CN116721649A - Robust feedforward type wide-narrow-band hybrid active noise control system and method - Google Patents
Robust feedforward type wide-narrow-band hybrid active noise control system and method Download PDFInfo
- Publication number
- CN116721649A CN116721649A CN202310665844.7A CN202310665844A CN116721649A CN 116721649 A CN116721649 A CN 116721649A CN 202310665844 A CN202310665844 A CN 202310665844A CN 116721649 A CN116721649 A CN 116721649A
- Authority
- CN
- China
- Prior art keywords
- subsystem
- noise
- broadband
- narrowband
- auxiliary
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000001914 filtration Methods 0.000 claims abstract description 121
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 73
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 73
- 230000001629 suppression Effects 0.000 claims abstract description 9
- 239000000654 additive Substances 0.000 claims description 24
- 230000000996 additive effect Effects 0.000 claims description 20
- 230000002194 synthesizing effect Effects 0.000 claims description 17
- 230000001902 propagating effect Effects 0.000 claims description 4
- 230000009467 reduction Effects 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000005299 abrasion Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K11/00—Methods or devices for transmitting, conducting or directing sound in general; Methods or devices for protecting against, or for damping, noise or other acoustic waves in general
- G10K11/16—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general
- G10K11/175—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound
- G10K11/178—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound by electro-acoustically regenerating the original acoustic waves in anti-phase
- G10K11/1785—Methods, e.g. algorithms; Devices
- G10K11/17853—Methods, e.g. algorithms; Devices of the filter
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K11/00—Methods or devices for transmitting, conducting or directing sound in general; Methods or devices for protecting against, or for damping, noise or other acoustic waves in general
- G10K11/16—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general
- G10K11/175—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound
- G10K11/178—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound by electro-acoustically regenerating the original acoustic waves in anti-phase
- G10K11/1781—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound by electro-acoustically regenerating the original acoustic waves in anti-phase characterised by the analysis of input or output signals, e.g. frequency range, modes, transfer functions
- G10K11/17821—Methods or devices for protecting against, or for damping, noise or other acoustic waves in general using interference effects; Masking sound by electro-acoustically regenerating the original acoustic waves in anti-phase characterised by the analysis of input or output signals, e.g. frequency range, modes, transfer functions characterised by the analysis of the input signals only
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K2210/00—Details of active noise control [ANC] covered by G10K11/178 but not provided for in any of its subgroups
- G10K2210/10—Applications
- G10K2210/128—Vehicles
- G10K2210/1281—Aircraft, e.g. spacecraft, airplane or helicopter
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K2210/00—Details of active noise control [ANC] covered by G10K11/178 but not provided for in any of its subgroups
- G10K2210/10—Applications
- G10K2210/128—Vehicles
- G10K2210/1282—Automobiles
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K2210/00—Details of active noise control [ANC] covered by G10K11/178 but not provided for in any of its subgroups
- G10K2210/30—Means
- G10K2210/301—Computational
- G10K2210/3027—Feedforward
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10K—SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR
- G10K2210/00—Details of active noise control [ANC] covered by G10K11/178 but not provided for in any of its subgroups
- G10K2210/30—Means
- G10K2210/301—Computational
- G10K2210/3028—Filtering, e.g. Kalman filters or special analogue or digital filters
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Soundproofing, Sound Blocking, And Sound Damping (AREA)
Abstract
The invention discloses a robust feedforward type wide-narrow-band hybrid active noise control system and method, and belongs to the technical field of active noise control. The system utilizes the first linear predictive filtering subsystem to realize the synthesis of the narrowband reference signal and the broadband reference signal, thereby overcoming the frequency offset problem; the second linear prediction filtering subsystem and the auxiliary filtering subsystem are utilized, so that the independence among the broadband secondary sound source synthesis subsystem, the narrowband secondary sound source synthesis subsystem and the secondary channel online identification subsystem can be improved, the accuracy and the speed of online identification of the secondary channel are improved, the dynamic performance of the whole system is improved, the influence of the introduced auxiliary noise on residual noise can be reduced, and the noise suppression performance of the whole system is improved. The invention also utilizes the secondary channel on-line identification module, can cope with the complex time variability of the secondary channel, does not need to install a non-acoustic sensor, reduces the requirement on physical space and hardware cost, and widens the application range of actual noise reduction.
