KR101696597B1 - Noise control method - Google Patents

Abstract

A noise control method according to an embodiment of the present invention includes: a first output step of receiving a reference signal to generate a first anti-noise signal and outputting a first sound of the first anti-noise signal; A first sensing step of sensing a first noise while the first sound is output; Generating a second anti-noise signal by adding the first anti-noise signal to the first anti-noise signal, and generating a second anti-noise signal by adding the second anti-noise signal to the first anti- A second output step of outputting a second sound to the second output step; A second sensing step of sensing a second noise while the second sound is output; And generating a third anti-noise signal by using the delay time error signal when the second noise is detected by comparing the additional signal with a signal based on the second noise to generate a delay time error signal, And a third output step of outputting a third sound for the third counter-noisy signal.

Description

{NOISE CONTROL METHOD}

The present invention relates to a noise control method.

Generally, there are noise source control, passive noise control, and active noise control (ANC).

The noise source control method is a method in which a noise source, which is the first sound source that causes noise, is removed or the noise is reduced. For example, the noise source control method is to remove any vibration sources that cause noise, remove them if they can be removed, or redesign or install isolation to reduce vibration as much as possible.

The passive noise control method is a method of reducing noise by placing a sound absorbing material or a sound insulating material on a path through which noise is transmitted from a noise source. For example, a passive noise control method is to install a sound barrier to block road noise.

The active noise control method is a method of reducing noise at a desired target area (or target point) by using a second control source or sound sources. For example, the active noise control method is to detect the noise measured at the microphone and output the opposite noise signal through the speaker to control the noise. The active noise control method includes an active noise control method using an open loop control, Filtered-X Least Mean Square (FXLMS), Filtered-X Normalized Least Mean Square (FXNLMS), Filtered-U Recursive Least Mean Square (FULMS) And active noise control method using multiple frequencies.

Here, open-loop control is effective when the system is very stable, but there are uncertainties in the majority of systems, which may result in poor repeatability and thus noise reduction performance.

In an environment where the temperature or pressure around the system greatly changes, the speed of sound changes. This is because the delay time of the system is changed. In this environment, in the conventional active noise control method, an error occurs between the noise signal and the opposite noise signal, so that the noise can not be reduced.

In addition, in the conventional active noise control method, an additional path transfer function model is updated by an LMS (Least Mean Square) method using an error signal measured in a microphone.

Specifically, in the conventional active noise control method, a noise signal is sensed in a first microphone, an error signal is sensed in a second microphone, a reference signal based on a sensed noise signal is filtered in an initial model of an additive path transfer function, A signal is generated, the reference noise signal is multiplied by an error signal and a convergence element to generate an improved adaptive filter, an improved adaptive filter and an adaptive filter are added to generate a new adaptive filter, a noise signal is filtered into a new adaptive filter, Signal.

However, if the delay time of the system changes in the environment where the temperature or pressure around the system changes greatly, the additional path transfer function changes due to the uncertainty in the system, and an error occurs between the additional path transfer function model and the actual transfer function. At this time, if the additional path transfer function changes within the stability condition of the system, the system can reliably reduce the noise. However, if the additional path transfer function changes outside the stability condition of the system, The stability of the noise control and the convergence of the noise are deteriorated, and further, the problem of the system can be diverted.

SEN M. KUO AND DENNIS R. MORGAN, "Active Noise Control: A Tutorial Review" PROCEEDINGS OF THE IEEE, VOL. 87, NO, 6, JUNE 1999

It is an object of the present invention to provide a noise control method capable of calculating a delay time of a system in an environment where ambient temperature or pressure varies and compensating an error between a noise signal and an opposite noise signal to control noise.

Further, in the embodiment, the delay time of the system is calculated in an environment in which the ambient temperature or pressure changes, and the noise control is performed by compensating the difference between the previously measured additional path transfer function model and the actual transfer function And a method thereof.

A noise control method according to an embodiment of the present invention includes: a first output step of receiving a reference signal to generate a first anti-noise signal and outputting a first sound of the first anti-noise signal; A first sensing step of sensing a first noise while the first sound is output; Generating a second anti-noise signal by adding the first anti-noise signal to the first anti-noise signal, and generating a second anti-noise signal by adding the second anti-noise signal to the first anti- A second output step of outputting a second sound to the second output step; A second sensing step of sensing a second noise while the second sound is output; And generating a third anti-noise signal by using the delay time error signal when the second noise is detected by comparing the additional signal with a signal based on the second noise to generate a delay time error signal, And a third output step of outputting a third sound for the third counter-noisy signal.

Wherein generating the first anti-noise signal in the first output stage is to apply the reference signal to a first adaptive filter to generate the first anti-noise signal, and in the third output stage, The generation of the opposite noise signal may comprise generating the reference noise signal by applying the reference signal to the updated path transfer function model updated by the delay time error signal and multiplying the reference noise signal by the signal by the second noise Generate a second adaptive filter by adding the generated enhancement adaptive filter and the first adaptive filter, and apply the reference signal to the second adaptive filter to generate the third anti-noise signal.

Generating a white noise signal before the first output step and outputting a fourth sound for the white noise signal; And generating an additive path transfer function initial model of the additive path transfer function model using the transfer signal by the fifth sound sensed during the output of the fourth sound and the white noise signal .

Here, generating the delay time error signal in the third output step may be a step of high-pass filtering the signal due to the second noise, and then generating the delay time error signal by comparing the signal with the additional signal .

Here, the generating of the first anti-noise signal in the first output step may include analyzing the frequency of the reference signal and selecting the frequency to be controlled among the frequencies of the reference signal to generate the first anti- Lt; / RTI >

Here, the additional signal may be a frequency signal other than an audible frequency.

