KR20060123732A - Active noise control system and method - Google Patents

Abstract

An Active Noise Control (ANC) for controlling a noise produced by a noise source may include an acoustic sensor (212) to sense a noise pattern and to produce a noise signal corresponding to the sensed noise pattern, an estimator (202) to produce a predicted noise signal by applying an estimation function to the noise signal, and an acoustic transducer (216) to produce a noise destructive pattern based on the predicted noise signal.

Description

ACTIVE NOISE CONTROL SYSTEM AND METHOD

Cross-file

This application takes precedence over US Provisional Application No. 60 / 503,471, filed September 17, 2003, with Israel Patent Application No. 121555, filed August 14, 1997, as of July 22, 1998. Part of US Patent Application No. 09 / 120,973, filed as

Field of invention

The present invention relates to active noise control.

Conventional passive noise control systems employ conventional mufflers such as "insulation" elements, silencers, vibration mounts, damping treatments, absorption treatments such as sealing tiles, and / or mufflers that are also used, for example, in the automotive industry. Will include. The dimensions and / or mass of this passive noise control system will typically depend on the acoustic pattern length of the noise to be reduced. In general, passive noise control systems implemented to reduce noise at relatively low frequencies are bulky, large, overweight, and / or expensive.

According to an embodiment of the present invention, an active noise control (ANC) system includes an acoustic sensor such as a microphone for detecting a noise pattern and generating a noise signal corresponding to the detected noise pattern, and an estimation function for the noise signal. And an acoustic transducer such as a speaker for generating a noise canceling pattern based on the predicted noise signal.

According to one exemplary embodiment of the present invention, the estimation function will comprise a non-linear estimation function such as, for example, a radial basis function (RBF).

The estimator may adaptively adapt one or more parameters of the estimation function based on the noise error at a given point. For example, the ANC system may include an error evaluator that evaluates the noise error based on the noise signal and the predicted noise signal. In addition or alternatively, the ANC system may include an error sensing acoustic sensor that detects noise errors at certain points.

The error evaluator includes, for example, a speaker transfer function module that generates an estimate of the noise cancellation pattern by applying a speaker transfer function to the predicted noise signal, and an estimate of the noise pattern at a given point, for example, by applying a modulation transfer function to the noise signal. And a subtractor for subtracting the estimate of the noise cancellation pattern from the estimate of the noise pattern.

According to some example embodiments, the estimator may adaptively change one or more parameters based on certain criteria. For example, the estimator may reduce, eg, minimize, the error value by adapting one or more parameters as applicable.

According to another example embodiment of the present invention, the ANC system detects a noise pattern and generates a corresponding primary noise signal, for example, a primary acoustic sensor such as a microphone, and a residual noise pattern to detect a residual noise pattern. At least one secondary acoustic sensor, such as a microphone, for example, which generates a corresponding at least one secondary noise signal and is further away from the noise source by a distance greater than the distance between the primary acoustic sensor and the noise source, and the primary noise signal And a controller for controlling the acoustic transducer based on the at least one secondary noise signal to generate a noise canceling pattern.

The controller includes, for example, a first estimator for generating a first order prediction signal by applying a first order estimation function to the first order noise signal, and at least one by applying at least one second order estimation function to the at least one second order noise signal. At least one secondary estimator for generating a second order prediction signal, respectively.

The primary estimator may repeatedly adapt one or more parameters of the primary estimation function based on the noise error. The at least one estimator may repeatedly adapt one or more parameters of the at least one second order estimation function based on the noise error, respectively.

The controller controls the acoustic transducer based on a combination of the first prediction signal and the at least one second prediction signal.

1 is a schematic diagram of an active noise control system according to an exemplary embodiment of the present invention.

2 is a schematic diagram of a controller according to some example embodiments of the present invention that may be used with the active noise control system of FIG. 1.

3 is a schematic diagram of an active noise control system according to another embodiment of the present invention.

4 is a schematic diagram of a controller according to another example embodiment of the present invention that may be used with the active noise control system of FIG. 3.

In the following the subject matter of the invention is described in detail and is clearly claimed in the conclusion of the present specification. However, the objects, features and advantages of the present invention, as well as the structure and operation of the present invention will be more clearly understood by referring to the following detailed description with reference to the accompanying drawings.

The components shown in the figures are not necessarily drawn to scale or scale, for simplicity and clarity of illustration. For example, the dimensions of some components may be enlarged relative to other components for clarity, or may be enlarged relative to other physical parts included in one functional block or component. Further, reference numerals may be repeated to indicate corresponding or similar elements in the figures. Moreover, some of the blocks shown in the figures may be combined into a single function.

In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be understood by those skilled in the art that the present invention may be practiced without these specific details. In other instances, well known methods, procedures, components, and circuits will not be described in detail so as not to obscure the present invention.