Description
Technical Field
The invention relates to a robust feedforward type wide-narrow-band hybrid active noise control system and method, and belongs to the technical field of active noise control.
Background
With the progress of electronic, electroacoustic, signal processing, and other technologies, active noise control (Active Noise Control, ANC) technology has been widely used in noise reduction in automobiles, aircraft, and other applications. Compared with the traditional passive noise reduction technology, the active noise control technology utilizes the sound wave destructive interference principle, has the advantages of good low-frequency noise suppression performance, small volume, low cost and the like, and is an advantageous supplement to the traditional passive noise control technology (L.Lu, K.Yin, R.C.de Lanare, Z.Zheng, Y.Yu, X.Yang, B.Chen, "A survey on active noise control in the past decade, part I: linear systems," Signal Process.183 (2021), 108039). The ANC system may be classified into a feedforward type, a feedback type, and a feedforward and feedback hybrid type according to the structure of the controller. The feedforward ANC system can be divided into a feedforward narrowband ANC system, a feedforward broadband ANC system and a feedforward broadband mixed ANC system according to the spectral characteristics of target noise.
The feedforward type broadband-narrowband mixed ANC system can reduce broadband and narrowband mixed noise or interference at the same time, and effectively solves the problem of a spark phenomenon existing in the feedforward type broadband ANC system when the broadband-narrowband mixed noise is restrained (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.). Then, the feedforward type wide-narrow-band hybrid ANC system needs to adopt a non-acoustic reference sensor and an acoustic reference sensor to obtain a reference signal, wherein when the required non-acoustic reference sensor (such as a tachometer and the like) obtains the reference signal, the obtained narrow-band reference frequency possibly does not coincide with the frequency of a real narrow-band component due to the conditions of self aging, abrasion and the like, namely, frequency offset occurs. The frequency offset may directly cause the performance of the narrowband controller in the wideband and narrowband hybrid ANC system for suppressing the target noise narrowband component to be reduced, thereby affecting the stability and convergence of the wideband and narrowband hybrid ANC system and limiting the practical application of the wideband and narrowband hybrid ANC system. In addition, in practical application, the complex time variability of the secondary channel also can seriously affect the stability of the system, and an effective secondary channel online identification method is designed aiming at the feedforward type wide-band and narrow-band mixed ANC system, so that the method has important research and application values, and further the practical application range of the method is widened.
In 2013, xiao et al designed a feed-forward wideband hybrid ANC system that dealt with frequency imbalance, which employed a band-pass filter bank consisting of several parallel adaptive traps to compensate for the frequency imbalance, thereby providing accurate inputs to a controller based on an amplitude-phase adjustment structure for synthesizing a secondary sound source (y.xiao and k.doi., "A robust hybrid active noise control system using IIR notch filters," int.j. Advanced Mechatronic Systems,5 (1), (2013): 69-77). The system then has strict requirements on the initial filter weight setting of the adaptive notch filter on the one hand, and on the other hand, the system can cope with the problem that the frequency compensation capability is insufficient when the frequency is suddenly changed or the frequency is suddenly changed. Recently, ma & Xiao et al developed a feed-forward broadband hybrid ANC system based on secondary channel on-line recognition that can cope with complex timeliness of secondary channels, improving secondary channel on-line recognition accuracy and speed while further reducing residual noise (Y.Ma, Y.Xiao, L.Ma, and k.khoraani, "A robust feedforward hybrid active noise control system with online secondary-path modeling," IET Signal process, 17 (1) (2023), e 12183.). However, the system has not been related to frequency offset compensation on the one hand, and the residual noise separation structure adopted by the system can separate broadband noise and narrowband noise in residual noise, but does not realize effective separation of broadband components related to target noise and broadband components related to reference signals, so that the narrowband controller, the broadband controller and the secondary channel online identification module still have coupling connection, and the convergence performance of the whole system is affected.
In order to solve the above problems of frequency offset and complex time variability of secondary channels, which limit the system performance, a more effective and practical feedforward wideband hybrid active noise control system needs to be provided.