Further, in the noise control method according to the embodiment, the first microphone may include a first sample, a second sample, , The first reference noise for the nth sample, the second reference noise, ... , The nth reference noise is detected, and the second microphone detects the first noise, the second noise, ... A sensing step of sensing n-th noises; The first reference noise, the second reference noise, and the second reference noise detected by the first microphone. A first reference signal based on the n-th reference noise, a second reference signal, , An n-th reference signal, the first noise detected by the second microphone, the second noise, ... A first error signal due to the n-th noise, a second error signal, A first additional signal, a second additional signal, and a second additional signal to calculate a change in delay time when an n-th error signal is transmitted. An additional signal generation step of generating an n-th additional signal; The first reference signal, the second reference signal, , The nth reference signal, the first error signal, the second error signal, ... , The n-th error signal and the first additional signal, the second additional signal, ... A second anti-noise signal, a second anti-noise signal, and a second anti-noise signal using the n-th additional signal. An opposite noise signal generating step for generating an nth reverse noise signal; And the first counter-noise signal, the second counter-noise signal, , The n-th reverse noise signal, the first additional signal, the second additional signal, ... A first sound for the first anti-noise signal, a second sound for the second anti-noise signal, And outputting the n-th sound for the n-th reverse noise signal.

Here, the opposite noise signal generating step may include: generating the first additional signal, the second additional signal, , The n-th additional signal, the first error signal, the second error signal, ... , And compares the n-th error signal to obtain a first delay time error signal, a second delay time error signal, ... A delay time error signal generation step of generating an nth delay time error signal; The first delay time error signal, the second delay time error signal, And generates an updated additive path transfer function model by updating the initial model of the additive path transfer function of the additive path transfer function model in real time using the nth delay time error signal, , ... And applying the n-th reference signal to the updated additive path transfer function model to generate a first reference noise signal, a second reference noise signal, A reference noise signal generation step of generating an n-th reference noise signal; The first error signal, the second error signal, ... , The nth error signal is the first reference noise signal, the second reference noise signal, A first enhancement adaptive filter, a second enhancement adaptive filter, , An nth improved adaptive filter is generated, and the first improved adaptive filter, the second improved adaptive filter, , The nth improved adaptive filter, the first adaptive filter, the second adaptive filter, A second adaptive filter, a third adaptive filter, and an n-th adaptive filter. An adaptive filter generating step of generating an (n + 1) -th adaptive filter; And the first reference signal, the second reference signal, , The n-th reference signal to the second adaptive filter, the third adaptive filter, A first addition signal, a second addition signal, and a second addition signal after being applied to the (n + 1) th adaptive filter. , An n-th additional signal is added to the first anti-noise signal, the second anti-noise signal, ... And a reference signal applying step of generating the n-th reverse noise signal, wherein in the generating of the adaptive filter, the second adaptive filter, the third adaptive filter, , Generating the (n + 1) -th adaptive filter comprises: generating the second error signal, , And generates the n-th error signal in a direction in which the squared value of the n-th error signal is minimized.

Here, before the first microphone detects the first reference noise and the second microphone detects the first noise, a white noise signal is generated and a sound for the white noise signal is output ; And generating the additional path transfer function initial model using the sound signal and the white noise signal sensed by the second microphone.

Here, the delay time error signal generation step may include: generating the first error signal, the second error signal, An n-th error signal, and outputs the first additional signal, the second additional signal, The first delay time error signal, the second delay time error signal, and the second delay time error signal, , And may generate the n-th delay time error signal.

Here, the opposite noise signal generating step may include: generating the first reference signal, the second reference signal, , Analyzes the frequency of the n-th reference signal, and outputs the analyzed first reference signal, second reference signal, , A frequency to be controlled is selected from the frequency of the nth reference signal, and the first anti-noise signal, the second anti-noise signal, , And generating the n-th reverse noise signal.

Here, the first additional signal, the second additional signal, ... , And the n-th additional signal may be a frequency signal other than an audible frequency.

According to the noise control method of the embodiment, when the first noise is sensed, an additional signal is generated to generate a delay time error signal. Therefore, an unstable supplementary path transfer function It is possible to control the noise even in an environment having

According to the noise control method of the embodiment, the initial model of the additional path transfer function is updated in real time by the delay time error signal, the second adaptive filter is generated by using the reference noise signal and the signal by the second noise, Signal is applied to the second adaptive filter to generate the third opposite noise signal, noise can be controlled in any environment in which the delay time of the additional path transfer function linearly changes according to the temperature or the pressure.

According to the noise control method according to the embodiment, since the white noise signal is generated and the sound for the white noise signal is output to generate the initial model of the additive path transfer function, the additive path transfer function It is possible to simply generate the initial model of the additive path transfer function.

According to the noise control method of the embodiment, since the signal by the second noise is high-pass filtered and then compared with the additional signal to generate the delay time error signal, the audible frequency band included in the signal by the second noise The change of the delay time of the additional path transfer function can be calculated only by the ultrasonic wave band included in the signal by the second noise. Thus, the calculation of the change in the delay time of the additive path transfer function can be further simplified.

According to the noise control method according to the embodiment, since the opposite noise signal is generated by analyzing the frequency of the reference signal and selecting the frequency to be controlled out of the frequency of the analyzed reference signal, only the desired frequency noise can be reduced, The process of calculating the change in the delay time using the additional signal can be simplified.

According to the noise control method according to the embodiment, since a frequency signal other than the audible frequency is used as the additional signal, it is possible to detect a change in the delay time without generating additional noises generated when the audible frequency signal is used as the additional signal, You can compensate.

According to the noise control method according to the embodiment, the first additional signal, the second additional signal, , An n-th adder generates a first delay time error signal, a second delay time error signal, ... , The n-th delay time error signal is generated, so that the noise can be controlled and the noise can be completely eliminated even in an environment having an unstable additional path transfer function in which the ambient temperature or pressure changes and the speed of sound changes.

According to the noise control method according to the embodiment, the initial model of the additional path transfer function includes a first delay time error signal, a second delay time error signal, ... And the n-th delay time error signal, noise can be controlled in any environment in which the delay time of the additional path transfer function linearly changes according to the temperature or the pressure.

According to the noise control method according to the embodiment, since the sound for the white noise signal is output to generate the initial model of the additive path transfer function, the additive path transfer function initial model Can be simply generated.

According to the noise control method according to the embodiment, the first error signal, the second error signal, , The nth error signal is subjected to high pass filtering, and then the first additional signal, the second additional signal, A first delay time error signal, a second delay time error signal, , And generates an n-th delay time error signal, the first error signal, the second error signal, , The audio frequency band included in the nth error signal is filtered to obtain a first error signal, a second error signal, ... , It is possible to calculate the change of the delay time of the additional path transfer function more simply by using only the ultrasonic wave band included in the nth error signal.