According to an embodiment of the present invention, active noise control (ANC) may be used to reduce the wave amplitude and noise energy of a source noise pattern comprising, for example, one or more acoustic waves through an ANC sound system, which ANC sound system One or more noise canceling patterns, including acoustic waves, associated with the source noise pattern may be generated such that a reduced noise region may be generated.

Embodiments of the present invention include ANC systems and methods that can be effectively implemented to reduce undesirable noise, such as at least overall low frequency noise, as described below.

According to some example embodiments of the present invention, some configurations of ANC methods and systems are disclosed in US Patent Application Nos. 09 / 120,973, filed July 22, 1998, entitled “ACTIVE ACOUSTIC NOISE REDUCTION SYSTEM”, and 10 2002. The European Patent Application No. 02023483.7, filed on May 21 entitled "ACTIVE ACOUSTIC NOISE REDUCTION SYSTEM" (published on April 28, 2004), published on April 28, 2004, the entire disclosure content of these patent applications Is incorporated herein by reference.

Reference is made to FIG. 1 which schematically illustrates an ANC system 100 according to an exemplary embodiment of the present invention.

The ANC system 100 includes an acoustic sensor, such as the microphone 102 shown for example MIC1, which senses the noise energy and / or wave amplitude of the noise pattern generated by the noise source. The microphone 102 may be any suitable microphone capable of generating an output noise signal 103 in response to the noise pattern sensed by the microphone 112. For example, the microphone 102 may include ARIO Electronics Co., Ltd., Taoyuan, Taiwan. The microphone of part number ECM6AP available from Ltd. is included. The noise signal 103 may, for example, comprise a sequence of N samples per second. For example, N would be 1000 samples per second, for example when microphone 103 is operating at a sampling rate of about 10 Hz.

The ANC system 100 also includes a sound transducer, such as, for example, a speaker 180, and a noise to reduce or cancel the noise energy and / or wave amplitude of the noise pattern, eg, within the noise-reduction region 110, as described in detail below. A controller 106 that controls the speaker 108 to produce a cancellation pattern. Speaker 108 may be any suitable speaker as is known in the art. For example, speaker 108 includes a speaker of part number AI 4.0 available from Cerwin-Vega Inc., California, USA.

According to some example embodiments of the present invention, the controller 106 may have anticipated destructive interference between the noise pattern and the noise cancellation pattern at a predetermined point 112 within the noise-reduction area 110 as described below. We will evaluate the noise error corresponding to. The noise error will be assessed by the controller 106 based on the noise signal 103 as described below, for example. In addition or alternatively, the noise error will be detected by an error-sampling microphone located at a predetermined position as described below. The controller 106 will control the speaker 108 to generate a noise cancellation pattern based on the noise signal 103 and / or the estimated noise error, for example, as described below.

According to some example embodiments of the present invention, it will be required to control the timing at which a noise cancellation pattern is generated in order to effectively control, for example, reduce noise in the noise-reduction area 110. For example, it is possible to control the timing of the noise cancellation pattern corresponding to the samples of the noise pattern, such that the noise cancellation is at substantially the same time as the sampled noise pattern reaching any point in the noise-reduction area 110, for example point 112. The pattern will be required to reach the same point.

According to an embodiment of the present invention, there may be a time delay between the time when the current sampled noise pattern reaches point 112 and the time when the noise cancellation pattern corresponding to the current sample of the noise pattern reaches that point 112. have. This time delay may be, for example, the time required for the microphone 102 to detect the noise pattern, the time required for the controller to process the noise signal 103, and / or the time required for the speaker 108 to generate the noise cancellation pattern. Will be generated from

Therefore, according to some example embodiments of the present invention, the controller 106 may estimate a sample of the noise pattern ("following sample") that follows the current sample based on the current sample and / or one or more previous samples of the noise pattern. will be. The controller 118 causes the noise canceling pattern to appear at that point 112 at the same time that the noise pattern reaches the point 112, such that the speaker 113 generates a noise canceling pattern based on the estimated subsequent samples. To reach, it will provide input to the speaker 113.

Acoustic patterns, such as noise patterns, will be characterized by their overall nonlinear function. Therefore, in accordance with an exemplary embodiment of the present invention, the controller 106 will use nonlinear estimation to estimate subsequent samples. Such a nonlinear estimate will, according to an exemplary embodiment of the present invention, provide a better estimate for subsequent samples compared to the corresponding linear estimate. However, according to another embodiment of the present invention, the controller 106 may use any other suitable estimate, such as a linear estimate, suitable for estimating subsequent samples.

According to one exemplary embodiment of the present invention, the controller 106 will include an estimator 121 that generates the predictive noise signal 114 by applying an estimation function to one or more samples of the noise signal 103. The speaker 113 will generate a noise cancellation pattern based on the predictive noise signal 114 as described below.