Disclosure of Invention
In order to solve the problems that the traditional feedforward type broadband and narrowband mixed ANC system has frequency offset, complex time variability of a secondary channel and the like caused by aging, abrasion and the like of a non-acoustic reference sensor, and severely restricts the convergence and stability of the feedforward type broadband and narrowband mixed ANC system, and further reduces the broadband and narrowband mixed noise suppression performance of the whole system, the invention provides a robust feedforward type broadband and narrowband mixed active noise control system and a method, and the technical scheme is as follows:
a first object of the present invention is to provide a robust wide and narrow band hybrid active noise control system that uses a reference microphone to collect reference signals, an error microphone to collect residual noise, and a secondary speaker to provide a secondary sound source, respectively; the actual primary channel in the acoustic space is a channel model of the reference signal propagating to the error microphone; the actual secondary channel in the acoustic space is a model of the channel that propagates the secondary sound source provided by the secondary speaker to the error microphone; the active noise control system comprises a first linear prediction filtering subsystem 1, a broadband secondary sound source synthesis subsystem 2, a narrowband secondary sound source synthesis subsystem 3, a second linear prediction filtering subsystem 4, an auxiliary filtering subsystem 5 and a secondary channel online identification subsystem 6;
The first linear prediction filtering subsystem 1 is respectively connected with the broadband secondary sound source synthesis subsystem 2, the narrowband secondary sound source synthesis subsystem 3 and the auxiliary filtering subsystem 5; the broadband secondary sound source synthesis subsystem 2 is respectively connected with the first linear prediction filtering subsystem 1 and the auxiliary filtering subsystem 5; the narrow-band secondary sound source synthesis subsystem 3 is respectively connected with the first linear prediction filtering subsystem 1 and the second linear prediction filtering subsystem 4; the second linear prediction filtering subsystem 4 is respectively connected with the narrow-band secondary sound source synthesis subsystem 3, the auxiliary filtering subsystem 5 and the secondary channel on-line identification subsystem 6; the auxiliary filtering subsystem 5 is respectively connected with the first linear prediction filtering subsystem 1, the broadband secondary sound source synthesis subsystem 2, the second linear prediction filtering subsystem 4 and the secondary channel on-line identification subsystem 6; the secondary channel online identification subsystem 6 is respectively connected with the second linear prediction filtering subsystem 4 and the auxiliary filtering subsystem 5;
the first linear prediction filtering subsystem 1 is used for synthesizing a broadband reference signal and a narrowband reference signal; the broadband secondary sound source synthesis subsystem 2 is used for synthesizing a broadband secondary sound source; the narrow-band secondary sound source synthesis subsystem 3 is used for synthesizing a narrow-band secondary sound source; the second linear prediction filtering subsystem 4 is configured to separate a narrowband residual noise component and a wideband residual noise component from residual noise; the auxiliary filtering subsystem 5 is configured to separate a wideband residual noise component related to the wideband reference signal and a wideband residual noise component related to additive noise in the auxiliary noise and the target signal from the wideband residual noise component; the secondary channel online identification subsystem 6 is used for estimating a time-varying secondary channel model in real time;
The narrow-band residual noise component separated by the second linear prediction filtering subsystem 4 is used as an error output of the narrow-band secondary sound source synthesis subsystem 3 and an auxiliary noise adjustment module input of the secondary channel on-line identification subsystem 6 respectively; meanwhile, the broadband residual noise component which is separated by the auxiliary filtering subsystem 5 and related to the broadband reference signal is respectively used as an error output of the broadband secondary sound source synthesizing subsystem 2 and an auxiliary noise adjusting module input of the secondary channel on-line identifying subsystem 6; meanwhile, the broadband residual noise component which is separated by the auxiliary filtering subsystem 5 and related to the auxiliary noise and additive noise in the target signal is used as the expected input of the secondary channel on-line identification subsystem 6; the method can improve the independence among the broadband secondary sound source synthesis subsystem 2, the narrowband secondary sound source synthesis subsystem 3 and the secondary channel online identification subsystem 6, improve the accuracy and the speed of online identification of the secondary channel, improve the dynamic performance of the whole system, reduce the influence of the introduced auxiliary noise on the residual noise and improve the noise suppression performance of the whole system.
