Wide-narrow band hybrid active noise control system and method with acoustic feedback compensation

Abstract

The invention discloses a wide-narrow band hybrid active noise control system and method with acoustic feedback compensation, and belongs to the technical field of active noise control. According to the system provided by the invention, the acoustic feedback compensation subsystem is adopted, so that the influence of acoustic feedback on a reference signal is effectively overcome, and the quality of the reference signal is improved; a wide-narrow band reference signal synthesis structure based on the combination of a front serial self-adaptive band-pass filter and a rear parallel self-adaptive band-pass filter is adopted to obtain a wide-band reference signal and a narrow-band reference signal to respectively provide reliable input for a wide-band secondary sound source synthesis subsystem and a narrow-band secondary sound source synthesis subsystem; therefore, the stability, convergence and noise reduction performance of the system are improved on the whole. The system provided by the invention can cope with non-stably changed wide-narrow band mixed target noise. The system provided by the invention does not need a non-acoustic reference microphone, so that not only is frequency deviation avoided, but also the hardware cost of the system is reduced, and the practical application range of the system is widened.

Description

Technical Field

[0001] The invention relates to a wide-narrowband hybrid active noise control system and method including acoustic feedback compensation, belonging to the technical field of active noise control. Background Art

[0002] Traditional passive noise reduction technology uses acoustic structures or materials to achieve noise reduction through sound absorption, sound barriers, vibration isolation, vibration reduction and other means. It usually has a good noise reduction effect on medium and high frequency noise. Active noise control (ANC) is based on the principle of destructive interference of sound waves. It has good noise reduction performance for low-frequency noise, as well as advantages such as small size and low cost (SM Kuo and DRMorgan, "Active noise control: a tutorial review," Proc. IEEE, vol. 87, no. 6, pp. 943-973, Jun. 1999.), which is a beneficial supplement to passive noise reduction technology. Rotating mechanical equipment such as cutting machines, fans, and engines not only generate a large amount of periodic interference in working conditions, but also broadband interference. The noise reduction performance of traditional narrowband or broadband feedforward ANC systems for this type of wide- and narrowband mixed noise is limited.

[0003] In 2011, scholars Xiao et al. proposed a wide-narrowband hybrid ANC system based on a sinusoidal noise canceller, which overcomes the "spark" noise problem of the traditional broadband ANC system in suppressing wide-narrowband hybrid target noise (Y. Xiao and J. Wang, "A new feedforward hybrid active noise control system," IEEE Signal Process. Letters, vol. 18, no. 10, pp. 591-594, Oct. 2011.). The system uses both an acoustic reference sensor and a non-acoustic reference sensor to obtain the reference signal and the narrowband frequency component in the reference signal respectively. However, the non-acoustic reference sensor (such as a tachometer, etc.) may age, wear, etc., which may cause the obtained narrowband reference frequency to be inconsistent with the actual narrowband reference frequency, that is, frequency offset occurs. The above frequency offset seriously affects the noise reduction performance of the wide-narrowband hybrid ANC system, and even restricts the convergence and stability of the wide-narrowband hybrid ANC system, which is not conducive to its practical application. In addition, the wide-narrowband hybrid ANC system ignores the problem of acoustic feedback, which seriously restricts the scope of application of the system.

[0004] In 2013, in order to solve the frequency offset problem, scholar Xiao improved the wide-narrowband hybrid ANC system based on the sinusoidal noise canceller and proposed a structure using a parallel IIR filter to compensate for the frequency offset. The simulation results show that the structure has a good noise reduction effect for smaller frequency offsets, thereby improving the noise reduction performance of the wide-narrowband ANC system (Y. Xiao and K. Doi, "Arobust hybrid active noise control system using IIR notch filters, Int. J. Advanced Mechatronic Systems," vol. 5, no. 1, pp. 69-77, 2013.). However, the wide-narrowband hybrid ANC system based on the parallel IIR filter cannot cope with the situation of large frequency shift offset, and ignores the acoustic feedback problem, which restricts its noise reduction level. In addition, the above-mentioned traditional wide-narrowband hybrid ANC systems all require non-acoustic sensors to obtain the narrowband component frequency in the target noise as prior information, which increases the complexity and cost of the hardware system. Many scholars have subsequently conducted in-depth research on the anti-interference ability and dynamic performance improvement of the wide-narrowband hybrid ANC system, but none of them fully considers the impact of acoustic feedback on the system. Therefore, it is urgent to develop a new wide- and narrow-band hybrid ANC system to effectively solve the problem of limited system noise reduction performance caused by acoustic feedback and frequency offset, while reducing hardware costs. Summary of the invention

[0005] In order to solve the problems of acoustic feedback and frequency offset in the conventional wide-narrowband hybrid active noise control system, which leads to reduced wide-narrowband hybrid noise suppression performance and stability of the overall system and high system complexity, the present invention provides a wide-narrowband hybrid active noise control system and method with acoustic feedback compensation. The technical solution of the present invention is as follows:

[0006] The first object of the present invention is to provide a wide-band and narrow-band hybrid active noise control system with acoustic feedback compensation, wherein the active noise control system uses a reference microphone to collect a reference signal, uses an error microphone to collect an error signal, and uses a secondary speaker to provide a secondary sound source; the actual primary channel in the acoustic space is a channel model from the reference signal to the error microphone; the actual acoustic feedback channel in the acoustic space is a channel model from the secondary sound source to the reference microphone; the actual secondary channel in the acoustic space is a channel model from the secondary sound source to the error microphone; the active noise control system comprises an acoustic feedback compensation subsystem (1), a wide-band and narrow-band reference signal synthesis subsystem (2), a wideband secondary sound source synthesis subsystem (3), a narrowband secondary sound source synthesis subsystem (4), and a wide-band and narrowband secondary sound source summing module (5);

[0007] The acoustic feedback compensation subsystem (1) is connected to the wide- and narrow-band reference signal synthesis subsystem (2) and the wide- and narrow-band secondary sound source summing module (5) respectively; the wide- and narrow-band reference signal synthesis subsystem (2) is connected to the wide- and narrow-band secondary sound source synthesis subsystem (3) and the narrow-band secondary sound source synthesis subsystem (4) respectively; the wide- and narrow-band secondary sound source synthesis subsystem (3) is connected to the wide- and narrow-band secondary sound source summing module (5); the narrow-band secondary sound source synthesis subsystem (4) is connected to the wide- and narrow-band secondary sound source summing module (5);