According to the noise control method according to the embodiment, the first reference signal, the second reference signal, , Analyzes the frequency of the nth reference signal, and outputs the analyzed first reference signal, second reference signal, ... , A frequency to be controlled is selected from frequencies of the n-th reference signal to generate a first anti-noise signal, a second anti-noise signal, , The n-th reverse noise signal is generated. Therefore, only the desired frequency noise can be reduced, and the process of calculating the delay time change using the n-th additional signal can be simplified.

According to the noise control method according to the embodiment, a frequency signal other than the audible frequency is referred to as a first additional signal, a second additional signal, , So that the audio frequency is used as the first additional signal, the second additional signal, ... , It is possible to compensate for the change of the delay time while sensing the additional noise generated when the n-th additional signal is used.

1 is a general schematic diagram of an algorithm for performing a noise control method according to an embodiment.
FIG. 2 is a schematic diagram for explaining an example of the active noise control algorithm shown in FIG. 1. FIG.
FIG. 3 is a first embodiment of the additional path transfer function model shown in FIG.
4 is a second embodiment of the additional path transfer function model shown in FIG.
5 is a schematic diagram of the initial model of the supplementary path transfer function shown in FIG.
6 is a schematic diagram of the supplementary path transfer function update algorithm shown in FIG.
FIG. 7 is a schematic diagram of the delay time analyzer shown in FIG. 2. FIG.
FIG. 8 illustrates an example of analyzing a change in delay time in the delay time analyzer shown in FIG.
9 is a flowchart for explaining a noise control method according to the embodiment.
10 is a flowchart for explaining a method of generating an initial model of an additional path transfer function.
11 is a flowchart of an algorithm for performing the noise control method according to the embodiment.

The following detailed description refers to the accompanying drawings which illustrate, by way of example, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the present invention may be different but need not be mutually exclusive. For example, certain features, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the invention in connection with one embodiment. It is also to be understood that the location or arrangement of the individual components within each disclosed embodiment may be varied without departing from the spirit and scope of the present invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is to be limited only by the appended claims, along with the full scope of equivalents to which such claims are entitled, if properly explained. In the drawings, like reference numerals refer to the same or similar functions throughout the several views.

<Embodiment>

1 is a general schematic diagram of an algorithm for a noise control method according to an embodiment.

Referring to FIG. 1, an algorithm for a noise control method according to an embodiment includes an active noise control (ANC) algorithm 100, a secondary path transfer function 200, And a sensing unit 300.

The active noise control algorithm unit 100 receives reference signals 10, an additional signal 20 and an error signal 60 and outputs an opposite noise signal 30 .

Specifically, the active noise control algorithm unit 100 can generate and output the first anti-noise signal 30 using the reference signal 10, and the first noise signal 60 due to the first noise is input It is possible to generate and output the second opposite noise signal 30 using the reference signal 10 and the additional signal 20. When the second noise signal 60 due to the second noise is input, The third counter-noise signal 30 may be generated and output using the reference signal 10 and the second error signal 60.

The additional path transfer function unit 200 receives the first to third opposite noise signals 30 and outputs the sound 40. Here, the additional path transfer function unit 200 may include an audio output means such as a speaker.

Specifically, the additive path transfer function unit 200 can output the first sound 40 for the first anti-noise signal 30 when the first anti-noise signal 30 is input, When the noise signal 30 is inputted, it is possible to output the second sound 40 for the second opposite noise signal 30, and when the third opposite noise signal 30 is inputted, (40) to the second sound source (30).

The sensing unit 300 may sense the first and second noises and may output the first error signal 60 and the second error signal 60. Here, the sensing unit 300 may include a sensor such as an error microphone or the like.

Specifically, the sensing unit 300 may sense the first noise while the first sound 40 is being output, output the first error signal 60, and may output the second error signal 60 while the second sound 40 is output. It is possible to detect the noise and output the second error signal 60. The reference signal 10 can be obtained by placing a microphone in a noise transmission path, by using an accelerometer capable of sensing vibration, or by using a signal generator or the like.

The additional signal 20 is a periodic signal repeated at a constant period. The additional signal 20 may be generated as a signal outside the audible frequency to calculate a change in the delay time.

The first to third anti-noise signals 30 are signals having an opposite phase (180 degrees) to the signal of the noise 50. The first to third counter-noise signals 30 may be output values of the algorithm automatically calculated for each sample to reduce the power of the noise.

The first to third sounds 40 are sounds for the first to third opposite noise signals 30. The first to third sounds 40 may overlap with the noise 50.

Noise 50 is generally a sound that makes you feel uncomfortable. Noise 50 may be noise generated by noise sources (engines, compressors, motors, fans, propellers, etc.).

The first and second error signals 60 are generated when the noise 50 is less than the noise 50 when the first to third sounds 40 and 50 are overlapped. The first and second error signals 60 are generated when the sensing unit 300 senses the first to third sounds 40.

In addition, another example of the algorithm for the noise control method according to the embodiment will be described. First, the first microphone senses first through n-th reference noises, and the sensing unit 300 senses can do. Here, the first microphone may be a microphone placed in the noise transmission path described above, and the sensing unit 300 may include an error microphone. At this time, the error microphone may be a second microphone.

Specifically, when the first microphone senses the first reference noise, the sensing unit 300 senses the first noise, the first microphone again senses the second reference noise, and the sensing unit 300 senses the second reference noise. After sensing the noise, the first microphone senses the n th reference noise, and the sensing unit 300 senses the n th noise. At this time, the above-described process can be repeated until the sensing unit 300 can not detect the noise of the fastest order among the second to n-th noises.

The active noise control algorithm unit 100 includes first through n-th reference signals 10, first through n-th error signals 60 and first through n-th reference signals 60, Th to (n) th signal 20, and generates and outputs the first to the (n) -th reverse noise signals 30.

