JP2544899B2 - Active noise control system - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an active noise control system for silencing noise by radiating sound waves having the same sound pressure and opposite phase to noise in a noise propagation passage such as an air conditioning duct.

[0002]

2. Description of the Related Art Generally, as one method for silencing noise propagating in an air-conditioning duct, a passive silencing method has been mainly used, such as a method of absorbing sound by a sound absorbing material stuck inside the duct. However, there are problems such as pressure loss and size. On the other hand, active muffling methods have been actively researched in which sound waves of the same sound pressure and opposite phase are simultaneously radiated into the duct with respect to the sound waves of noise propagating in the duct, and the sound waves are muted by the interference of both sound waves. However, many problems still remain.

FIG. 6 is a structural explanatory view showing a conventional active noise control system. That is, the information of noise propagating in the duct 61 is taken in as an analog electric signal by the first microphone 62, and further amplified by the microphone amplifier 65.

The analog electric signal amplified by the microphone amplifier 65 is first passed through a frequency cutoff filter 68. After that, the analog-digital converter 7
1 into a digital electric signal.

In this way, the digital electric signal containing only the signal in the low frequency region to be silenced is input to the arithmetic unit 79. In addition, the duct 61 due to the operation of the system
From the noise source, the muffling speaker 63
It is detected by the second microphone 64 installed further downstream. Duct 6 obtained from this second microphone 64
The analog electric signal, which is the information of the muffling situation in 1, is also the first
The analog electric signal obtained from the microphone 62 is passed through the microphone amplifier 67, the frequency cutoff filter 70 for analog signal processing, and the analog-digital converter 73.

From the second microphone 64, information on how much the propagation noise in the duct 61 has been silenced by the operation of the system is input. The calculation unit 79 takes in the information and calculates an optimum coefficient based on the adaptive control algorithm 78 so that the signal always approaches zero, and uses it as a filter coefficient of the silence signal generation filter 77 in the input signal from the first microphone 62. Perform a convolution operation.

In this way, the arithmetic unit 79 calculates various coefficients as the first coefficient.
The input signal from the microphone 62 is subjected to a convolution operation to create a mute digital electric signal. This mute digital electric signal is converted into an analog electric signal by the digital-analog converter 72, and passed through the frequency cutoff filter 69 by analog signal processing. Finally, the muffling analog electric signal containing only the signal in the low frequency region to be muted. Turn it into a signal.

This muffling analog electric signal is amplified by the power amplifier 66 to drive the muffling speaker 63 and radiate the muffling sound wave into the duct 61. In this way, various signal processing is performed, and the sound wave for silencing having the same sound pressure and the opposite phase to the noise propagating in the duct 61 is radiated into the duct 61. The sound deadening sound waves radiated interfere with the noise sound waves propagating in the duct 61 and cancel each other out, resulting in a sound deadening effect.

When the noise propagating in the duct 61 is actively silenced, the signal of the noise in the duct 61 from the first microphone 62 and the signal of the information of the silencing condition from the second microphone 64 are taken in Various processing is performed on one signal to create a muffling signal for driving the muffling speaker 63. Among the various processes, the one that is indispensable is "to cut off signals in the frequency domain that is not subject to noise reduction in order to process only the signals in the frequency domain that are subject to noise reduction."
There is a process called.

Even when the muffling signal generated by the arithmetic unit 79 is reproduced by the muffling speaker 63, in order to reproduce only the sound wave in the frequency region targeted for muffling, the signal in the frequency region not targeted for muffling is selected. The process of shutting off is required.

The frequency cut-off filters 68, 69, 70 perform the processing, and the input signals from the first microphone 62 and the second microphone 64 and the output signal from the arithmetic unit 79 are processed by the frequency cut-off filters 68, 69, 70. Must pass 70.