[0008] The acoustic feedback compensation subsystem (1) is used to synthesize an acoustic feedback signal and compensate for the influence of the actual acoustic feedback signal on the actual reference signal; the acoustic feedback compensation subsystem (1) comprises: an acoustic feedback channel estimation model (11) and a summation operation, which can provide an estimate of the actual reference signal, which is then used as an input of the wide-narrowband reference signal synthesis subsystem (2);

[0009] The acoustic feedback channel estimation model (11) The actual acoustic feedback channel F(z) is obtained by offline identification, which is represented by a linear finite impulse response filter, whose coefficients and lengths are and Output of acoustic feedback compensation subsystem (1) Expressed as

[0010]

[0011] Among them, x r (n) is the reference signal provided by the reference microphone; y(n) is the secondary sound source; n is the time, n≥0;

[0012] The wide-narrowband reference signal synthesis subsystem (2) is used to synthesize a narrowband reference signal and a wideband reference signal; the wide-narrowband reference signal synthesis subsystem (2) comprises: a pre-serial type adaptive bandpass filter group (21) and a post-parallel type adaptive bandpass filter group (22);

[0013] The front serial adaptive bandpass filter group (21) is composed of q adaptive bandpass filters in a serial manner, and the value of q is not less than the number of narrowband frequency components in the target noise; each adaptive bandpass filter is composed of a second-order IIR notch filter; the front serial adaptive bandpass filter group (21) provides a high signal-to-noise ratio input for the rear parallel adaptive bandpass filter group (22);

[0014] The post-parallel adaptive bandpass filter group (22) is composed of q adaptive bandpass filters in parallel, providing q narrowband reference signals Using the output of the acoustic feedback compensation subsystem (1) Subtract the sum of all narrowband reference signals output by the post-parallel adaptive bandpass filter module (22) to obtain a broadband reference signal x w (n);

[0015] The reference input of the broadband secondary sound source synthesis subsystem (3) is the broadband reference signal x separated by the broadband and narrowband reference signal synthesis subsystem (2). w (n); at the same time, the q reference inputs of the narrowband secondary sound source synthesis subsystem (4) are q narrowband reference signals provided by the wide-narrowband reference signal synthesis subsystem (2)

[0016] The coefficients related to the center frequencies of the q adaptive bandpass filters in the post-parallel adaptive bandpass filter module (22) are The initial values ​​are all -2, 0 or [0,1,…,q-1]π / (2q), and no non-acoustic sensor is required to obtain prior information on the narrowband frequency components in the reference signal;

[0017] The wide- and narrow-band secondary sound source summing module (5) includes a first-order delay operation and a summing operation, and sums the narrow-band secondary sound source and the wide-band secondary sound source to provide a secondary sound source for the secondary speaker.

[0018] In one embodiment of the present invention, in the wide-narrowband reference signal synthesis subsystem (2), the z-domain model H of the i-th adaptive bandpass filter i (z) is:

[0019]

[0020] Where ρ is the polar radius parameter, ρ∈[0,1); c i is time-varying, c i (n) = -2cos[ω i (n)] is the coefficient related to the center frequency of the i-th adaptive bandpass filter, ω i (n) is the center frequency of the i-th adaptive bandpass filter; z -1 represents a first-order delay operation; z -2 represents a second-order delay operation;

[0021] The output of the i-th adaptive bandpass filter of the pre-serial adaptive bandpass filter group (21) is:

[0022] y d,i (n) = y c,i-1 (n)-y c,i (n)

[0023] The output of the i-th second-order IIR notch filter of the pre-serial adaptive bandpass filter bank (21) is:

[0024] y c,i (n)=-ρc i (n)y c,i (n-1)-ρ 2 y c,i (n-2)+y c,i-1 (n)+c i (n)y c,i-1 (n-1)+y c,i-1 (n-2)

[0025] Among them, y c,i (n) is the output of the i-th second-order IIR notch filter, y c,0 (n) = x(n);

[0026] The input and output of the i-th IIR notch filter of the post-parallel adaptive bandpass filter bank (22) are respectively:

[0027]

[0028] y b,i (n)=-ρc i (n)y b,i (n-1)-ρ 2 y b,i (n-2)+x d,i (n)+c i (n)x d,i (n-1)+x d,i (n-2)

[0029] The output of the i-th adaptive bandpass filter of the post-parallel adaptive bandpass filter bank (22) is:

[0030] x i (n) = x d,i (n)-y b,i (n)

[0031] Among them, x i (n) serving as a narrowband reference signal input for the narrowband secondary sound source synthesis subsystem (4);

[0032] The broadband reference signal provided by the broadband and narrowband reference signal synthesis subsystem (2) is:

[0033]

[0034] Among them, x w (n) is used as a broadband reference signal input for the broadband secondary sound source synthesis subsystem (3).

[0035] In one embodiment of the present invention, in the wide-narrowband reference signal synthesis subsystem (2), a normalized smooth gradient algorithm is used to calculate the coefficient c related to the center frequency of the i-th adaptive bandpass filter. i (n) To update, that is:

[0036]

[0037] g i (n) = x d,i (n-1)-ρy b,i (n-1);

[0038]

[0039] In the formula, μ c c i (n); ε is a non-zero positive number; λ is the forgetting factor, λ∈(0,1); when the system reaches a steady state, the coefficient c i (n) converges to -2cos(ω i,s ), where ω i,s is the reference signal x r The true frequency value of the i-th narrowband component in (n).

[0040] In one embodiment of the present invention, the broadband secondary sound source synthesis subsystem (3) comprises: a broadband controller (31) and a first filter-X minimum mean square algorithm module (32) with a first-order delay; the broadband controller (31) adopts a linear finite impulse response filter, and the first filter-X minimum mean square algorithm module (32) with a first-order delay is used to update the broadband controller (31);

[0041] The broadband controller (31) adopts a linear finite impulse response filter, the coefficient of which is The length is L; the coefficient The update formula is:

[0042]

[0043] In the formula, μ w is the coefficient The update step size is ; e(n) is the residual noise; is the broadband reference signal x w (n) Estimation model via secondary channel Output.