Hereinafter, a controller 200 according to some example embodiments of the present invention will be described with reference to FIG. 2. Although the present invention is not limited to this configuration, the controller 200 may be implemented by the ANC system 100 (FIG. 1).

According to one exemplary embodiment of the present invention, the controller 200 includes an estimator 202 that receives a noise signal 210 that includes, for example, a plurality of samples of a sensed noise pattern from an acoustic sensor, such as a microphone 212. Will include. The estimator 202 applies the sample function MIC (n) by applying the estimation function F to the n th sample received from the microphone 212 and one or more other samples previously received from the microphone 212 as described below. Will generate a predictive noise signal 230 having a value y (n) corresponding to The controller 202 will control the acoustic transducer, such as the speaker 216, to generate the noise cancellation pattern 218 based on the output 230.

According to some example embodiments of the present invention, estimator 202 will implement a nonlinear estimation algorithm as described below.

According to some example embodiments of the present invention, estimator 202 may implement a radial basis function (RBF) algorithm, as described below.

The estimator 202 will implement an RBF algorithm to estimate the value of the subsequent sample of the noise signal based on the value of one or more samples of the noise signal received from the microphone 212. For example, RBF algorithm is valid represented by, for example, the heart, v _k of the function parameter represented by c _k As can correspond to a combination of a set of K radial n- dimensional functions, each function is known in the art One or more parameters, such as the radius parameter and / or the strength of the function represented by w _k , will be different. For example, estimator 202 was authored by S. Haykin and published in the third edition of "Adaptive Filter Theroy", Prentice Hall, pp. It will implement an RBF algorithm similar to the algorithm disclosed in 863-565.

According to some example embodiments of the present invention, estimator 202 will produce predicted noise 230 according to the following equation:

Where L represents the number of predetermined noise signal samples to be implemented for an estimate of y (n).

According to some example embodiments of the present invention, estimator 202 repeatedly applies one or more parameters of the estimation function F, such as one or more of parameters c _k , v _k and w _k , based on a predetermined criterion as described below. Will change it if possible.

According to some example embodiments of the present invention, the estimator 202 repeatedly applies one or more of the parameters c _k , v _k and w _k based on the estimated noise error at a predetermined point, such as point 112, as described below. Will change it if possible.

In accordance with some example embodiments of the present invention, the controller 200 also includes an error evaluation module 203 to evaluate noise errors based on, for example, the noise signal 210 and the predicted noise signal 230 as described below. Will include.

According to some example embodiments of the present invention, the error evaluation module 203 may, for example, modify a modulation transfer function (MTF) module 204 to apply a predetermined modulation transfer function (MTF) to the noise signal 210. And produces an output 241 having a value corresponding to an estimate d (n) of the n-th sample of the noise pattern at a given point. The MTF is, as is known in the art, for example, based on the characteristics of the microphone 212 and / or the geometric and / or physical of the path and / or medium (eg, air) between the microphone 212 and the predetermined location. Will be determined based on the characteristics. The MTF module 204 may include any suitable hardware and / or software as known in the art to apply a given MTF to the noise signal 210.

According to an exemplary embodiment of the present invention, the error evaluation module 203 also includes a Speaker Transfer Function (STF) module 206 to apply the STF to the predicted noise signal 230, thereby Generate an output 249 having a value corresponding to the estimate of the noise cancellation pattern 218 generated in response to the predicted noise signal 230. The STF will be determined based on, for example, the characteristics of the speaker 216 as known in the art. The STF module 206 may include any hardware and / or software suitable for applying a given STF to the predictive noise signal 230 as is known in the art. For example, the value z (n) of the output 249 may be calculated using Equation 2:

Here, S refers to a certain STF frequency parameter vector as known in the art.

Corresponding Equation 1 to Equation 2 gives the following Equation 3:

According to an exemplary embodiment of the present invention, the error evaluation module 203 also includes a subtractor 208 that can be implemented in any suitable hardware and / or software as known in the art. The subtractor 208 subtracts, for example, the value of the estimated noise cancellation pattern of the output STF 249 from the estimated value of the noise pattern of the output 241, and includes an evaluation noise error e (n) corresponding to the sample MIC (n). Produces an output 245.

According to an exemplary embodiment of the present invention, the estimator 202 is adapted to iteratively change the value of one or more parameters of the parameters v _k , c _k and w _k based on the value of the noise error, for example, as described below. We will implement an adaptive algorithm.