Duct 6 obtained with the first microphone 62
Information of noise propagating in the first microphone 6 and the second microphone 6
The information on the muffling situation obtained in 4 is an analog electric signal. Conventionally, as shown in FIG. 6, only analog signal processing is performed on the analog electric signals captured by the microphones 62 and 64 and amplified by the amplifiers 65 and 67 to block signals in the frequency domain that are not subject to noise reduction. I've been Further, in order to reproduce only the sound wave in the frequency range to be muted by the muffling speaker 63, the signal in the frequency range not to be muted must be cut off. The mute digital electric signal is converted to a mute analog electric signal by the digital-analog converter 72 and then subjected to analog signal processing.

That is, the frequencies which are not the target of the noise reduction are cut off by passing only the frequency cut-off filters 68, 69 and 70 by analog signal processing. By the way, the conventional passive noise reduction method exerts a sufficient noise reduction effect on noise in the middle and high frequency regions, but does not achieve the same sufficient noise reduction effect on noise in the low frequency region. It was difficult. (Excessive use of the muffling duct using glass wool increases the air resistance inside the duct, which increases the load on the air conditioner. Also, unless the muffling duct is quite large, it is not possible to effectively absorb the sound in the low frequency range. Therefore, it is considered that the particularly effective frequency range for actively muting noise is the low frequency range.

However, when the signal in the middle / high frequency region is cut off by analog signal processing (when only the signal in the low frequency region is taken in), "time required for processing = time delay"
And “frequency cutoff accuracy = frequency cutoff characteristic” have contradictory relationships. When the frequency cutoff characteristic is made steep as shown in FIG. 5 to accurately cut off a signal of f ′ Hz or higher, the time delay becomes large, and conversely, when the time delay is made small, the frequency cutoff characteristic becomes gentle as shown in FIG. Since the signal of f'Hz or more is also included, the accuracy of frequency cutoff deteriorates.

After the sound wave of noise is detected by the first microphone 62, the speaker 61 for silencing is further placed in the duct 61.
Propagate to the installation position of. In the system, various signal processing is performed while the noise sound wave propagates from the first microphone 62 to the noise reduction speaker 63, and the noise reduction sound wave is emitted from the noise reduction speaker 63. Therefore, if the time required for signal processing in the system becomes longer (the delay time becomes longer), the distance between the first microphone 62 and the muffling speaker 63 must be increased. (It is meaningless if the sound wave to be silenced passes through the installation position of the silencing speaker 63 during the signal processing.) As in the prior art, only the signal in the low frequency range is obtained by the method by analog signal processing. In order to take out accurately, the frequency cutoff characteristic must be made steep and the time delay becomes large, so that the distance between the first microphone 62 and the muffling speaker 63 must be made large, and the system becomes large.

[0016]

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and makes it possible to obtain an accurate frequency cutoff with a small time delay, perform highly accurate noise control, and perform duct operation. An object of the present invention is to provide an active noise control system in which a compact system can be provided by reducing a distance between a sensor for detecting internal noise and a muffling speaker.

[0017]

In order to achieve the above object, the present invention provides a means for cutting off signals in the middle and high frequency regions, which are not subject to noise reduction, by using not only analog signal processing but also frequency processing by digital signal processing. By also using a cutoff filter, the time delay is reduced and frequency cutoff is performed accurately.

[0018]

With the above means, if a frequency cutoff filter by analog signal processing and a frequency cutoff filter by digital signal processing are used together to cut off signals in the middle and high frequency regions that are not subject to noise reduction, only analog signal processing is required. The frequency cutoff can be performed with high accuracy without a large time delay as when the frequency cutoff is performed.

[0019]

Embodiments of the present invention will now be described in detail with reference to the drawings. The overall flow of the active noise control system of the present invention will be described with reference to FIG. First microphone (first
Information of noise propagating in the duct 1 is taken in by a sensor 2 as an analog electric signal and further amplified through a microphone amplifier 5.

Since the present system targets the noise in the low frequency range to be silenced, it is necessary to pass the analog electric signal of this noise through a frequency cutoff filter for blocking the signals in the medium and high frequency ranges which are not the target. .