[0044] In one embodiment of the present invention, the broadband secondary sound source synthesis subsystem (3) is used to synthesize a broadband secondary sound source, and then used to suppress the noise component related to the broadband reference signal in the target noise; the synthesized broadband secondary sound source is

[0045]

[0046] In one embodiment of the present invention, the narrowband secondary sound source synthesis subsystem (4) comprises: a narrowband controller (41) and a second filter-X minimum mean square algorithm module (42) with a first-order delay; the narrowband controller (41) adopts an amplitude phase adjustment structure and uses the second filter-X minimum mean square algorithm module (42) with a first-order delay to update the narrowband controller (41);

[0047] The coefficients of the narrowband controller (41) The update formula is:

[0048]

[0049] In the formula, μ f is the coefficient The update step size is a positive value; is the narrowband reference signal x i (n) Estimation model via secondary channel Output.

[0050] In one embodiment of the present invention, the narrowband secondary sound source synthesis subsystem (4) is used to synthesize a narrowband secondary sound source, and then used to suppress the noise component related to the narrowband reference signal in the target noise; the synthesized narrowband secondary sound source is

[0051]

[0052] In one embodiment of the present invention, the first filter-X minimum mean square algorithm module (32) with first-order delay and the second filter-X minimum mean square algorithm module (42) with first-order delay, the secondary channel estimation model The secondary channel estimation model is obtained by offline identification of the actual secondary channel S(z); A linear finite impulse response filter is used, whose coefficients are Length is

[0053] In one embodiment of the present invention, the wide-narrowband secondary sound source summing module (5) includes a first-order delay operation and a summing operation; the wide-narrowband secondary sound source summing module (5) is used to synthesize the secondary sound source, that is:

[0054] y(n)=y w (n-1)+y f (n-1)

[0055] In the acoustic space, the target noise and the signal y(n) sent to the secondary speaker and after passing through the actual secondary channel model S(z) interfere with each other in the local noise reduction area to achieve noise reduction.

[0056] The second object of the present invention is to provide a wide-narrowband hybrid active noise control method with acoustic feedback compensation, the method adopts the wide-narrowband hybrid active noise control system with acoustic feedback compensation, the method comprises:

[0057] Step 1: Offline identification of actual acoustic feedback channel and actual secondary channel:

[0058] The actual acoustic feedback channel F(z) is modeled using offline identification technology to obtain the acoustic feedback channel estimation model. The coefficients and lengths are and At the same time, the offline identification technology is used to realize the modeling of the actual secondary channel S(z) and obtain the secondary channel estimation model The coefficient and length of and

[0059] Step 2: System initial value setting:

[0060] The lengths and update steps of the broadband controller (31) and the narrowband controller (41) are respectively set; the initial values ​​of the broadband controller (31) and the narrowband controller (41) are respectively set to zero; the polar radius parameters ρ and c are set. i (n) The update of μ required c , ε, λ; set coefficient The initial value of

[0061] Step 3: System status at time n:

[0062] At time n, the secondary sound source y(n) is obtained, and then the output of the acoustic feedback compensation subsystem (1) is obtained. The wide-band and narrow-band reference signal synthesis subsystem (2) synthesizes a wideband reference signal x w (n) and narrowband reference signal The broadband secondary sound source synthesis subsystem (3) and the narrowband secondary sound source synthesis subsystem (4) provide broadband secondary sound sources and narrowband secondary sound sources respectively;

[0063] Step 4: System status update:

[0064] The coefficients c of the post-parallel adaptive bandpass filter bank (22) are updated using the normalized smoothed gradient algorithm. i (n), and obtain the corresponding coefficient c i(n) at time n+1; at the same time, using the first filter-X minimum mean square algorithm module (32) with a first-order delay to update the coefficient of the broadband controller (31) to obtain the coefficient of the corresponding broadband controller (31) at time n+1; at the same time, using the second filter-X minimum mean square algorithm module (42) with a first-order delay to update the coefficient of the narrowband controller (41) to obtain the coefficient of the corresponding narrowband controller (41) at time n+1;

[0065] Step 5: Return to step 3, and repeat steps 3 to 4 until the system converges and reaches a steady state, thus completing noise control.

[0066] The beneficial effects of the present invention are:

[0067] 1. The system of the present invention adopts a front serial adaptive bandpass filter group and a rear parallel adaptive bandpass filter group to obtain a narrowband reference signal and a broadband reference signal without using a non-acoustic reference microphone, which not only avoids frequency offset but also reduces the hardware cost of the system;

[0068] Second, the present invention adopts an acoustic feedback compensation subsystem to effectively overcome the influence of acoustic feedback on the reference signal, improve the quality of the reference signal, and thus improve the stability and convergence of the system as a whole;

[0069] 3. The present invention adopts a post-parallel adaptive bandpass filter group to adaptively provide narrowband frequency information in the target noise, providing an important reference for health monitoring or fault diagnosis of rotating mechanical equipment;

[0070] 4. The present invention adopts an acoustic feedback compensation subsystem, a wide-band and narrow-band reference signal synthesis subsystem, a broadband secondary sound source synthesis subsystem and a narrow-band secondary sound source synthesis subsystem. The four are organically combined to improve the wide-band and narrow-band noise suppression performance of the system as a whole, so that the residual error mean square error of the system gradually approaches the ambient noise level, which is beneficial to noise reduction applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0072] Figure 1 This is a schematic diagram of a wide-narrowband hybrid active noise control system with acoustic feedback compensation according to the first embodiment.

[0073] FIG2( a ) is a dynamic variation curve of the mean square residual error in the system of Example 3;

[0074] FIG2( b ) is a curve showing the dynamic change of coefficients related to the center frequency of the adaptive bandpass filter in the system of Embodiment 3;

[0075] FIG3( a ) is a dynamic change curve of target noise and residual noise in the system of Example 4;

[0076] FIG3( b ) is a dynamic change curve of coefficients related to the center frequency of the adaptive bandpass filter in the system at 1400 rpm in the fourth embodiment;

[0077] FIG3( c ) is a curve showing the dynamic change of coefficients related to the center frequency of the adaptive bandpass filter in the system at 1600 rpm in the fourth embodiment.