Optionally, the first linear prediction filtering subsystem 1 includes the first delay element 11 and the first linear prediction filter 12, the first delay element 11 and the first linear prediction filter 12 are connected in series, and the order of the first delay element 11 is D 1 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the first linear prediction filter 12 are respectivelyAnd L 1 The coefficient is updated by using a least mean square algorithm, and the updating formula is as follows:
h 1,j (n+1)＝h 1,j (n)+μ 1 x w (n)x r (n-D 1 -j)
wherein mu 1 Taking a positive value for the update step of the first linear prediction filter 12; x is x w (n) wideband reference signal separated for the first linear prediction subsystem 1, x r (n) a reference signal provided for a reference microphone; n is the moment, n is more than or equal to 0;
the narrowband reference signals and the broadband reference signals synthesized by the first linear prediction filtering subsystem 1 are respectively:
x w (n)＝x r (n)-x f (n)
optionally, the broadband secondary sound source synthesis subsystem 2 includes the broadband controller 21 and the first filter-X least mean square algorithm module 22;
the broadband controller 21 adopts a linear filter with coefficients and lengths of respectivelyAnd L w ；
The first filter-X least mean square algorithm module 22 uses the wideband reference signal X separated by the auxiliary filter subsystem 5 w (n) related wideband residual noise component y h (n) as an error output and for updating coefficients of the wideband controller 21; the coefficient update formula of the broadband controller 21 is:
wherein mu w Taking a positive value as the update step length of the broadband controller 21;for wideband reference signal x w (n) output of a secondary channel estimation model in the first filter-X least mean square algorithm module 22;
the broadband secondary sound source obtained by the broadband secondary sound source synthesis subsystem 2 is
Optionally, the narrowband secondary sound source synthesis subsystem 3 includes the narrowband controller 31 and the second filter-X least mean square algorithm module 32;
the narrowband controller 31 employs linear filters with coefficients and lengths of respectivelyAnd L f ；
The second filter-X least mean square algorithm module 32 uses the narrow band residual noise component y separated by the second linear predictive filtering subsystem 4 LP (n) as an error output and for updating coefficients of the narrowband controller 31; the coefficient update formula of the narrowband controller 31 is:
wherein mu f Taking a positive value as an update step length of the narrowband controller 31;for narrowband reference signal x f (n) output of a secondary channel estimation model in the second filter-X least mean square algorithm module 32;
The narrowband secondary sound source obtained by the narrowband secondary sound source synthesis subsystem 3 is
Optionally, the second linear prediction filtering subsystem 4 includes the second delay element 41 and the second linear prediction filter 42, where the second delay element 41 and the second linear prediction filter 42 are connected in series, and the order of the second delay element 41 is D 2 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the second linear prediction filter 42 are respectivelyAnd L 2 The coefficient is updated by using a least mean square algorithm, and the updating formula is as follows:
h 2,j (n+1)＝h 2,j (n)+μ 2 e LP (n)e(n-D 2 -j)
wherein mu 2 Taking a positive value for the update step of the second linear prediction filter 42; e, e LP (n) a wideband residual noise component separated for the second linear prediction subsystem 4; e (n) is the residual noise provided by the error microphone;
the narrowband residual noise component and the wideband residual noise component separated from the residual noise by the second linear prediction filtering subsystem 4 are respectively:
e LP (n)＝e(n)-y LP (n)
broadband residual noise component e separated by the second linear predictive filtering subsystem 4 LP (n) serving as a desired input to the auxiliary filtering subsystem 5.
Optionally, the auxiliary filtering subsystem 5 includes the linear filter 51 and the least mean square algorithm module 52;
The coefficients and lengths of the linear filter 51 are respectivelyAnd L 3 The coefficient of the linear filter 51 is updated by the least mean square algorithm module 52, and the update formula is:
h 3,j (n+1)＝h 3,j (n)+μ 3 e h (n)x w (n-j)
wherein mu 3 The update step length of the linear filter 51 is a positive value; e, e h (n) a wideband residual noise component related to additive noise in the auxiliary noise and target signal separated by the auxiliary filtering subsystem 5;
the wideband residual noise component related to the wideband reference signal separated by the auxiliary filtering subsystem 5 is
Broadband residual noise component e related to additive noise in the auxiliary noise and target signal separated by the auxiliary filtering subsystem 5 h (n) as a desired input to the secondary channel on-line recognition subsystem 6.