Conventionally, only the frequency cut-off by analog signal processing has been performed on the analog electric signal amplified by the microphone amplifier. (See FIG. 6) However, if the frequency cutoff characteristic is sharpened in order to accurately cut the frequency in the middle and high frequency regions by the frequency cutoff filter by analog signal processing, a large time delay occurs.

Therefore, in the present embodiment, two types of frequency cutoff filters 8, 9, 10 by analog signal processing and frequency cutoff filters 14, 15, 16 by digital signal processing are used to cut off signals in the middle and high frequency regions. The frequency cutoff filter of is used.

The frequency cutoff filters 14, 15 and 16 based on digital signal processing store an input signal from a past to the present for a certain fixed time in a memory and perform a convolution calculation of a frequency cutoff coefficient to the middle and high frequencies. Since the latest signal subjected to frequency cutoff processing is calculated, steep frequency cutoff characteristics can be obtained without much time delay.

The analog electric signal amplified by the microphone amplifier 5 is first passed through a first frequency cutoff filter 8 having a gentle frequency cutoff characteristic so as to have no time delay. After that, it is converted into a digital electric signal through the analog-digital converter 11.

The digital electric signal whose frequency is cut off gently is passed through the second frequency cut-off filter 14 by digital signal processing which can obtain a sharp frequency cut-off characteristic without time delay. In this way, the digital electric signal including only the signal in the low frequency region to be silenced is input to the arithmetic unit 19.

Further, the situation of noise reduction in the duct 1 due to the operation of the system is detected by the second microphone (second sensor) 4 installed downstream of the noise reduction speaker 3 as seen from the noise source. Similarly to the analog electric signal obtained from the first microphone 2, the analog electric signal obtained from the second microphone 4 which is the information on the muffling condition in the duct 1 is also cut off by the microphone amplifier 7 and the analog signal processing. It is passed through a filter 10, an analog-digital converter 13 and a frequency cutoff filter 16 by digital signal processing.

Then, the digital electric signal of only the signal in the low frequency region to be muted is input to the arithmetic unit 19.
In the calculation unit 19, the muffling speaker 3 creates a muffling signal for emitting sound waves having the same sound pressure and opposite phase to the noise propagated in the duct 1. However, simply the first microphone 2
The noise in the duct 1 cannot be silenced only by making the noise signal obtained from the above into the same sound pressure antiphase. Therefore, in the calculation unit 19, the “sound-reducing speaker 3 radiates a sound wave having the same sound pressure and an opposite phase to the sound wave of the noise in the duct 1 at the place where the speaker 3 is installed to cause the sound wave interference. The calculation described below is performed so as to create a silencing signal such that the noise in 1 becomes zero.

In the system, as shown in FIG. 2, when a sound wave propagates in a duct or a signal passes through a circuit, it is affected by its own transfer function such as a time delay and a change in frequency characteristic (later. explain in detail). Therefore,
Correction must be made for transfer functions that have a particularly adverse effect on noise control. Therefore, the calculation unit 19 performs a convolution calculation on the noise digital electric signal obtained from the first microphone 2 with the coefficient for correcting the transfer functions measured in advance before the system is operated.

From the second microphone 4, information about how much the noise propagating in the duct 1 has been silenced by the operation of the system is input. The calculation unit 19 takes in the information and calculates an optimum coefficient based on the adaptive control algorithm 18 such that the signal always approaches zero, and uses it as the filter coefficient of the mute signal generation filter 17 for the input signal from the first microphone 2. Perform a convolution operation.