[0078] In the figure: 1. Acoustic feedback compensation subsystem; 2. Wide- and narrow-band reference signal synthesis subsystem; 3. Wide-band secondary sound source synthesis subsystem; 4. Narrow-band secondary sound source synthesis subsystem; 5. Wide- and narrow-band secondary sound source summation module; 11. Acoustic feedback channel estimation model; 21. Pre-series adaptive bandpass filter group, 22. Post-parallel adaptive bandpass filter group, 31. Wideband controller, 32. First filter with first-order delay - X minimum mean square algorithm module, 41. Narrowband controller, 42. Second filter with first-order delay - X minimum mean square algorithm module. DETAILED DESCRIPTION

[0079] The present invention is described in detail below. In the following paragraphs, different aspects of the embodiments are defined in more detail. Each aspect so defined can be combined with any other aspect or aspects, unless explicitly stated not to be combined. In particular, any feature that is considered to be preferred or advantageous can be combined with one or more other features that are considered to be preferred or advantageous.

[0080] The terms "first", "second" and the like appearing in the present invention are only for the convenience of description to distinguish different components with the same name, and do not indicate a priority or primary and secondary relationship.

[0081] In addition, when an element is referred to as being "on" another element, the element may be directly on the other element, or may be indirectly on the other element with one or more intermediate elements interposed therebetween. In addition, when an element is referred to as being "connected to" another element, the element may be directly connected to the other element, or may be indirectly connected to the other element with one or more intermediate elements interposed therebetween. Hereinafter, the same reference numerals represent the same elements.

[0082] Embodiment 1:

[0083] The present embodiment provides a wide-band and narrow-band hybrid active noise control system with acoustic feedback compensation, wherein the active noise control system respectively uses a reference microphone to collect a reference signal, uses an error microphone to collect an error signal, and uses a secondary speaker to provide a secondary sound source; the actual primary channel in the acoustic space is a channel model from the reference signal to the error microphone; the actual acoustic feedback channel in the acoustic space is a channel model from the secondary sound source to the reference microphone; the actual secondary channel in the acoustic space is a channel model from the secondary sound source to the error microphone; the active noise control system comprises an acoustic feedback compensation subsystem 1, a wide-band and narrow-band reference signal synthesis subsystem 2, a wide-band and narrow-band secondary sound source synthesis subsystem 3, a narrow-band and narrow-band secondary sound source synthesis subsystem 4, and a wide-band and narrow-band secondary sound source summing module 5;

[0084] The acoustic feedback compensation subsystem 1 is connected to the wide- and narrow-band reference signal synthesis subsystem 2 and the wide- and narrow-band secondary sound source summing module 5 respectively; the wide- and narrow-band reference signal synthesis subsystem 2 is connected to the wide- and narrow-band secondary sound source synthesis subsystem 3 and the narrow-band secondary sound source synthesis subsystem 4 respectively; the wide- and narrow-band secondary sound source synthesis subsystem 3 is connected to the wide- and narrow-band secondary sound source summing module 5; the narrow-band secondary sound source synthesis subsystem 4 is connected to the wide- and narrow-band secondary sound source summing module 5;

[0085] Figure 1 As shown, the target noise in the acoustic space is

[0086] p(n)=p0(n)+v p (n)

[0087] Among them, p0(n) is the corresponding reference signal x in the acoustic space s (n) The signal propagated to the error microphone through the actual primary channel P(z); v p (n) has a mean of zero and a variance of Additive Gaussian white noise; n is the moment, n≥0.

[0088] Acoustic feedback signal y in acoustic space f (n) is the signal transmitted from the secondary sound source y(n) to the reference microphone through the actual sound feedback channel F(z). The reference microphone obtains the reference signal x r (n) = x s (n)+y f (n), where x s (n) includes broadband component x B (n) and the narrowband component x N (n).

[0089] The acoustic feedback compensation subsystem 1 is used to synthesize the acoustic feedback signal and compensate for the influence of the actual acoustic feedback signal on the actual reference signal; the acoustic feedback compensation subsystem 1 includes: an acoustic feedback channel estimation model 11 and a summation operation, which can provide an estimate of the actual reference signal, and then serve as an input of the wide-narrowband reference signal synthesis subsystem 2;

[0090] The acoustic feedback channel estimation model The actual acoustic feedback channel F(z) is obtained by offline identification, which is represented by a linear finite impulse response filter, whose coefficients and lengths are and Output of acoustic feedback compensation subsystem 1 Expressed as

[0091]

[0092] The wide-narrowband reference signal synthesis subsystem 2 is used to synthesize a narrowband reference signal and a wideband reference signal; the wide-narrowband reference signal synthesis subsystem 2 comprises: a pre-serial type adaptive bandpass filter group 21 and a post-parallel type adaptive bandpass filter group 22;

[0093] The front serial adaptive bandpass filter group 21 is composed of q adaptive bandpass filters in a serial manner, and the value of q is not less than the number of narrowband frequency components in the target noise; each adaptive bandpass filter is composed of a second-order IIR notch filter; the front serial adaptive bandpass filter group 21 provides a high signal-to-noise ratio input for the rear parallel adaptive bandpass filter group 22;

[0094] The post-parallel adaptive bandpass filter group 22 is composed of q adaptive bandpass filters in parallel to provide q narrowband reference signals. Using acoustic feedback to compensate the output of subsystem 1 Subtract the sum of all narrowband reference signals output by the post-parallel adaptive bandpass filter module 22 to obtain the broadband reference signal x w (n);

[0095] The reference input of the broadband secondary sound source synthesis subsystem 3 is the broadband reference signal x separated by the broadband and narrowband reference signal synthesis subsystem 2. w (n); At the same time, the q reference inputs of the narrowband secondary sound source synthesis subsystem 4 are q narrowband reference signals provided by the wide-narrowband reference signal synthesis subsystem 2

[0096] The coefficients related to the center frequencies of the q adaptive bandpass filters in the post-parallel adaptive bandpass filter module 22 The initial values ​​are all -2, 0 or [0,1,…,q-1]π / (2q), and no non-acoustic sensor is required to obtain prior information on the narrowband frequency components in the reference signal;

[0097] The wide- and narrow-band secondary sound source summing module 5 includes a first-order delay operation and a summing operation, and sums the narrow-band secondary sound source and the wide-band secondary sound source to provide a secondary sound source for the secondary speaker.