Optionally, the secondary channel online identification subsystem 6 includes: a secondary channel on-line recognition module 61 and an auxiliary noise adjustment module 62;
the secondary channel on-line recognition module 61 includes a secondary channel estimation modelThe secondary channel on-line identification module 61 uses e h (n) is an auxiliary noise v generated by the Gaussian white noise v (n) passing through the auxiliary noise adjusting module 62 as a desired input 0 (n) as a reference input, estimating a time-varying secondary channel in real time using a least mean square algorithm;
The secondary channel on-line identification module 61 is a secondary channel estimation modelCoefficient and length of (a) are respectivelyAnd->The coefficient update formula is as follows:
e s (n)＝e h (n)-y s (n)
wherein mu s The updating step length of the secondary channel estimation model is taken as a positive value; y is s (n) is the output of a secondary channel estimation model of the secondary channel online recognition module 61;
the auxiliary noise v 0 (n) is:
v 0 (n)＝v(n)G(n)
G(n)＝max{G N (n),G B (n)}
G N (n)＝abs[y LP (n-1)]
G B (n)＝abs[y h (n-1)]
wherein G (n) is a gain adjustment factor of the auxiliary noise adjustment module 62; g N (n) and G B (n) is y respectively LP (n-1) and y h (n-1) a low pass filtered output; abs []Calculating for absolute value; v (n) is zero in mean and variance isAdditive white gaussian noise of (c).
Alternatively, the synthesized secondary sound source is:
y(n)＝y w (n)+y f (n)-v 0 (n)
and y (n) is output to the secondary speaker, and interference with the target noise is canceled in the acoustic space.
A second object of the present invention is to provide an active noise control method, implemented based on the aforementioned robust feedforward type wide and narrow band hybrid active noise control system, comprising:
step one: setting system parameters;
the first linear prediction filter 12, the wide-band controller 21, the narrow-band controller 31, the second linear prediction filter 42, the linear filter 51, and the secondary channel estimation model are respectively set Length and step length of (a); the orders of the first delay links 11 and 41 are respectively set; setting a forgetting factor of the auxiliary noise adjustment module 62; the first linear prediction filter 12, the wideband controller 21, the narrowband controller 31, the second linear prediction filter 42, the linear filter 51, the secondary channel estimation model +.>And the initial values of the gain adjustment factors G (n) of the auxiliary noise adjustment module 62 are all zero; setting auxiliary noise v (n);
step two: synthesizing a wideband reference signal and a narrowband reference signal;
at time n, reference signal x obtained with reference microphone r (n) separating the wideband reference signal and the narrowband reference signal after passing through the first linear predictive filtering subsystem 1, respectively; the separated broadband reference signals are respectively provided for the broadband secondary sound source synthesis subsystem 2 and the auxiliary filtering subsystem 5; the separated narrowband reference signal is provided to a narrowband secondary sound source synthesis subsystem 3;
step three: at time n, first, the broadband reference signal gets the broadband secondary sound source y via the broadband secondary sound source synthesis subsystem 2 w (n) obtaining a narrowband secondary sound source y from the narrowband reference signal via the narrowband secondary sound source synthesis subsystem 3 f (n); second, the auxiliary noise v is obtained by the auxiliary noise adjustment module 62 0 (n) and then y w (n) and y f (n) superposing and synthesizing to obtain a secondary sound source y (n); finally, the residual noise e (n) is separated by the second linear prediction filtering subsystem 4 to obtain a narrow-band residual noise component y LP (n) and wideband residual noise component e LP (n)；
Step four: at time n, wideband residual noise component e LP (n) obtaining wideband residual noise components e related to additive noise in the auxiliary noise and the target signal respectively after passing through the auxiliary filtering subsystem 5 h (n) wideband residual noise component y related to wideband reference signal h (n)；e h (n) as a desired input to the secondary channel online recognition module 61;
step five: update control system
According to the reference signal x r (n) and the wideband reference signal x w (n) calculating coefficients of the updated first linear prediction filter 12 at time n+1;
according to the broadband reference signal x w (n) and y h (n) calculating coefficients of the updated wideband controller 21 at time n+1;
from narrowband reference signal x f (n) and narrowband residual noise y LP (n) calculating coefficients of the updated narrowband controller 31 at time n+1;
based on the residual noise e (n) and the wideband residual noise component e LP (n) calculating coefficients of the updated second linear prediction filter 42 at time n+1;
according to the broadband reference signal x w (n) and e h (n) calculating coefficients of the update linear filter 51 at time n+1;
According to the auxiliary noise v 0 (n) and e h (n) computing and updating the secondary channel estimation model in the secondary channel on-line recognition module 61A coefficient at time n+1;
from the narrowband residual noise component y LP (n) and y h (n) calculating a gain adjustment factor at time n+1 for updating the auxiliary noise adjustment module 62;
step six: and returning to the second step, repeating the second step to the fifth step until the system converges and reaches a steady state, and realizing active noise control.