Specifically, the active noise control algorithm unit 100 receives the first reference signal 10, the first additional signal 20, and the first error signal 60 and generates a first anti-noise signal 30 The second reference signal 10, the second additional signal 20 and the second error signal 60 and generates and outputs the second opposite noise signal 30, The n-th additional signal 20 and the n-th error signal 60, and generates and outputs the n-th reverse noise signal 30.

The additional path transfer function unit 200 receives the first to the n-th opposite noise signals 30 and outputs the first to n-th sounds 40.

Specifically, the additive path transfer function unit 200 outputs the first sound 40 for the first anti-noise signal 30 when the first anti-noise signal 30 is input, When the n th reverse noise signal 30 is input, the second sound 40 for the second opposite noise signal 30 is output, and when the n th reverse noise signal 30 is input, The n-th sound 40 can be output.

The sensing unit 300 may sense the first and second noises and may output the first error signal 60 and the second error signal 60.

Specifically, the sensing unit 300 may sense the first noise while the first sound 40 is being output, output the first error signal 60, and may output the second error signal 60 while the second sound 40 is output. It is possible to detect the noise and output the second error signal 60.

2 is a schematic diagram for explaining an example of the active noise control algorithm unit shown in FIG. 1 and 2, an active noise control algorithm unit 100 according to an embodiment includes adaptive filters 110, a group delay analyzer 120, an additive path transfer function model a secondary path transfer function model 130 and an update algorithm 140.

The adaptive filter 110 receives the reference signal 10 and can output the filtered reference signal. In detail, the adaptive filter 110 may filter the input reference signal 10 to generate a filtered reference signal, and output the generated filtered reference signal. Signal 30 as shown in FIG. Here, an additional signal 20 may be added to the opposite noise signal 30.

The delay time analyzer 120 receives the error signal 60 and the additional signal 20 and outputs a group delay error signal 70. [ Here, the delay time error signal 70 may be a phase error changed by a temperature, a pressure, or the like around the system.

Specifically, the delay time analyzer 120 compares the error signal 60 with the additional signal 20 to calculate the delay time change of the system due to the temperature, pressure, etc. around the system, and outputs the delay time error signal 70 And outputs the generated delay time error signal 70. The delay time analyzer 120 generates a delay time error signal 70 using a method of analyzing the correlation between the additive signal 20 and corresponding frequency components of the error signal 60 . Here, analyzing the correlation may be using a correlation analyzer.

The additive path transfer function model 130 receives the delay time error signal 70 and the reference signal 10 and can output the reference noise signal 80. Specifically, the additive path transfer function model 130 is updated in real time to compensate for the effect of the additive path transfer function 200 on the algorithm using the delay time error signal 70. At this time, the reference signal 10 may be applied to the updated additive path transfer function model 130 to generate the reference noise signal 80.

The update algorithm 140 can receive the reference noise signal 80 and the error signal 60 and output the improved adaptive filter 90. Specifically, the update algorithm 140 may generate the improved adaptive filter 90 by multiplying the reference noise signal 80 with the error signal 60. [ Here, the reference noise signal 80 and the error signal 60 may be multiplied by a constant called a step-size.

The adaptive filter 110 may further receive the improved adaptive filter 90 and output a filtered reference signal. In particular, the adaptive filter 110 may generate the second adaptive filter 110 by adding the enhancement adaptive filter 90 and the first adaptive filter 110. Here, the first adaptive filter 110 is set to a value in advance, and may be a value of 0, and the second adaptive filter 110 may be an update of the first adaptive filter 110. The second adaptive filter 110 may filter the reference signal 10 to generate a filtered reference signal and output the generated filtered reference signal.

Here, the additional signal 20 may be a signal other than the audible frequency. Specifically, since the audible frequency is about 20 Hz to 20.000 Hz, the additional signal 20 may be an ultrasonic wave of 20 Hz or less or an ultrasonic wave of 20,000 Hz or more. However, general speakers can not reproduce sound below 20Hz due to limitations in dynamic characteristics. Therefore, the additional signal 20 may be an ultrasonic wave except for an extremely low frequency.

In the case of generating the additional signal 20 at the audible frequency and calculating the delay time error signal 70 using the additional signal 20 and the opposite noise signal 30, Additional noise may occur. Furthermore, if the frequency is a component included in the frequency band of the noise spectrum, a problem may occur in the stability and convergence of the algorithm for performing the noise control method.

In contrast, in the noise control method according to the embodiment, the additional signal 20 is generated from a signal other than the audible frequency, and the delay time error signal 70 is calculated using the additional signal 20 and the opposite noise signal 30 The additional noise due to the additional signal 20 is not heard. Since the normal noise is generated in the audio frequency band, there is an advantage that the stability and the convergence of the algorithm for performing the noise control method do not occur.

FIG. 3 is a first embodiment of the additional path transfer function model shown in FIG.

2 and 3, the additional path transfer function model 130 includes an initial additional path transfer function initial model 131 and a pure delay filter 132 in the form of a transversal filter in a general active noise control method. ).

The additive path transfer function initial model 131 may be generated prior to leaving the factory or after installing the active noise control system. Specifically, after generating a white noise signal in an audible frequency band, it can be detected and generated by an error microphone.

The additive path transfer function initial model 131 may be expressed as a finite length filter that simulates the impulse response of the additive path transfer function 200.

The pure delay filter 132 may be updated in real time according to the delay time error signal 70 input from the delay time analyzer 120.

Specifically, the coefficient of the adaptive filter 110 is updated every sample using the reference signal 10 and the error signal 60 so that the squared value of the error signal 60 is minimized. In this case, can do.

FIG. 4 is a second embodiment of the additional path transfer function model shown in FIG. 2, FIG. 5 is a schematic diagram of the initial model of the additive path transfer function shown in FIG. 4, Figure 7 is a schematic diagram of a function update algorithm.

1 to 6, the additional path transfer function model 130 may be processed in parallel for each frequency. Specifically, the sine and cosine signals at one frequency are filtered by the additive path transfer function initial model 133 and the additive path transfer function update algorithm 134. The delay time error signal 70 is input to each additional path transfer function update algorithm 134.

The additive path transfer function initial model 133 may be generated prior to shipment from the factory or after installation of the active noise control system. Specifically, after generating a white noise signal in an audio frequency band, it can be detected and generated by an error microphone.