[0098] In the wide-narrowband reference signal synthesis subsystem 2, the z-domain model H of the i-th adaptive bandpass filter i (z) is:

[0099]

[0100] Where ρ is the polar radius parameter, ρ∈[0,1); c i is time-varying, c i (n) = -2cos[ω i (n)] is the coefficient related to the center frequency of the i-th adaptive bandpass filter, ω i (n) is the center frequency of the i-th adaptive bandpass filter; z -1 represents a first-order delay operation; z -2 represents a second-order delay operation;

[0101] The output of the i-th adaptive bandpass filter of the pre-serial adaptive bandpass filter group 21 is:

[0102] y d,i (n) = y c,i-1 (n)-y c,i (n)

[0103] The output of the i-th second-order IIR notch filter of the pre-serial adaptive bandpass filter bank 21 is:

[0104] y c,i (n)=-ρc i (n)y c,i (n-1)-ρ 2 y c,i (n-2)+y c,i-1 (n)+c i (n)y c,i-1 (n-1)+y c,i-1 (n-2)

[0105] Among them, y c,i (n) is the output of the i-th second-order IIR notch filter, y c,0 (n) = x(n);

[0106] The input and output of the i-th IIR notch filter of the post-parallel adaptive bandpass filter bank 22 are respectively:

[0107]

[0108] y b,i (n)=-ρc i (n)y b,i (n-1)-ρ 2 y b,i (n-2)+x d,i (n)+c i (n)x d,i (n-1)+x d,i (n-2)

[0109] The output of the i-th adaptive bandpass filter of the post-parallel adaptive bandpass filter group 22 is:

[0110] x i (n) = x d,i (n)-y b,i (n)

[0111] Among them, x i (n) used as a narrowband reference signal input for the narrowband secondary sound source synthesis subsystem 4;

[0112] The broadband reference signal provided by the broadband and narrowband reference signal synthesis subsystem 2 is:

[0113]

[0114] Among them, x w (n) Used as the broadband reference signal input of the broadband secondary sound source synthesis subsystem 3.

[0115] In the wide-narrowband reference signal synthesis subsystem 2, the coefficient c related to the center frequency of the i-th adaptive bandpass filter is calculated using the normalized smooth gradient algorithm. i (n) To update, that is:

[0116]

[0117] g i (n) = x d,i (n-1)-ρy b,i (n-1);

[0118]

[0119] In the formula, μ c c i(n); ε is a non-zero positive number; λ is the forgetting factor, λ∈(0,1); when the system reaches a steady state, the coefficient c i (n) converges to -2cos(ω i,s ), where ω i,s is the reference signal x r The true frequency value of the i-th narrowband component in (n).

[0120] The broadband secondary sound source synthesis subsystem 3 includes: a broadband controller 31 and a first filter-X minimum mean square algorithm module 32 with a first-order delay; the broadband controller 31 adopts a linear finite impulse response filter, and the broadband controller 31 is updated by using the first filter-X minimum mean square algorithm module 32 with a first-order delay;

[0121] The broadband controller 31 adopts a linear finite impulse response filter, whose coefficient is The length is L; the coefficient The update formula is:

[0122]

[0123] In the formula, μ w is the coefficient The update step size is ; e(n) is the residual noise; is the broadband reference signal x w (n) Estimation model via secondary channel Output.

[0124] The broadband secondary sound source synthesis subsystem 3 is used to synthesize broadband secondary sound sources, and then used to suppress the noise component related to the broadband reference signal in the target noise; the synthesized broadband secondary sound source is

[0125]

[0126] The narrowband secondary sound source synthesis subsystem 4 includes: a narrowband controller 41 and a second filter-X minimum mean square algorithm module 42 with a first-order delay; the narrowband controller 41 adopts an amplitude phase adjustment structure and uses the second filter-X minimum mean square algorithm module 42 with a first-order delay to update the narrowband controller 41;

[0127] Coefficients of narrowband controller 41 The update formula is:

[0128]

[0129] In the formula, μ f is the coefficient The update step size is a positive value; is the narrowband reference signal x i(n) Estimation model via secondary channel Output:

[0130] The narrowband secondary sound source synthesis subsystem 4 is used to synthesize a narrowband secondary sound source, and then used to suppress the noise component related to the narrowband reference signal in the target noise; the synthesized narrowband secondary sound source is

[0131]

[0132] The secondary channel estimation model of the first filter-X minimum mean square algorithm module 32 with first-order delay and the second filter-X minimum mean square algorithm module 42 with first-order delay The secondary channel estimation model is obtained by offline identification of the actual secondary channel S(z); A linear finite impulse response filter is used, whose coefficients are Length is

[0133] The wide-narrowband secondary sound source summing module 5 includes a first-order delay operation and a summing operation; the wide-narrowband secondary sound source summing module 5 is used to synthesize the secondary sound source, that is:

[0134] y(n)=y w (n-1)+y f (n-1)

[0135] In the acoustic space, the target noise and the signal y(n) sent to the secondary speaker and after passing through the actual secondary channel model S(z) interfere with each other in the local noise reduction area to achieve noise reduction.

[0136] In summary, the system provided by the present invention utilizes an acoustic feedback compensation subsystem, a wide- and narrow-band reference signal synthesis subsystem, a broadband secondary sound source synthesis subsystem, and a narrow-band secondary sound source synthesis subsystem to improve the overall performance of the system in dealing with wide- and narrow-band noise; the acoustic feedback compensation subsystem is used to effectively overcome the influence of acoustic feedback on the reference signal, improve the quality of the reference signal, and thus improve the overall stability and convergence of the system. The system provided by the present invention does not need to utilize a non-acoustic reference microphone, which not only avoids frequency offset, but also reduces the hardware cost of the system and broadens its practical application range.

[0137] Embodiment 2:

[0138] This embodiment provides a wide-narrowband hybrid active noise control method with acoustic feedback compensation. The method uses the above-mentioned wide-narrowband hybrid active noise control system with acoustic feedback compensation. The method includes:

[0139] Step 1: Offline identification of actual acoustic feedback channel and actual secondary channel:

[0140] The actual acoustic feedback channel F(z) is modeled using offline identification technology to obtain the acoustic feedback channel estimation model. The coefficients and lengths are and At the same time, the offline identification technology is used to realize the modeling of the actual secondary channel S(z) and obtain the secondary channel estimation model The coefficient and length of and

[0141] Step 2: System initial value setting:

[0142] Set the length and update step of the broadband controller 31 and the narrowband controller 41 respectively; set the initial values ​​of the broadband controller 31 and the narrowband controller 41 to zero respectively; set the polar radius parameters ρ, c i (n) The update of μ required c , ε, λ; set coefficient The initial value of

[0143] Step 3: System status at time n:

[0144] At time n, the secondary sound source y(n) is obtained, and then the output of the acoustic feedback compensation subsystem 1 is obtained The wide-band and narrow-band reference signal synthesis subsystem 2 synthesizes a wideband reference signal x w (n) and narrowband reference signal The broadband secondary sound source synthesis subsystem 3 and the narrowband secondary sound source synthesis subsystem 4 provide broadband secondary sound sources and narrowband secondary sound sources respectively;

[0145] Step 4: System status update:

[0146] The coefficients c of the post-adaptive bandpass filter bank 22 are updated using the normalized smooth gradient algorithm. i (n), and obtain the corresponding coefficient c i (n) coefficients at time n+1; at the same time, the coefficients of the broadband controller 31 are updated using the first filter-X minimum mean square algorithm module 32 with a first-order delay to obtain the coefficients of the corresponding broadband controller 31 at time n+1; at the same time, the coefficients of the narrowband controller 41 are updated using the second filter-X minimum mean square algorithm module 42 with a first-order delay to obtain the coefficients of the corresponding narrowband controller 41 at time n+1;

[0147] Step 5: Return to step 3, and repeat steps 3 to 4 until the system converges and reaches a steady state, thus completing noise control.

[0148] Example 3: Verification under simulated noise and secondary channel conditions

[0149] Assume that the reference signal includes narrowband frequency components and broadband additive Gaussian white noise. The normalized angular frequencies of the three narrowband frequency components are ω1 = 0.1π, ω2 = 0.2π, and ω3 = 0.3π; the corresponding discrete Fourier coefficients are a1 = 2.0, b1 = -1.0, a2 = 1.0, b2 = -0.5, a3 = 0.5, and b3 = 0.1; the broadband additive Gaussian white noise has a mean of zero and a variance of 0.2 times the energy of the narrowband frequency components. Additive Gaussian white noise v p The variance of (n) is 0.1. The actual primary channel P(z) adopts a FIR model, whose length and cutoff frequency are 41 and 0.4π respectively; the actual acoustic feedback channel F(z) also adopts a FIR model, whose length and cutoff frequency are 32 and 0.4π respectively; the actual secondary channel S(z) also adopts a FIR model, whose length and cutoff frequency are 21 and 0.4π respectively. The actual secondary channel is identified offline using a FIR filter with a length of 31; the actual acoustic feedback channel is identified offline using a FIR filter with a length of 35. The broadband controller adopts a FIR model with a length of 51; the polar radius parameter is ρ=0.975; μ c , ε, and λ are 0.0005, 0.08, and 0.997 respectively; the update steps of the narrowband controller and the broadband controller are 0.008 and 0.001 respectively; the number of runs is 100; and the simulation data length is 10000. The coefficients related to the center frequency of the three adaptive bandpass filters The initial value is set to [0,π / 6,π / 3].

[0150] FIG2(a) shows a dynamic change curve of the mean square residual error in the system of the present embodiment; FIG2(b) shows a dynamic change curve of the coefficient related to the center frequency of the adaptive bandpass filter in the system of the present embodiment. As shown in FIG2(a), when the system converges and reaches a steady state, the energy of the system residual error is about 0.11, which is close to the variance of the additive Gaussian white noise in the target noise of 0.10, indicating that the system of the present invention has good wide-band and narrow-band noise suppression performance, and indirectly reflects that it effectively overcomes the influence of acoustic feedback; as shown in FIG2(b), the system of the present invention does not need non-acoustic sensors to obtain prior information of narrow-band frequency components. In this case, it can still effectively track the narrow-band frequency components in the reference signal, effectively avoid offset compensation, and reduce the hardware cost of the system, which is beneficial to system design.

[0151] Example 4: Verification under actual noise and secondary channel conditions

[0152] The actual noise comes from the noise of a large cutting machine. In order to simulate the non-stationary characteristics of the target noise, the target noise is divided into two parts. The speed corresponding to the first half is 1400rpm, and the speed corresponding to the second half is 1600rpm. The actual primary channel P(z) adopts a linear FIR model, whose length and cutoff frequency are 61 and 0.45π respectively; the actual secondary channel, the actual acoustic feedback channel, and the actual secondary channel all adopt the actual IIR model widely used in the field (SMKuo and DRMorgan, ActiveNoiseControlSystems-Algorithms and DSPImplementation, New York: Wiley, 1996.). The actual secondary channel is identified offline using an FIR filter with a length of 64; the actual acoustic feedback channel is identified offline using an FIR filter with a length of 64. The FIR estimation model length of the broadband control filter is 128; the polar radius parameter is ρ=0.96; μ c , ε, λ are 0.015, 0.05, 0.98 respectively; the update steps of the narrowband controller and the broadband controller are 0.5 and 0.1 respectively; q is 5; the actual data length is 30000. The coefficients related to the center frequency of the five adaptive bandpass filters The initial value is set to [0, 0.1π, 0.2π, 0.3π, 0.4π].

[0153] Fig. 3 (a) is a dynamic change curve of target noise and residual noise in the system of the present embodiment; Fig. 3 (b) is a dynamic change curve (first half) of coefficients related to the center frequency of the adaptive bandpass filter in the system of the present embodiment under the condition of 1400rpm; Fig. 3 (c) is a dynamic change curve (second half) of coefficients related to the center frequency of the adaptive bandpass filter in the system of the present embodiment under the condition of 1600rpm. As shown in Fig. 3 (a), when the system reaches a steady state, the noise reduction amount of the first half is 20.51dB, and the noise reduction amount of the second half is 20.61dB, indicating that the system of the present invention still has good non-stationary wide-band and narrow-band noise suppression performance under the condition of actual noise and acoustic channel, and also indirectly reflects that it can overcome the influence of acoustic feedback; as shown in Fig. 3 (b) and Fig. 3 (c), the system of the present invention can effectively track the actual narrow-band frequency components of non-stationary changes in the reference signal, effectively solve the frequency offset problem, and reduce the hardware cost of the system, broadening its practical application range.

[0154] The above-mentioned Embodiment 3 and Embodiment 4 have verified the effectiveness and practicability of a wide- and narrow-band hybrid active noise control system and method with acoustic feedback compensation provided by the present invention from both theoretical and experimental perspectives, and will further promote the practical application of active noise control technology.

