm represented by the FIR digital filter, and J is the number of taps of the adaptive filter 302.

[0028] In the above equation (1), of the right-hand side "Σw _mi x (n
term -Ji) "is (represents the output when the input vibration signal to the coefficient w _{m)," Σc lmj (Σw mi x (nji} ) filter for each output channel in the adaptive filter 302 section) "is m-th Represents the signal when the signal energy input to the speaker 204 is output as sound energy from the speaker 204 and reaches the l-th microphone 205 via the noise transfer function c _lm in the acoustic space.
entire right side of _{lmj (Σw mi x (nji)} ) ", since the arrival signal to the l-th microphone 205 are added together for all the speaker 204, represents the sum of the secondary sound reaching the l-th microphone 205 .

Then, the evaluation function J _e is

[0030]

[Equation 2]

Let us say.

Then, in order to obtain the filter coefficient w _m that minimizes the evaluation function J _e , the LMS adaptive algorithm 30
1 is used. That is, the evaluation function J _{e is set} to each filter coefficient w.
_The filter coefficient w _mi is updated with the instantaneous gradient value approximated by _mi .

Therefore, from the equation (2),

[0034]

(Equation 3)

From equation (1),

[0036]

[Equation 4]

Therefore, the right side of the equation (4) is r _lm (n
-i), the filter coefficient rewriting formula is the following formula (5).
Given in.

[0038]

(Equation 5)

Here, μ is a constant called a step size, and is a parameter that influences the convergence speed and convergence accuracy of the adaptive filter.

That is, in order to pursue the convergence speed, it is sufficient to increase the parameter μ within a converging range, but in consideration of the accuracy after convergence, it is necessary to comprehensively select the parameter μ. In particular, for noise generated from a mechanism such as a sewing machine that reciprocally rotates at any time, it is necessary to suppress the noise in a short time, so it is considered that a convergence speed is required.

As described above, the estimated filter coefficient w _1n
w _2n ... W _Mn is transmitted to the adaptive filter 302 of FIG.
Output digital signals y ₁ (n) and y ₂ (n) are generated. The digital signals y ₁ (n) and y ₂ (n) are D / A converted by the D / A converter 308, the low pass filter LPF 305 and the amplifier 304.
A-converted, amplified, and applied as a drive signal for the speakers 204a and 204b, and a control sound for canceling the noise of the sewing machine is generated from each speaker 204.

The coefficient c _lmj of the digital filter 303 shown in FIG. 3 is obtained in advance. That is, the electric circuit, the microphones 205a and 205b, and the speaker 204a,
The mutual transfer function in the acoustic space from the control speakers 204a and 204b including the microphone 204b to the microphones 205a and 205b that reduce noise is obtained. An FIR filter is mainly used as the digital filter 303.

FIG. 6 shows a block diagram for setting the transfer function of this embodiment. Here, for the time being, a drive signal (white noise is used in this embodiment) is input to the speaker 204a, and the sound generated from the speaker 204a is transferred to the residual noise detecting microphones 205a and 205a via the transfer functions c ₁₁ and c ₂₁ , respectively. Picked up by 205b. In addition, the sound picked up by the residual noise detection microphone 205 is used as a desired signal, and the speaker 2
The same signal that drives 04a is used as the input signal of the adaptive filter, and this desired signal is estimated. When the estimation error becomes sufficiently small, the coefficients of the adaptive filter become the estimated values of the transfer functions c ₁₁ and c ₂₁ . Furthermore, the estimated values of the transfer functions c ₁₂ and c ₂₂ can be obtained by the same method as described above. In this way, the transfer function c indicating the characteristics and mutual interference characteristics of the mechanical system components and the control electrical system components existing in the noise propagation path.
_11, c _21, c _12, set the c ₂₂ advance in the storage unit such as a RAM connected to the noise controller 201 of these transfer functions (not shown). Then, the noise control unit 20
When 1 is operated, the transfer function stored in the storage unit is read from the storage unit and applied to the noise control unit 201.

Next, the operation of this embodiment will be described.

When the sewing machine motor is started, the vibration detection sensor 202 detects the vibration signal x and supplies it to the noise control unit 201.

In the noise control unit 201, the vibration signal x is sent to the amplifier AMP304, the high pass filter HPF307,
A digital filter 303 and an adaptive filter 302 via a low pass filter LPF 305 and an A / D converter 306.
Is supplied to. Of these, the digital filter 303 uses the input vibration signal x to calculate the filtered signal r _{lm according} to the equation (4) in correspondence with the transfer function c _lm between the microphone and the speaker, and the like. The signal r _lm is output to the LMS adaptive zugorism 301.

On the other hand, the microphones 205a and 205b detect noise remaining at the observation position in the acoustic space and output error signals e ₁ and e _{2 corresponding} to the noise to the noise control unit 201. In the noise control unit 201, the input signals e ₁ and e _{2 are} input.
Are amplifier AMP304 and low-pass filter LP, respectively
It is sent to the LMS adaptive algorithm 301 via the F305 and the A / D converter 306.