The additive path transfer function initial model 133 can classify the additive path transfer function 200 into frequency-specific size and phase information. The additive path transfer function initial model 133 can obtain a corresponding gain value from the magnitude and phase information of each frequency using the classified frequency-dependent sine and cosine signals, multiply and add them to obtain the reference noise signal 80 . Each gain value can be calculated in the pre-measured supplementary path transfer function 200, and its value is fixed.

Specifically, the size and phase of the additional path transfer function initial model 133 may be filtered and calculated as shown in FIG. The system initial delay time can be obtained when the additional path transfer function initial model 133 is obtained by measuring in advance.

When the algorithm is activated, the delay time error signal 70, which is the difference between the current and the initial delay time, is applied to the supplementary path transfer function update algorithm 134. The delay time error signal 70 is multiplied by the gain value of the previously calculated frequency, and compensates for the phase of the value.

For example, it can be calculated using [Equation 1] and [Equation 2].

[Equation 1]

&Quot; (2) &quot;

Here, f _k is the k-th frequency, n is a sample (sample) and, Δ is the delay time error signal 70 is calculated in the delay time analyzer 120.

Here, the delay time error signal 70 is multiplied by the gain value of the previously calculated frequency, and the delay time error signal 70 can be applied in the same manner as the additional path transfer function initial model 133 in order to compensate for the phase value.

The supplemental path transfer function update algorithm 134 can predict the arrival time of the sound from the physical distance since the sound is transmitted in the form of a plane wave in a simple structure such as a duct structure. Similarly, in such a system, if the amount of change in temperature is known, the phase change of the additional path transfer function 200 can be predicted.

In an ideal system, the frequency-dependent delay time is the same, and the phase in the frequency domain is linear. The additional signal 20 uses a frequency signal other than the audible frequency that can not be heard by a person, so that additional noise can not be heard but can be measured from the sensor. Therefore, additional noise due to the additional signal 20 does not occur, so that the performance of noise control as an original purpose may not be deteriorated.

Here, the additional path transfer function model according to the second embodiment can exhibit excellent performance when the noise has periodic or quasi-periodic characteristics.

FIG. 7 is a schematic diagram of the delay time analyzer shown in FIG. 2, and FIG. 8 shows an example of analyzing a change in delay time in the delay time analyzer shown in FIG.

Referring to FIGS. 7 and 8, the delay time analyzer 120 may include a high pass filter (HPF) (not shown). When the high-pass filter is included in the delay time analyzer 120, the error signal 60 can be high-pass filtered. In detail, the error signal 60 transmitted to the delay time analyzer 120 may include an audible frequency band according to the noise and an ultrasonic band corresponding to the additional signal 20. Therefore, when the error signal 60 is subjected to high-pass filtering, the audio frequency band of the error signal 60 is filtered and only the ultrasonic band is left. Therefore, a change in the delay time can be calculated using a method of analyzing the correlation between the ultrasonic band of the error signal 60 and the additional signal 20.

Since the delay time analyzer 120 includes the high pass filter and high pass filters the error signal 60 and calculates the change in the delay time of the additional path transfer function 200, the error signal 60 May be filtered to calculate the change in delay time of the additional path transfer function 200 only in the ultrasonic wave band included in the error signal 60. [ Therefore, there is an advantage that the calculation of the change in the delay time of the additional path transfer function 200 can be further simplified.

The delay time analyzer 120 analyzes the frequency components of the additional signal 20 and the error signal 60 and measures the phase difference between the additional signal 20 and the error signal 60 to generate a delay time error signal 70) can be calculated. Therefore, the process of calculating the delay time error signal 70 using the additional signal 20 can be simplified.

Hereinafter, a noise control method according to the embodiment will be described.

9 is a flowchart for explaining a noise control method according to the embodiment.

1 to 5 and 9, a noise control method according to an embodiment includes a first output step S10, a first sensing step S20, a second output step S30, a second sensing step S40 And a third output step S50. First, the first output step S10 receives the reference signal 10 to generate a first anti-noise signal 30, and outputs a first sound 40. [ Specifically, the reference signal 10 is input, the first adaptive filter 110 is generated using the additive path transfer function model 130, and the reference signal 10 is applied to the first adaptive filter 110 It is possible to generate the first anti-noise signal 30 and output the first sound 40 for the first anti-noise signal 30 using an acoustic output means such as a speaker.

The first sensing step S20 senses the first noise while the first sound 40 is output. In particular, if the noise is less compensated for by the superposition of the first sound 40 and the noise 50, the first noise can be detected by the error microphone.

The second output step S30 generates a second counter-noise signal 30 and outputs a second sound 40 when a first noise is sensed. Specifically, when the error microphone detects the first noise, it generates the additional signal 20 to calculate the change in the delay time, and adds the additional signal 20 to the first counter-noisy signal 30, To produce a noise signal (30), and to output a second anti-noise signal (30). Here, the additional signal 20 may be a signal other than the audible frequency.

The second sensing step S40 senses the second noise while the second sound 40 is output. Specifically, if the noise 50 is less compensated for by the superposition of the second sound 40 and the noise 50, the second noise can be detected by the error microphone.

The third output step (S50) generates a delay time error signal (70) and outputs a third sound (40) when a second noise is sensed. Specifically, when the error microphone senses the second noise, the delay time analyzer 120 compares the second signal 20 with the second noise signal 60 to generate a delay time error signal 70, The additive path transfer function initial models 131 and 133 are updated in real time by the delay time error signal 70 and are supplied to the additional path transfer function model 130 updated by the delay time error signal 70 with reference signals 10 ) To generate a reference noise signal 80, multiplying the reference noise signal 80 by a second noise signal 60 and a convergence element to generate an improved adaptive filter 90, A second adaptive filter 110 is added by adding a first adaptive filter 90 and a first adaptive filter 110 and a reference signal 10 is applied to a second adaptive filter 110 to generate a third opposite noise signal 30 , And output a third sound (40) for the third counter-noisy signal (30). The generation of the delay time error signal 70 in the third output stage is achieved by high pass filtering the signal 60 by the second noise and then comparing the delay time error signal 70 with the additional signal 20 Lt; / RTI &gt;

Here, generating the first anti-noise signal 30 in the first output stage may include analyzing the frequency of the reference signal 10, selecting a frequency to be controlled among the frequencies of the analyzed reference signal 10, Noise signal 30 may be generated.