According to an exemplary embodiment of the present invention, the value of the noise error e (n) corresponding to the n-th sample of the noise signal 210 will be estimated using the following equation:

e (n) = d (n)-z (n)

Substituting Equation 3 into Equation 4 yields the following Equation 5:

According to some example embodiments of the invention, the estimator 202 iteratively changes one or more of the parameters v _k , c _k, and w _k , such as, for example, the arithmetic mean E [(e (n (n)) of the square root of the noise error. )) ² ] decrease, eg minimize. For example, the estimator 202 may repeatedly apply one or more parameters of the estimation function such that the partial deviation of E [(e (n)) ² ] for one or more parameters is equal to zero, respectively, as described below. You can change it.

According to some example embodiments of the present invention, the arithmetic mean of the square root of the estimated noise error will be calculated using the following equation:

From here,

The partial deviation of equation (6) for the parameters v _k , c _k and w _k will be calculated using the following equations (8) to (10), respectively:

The minimum value of E [(e (n)) ² ] will be determined by the following Equations 11-13:

Applying the condition of Equation 11 to Equation 8, it will be among the parameters w _k applied to the possibly changed values w _k (n + 1) and the current value w _k (n), the following relationship is formed:

Where _k is the determined convergence parameter corresponding to w _k .

Applying the conditions of equation (12) to Equation (9) is changed to enable the application of the parameter c _k c _k value between the (n + 1) and the current value c _k (n) to form the following relationship:

Where _c is the determined convergence parameter corresponding to c _k .

Applying the condition of Equation 13 in Equation 10 the parameter v is changed to enable application of the value of _k v, the following relation is formed between a _k (n + 1) and the current value v _k (n):

According to some example embodiments of the present invention, adaptive estimator 202 implements one or more of Equations 14-16 to iteratively modify one or more parameters w _k , c _k and v _k , respectively. something to do.

Some example embodiments of the present invention relate to an ANC system, such as system 100 (FIG. 1), which implements the error estimation module of module 203, for example, to evaluate noise errors at certain points, such as point 112 (FIG. 1). . However, one of ordinary skill in the art appreciates that in accordance with other embodiments of the present invention, one or more other suitable modules may be implemented to evaluate noise errors. For example, the error sensing microphone 239 may be located at a certain point, and the output 240 of the error sensing microphone 239 corresponding to the sensed noise error at the predetermined point may be provided to the estimator 202.

Some example embodiments of the present invention provide a controller, such as a controller, for controlling an acoustic transducer, such as speaker 108 (FIG. 1), based on a noise signal of a noise pattern received from an acoustic sensor, such as microphone 102 (FIG. 1). ANC system including controller 106 (FIG. 1), such as ANC system 100 (FIG. 1). However, another embodiment of the present invention is directed to an ANC system including a controller capable of controlling an acoustic transducer based on one or more noise signals of noise patterns received from two or more acoustic sensors as described below.

The following is a description with reference to FIG. 3 which schematically illustrates an ANC system 300 according to another example embodiment of the present invention.

The ANC system 300 may include a primary acoustic sensor, such as microphone MIC1 302, for sampling the wave energy of the noise energy and / or noise pattern generated by the noise source 304. Microphone 302 may include any suitable microphone as described, for example, with reference to microphone 102 (FIG. 1).

The ANC system 300 also provides noise converters, such as speakers 308 and noise to reduce or cancel wave amplitudes of noise energy and / or noise patterns within the noise-reduction area 310, as described below. And a controller 306 that can control the speaker 308 to produce a cancellation pattern. Speaker 308 may include any suitable speaker as described with reference to speaker 108 (FIG. 1).

According to some example embodiments of the present invention, the controller 306 may combine a noise pattern and a noise cancellation pattern, such as a noise pattern and a noise cancellation pattern, at a predetermined point 312 in the noise-reduction area 310 as described below. The noise error corresponding to the difference between the patterns can be evaluated. The controller 306 will control the speaker 308 to produce a noise cancellation pattern, such as to reduce or minimize the noise error, as described below.

According to an exemplary embodiment of the present invention, in order to enable the ANC system 300 to achieve an effective degree of noise reduction as described below, noise error at the associated frequency of the primary microphone 302 and, for example, the noise pattern A relatively good agreement will be required between the estimates. For example, the higher the correlation between the noise pattern sampled by the microphone 302 and the noise error, the higher the level of noise control that can be achieved by the ANC system 300, such as the level of noise reduction. The match between the noise pattern sampled by the microphone 302 and the noise error will depend, for example, on the geometry of the path between the microphone 302 and point 312. In addition or alternatively, the match between the noise pattern and the noise error received by the microphone 302 will depend on the aerodynamic properties of the path, such as surface roughness. For example, if there is no “eye contact” between the microphone 302 and the point 312 and / or a path with a relatively rough surface, the match between the signal received by the microphone 302 and the estimated noise error Will be reduced. Also, if the structure of a device implementing one or more components of the ANC system 300, for example, does not have an aerodynamically optimized design due to price and size constraints, then the action of the ANC system 300 to reduce noise is to Uneven airflow and / or friction between and the path material will be disturbed by acoustic signals formed along the path between the microphone 302 and the point 312. Uneven airflow is characterized by the probabilistic formation of vortices with significant sibilance, and friction between air and relatively rough surfaces is characterized by the conversion of kinetic energy into heat and noise energy.