The invention has the beneficial effects that:
1. according to the invention, a non-acoustic sensor is not required, so that the requirements on space and the hardware cost of a system are reduced;
2. the invention separates out the narrow-band reference signal by utilizing the first linear predictive filtering subsystem, thereby providing accurate input for the narrow-band sound source synthesis subsystem and effectively solving the problem of frequency offset;
3. the invention utilizes the broadband residual noise component related to the broadband reference signal and the broadband residual noise component related to the auxiliary noise and the additive noise in the target signal separated by the auxiliary filtering subsystem, reduces the coupling relation among the broadband secondary sound source synthesis subsystem, the narrowband secondary sound source synthesis subsystem and the secondary channel on-line identification subsystem, improves the accuracy and the speed of the on-line identification of the secondary channel, and simultaneously improves the dynamic performance of the whole system;
4. The invention utilizes the narrow-band residual noise component provided by the second linear prediction filtering subsystem and the broadband residual noise component related to the broadband reference signal provided by the auxiliary filtering subsystem to be used for auxiliary noise amplitude adjustment together, thereby reducing the influence of auxiliary noise on the residual noise, improving the noise suppression performance of the whole system and theoretically realizing that the residual noise after the system reaches a steady state tends to the environmental level;
5. the invention utilizes the secondary channel on-line identification module to effectively cope with the complex time variability of the secondary channel under the actual condition, and widens the actual application range.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a robust feedforward wide-narrowband hybrid active noise control system in accordance with a first embodiment of the present invention.
Fig. 2 (a) is a graph showing the dynamic change of the mean square residual error in the third embodiment of the present invention.
Fig. 2 (b) is a graph showing the dynamic change of the mean square error of the secondary channel estimation according to the third embodiment of the present invention.
Fig. 3 is a graph showing the dynamic change of the target noise and the residual noise in the fourth embodiment of the present invention.
In the figure: 1 is a first linear prediction filtering subsystem, 2 is a broadband secondary sound source synthesis subsystem, 3 is a narrowband secondary sound source synthesis subsystem, 4 is a second linear prediction filtering subsystem, 5 is an auxiliary filtering subsystem, and 6 is a secondary channel online identification subsystem; 11 is a first delay element, 12 is a first linear prediction filter, 21 is a wideband controller, 22 is a first filtering-X least mean square algorithm module, 31 is a narrowband controller, 32 is a second filtering-X least mean square algorithm module, 41 is a second delay element, 42 is a second linear prediction filter, 51 is a linear filter, 52 is a least mean square algorithm module, 61 is a secondary channel online identification module, and 62 is an auxiliary noise adjustment module.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings.
Embodiment one:
the embodiment provides a robust feedforward type wide-narrow-band mixed active noise control system, referring to a system schematic diagram shown in fig. 1, the active noise control system respectively adopts a reference microphone to collect reference signals, adopts an error microphone to collect residual noise, and adopts a secondary loudspeaker to provide a secondary sound source; the actual primary channel in the acoustic space is a channel model of the reference signal propagating to the error microphone; the actual secondary channel in the acoustic space is a model of the channel that propagates the secondary sound source provided by the secondary speaker to the error microphone; the active noise control system comprises a first linear prediction filtering subsystem 1, a broadband secondary sound source synthesis subsystem 2, a narrowband secondary sound source synthesis subsystem 3, a second linear prediction filtering subsystem 4, an auxiliary filtering subsystem 5 and a secondary channel online identification subsystem 6;
The first linear prediction filtering subsystem 1 is respectively connected with the broadband secondary sound source synthesis subsystem 2, the narrowband secondary sound source synthesis subsystem 3 and the auxiliary filtering subsystem 5; the broadband secondary sound source synthesis subsystem 2 is respectively connected with the first linear prediction filtering subsystem 1 and the auxiliary filtering subsystem 5; the narrow-band secondary sound source synthesis subsystem 3 is respectively connected with the first linear prediction filtering subsystem 1 and the second linear prediction filtering subsystem 4; the second linear prediction filtering subsystem 4 is respectively connected with the narrow-band secondary sound source synthesis subsystem 3, the auxiliary filtering subsystem 5 and the secondary channel on-line identification subsystem 6; the auxiliary filtering subsystem 5 is respectively connected with the first linear prediction filtering subsystem 1, the broadband secondary sound source synthesis subsystem 2, the second linear prediction filtering subsystem 4 and the secondary channel on-line identification subsystem 6; the secondary channel on-line identification subsystem 6 is respectively connected with the second linear prediction filtering subsystem 4 and the auxiliary filtering subsystem 5;