As described above, in the noise control method according to the embodiment, when the first noise is sensed, the additional signal 20 is generated to generate the delay time error signal 70, so that the ambient temperature, pressure, There is an advantage in that the noise can be controlled even in an environment having an unstable supplementary path transfer function 200 in which the signal is transferred.

In addition, in the noise control method according to the embodiment, the additional path transfer function initial models 131 and 133 are updated in real time by the delay time error signal 70, and the reference noise signal 80 and the signal 60 to generate the second adaptive filter 110 and apply the reference signal 10 to the second adaptive filter 110 to generate the third counter-noisy signal 30, There is an advantage that the noise can be controlled in any environment in which the delay time of the additional path transfer function 200 changes linearly with frequency.

In addition, since the noise control method according to the embodiment generates the delay time error signal 70 by comparing the second signal 60 with the additional signal 20 after high-pass filtering the second noise signal 60, The acoustic frequency band included in the signal 60 due to the second noise can be filtered and the change in the delay time of the additional path transfer function 200 can be calculated only in the ultrasonic wave band included in the signal 60 based on the second noise. Therefore, there is an advantage that the calculation of the change in the delay time of the additional path transfer function 200 can be further simplified.

Since the noise control method according to the embodiment analyzes the frequency of the reference signal 10 and generates the opposite noise signal 30 by selecting a frequency to be controlled from the frequency of the analyzed reference signal 10, Only the frequency noise can be reduced, and the process of calculating the change in the delay time using the additional signal 20 can be simplified.

Since the noise control method according to the embodiment uses a frequency signal other than the audible frequency as the additional signal 20, it does not generate additional noise generated when the audio frequency signal is used as the additional signal 20, It has the advantage of being able to detect changes and compensate them at the same time.

10 is a flowchart for explaining a method of generating an initial model of an additional path transfer function. Here, the additional path transfer function initial models 131 and 133 may be generated off-line before the active noise control is activated.

1 to 5 and 10, a method of generating the additional path transfer function initial models 131 and 133 before the first output step first generates a white noise signal and outputs a fourth sound ( S101). Specifically, a white noise signal in an audio frequency band can be generated, and a fourth sound for a generated white noise signal can be output.

While the fourth sound is being output, the fifth sound is sensed and the additional path transfer function initial models 131 and 133 are generated (S102). Specifically, the additional path transfer function initial models 131 and 133 are generated using the fifth sound transmission signal and the white noise signal sensed by the error microphone while the fourth sound is output.

As described above, according to the noise control method according to the embodiment, since the white noise signal is generated and the sound for the white noise signal is output to generate the additional path transfer function initial models 131 and 133, It is possible to generate the initial model of the additive path transfer function according to the additive path transfer function more simply by the transfer signal alone. Here, the additional path transfer function initial models 131 and 133 can also be generated by the LMS (Least Mean Square) method.

11 is a flowchart of an algorithm for performing the noise control method according to the embodiment.

1 to 5 and 11, a noise control method according to an embodiment may include a sensing step, an additional signal generating step, an opposite noise signal generating step, and an outputting step. First, in the sensing step, the first microphone senses the first reference noise for the first sample, and the second microphone senses the first noise. Here, the first microphone may be a microphone for sensing the noise 50, and the second microphone may be the above-described error microphone.

In the additional signal generation step, when the first reference signal 10 by the first reference noise sensed by the first microphone and the first error signal 60 by the first noise sensed by the second microphone are transmitted, Thereby generating the additional signal 20. Specifically, the first additional signal 20 may be generated by a signal generator or the like. Here, the first additional signal 20 may be a signal other than the audible frequency.

The opposite noise signal generation step generates the first anti-noise signal 30 using the first reference signal 10, the first error signal 60 and the first additional signal 20. Specifically, the counter-noise signal generation step includes a delay time analyzer 120 for comparing the first additional signal 20 with the first error signal 60 to generate a first delay time error signal 70, The additional path transfer function model 130 updates the additional path transfer function initial models 131 and 133 in real time using the first delay time error signal 70 to generate the updated additional path transfer function model 130, Generating a reference noise signal (80) by applying the first reference signal (10) to the updated additive path transfer function model (130), generating a first error signal (60) and a second reference signal The first adaptive filter 90 is multiplied by the first reference noise signal 80 to generate a first adaptive filter 90 and a second adaptive filter 110 is added to the first adaptive filter 90 and the first adaptive filter 110, After applying the first reference signal 10 to the second adaptive filter 110, And applying a reference signal to add the signal (20) to generate a first anti-noise signal (30). Here, the delay time error signal generating step may generate the first delay time error signal 70 by comparing the first error signal 60 with the first additional signal 20 after high-pass filtering the first error signal 60. The opposite noise signal generating step may be generated by analyzing the frequency of the first reference signal 10 and selecting a frequency to be controlled among the frequencies of the first reference signal 10 analyzed.

The output stage outputs the first sound 40 for the first anti-noise signal 30.

Also, the sensing step senses a second reference noise for the second sample from the first microphone, and the second noise senses the second microphone.

In the additional signal generation step, when the second reference signal 10 based on the second reference noise sensed by the first microphone and the second error signal 60 based on the second noise sensed by the second microphone are transmitted, Signal 20 is generated.

The opposite noise signal generating step generates the second opposite noise signal 30 using the second reference signal 10, the second error signal 60, and the second additional signal 20. Specifically, the counter-noise signal generating step may include a delay time analyzer 120 for comparing the second additional signal 20 with the second error signal 60 to generate a second delay time error signal 70, The additional path transfer function model 130 generates the updated additional path transfer function model by updating the additional path transfer function initial models 131 and 133 in real time using the second delay time error signal 70 A reference noise signal generation step of applying the second reference signal 10 to the updated additional path transfer function model 130 to generate a second reference noise signal 80, a second error signal 60 and a convergence element The second adaptive filter 90 is multiplied by the second reference noise signal 80 to generate the second adaptive adaptive filter 90 and the second adaptive filter 90 and the second adaptive filter 110 are added to generate the third adaptive filter 110 After applying the adaptive filter generation step and the second reference signal 10 to the third adaptive filter 110, In addition to the 20 second opposed it contains a reference signal applied to produce a noise signal (30).