According to an exemplary embodiment of the present invention, the noise error will be evaluated using the MTF as described below with reference to FIG. The MTF may be predetermined based on, for example, one or more path characteristics and / or one or more expected noise pattern characteristics between microphone 302 and point 312. However, one or more path characteristics and / or expected noise pattern characteristics may be different from the expected characteristics. As a result, the correlation between the noise error evaluated based on the given predetermined MTF and the actual noise at point 312 may not be satisfactorily accurate.

According to some example embodiments of the present invention, the ANC system 300 also includes at least one secondary acoustic sensor, such as MIC21, for sampling the wave amplitude of the noise energy and / or noise pattern generated by the noise source 304. At least one secondary microphone 392 is shown. The secondary microphone 392 is separated from the noise source 304 by the distance d1 between the primary microphone 302 and the noise source 304, and the distance d1 is greater than the distance d2. For example, microphone 392 would be located along the path between microphone 302 and point 312. The distance d1-d2 between the microphone 392 and the microphone 302 will be large enough to allow the microphone 392 to sample a residual noise pattern that may not be received by the microphone 302, such as the noise pattern formed by the path. Microphone 392 may be any suitable microphone as described with reference to microphone 102 (FIG. 1).

According to some example embodiments of the present invention, the controller 306 may generate a noise canceling pattern based on a noise pattern detected by the microphone 302 and / or a residual noise pattern detected by the microphone 392 as described below. 308 will be controlled.

Hereinafter, a controller 400 according to another exemplary embodiment of the present invention will be described with reference to FIG. 4. The present invention is not limited to this feature, and the controller 400 may be implemented by the ANC system 300 (FIG. 3).

According to one exemplary embodiment of the present invention, the controller 400 will include a reference estimator 408 to receive from the primary microphone 402 a primary noise signal 412 comprising, for example, a plurality of samples. The reference estimator 408 applies the first order estimation function F ₁ to the sample MIC1 (n) and one or more other samples previously received from the microphone 402 as described below, for the n-th sample received from the microphone 402. Will generate a first order prediction signal 414 with a value y ₁ (n) corresponding to MIC1 (n).

According to one exemplary embodiment of the present invention, the controller 400 may also receive at least one secondary noise signal for receiving from the at least one secondary microphone 404 at least one secondary noise signal, each comprising a plurality of samples, for example. Will include an estimator 410. Quadratic estimator 410 receives the n-th received from microphone 404 by applying quadratic estimation function F ₂ to sample MIC21 (n) and one or more other samples previously received from microphone 404, as described below. Will generate a second order prediction signal 422 with a value y ₂ (n) corresponding to sample MIC21 (n).

Controller 400 controls an acoustic transducer, such as speaker 406, to generate noise canceling pattern 418 based on the combination of signals 414 and 422. For example, controller 400 also includes adder 424 as known in the art to provide speaker 406 with an input 426 corresponding to the sum of signals 422 and 414.

According to some example embodiments of the present invention, estimator 408 will generate signal 414 according to the following equation:

Here, W ₁ represents a predetermined prediction filter (PF) vector of length L ₁ corresponding to the estimation function F ₁ .

According to some example embodiments of the present invention, estimator 410 will generate signal 422 according to Equation 18:

Here, W ₂ represents a predetermined prediction filter (PF) vector of length L ₂ corresponding to the estimation function F ₂ .

According to some example embodiments of the present invention, based on certain criteria as described below, the estimator 408 repeatedly adapts the vector W ₁ and / or the estimator 410 replaces the vector W ₂ . Change it to be applicable repeatedly.

According to some example embodiments of the present invention, the estimator 408 is configured to determine the signal y ₁ (n) for the noise pattern and the noise cancellation pattern 418 at a predetermined point, such as point 312 (FIG. 3). The vector W ₁ will be iteratively adaptable based on a combination between estimates of the contribution, for example a noise error corresponding to the difference.

According to some example embodiments of the present invention, the controller 400 will also include a first evaluation module 430 to evaluate the noise error based on signals 412 and 414 as described below.