[0155] In summary, the present invention utilizes an acoustic feedback compensation subsystem to effectively overcome the influence of acoustic feedback on the reference signal and improve the quality of the reference signal; a wide-band and narrow-band reference signal synthesis structure based on a combination of a front serial adaptive bandpass filter and a rear parallel adaptive bandpass filter is used to obtain a broadband reference signal and a narrow-band reference signal to provide reliable inputs for a broadband secondary sound source synthesis subsystem and a narrow-band secondary sound source synthesis subsystem, respectively, and then used to synthesize secondary sound sources, thereby improving the stability, convergence and noise reduction performance of the system as a whole. The system provided by the present invention is capable of coping with non-stationarily changing wide-band and narrow-band mixed target noise. The system provided by the present invention does not require a non-acoustic reference microphone, which not only avoids frequency offset, but also reduces the hardware cost of the system and broadens its practical application range.

[0156] Some steps in the embodiments of the present invention may be implemented using software, and the corresponding software program may be stored in a readable storage medium, such as a CD or a hard disk.

[0157] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A wide- and narrow-band hybrid active noise control system with acoustic feedback compensation, characterized in that: The active noise control system adopts a reference microphone to collect a reference signal, adopts an error microphone to collect an error signal, and adopts a secondary speaker to provide a secondary sound source; the actual primary channel in the acoustic space is a channel model from the reference signal to the error microphone; the actual acoustic feedback channel in the acoustic space is a channel model from the secondary sound source to the reference microphone; the actual secondary channel in the acoustic space is a channel model from the secondary sound source to the error microphone; the active noise control system comprises an acoustic feedback compensation subsystem (1), a wide- and narrow-band reference signal synthesis subsystem (2), a wide-band secondary sound source synthesis subsystem (3), a narrow-band secondary sound source synthesis subsystem (4), and a wide- and narrow-band secondary sound source summing module (5); The acoustic feedback compensation subsystem (1) is connected to the wide- and narrow-band reference signal synthesis subsystem (2) and the wide- and narrow-band secondary sound source summing module (5) respectively; the wide- and narrow-band reference signal synthesis subsystem (2) is connected to the wide- and narrow-band secondary sound source synthesis subsystem (3) and the narrow-band secondary sound source synthesis subsystem (4) respectively; the wide- and narrow-band secondary sound source synthesis subsystem (3) is connected to the wide- and narrow-band secondary sound source summing module (5); the narrow-band secondary sound source synthesis subsystem (4) is connected to the wide- and narrow-band secondary sound source summing module (5); The acoustic feedback compensation subsystem (1) is used to synthesize an acoustic feedback signal and compensate for the influence of the actual acoustic feedback signal on the actual reference signal; The acoustic feedback compensation subsystem (1) comprises: an acoustic feedback channel estimation model (11) and a summation operation, which can provide an estimate of the actual reference signal, which is then used as an input of the wide-narrowband reference signal synthesis subsystem (2); The acoustic feedback channel estimation model (11) The actual acoustic feedback channel F(z) is obtained by offline identification, which is represented by a linear finite impulse response filter, whose coefficients and lengths are and Output of acoustic feedback compensation subsystem (1) Expressed as Among them, x r (n) is the reference signal provided by the reference microphone; y(n) is the secondary sound source; n is the time, n≥0; The wide-narrowband reference signal synthesis subsystem (2) is used to synthesize a narrowband reference signal and a wideband reference signal; the wide-narrowband reference signal synthesis subsystem (2) comprises: a pre-serial type adaptive bandpass filter group (21) and a post-parallel type adaptive bandpass filter group (22); The front serial adaptive bandpass filter group (21) is composed of q adaptive bandpass filters in a serial manner, and the value of q is not less than the number of narrowband frequency components in the target noise; each adaptive bandpass filter is composed of a second-order IIR notch filter; the front serial adaptive bandpass filter group (21) provides a high signal-to-noise ratio input for the rear parallel adaptive bandpass filter group (22); The post-parallel adaptive bandpass filter group (22) is composed of q adaptive bandpass filters in parallel, providing q narrowband reference signals Using the output of the acoustic feedback compensation subsystem (1) Subtract the sum of all narrowband reference signals output by the post-parallel adaptive bandpass filter module (22) to obtain a broadband reference signal x w (n); The reference input of the broadband secondary sound source synthesis subsystem (3) is the broadband reference signal x separated by the broadband and narrowband reference signal synthesis subsystem (2). w (n); at the same time, the q reference inputs of the narrowband secondary sound source synthesis subsystem (4) are q narrowband reference signals provided by the wide-narrowband reference signal synthesis subsystem (2) The coefficients related to the center frequencies of the q adaptive bandpass filters in the post-parallel adaptive bandpass filter module (22) are The initial values ​​are all -2, 0 or [0,1,…,q-1]π / (2q), and no non-acoustic sensor is required to obtain prior information on the narrowband frequency components in the reference signal; The wide- and narrow-band secondary sound source summing module (5) includes a first-order delay operation and a summing operation, and sums the narrow-band secondary sound source and the wide-band secondary sound source to provide a secondary sound source for the secondary speaker.

2. A wide- and narrow-band hybrid active noise control system with acoustic feedback compensation as claimed in claim 1, characterized in that: In the wide-narrowband reference signal synthesis subsystem (2), the z-domain model H of the i-th adaptive bandpass filter i (z) is: Where ρ is the polar radius parameter, ρ∈[0,1); c i is time-varying, c i (n) = -2cos[ω i (n)] is the coefficient related to the center frequency of the i-th adaptive bandpass filter, ω i (n) is the center frequency of the i-th adaptive bandpass filter; z -1 represents a first-order delay operation; z -2 represents a second-order delay operation; The output of the i-th adaptive bandpass filter of the pre-serial adaptive bandpass filter group (21) is: and d,i (n)=y c,i-1 (n)-y c,i (n) The output of the i-th second-order IIR notch filter of the pre-serial adaptive bandpass filter bank (21) is: y c,i (n)=-ρc i (n)y c,i (n-1)-ρ 2 y c,i (n-2)+y c,i-1 (n)+c i (n)y c,i-1 (n-1)+y c,i-1 (n-2) Among them, y c,i (n) is the output of the i-th second-order IIR notch filter, y c,0 (n) = x(n); The input and output of the i-th IIR notch filter of the post-parallel adaptive bandpass filter bank (22) are respectively: y b,i (n)=-ρc i (n)y b,i (n-1)-ρ 2 y b,i (n-2)+x d,i (n)+c i (n)x d,i (n-1)+x d,i (n-2) The output of the i-th adaptive bandpass filter of the post-parallel adaptive bandpass filter bank (22) is: x i (n)=x d,i (n)-y b,i (n) Among them, x i (n) serving as a narrowband reference signal input for the narrowband secondary sound source synthesis subsystem (4); The broadband reference signal provided by the broadband and narrowband reference signal synthesis subsystem (2) is: Among them, x w (n) is used as a broadband reference signal input for the broadband secondary sound source synthesis subsystem (3).