The output stage outputs a second sound (40) for the second counter-noisy signal (30).

And, in the sensing step, the first microphone senses the n-th reference noise for the n-th sample, and the second microphone senses the n-th noise.

In the additional signal generation step, when the n-th reference signal 10 by the n-th reference noise sensed by the first microphone and the n-th error signal 60 by the n-th noise sensed by the second microphone are transmitted, Thereby generating the additional signal 20.

The opposite noise signal generating step generates the n th reverse noise signal 30 using the n th reference signal 10, the n th error signal 60, and the n th additive signal 20. Specifically, the counter-noise signal generating step includes a delay time analyzer 120 for comparing the n-th additional signal 20 with the n-th error signal 60 to generate a delay time error signal 70, The additive path transfer function model 130 updates the additive path transfer function initial models 131 and 133 in real time using the n th delay time error signal 70 to generate an updated additive path transfer function model A reference noise signal generating step of applying the n-th reference signal 10 to the updated additive path transfer function model 130 to generate an n-th reference noise signal 80, an n-th error signal 60 and a convergence element The nth adaptive filter 90 is multiplied by the nth reference noise signal 80 to generate an nth improved adaptive filter 90 and an nth improved adaptive filter 90 and an nth adaptive filter 110 are added to the After the adaptive filter generation step and the n-th reference signal 10 to be generated are applied to the (n + 1) -th adaptive filter 110, (20) to generate an n-th reverse noise signal (30).

The output step outputs the n-th sound 40 for the n-th reverse noise signal 30.

Here, the n-th reference signal 10 by the n-th reference noise sensed by the first microphone is filtered into the updated additive path transfer function model 130 to generate the n-th reference noise signal 80, The noise signal 80 is sequentially stored in a buffer having a length L of the (n + 1) -th adaptive filter 110 (thus, an L-length vector can be obtained for every sample) And is multiplied by the n-th error signal 60 due to the n-th noise sensed by the second microphone. At this time, the multiplication result is multiplied by a constant called a step-size, and the coefficient of the n-th adaptive filter 110 used in the previous sample (the coefficient value is L because of the L-length vector) The coefficient of the adaptive filter 110 to be used is obtained. The n-th reference signal 10 is filtered to the (n + 1) -th adaptive filter 110.

In the adaptive filter generation step, the second adaptive filter 110, the third adaptive filter 110, , Generating the (n + 1) -th adaptive filter 110 includes generating a second error signal 60, ... , And the n-th error signal 60 in the direction in which the square value is minimized.

And generating a white noise signal and outputting a sound for the generated white noise signal before the first microphone senses the first reference noise and the second microphone senses the first noise, And generating an initial model of the additive path transfer function using the acoustic signal and the white noise signal sensed by the additive path transfer function generating unit.

As described above, the noise control method according to the embodiment includes the first additional signal 20, the second additional signal 20, , The n-th adder 20 generates a first delay time error signal 70, a second delay time error signal 70, , And the n-th delay time error signal 70, noise can be controlled even in an environment having an unstable additional path transfer function in which the ambient temperature or pressure changes and the speed of sound changes, There is an advantage.

In addition, in the noise control method according to the embodiment, the additional path transfer function initial models 131 and 133 calculate the first delay time error signal 70, the second delay time error signal 70, And the nth delay time error signal 70, the noise can be controlled in any environment in which the delay time of the additional path transfer function 200 varies linearly with frequency depending on the temperature or the pressure .

In addition, since the additional-path transfer function initial models 131 and 133 described with reference to FIG. 11 are generated in the same manner as described with reference to FIG. 10, the noise control method according to the embodiment is not limited to the above- There is an advantage that the additional path transfer function initial models 131 and 133 according to the function 200 can be simply generated.

In addition, the noise control method according to the embodiment includes a first error signal 60, a second error signal 60, The n-th error signal 60, and then outputs the first additional signal 20, the second additional signal 20, A first delay time error signal 70, a second delay time error signal 70, ..., , And the n-th delay time error signal 70, the first error signal 60, the second error signal 60, ... , The audio frequency band included in the nth error signal 60 is filtered to obtain a first error signal 60, a second error signal 60, ... , It is possible to calculate the change of the delay time of the additional path transfer function 200 more simply by using only the ultrasonic wave band included in the nth error signal 60. [

The noise control method according to the embodiment includes a first reference signal 10, a second reference signal 10, , Analyzes the frequency of the nth reference signal (10), and outputs the analyzed first reference signal (10), second reference signal (10), ... , The frequency of the nth reference signal (10) to be controlled is selected so as to generate a first anti-noise signal (30), a second anti-noise signal (30), ... , The n-th reverse noise signal 30 is generated. Therefore, only the desired frequency noise can be reduced, and the process of calculating the change in the delay time using the n-th additional signal 20 can be simplified .

In the noise control method according to the embodiment, a frequency signal other than an audible frequency is divided into a first additional signal 20, a second additional signal 20, , The n-th additional signal 20 is used as the first additional signal 20, the second additional signal 20, ... There is an advantage that it is possible to simultaneously compensate for the change of the delay time without generating additional noises generated when the n-th additional signal 20 is used.

As described above, by using the noise control method according to the embodiment, it is possible to effectively control the noise even in an environment where ambient temperature or pressure can be greatly changed. Also, in an environment where the delay time of the system does not greatly change, the algorithm will be applied as it is in the conventional manner. If the delay time greatly changes, the phase error according to the algorithm will be compensated for and the noise can be reduced stably.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those of ordinary skill in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. It can be understood that The features, structures, effects and the like described in the embodiments are included in at least one embodiment of the present invention and are not necessarily limited to only one embodiment. Furthermore, the features, structures, effects and the like illustrated in the embodiments can be combined and modified by other persons skilled in the art to which the embodiments belong. Therefore, it should be understood that the present invention is not limited to these combinations and modifications.