According to some example embodiments of the present invention, the first evaluation module 430 will include a synthesizer 434, for example to synthesize signals 412 and 420. For example, synthesizer 434 includes a first 1 / 2MTF module 436 to apply a predetermined first MTF (MTF ₁ ) to signal 412 and divide the result by two. Synthesizer 434 also includes a second 1 / 2MTF module 438 to apply a predetermined second MTF (MTF ₂ ) to signal 420 and divide the result by two. For example, MTF ₁ will be determined as known in the art based on the characteristics of the microphone 402 and / or based on the geometric and / or physical properties of the path between the microphone 402 and a particular point, where MTF ₂ is the microphone. It will be determined as known in the art based on the characteristics of 404 and / or based on the geometric and / or physical properties of the path between microphone 404 and a particular point. Combiner 434 is also output corresponding to the average between the estimate of the noise pattern of the n- th sample at a specific point by using an estimate of the n- th sample of the noise pattern at the specific point using the MTF _1, MTF ₂ d ( n) adder 440 to generate 434.

For example, d (n) may be calculated using the following equation (19):

Here, M ₁ represents a predetermined number of samples of MTF ₁ , and M ₂ represents a predetermined number of samples of MTF ₂ .

According to an exemplary embodiment of the invention, the first evaluation module 430 also applies an STF to the signal 414 to output an estimate of the primary portion of the noise cancellation pattern corresponding to the first order prediction signal 414. It will include an STF module 450 that generates 452. The STF will be determined based on the characteristics of the speaker 406 as known in the art, for example. The STF module 450 may include any suitable hardware and / or software as known in the art to apply a given STF to the signal 414. For example, the value z ₁ (n) of the output 452 may be calculated using the following equation (20):

Substituting Equation 17 into Equation 20 yields the following Equation 21:

According to an exemplary embodiment of the present invention, the first evaluation module 430 also includes a subtractor 454 implemented by any suitable hardware and / or software as known in the art. Subtractor 454 subtracts the value of output 452 from the value of output 442 to produce an output 455 that includes an estimated noise error e ₁ (n) corresponding to samples MIC1 (n) and MIC21 (n).

According to an exemplary embodiment of the present invention, the first order estimator 408 may implement an adaptive algorithm to repeatedly adapt the value of the vector W ₁ based on the value of e ₁ (n) as described below. will be.

According to an exemplary embodiment of the present invention, the noise error e1 (n) corresponding to the n-th sample received from the microphones 402, 404 will be evaluated using the following equation (22):

Substituting Equation 21 into Equation 22 yields the following Equation 23:

According to some example embodiments of the present invention, the estimator 408 will iteratively change the value of the vector W ₁ to reduce, eg, minimize, the estimated noise error e ₁ (n). For example, the estimator 408 may repeatedly adapt the value of the vector W ₁ repeatedly using equation 24:

Here, W ₁ (n + ₁₎ represents the value is changed to enable application of the W _1, W ₁ (n) denotes the current value of the W _1, μ ₁ represents the predetermined convergence parameter corresponding to W ₁ . For example, μ ₁ will be determined according to the condition of Equation 25:

According to some example embodiments of the present invention, estimator 410 corresponds to a combination, e.g., a difference, between the estimated noise error e ₁ (n) and an estimate of the contribution of y ₂ (n) to noise canceling pattern 418. Based on the estimated residual noise error, the value of the vector W ₂ of the estimation function F ₂ will be repeatedly changed to be applicable.

According to some example embodiments of the present invention, the controller 400 may also evaluate at least one second order to evaluate the residual noise error based on the signal 422 and the estimated noise error e ₁ (n) as described below. Module 432.

According to an exemplary embodiment of the invention, the secondary evaluation module 430 also applies an STF to the signal 422 to output 462 representing an estimate of the secondary portion of the noise cancellation pattern corresponding to the signal 422. It will include a STF module 460 to generate. The STF module 460 may include any suitable hardware and / or software as known in the art to apply a given STF to the signal 422. The STF will be determined based on, for example, the characteristics of the speaker 406 as known in the art. For example, the value z ₂ (n) of the output 462 may be calculated using Equation 26:

Substituting Equation 26 into Equation 18 yields the following equation:

According to one embodiment of the present invention, secondary evaluation module 432 also includes a subtractor 464 implemented by any suitable hardware and / or software as known in the art. Subtractor 464 subtracts the value of output 462 from the value of output 452 to produce an output 466 that includes an estimated residual noise error e ₂ (n) corresponding to samples MIC1 (n) and MIC21 (n).

According to one exemplary embodiment of the present invention, estimator 410 will implement an adaptive algorithm to repeatedly adapt the value of vector W ₂ based on the value of e ₂ (n) as described below.

According to an exemplary embodiment of the present invention, the residual noise error e ₂ (n) corresponding to the n-th sample received from the microphones 402, 404 will be evaluated using Equation 28:

Substituting Equations 23 and 27 into Equation 28 will yield the following Equation 29:

According to some example embodiments of the present invention, estimator 410 will iteratively change the value of vector W ₂ to reduce, eg, minimize, the estimated residual noise error e ₂ (n). For example, the estimator 410 may repeatedly change one or more components of the vector W ₁ using the following equation (30).