In the LMS adaptive algorithm 301, the filter coefficient update calculation based on the equation (5) is performed using each input signal. That is, the filter coefficient w _mi (n) at the current sampling time n is equal to the evaluation function, that is, the error signal e _l (n) corresponding to the noise from the microphones 205a and 205b, the filter coefficient in the direction in which the root mean square is minimized. The filter coefficient w _mi (n + 1) which is calculated for each time and should be set at the sampling time (n + 1) is obtained. So L
The filter coefficient w _mi (n + 1) updated by the MS adaptive algorithm 301 is transferred to the adaptive filter 302. Therefore, the coefficient of each filter in the adaptive filter 302 is
At the sampling time (n + 1), the newly calculated filter coefficient w _mi can be used. In this way, the LMS adaptive algorithm 301 repeats the update of the filter coefficient at every predetermined sampling time so as to minimize the evaluation function J _e .

Then, each filter in the adaptive filter 302 performs the vector operation of the input synchronizing signals x and w _mi by the filter coefficient set at that time to obtain the output values y ₁ and y ₂ , Using this value as a drive signal, the amplifier AMP304, the low pass filter LPF305, A /
It outputs to speaker 204a, 204b via D converter 308, respectively.

As a result, the speakers 204a, 204
b generates a control sound according to the input signals y ₁ and y ₂ ,
The generated sound is propagated in the acoustic space corresponding to the preset transfer function c _lm to form a control sound. Therefore, after the control is converged, the noise at the observation position in the acoustic space and the control sound interfere with each other and are almost cancelled, and the residual noise is significantly reduced. At this time, even if a change in noise occurs due to the rotation of the motor, this change is reflected by the vibration signal x
Is reliably followed and detected, the noise is reliably reduced.

FIG. 4 shows the result of an experiment using an actual sewing machine based on the above-described noise control device for the sewing machine. The dotted line and the solid line in FIG. 4 are the frequency characteristics measured in the vicinity of 2 cm from the microphone 205a when the control is OFF and the control is ON at the rotation of 3000 rpm, respectively.

FIG. 5 shows the result of an experiment using an actual sewing machine based on the above-mentioned noise control device for a sewing machine. The dotted line and the solid line in FIG. 5 are the microphone 205 when the control is OFF and the control is ON at the rotation of 3000 rpm, respectively.
It is a frequency characteristic measured in the vicinity of 2 cm from b.

From the results shown in FIGS. 4 and 5, it can be seen that the effect of the control by the noise control device of the sewing machine has been sufficiently demonstrated.

The present invention is not limited to the above-described embodiment, but various modifications can be made without departing from the spirit of the invention.

That is, for example, the vibration detection sensor 202 for detecting vibration shown in FIG. 2 is fixedly arranged on the bed portion under the sewing machine main body, but it can be made movable and can be changed to any position on the sewing machine main body. You may do it. Further, the number is not limited to one, and a plurality of vibration detection sensors may be used at the same time. A vibration detection sensor may be provided on a member that directly transmits vibration such as a sewing machine table that supports the sewing machine main body 203.

In the above embodiment, the case where two speakers as the control sound source and two microphones for noise detection are provided respectively has been described, but the number is not necessarily limited to this number. It is also possible to provide three addresses for each, and it is of course possible to provide more than that.

[0057]

As is apparent from the above description, according to the noise control device of the sewing machine of the present invention, when the sewing machine is driven, the vibration signal detected by the vibration detecting means and the plurality of noise detecting means detect the vibration signal. The control means controls the plurality of control sound generating means based on the generated noise signals at the plurality of positions, so that each control sound generating means minimizes the noise so as to cancel the noise at the plurality of positions. Since the sound is generated, the noise generated from the sewing machine main body can be simultaneously and actively reduced at a plurality of observation positions. Further, it has excellent effects such as not lowering the mechanical strength of the sewing machine.

[Brief description of drawings]

FIG. 1 is a block diagram showing a functional configuration of the present invention.

FIG. 2 is a schematic configuration diagram of an example.

FIG. 3 is a block diagram showing a configuration of a noise control device.

FIG. 4 is a graph showing the result of actual noise control of the sewing machine.

FIG. 5 is a diagram showing a result of actual noise control of the sewing machine in a graph format.

FIG. 6 is a conceptual diagram for setting a transfer function of the noise control device.

[Explanation of symbols]

 201 Noise Control Unit 202 Vibration Detection Sensor 203 Sewing Machine Main Body 204a Speaker 204b Speaker 205a Microphone 205b Microphone 301 LMS Adaptive Algorithm 302 Adaptive Filter

Claims (4)

[Claims] 1. A vibration detecting means for detecting vibration generated by driving the sewing machine, a plurality of noise detecting means for detecting noise generated by driving the sewing machine at a plurality of observation positions, and a noise detecting means for detecting noise generated by driving the sewing machine. Based on the signals detected by the plurality of control sound generation means capable of generating the control sound for canceling and the vibration detection means and the signals detected by each of the noise detection means, each of the noise detection means detects at each of the plurality of observation positions. A noise control device for a sewing machine, comprising: a control unit that controls the control sound generation unit so that the generated noise is minimized. 2. The noise control device for a sewing machine according to claim 1, wherein the vibration detecting means is provided in the sewing machine main body. 3. A transfer function setting means for presetting a transfer function indicating a characteristic or a mutual interference characteristic of a mechanical system part of the sewing machine, an electric system part for control, etc., the control means comprising: the vibration detecting means; 2. The noise control device for a sewing machine according to claim 1, wherein the control sound generating means is controlled based on the transfer function together with each detection signal of the noise detecting means. 4. A storage means for storing the transfer function set by the transfer function setting means is provided, and when the control means is operated, the transfer function stored in the storage means is read and applied to the control means. The noise control device for a sewing machine according to claim 3, wherein the noise control device is configured to perform.