100: Active Noise Control Algorithm
110: adaptive filter
120: delay time analyzer
130: Additional path transfer function model
140: Update Algorithm
200: additional path transfer function

Claims (12)

A first output step of receiving a reference signal to generate a first counter-noise signal, and outputting a first sound corresponding to the first counter-noise signal;
A first sensing step of sensing a first noise while the first sound is output;
Generating a second anti-noise signal by adding the first anti-noise signal to the first anti-noise signal, and generating a second anti-noise signal by adding the second anti-noise signal to the first anti- A second output step of outputting a second sound to the second output step;
A second sensing step of sensing a second noise while the second sound is output; And
A second delay time error signal generating unit that generates a third reverse noise signal by comparing the additional signal with the second noise signal to generate a delay time error signal when the second noise is detected, And a third output step of outputting a third sound for the third opposite noise signal. The method according to claim 1,
Generating the first anti-noise signal in the first output stage comprises:
Applying the reference signal to a first adaptive filter to generate the first anti-noise signal,
Generating the third counter-noise signal in the third output stage comprises:
A reference noise signal is generated by applying the reference signal to an additional path transfer function model updated by the delay time error signal, and an improved adaptive filter generated by multiplying the reference noise signal by a signal by the second noise, 1 adaptive filter to generate a second adaptive filter and applying the reference signal to the second adaptive filter to generate the third anti-noise signal. 3. The method of claim 2,
Prior to the first output stage,
Generating a white noise signal and outputting a fourth sound for the white noise signal; And
And generating an initial model of an additional path transfer function of the additional path transfer function model using the fifth sound sensed during transmission of the fourth sound and the white noise signal, Way. The method according to claim 1,
Generating the delay time error signal in the third output stage comprises:
Filtering the signal by the second noise and then comparing the signal with the additional signal to generate the delay time error signal. The method according to claim 1,
Generating the first anti-noise signal in the first output stage comprises:
Analyzing the frequency of the reference signal,
Wherein a frequency to be controlled is selected from frequencies of the reference signal to generate the first anti-noise signal. The method according to claim 1,
Wherein the additional signal is a frequency signal other than an audible frequency. The first microphone includes a first sample, a second sample, ... , The first reference noise for the nth sample, the second reference noise, ... , The nth reference noise is detected, and the second microphone detects the first noise, the second noise, ... A sensing step of sensing n-th noises;
The first reference noise, the second reference noise, and the second reference noise detected by the first microphone. A first reference signal based on the n-th reference noise, a second reference signal, , An n-th reference signal, the first noise detected by the second microphone, the second noise, ... A first error signal due to the n-th noise, a second error signal, A first additional signal, a second additional signal, and a second additional signal to calculate a change in delay time when an n-th error signal is transmitted. An additional signal generation step of generating an n-th additional signal;
The first reference signal, the second reference signal, , The nth reference signal, the first error signal, the second error signal, ... , The n-th error signal and the first additional signal, the second additional signal, ... A second anti-noise signal, a second anti-noise signal, and a second anti-noise signal using the n-th additional signal. An opposite noise signal generating step for generating an nth reverse noise signal; And
The first anti-noise signal, the second anti-noise signal, , The n-th reverse noise signal, the first additional signal, the second additional signal, ... A first sound for the first anti-noise signal, a second sound for the second anti-noise signal, And outputting the n-th sound for the n-th reverse noise signal. 8. The method of claim 7,
Wherein the opposite noise signal generation step comprises:
The first additional signal, the second additional signal, ... , The n-th additional signal, the first error signal, the second error signal, ... , And compares the n-th error signal to obtain a first delay time error signal, a second delay time error signal, ... A delay time error signal generation step of generating an nth delay time error signal;
The first delay time error signal, the second delay time error signal, And generates an updated additive path transfer function model by updating the initial model of the additive path transfer function of the additive path transfer function model in real time using the nth delay time error signal, , ... And applying the n-th reference signal to the updated additive path transfer function model to generate a first reference noise signal, a second reference noise signal, A reference noise signal generation step of generating an n-th reference noise signal;
The first error signal, the second error signal, ... , The nth error signal is the first reference noise signal, the second reference noise signal, A first enhancement adaptive filter, a second enhancement adaptive filter, , An nth improved adaptive filter is generated, and the first improved adaptive filter, the second improved adaptive filter, , The nth improved adaptive filter, the first adaptive filter, the second adaptive filter, A second adaptive filter, a third adaptive filter, and an n-th adaptive filter. An adaptive filter generating step of generating an (n + 1) -th adaptive filter; And
The first reference signal, the second reference signal, , The n-th reference signal to the second adaptive filter, the third adaptive filter, A first addition signal, a second addition signal, and a second addition signal after being applied to the (n + 1) th adaptive filter. , An n-th additional signal is added to the first anti-noise signal, the second anti-noise signal, ... And applying a reference signal to generate the n-th reverse noise signal,
Wherein the second adaptive filter, the third adaptive filter, the third adaptive filter, , Generating the (n + 1) -th adaptive filter comprises: generating the second error signal, And generates the n-th error signal in a direction in which the square value of the n-th error signal is minimized. 9. The method of claim 8,
Prior to the step of the first microphone sensing the first reference noise and the second microphone sensing the first noise,
Generating a white noise signal and outputting the sound for the white noise signal; And
And generating the supplementary path transfer function initial model using the sound signal and the white noise signal sensed by the second microphone. 9. The method of claim 8,
Wherein the delay time error signal generation step comprises:
The first error signal, the second error signal, ... An n-th error signal, and outputs the first additional signal, the second additional signal, The first delay time error signal, the second delay time error signal, and the second delay time error signal, , And generates the n-th delay time error signal. 9. The method according to claim 7 or 8,
Wherein the opposite noise signal generation step comprises:
The first reference signal, the second reference signal, , Analyzes the frequency of the n-th reference signal,
The analyzed first reference signal, second reference signal, ... , A frequency to be controlled is selected from the frequency of the nth reference signal, and the first anti-noise signal, the second anti-noise signal, , The n-th reverse noise signal is generated
The noise control method comprising: 9. The method according to claim 7 or 8,
The first additional signal, the second additional signal, ... , And the n-th additional signal is a frequency signal other than an audible frequency.

Images