Here, W ₂ (n + 1) denotes the value is changed to enable application of the W _2, W ₂ (n) denotes the current value of the W _2, μ ₂ indicates a predetermined convergence parameter corresponding to W _2. For example, μ ₂ will be determined according to the condition of Equation 31:

Some of the embodiments described above are based on a combination of a primary acoustic signal such as the microphone 402's primary noise signal and a secondary acoustic sensor such as the microphone 404's secondary noise signal to generate a noise canceling pattern. An ANC system that implements a controller such as a controller 400 capable of controlling 406 is shown. However, it will be understood by those skilled in the art that according to another embodiment of the present invention these systems may be modified to implement one or more additional secondary acoustic sensors. For example, controller 400 may be modified to include a plurality of additional secondary estimators, respectively, to receive one or more primary noise signals of one or more additional secondary microphones. For example, the i-th estimator of the additional quadratic estimator will produce an output y _i (n) that corresponds, for example, to:

Where W _i represents a predetermined predictive filter (PF) vector of length L _i corresponding to the i-th estimator and MICi represents the output of the i-th additional secondary microphone.

Controller 400 may also be modified to include one or more additional residual noise error evaluators to evaluate residual noise errors similar to evaluator 410. For example, the i-th residual error evaluator will evaluate the i-th noise error e _i (n) using Equation 33:

According to some example embodiments of the present invention, the i-th evaluator of the additional evaluator will iteratively change the value of the vector W _i using equation 34:

Here, W _i (n + 1) represents the value is changed to enable application of the W _i, W _i (n) denotes the current value of W _i, μ ₁ represents the predetermined convergence parameter corresponding to W _i . For example, μ _i will be determined according to the condition of Equation 35:

Some of the embodiments described above may include a controller such as a controller 400 that includes one or more estimators, such as 408 and / or 410, to apply an adaptive linear estimation algorithm to one or more respective noise signals, such as outputs 412 and / or 420. Represents an ANC system that implements However, according to another embodiment of the present invention, these systems may be modified to implement one or more estimators to apply an adaptive nonlinear estimation algorithm to one or more respective noise signals. For example, controller 400 may be modified to implement one or more RBF estimation algorithms similar to controller 200 (FIG. 2).

Embodiments of the invention will be implemented by software, hardware or a combination of software and / or hardware that will be suitable for a particular application or conform to a particular design requirement. Embodiments of the present invention include modules, units and sub-units that are to be separated from one another or in whole or in part to be combined with one another, and include special processors or devices, multi-use processors or devices, or general purpose processors as known in the art. Or using a device. Some embodiments of the present invention include buffers, registers, storage units and / or memory devices for temporary storage or long term storage of data and / or to facilitate operation of certain embodiments.

Although specific features of the invention have been illustrated and described herein, those skilled in the art will appreciate that many modifications, substitutions, changes, and equivalents may be made. Therefore, it is to be understood that the appended claims are intended to cover all such modifications and variations as fall within the true spirit of the invention.

Claims (31)

An active noise control system for controlling noise generated by a noise source, An acoustic sensor that detects a noise pattern and generates a noise signal corresponding to the detected noise pattern; An estimator for generating a predictive noise signal by applying an estimation function to the noise signal; And An acoustic transducer that generates a noise destructive pattern based on the predicted noise signal Active noise control system comprising a. The method of claim 1, And the estimator is capable of adaptively changing one or more parameters of the estimation function based on the noise error at a given point. The method of claim 2, And said noise error comprises an expected cancellation interference between said noise pattern and said noise cancellation pattern at said predetermined point. The method according to claim 2 or 3, And an error sensing microphone for detecting the noise error at the predetermined point. The method according to claim 2 or 3, And an error evaluator to evaluate the noise error based on the noise signal and the predicted noise signal. The method of claim 5, The error evaluator, A speaker transfer function module for generating an estimate of the noise cancellation pattern by applying a speaker transfer function to the predicted noise signal; A modulation transfer function module for generating an estimate of the noise pattern at the predetermined point by applying a modulation transfer function to the noise signal; And A subtractor for subtracting the estimate of the noise cancellation pattern from the estimate of the noise pattern Active noise control system comprising a. The method according to any one of claims 2 to 6, The estimator is capable of adaptively changing the one or more parameters based on a predetermined criterion. The method of claim 7, wherein And the estimator can reduce the error value by adapting the one or more parameters as applicable. The method of claim 8, And the adaptive estimator is capable of minimizing the error value by applicably changing the one or more parameters. The method according to any one of claims 2 to 9, Wherein said at least one parameter comprises at least one parameter selected from the group consisting of a center parameter, an effective radius parameter, and an intensity parameter. The method of claim 10, The estimator may change the center parameter to be applicable based on the following equation,