3. A wide-narrowband hybrid active noise control system with acoustic feedback compensation as claimed in claim 2, characterized in that: In the wide-narrowband reference signal synthesis subsystem (2), a normalized smooth gradient algorithm is used to calculate the coefficient c related to the center frequency of the i-th adaptive bandpass filter. i (n) To update, that is: g i (n)=x d,i (n-1)-ρy b,i (n-1); In the formula, μ c c i (n); ε is a non-zero positive number; λ is the forgetting factor, λ∈(0,1); when the system reaches a steady state, the coefficient c i (n) converges to -2cos(ω i,s ), where ω i,s is the reference signal x r The true frequency value of the i-th narrowband component in (n).

4. A wide- and narrow-band hybrid active noise control system with acoustic feedback compensation as claimed in claim 3, characterized in that: The broadband secondary sound source synthesis subsystem (3) comprises: a broadband controller (31) and a first filter-X minimum mean square algorithm module (32) with a first-order delay; the broadband controller (31) adopts a linear finite impulse response filter, and the first filter-X minimum mean square algorithm module (32) with a first-order delay is used to update the broadband controller (31); The broadband controller (31) adopts a linear finite impulse response filter, the coefficient of which is The length is L; the coefficient The update formula is: In the formula, μ w is the coefficient The update step size is ; e(n) is the residual noise; is the broadband reference signal x w (n) Estimation model via secondary channel Output.

5. A wide-narrowband hybrid active noise control system with acoustic feedback compensation as claimed in claim 4, characterized in that: The broadband secondary sound source synthesis subsystem (3) is used to synthesize a broadband secondary sound source, and then used to suppress the noise component related to the broadband reference signal in the target noise; the synthesized broadband secondary sound source is 6. A wide-narrowband hybrid active noise control system with acoustic feedback compensation as claimed in claim 5, characterized in that: The narrowband secondary sound source synthesis subsystem (4) comprises: a narrowband controller (41) and a second filter-X minimum mean square algorithm module (42) with a first-order delay; the narrowband controller (41) adopts an amplitude phase adjustment structure and uses the second filter-X minimum mean square algorithm module (42) with a first-order delay to update the narrowband controller (41); The coefficients of the narrowband controller (41) The update formula is: In the formula, μ f is the coefficient The update step size is a positive value; is the narrowband reference signal x i (n) Estimation model via secondary channel Output.

7. A wide- and narrow-band hybrid active noise control system with acoustic feedback compensation as claimed in claim 6, characterized in that: The narrowband secondary sound source synthesis subsystem (4) is used to synthesize a narrowband secondary sound source, and then used to suppress the noise component related to the narrowband reference signal in the target noise; the synthesized narrowband secondary sound source is 8. A wide- and narrow-band hybrid active noise control system with acoustic feedback compensation as claimed in claim 7, characterized in that: The first filter-X minimum mean square algorithm module (32) with first-order delay and the second filter-X minimum mean square algorithm module (42) with first-order delay, the secondary channel estimation model The secondary channel estimation model is obtained by offline identification of the actual secondary channel S(z); A linear finite impulse response filter is used, whose coefficients are Length is 9. A wide-narrowband hybrid active noise control system with acoustic feedback compensation as claimed in claim 8, characterized in that: The wide- and narrow-band secondary sound source summing module (5) comprises a first-order delay operation and a summing operation; the wide- and narrow-band secondary sound source summing module (5) is used to synthesize the secondary sound source, namely: y(n)=y w (n-1)+y f (n-1) In the acoustic space, the target noise and the signal y(n) sent to the secondary speaker and after passing through the actual secondary channel model S(z) interfere with each other in the local noise reduction area to achieve noise reduction.

10. A wide- and narrow-band hybrid active noise control method with acoustic feedback compensation, characterized in that: The method adopts the wide-narrowband hybrid active noise control system with acoustic feedback compensation as claimed in any one of claims 1 to 9, and the method comprises: Step 1: Offline identification of actual acoustic feedback channel and actual secondary channel: The actual acoustic feedback channel F(z) is modeled using offline identification technology to obtain the acoustic feedback channel estimation model. The coefficients and lengths are and At the same time, the offline identification technology is used to realize the modeling of the actual secondary channel S(z) and obtain the secondary channel estimation model The coefficient and length of and Step 2: System initial value setting: The lengths and update steps of the broadband controller (31) and the narrowband controller (41) are respectively set; the initial values ​​of the broadband controller (31) and the narrowband controller (41) are respectively set to zero; the polar radius parameters ρ and c are set. i (n) The update of μ required c , ε, λ; set coefficient The initial value of Step 3: System status at time n: At time n, the secondary sound source y(n) is obtained, and then the output of the acoustic feedback compensation subsystem (1) is obtained. The wide-band and narrow-band reference signal synthesis subsystem (2) synthesizes a wideband reference signal x w (n) and narrowband reference signal The broadband secondary sound source synthesis subsystem (3) and the narrowband secondary sound source synthesis subsystem (4) provide broadband secondary sound sources and narrowband secondary sound sources respectively; Step 4: System status update: The coefficients c of the post-parallel adaptive bandpass filter bank (22) are updated using the normalized smoothed gradient algorithm. i (n), and obtain the corresponding coefficient c i (n) at time n+1; at the same time, using the first filter-X minimum mean square algorithm module (32) with a first-order delay to update the coefficient of the broadband controller (31) to obtain the coefficient of the corresponding broadband controller (31) at time n+1; at the same time, using the second filter-X minimum mean square algorithm module (42) with a first-order delay to update the coefficient of the narrowband controller (41) to obtain the coefficient of the corresponding narrowband controller (41) at time n+1; Step 5: Return to step 3, and repeat steps 3 to 4 until the system converges and reaches a steady state, thus completing noise control.

Images