In this way, the arithmetic unit 19 calculates various coefficients as the first coefficient.
The input signal from the microphone 2 is subjected to a convolution calculation to create a mute digital electric signal. After that, the sound deadening speaker 3 is driven by the sound deadening digital electric signal to radiate the sound deadening sound wave. However, since only the sound wave in the low frequency region to be muted is radiated into the duct 1, it is not a sound deadening target. Signals in the frequency domain of must be cut off. Therefore, this noise-eliminating digital electric signal is passed through a steep frequency cutoff filter 15 by digital signal processing. Further, it is converted into an analog electric signal by the digital-analog converter 12 and passed through the gradual frequency cutoff filter 9 by analog signal processing to finally become a muffling analog electric signal containing only the signal in the low frequency region to be muted. In some cases, such as when high-frequency electrical noise is not a problem, the frequency cutoff filter 9 based on analog signal processing may be omitted.

The mute analog electric signal is amplified by the power amplifier 6 to drive the muffling speaker 3 to radiate the muffling sound wave into the duct 1. In this way, various signal processing is performed, and the sound wave for silencing having the same sound pressure and the opposite phase to the noise propagating in the duct 1 is radiated into the duct 1.
The emitted sound wave for silencing interferes with the noise sound wave propagating in the duct 1 and cancels each other, resulting in a highly accurate sound deadening effect.

Now, the in-system transfer function will be described in detail. A transfer function that needs special consideration in an active noise control system, a transfer function “D, E” in a circuit through which a signal obtained from the first microphone 2 passes before being emitted from the muffling speaker 3, a muffling speaker The transfer function "B" in the duct 1 through which the sound wave radiated from the sound wave 3 reaches the first microphone 2, and the duct through which the sound wave radiated from the muffling speaker 3 reaches the second microphone 4. The transfer function "C" in 1 will be described with reference to FIG.

First, the function "D, E" will be described.
The information on the noise in the duct 1 includes the first microphone 2, the microphone amplifier 5, the frequency cutoff filter 8 by analog signal processing, the analog-digital converter 11, the frequency cutoff filter 14 by digital signal processing, and the calculation unit 1.
9. Digital signal processing frequency cut-off filter 1
5, the digital-analog converter 12, the power amplifier 6, and the muffling speaker 3 are processed in this order and radiated into the duct 1 as a muffling sound wave.

Since a processing time is required when processing the signal, a time delay occurs until the signal detected by the first microphone 2 is emitted from the muffling speaker 3. If the sound wave to be silenced passes through the noise elimination speaker 3 during the delay time generated by this processing, there is no effect of noise reduction. Therefore, this transfer function “D, E” is important for determining the distance between the first microphone 2 and the muffling speaker 3.

Next, regarding the transfer function "B". The sound deadening sound waves emitted from the sound deadening speaker 3 propagate not only in the downstream direction of the duct 1 (direction of the second microphone 4) but also in the upstream direction (direction of the first microphone 2). Therefore, the first microphone 2 detects not only the sound wave of the noise information in the duct 1 but also the sound wave for muff emitted by the sound muffling speaker 3. This causes howling and measurement errors, which adversely affects the convergence accuracy in coefficient calculation.

Finally, regarding the transfer function "C". The muffling sound wave emitted from the muffling speaker 3 interferes with the noise in the duct 1 and propagates in the downstream direction of the duct 1. The interference sound wave is detected by the second microphone 4 and the coefficient is calculated by the calculator 19. Therefore, unless the time delay generated in this path is taken into consideration, the convergence accuracy in the coefficient calculation is adversely affected.

The transfer functions "D, E" can be avoided by considering the distance between the first microphone 2 and the muffling speaker 3. Further, the transfer function “B” and the transfer function “C” can be avoided by electrically processing in the system.

To electrically process the transfer functions “B” and “C”, specifically, the following processes are performed as “H” and “I” in FIG. Regarding the transfer function “B”, the transfer function from the muffling speaker 3 to the first microphone 2 is measured in advance. It is convoluted with the muffling signal generated by the calculation unit 19 to artificially generate a signal as if the muffling signal propagated from the muffling speaker 3 to the first microphone 2. A process “H” is performed in which the artificially created signal is subtracted from the signal of the propagation noise in the duct 1 obtained from the first microphone 2.