Where c _k (n + 1) represents an applicable modified value of the center parameter, c _k (n) represents the current value of the center parameter, w _k represents the intensity parameter, and L represents the noise Represents a predetermined number of samples of the signal, STF represents a predetermined speaker transfer function, S represents a predetermined speaker transfer function frequency parameter, μ _c represents a predetermined convergence parameter corresponding to the center parameter, and v _k represents The effective radius parameter, e (n) represents the noise error, f _k represents a predetermined function, and x (n) represents the n-th sample of the noise signal. . The method of claim 10, The estimator may adaptively change the effective radius parameter based on the following equation,

Where v _k (n + 1) represents an applicable changed value of the effective radius parameter, v _k (n) represents the current value of the effective radius parameter, w _k represents the strength parameter, and L is Represents a predetermined number of samples of the noise signal, STF represents a predetermined speaker transfer function, S represents a predetermined speaker transfer function frequency parameter, μ _v represents a predetermined convergence parameter corresponding to the effective radius parameter, c _k denotes the center parameter, e (n) denotes the noise error, f _k denotes a predetermined function, and x (n) denotes the n-th sample of the noise signal Control system. The method of claim 10, The estimator may change the intensity parameter to be applicable based on the following equation,

Where w _k (n + 1) represents an applicable modified value of the intensity parameter, w _k (n) represents the current value of the intensity parameter, w _k represents the intensity parameter, and L represents the noise Represents a predetermined number of samples of the signal, STF represents a predetermined speaker transfer function, S represents a predetermined speaker residual function frequency parameter, μ _w represents a predetermined convergence parameter corresponding to the intensity parameter, and f _k represents Representing a predetermined function, wherein x (n) represents an n-th sample of the noise signal. The method according to any one of claims 1 to 13, And said estimation function comprises a non-linear estimation function. The method of claim 14, And the nonlinear estimation function comprises a radial basis function. The method according to any one of claims 1 to 15, And said acoustic sensor comprises a microphone. The method according to any one of claims 1 to 16, And said acoustic transducer comprises a speaker. An active noise control system for controlling noise generated by a noise source, A primary acoustic sensor for detecting a noise pattern and generating a corresponding primary noise signal; Detecting a residual noise pattern to generate at least one secondary noise signal corresponding to the sensed residual noise pattern, respectively, wherein the at least one distance from the noise source is greater than the distance between the primary acoustic sensor and the noise source Secondary acoustic sensors; And A controller for controlling an acoustic transducer to generate a noise cancellation pattern based on the primary noise signal and at least one secondary noise signal Active noise control system comprising a. The method of claim 18, The controller, A first estimator for generating a first order prediction signal by applying a first order estimation function to the first noise signal; And At least one second estimator for generating at least one second prediction signal by applying at least one second estimation function to at least one second noise signal Active noise control system comprising a. The method of claim 19, And the primary estimator is capable of iteratively changing one or more parameters of the primary estimation function based on the noise error. The method of claim 19 or 20, And said at least one secondary estimator is capable of iteratively changing one or more parameters of at least one said second estimation function, respectively, based on a noise error. The method according to any one of claims 19 to 21, And the controller is capable of controlling the acoustic transducer based on the first predictive signal and the at least one second predictive signal. The method of claim 22, And the controller is capable of controlling the acoustic transducer based on the sum of the first prediction signal and the at least one second prediction signal. The method according to any one of claims 20 to 23, wherein And said controller comprises a noise error evaluator to evaluate said noise error. The method of claim 24, And the noise error evaluator is capable of evaluating the noise error based on the first noise signal, at least one second noise signal, and the first prediction signal. The method of claim 25, The noise error evaluator, A speaker transfer function module for generating an estimate of the primary portion of the noise cancellation pattern corresponding to the first predictive signal by applying a speaker transfer function to the first predictive signal; A modulation transfer function module for generating an estimate of the noise pattern by applying a modulation transfer function to the combination of the primary noise signal and the at least one secondary noise signal; And A subtractor for subtracting an estimate of the primary portion of the noise cancellation pattern from the estimate of the noise pattern Active noise control system comprising a. The method according to any one of claims 24 to 26, The controller comprises at least one residual noise evaluator to evaluate at least one residual noise. The method of claim 27, And at least one residual noise evaluator can evaluate the residual noise based on the noise error and at least one second prediction signal, respectively. The method of claim 28, The residual error evaluator, A speaker transfer function module for generating an estimate of a secondary portion of the noise cancellation pattern corresponding to the second prediction signal by applying a speaker transfer function to the second prediction signal; And A subtractor for subtracting an estimate of the secondary portion of the noise cancellation pattern from the noise error Active noise control system comprising a. The method according to any one of claims 18 to 29, wherein And at least one of the primary acoustic sensor and the at least one secondary acoustic sensor comprises a microphone. The method according to any one of claims 18 to 30, And said acoustic transducer comprises a speaker.

Images