Regarding the transfer function "C", the transfer function from the muffling speaker 3 to the second microphone 4 is measured in advance, and the delay time in this path is read. A process "I" of convolving this delay time with the signal of the propagation noise in the duct 1 obtained from the first microphone 2 is performed.
That is, the process "I" is performed in which the muffling signal generating filter 17 is updated with the coefficient for the time required from the emission of the muffling sound wave to the detection of the sound wave in the muffling state. It should be noted that the processes “H” and “I” are both performed by the arithmetic unit 19.

[0040]

As described above, according to the present invention, when a frequency cutoff filter by analog signal processing and a frequency cutoff filter by digital signal processing are used together and only a signal in a low frequency region of interest is taken out, processing is performed. The time delay when performing is reduced, and processing can be performed with a sharp frequency cutoff characteristic. Since the time delay can be shortened while performing the silencing with high accuracy, the distance between the first microphone and the silencing speaker is shortened, and the entire system becomes compact.

[Brief description of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a transfer function of each part in the embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a process for correcting a transfer function that needs special consideration among the transfer functions of the respective parts in FIG.

FIG. 4 is a characteristic diagram showing a gradual frequency cutoff characteristic in which f ′ is a cutoff frequency.

FIG. 5 is a characteristic diagram showing a steep frequency cutoff characteristic with a cutoff frequency f ′.

FIG. 6 is a block diagram showing a conventional active noise control system.

[Explanation of symbols]

1 ... Duct, 2 ... 1st microphone, 3 ... Silent speaker, 4 ... 2nd microphone, 5 ... Microphone amplifier,
6 ... Power amplifier, 7 ... Microphone amplifier, 8 ... Frequency cutoff filter by analog signal processing, 9 ... Frequency cutoff filter by analog signal processing, 10 ... Frequency cutoff filter by analog signal processing, 11 ... Analog-digital converter, 12 ... Digital-analog converter,
Reference numeral 13 ... Analog-digital converter, 14 ... Frequency cutoff filter by digital signal processing, 15 ... Frequency cutoff filter by digital signal processing, 16 ... Frequency cutoff filter by digital signal processing, 17 ... Silence signal generation filter, 18 ... Adaptive control algorithm , 19 ... Operation unit.

Claims (1)

(57) [Claims] 1. An active noise control system for simultaneously radiating sound waves having the same sound pressure and antiphase to the sound wave of noise propagating in the duct into the duct, and suppressing the sound propagating in the duct by the interference of both sound waves. A first sensor that detects propagating noise and converts an acoustic signal into an analog electric signal; a second sensor that detects a muffling situation in the duct and converts an acoustic signal into an analog electric signal; the first sensor and the first sensor (2) A frequency cutoff filter by analog signal processing that has a gentle frequency cutoff characteristic that allows passage of only a signal in the frequency range to be silenced from the analog electric signal from the sensor, and an analog from the frequency cutoff filter by the analog signal processing An analog-digital converter for converting an electric signal into a digital electric signal, and the analog-digital converter Frequency cut-off filter by digital signal processing having a steep frequency cut-off characteristic that passes only the signal in the frequency domain to be muted from the digital electric signal of, and muting from digital electric signal from the frequency cut-off filter by the digital signal processing And a frequency by digital signal processing having a steep frequency cutoff characteristic that passes only a signal in the frequency range to be silenced from the silencing digital electrical signal created by the computing unit. A cutoff filter, a digital-analog converter for converting a muffling digital electric signal from the frequency cutoff filter by the digital signal processing into an analog electric signal, and a target of muffling from the muffling analog electric signal from the digital-analog converter Frequency domain A frequency cutoff filter by analog signal processing having a gentle frequency cutoff characteristic that passes only, and converts the mute analog electric signal from the frequency cutoff filter by the analog signal processing into an acoustic signal, and a sound wave for muffling in the duct. An active noise control system comprising: a radiating speaker.