the first linear predictive filtering subsystem 1 is used for synthesizing a broadband reference signal and a narrowband reference signal; the broadband secondary sound source synthesis subsystem 2 is used for synthesizing a broadband secondary sound source; the narrow-band secondary sound source synthesis subsystem 3 is used for synthesizing a narrow-band secondary sound source; the second linear prediction filtering subsystem 4 is used for separating a narrow-band residual noise component and a wide-band residual noise component from residual noise; the auxiliary filtering subsystem 5 is configured to separate a wideband residual noise component related to the wideband reference signal and a wideband residual noise component related to the auxiliary noise and additive noise in the target signal from the wideband residual noise component; the secondary channel online identification subsystem 6 is used for estimating a time-varying secondary channel model in real time;
The narrow-band residual noise component separated by the second linear prediction filtering subsystem 4 is used as error output of the narrow-band secondary sound source synthesis subsystem 3 and input of an auxiliary noise adjustment module in the secondary channel on-line identification subsystem 6 respectively; meanwhile, the broadband residual noise component which is separated by the auxiliary filtering subsystem 5 and related to the broadband reference signal is respectively used as the error output of the broadband secondary sound source synthesis subsystem 2 and the input of an auxiliary noise adjustment module in the secondary channel on-line identification subsystem 6; meanwhile, the broadband residual noise component which is separated by the auxiliary filtering subsystem 5 and related to the auxiliary noise and additive noise in the target signal is used as the expected input of the secondary channel on-line identification subsystem 6; the method can improve the independence among the broadband secondary sound source synthesis subsystem 2, the narrowband secondary sound source synthesis subsystem 3 and the secondary channel online identification subsystem 6, improve the accuracy and the speed of online identification of the secondary channel, improve the dynamic performance of the whole system, reduce the influence of the introduced auxiliary noise on the residual noise and improve the noise suppression performance of the whole system.
The actual primary channel P (z) represents the model of the acoustic space from the reference microphone to the error microphone; the actual secondary channel S (z) represents the model of the acoustic space from the secondary speaker to the error microphone.
The target noise is:
p(n)＝p 0 (n)+v p (n)
wherein p is 0 (n) is the reference signal x in the acoustic space r (n) a signal propagating through the actual primary path P (z) to the error microphone; v p (n) mean is zero and variance isAdditive white gaussian noise of (2); n is the moment, n is more than or equal to 0.
The first linear prediction filtering subsystem 1 comprises a first delay element 11 and a first linear prediction filter 12, wherein the first delay element 11 and the first linear prediction filter 12 are connected in series, and the order of the first delay element 11 is D 1 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the first linear prediction filter 12 are respectivelyAnd L 1 The coefficient is updated by using a least mean square algorithm, and the updating formula is as follows:
h 1,j (n+1)＝h 1,j (n)+μ 1 x w (n)x r (n-D 1 -j)
wherein mu 1 Taking a positive value for the update step of the first linear prediction filter 12; x is x w (n) wideband reference signal separated for first linear prediction subsystem 1Number, x r (n) a reference signal provided for a reference microphone;
the narrowband reference signal and the broadband reference signal synthesized by the first linear prediction filtering subsystem 1 are respectively:
x w (n)＝x r (n)-x f (n)
the broadband secondary sound source synthesis subsystem 2 comprises a broadband controller 21 and a first filtering-X least mean square algorithm module 22;
the broadband controller 21 employs linear filters with coefficients and lengths of respectively And L w ；
The first filter-X least mean square algorithm module 22 uses the wideband reference signal X separated by the auxiliary filter subsystem 5 w (n) related wideband residual noise component y h (n) as an error output and for updating coefficients of the wideband controller 21; the coefficient update formula of the broadband controller 21 is:
wherein mu w The update step length of the broadband controller 21 is a positive value;for wideband reference signal x w (n) output of the secondary channel estimation model in the first filtered-X least mean square algorithm module 22;
the broadband secondary sound source obtained by the broadband secondary sound source synthesis subsystem 2 is
The narrow-band secondary sound source synthesis subsystem 3 comprises a narrow-band controller 31 and a second filtering-X least mean square algorithm module 32;
the narrowband controller 31 employs linear filters having coefficients and lengths of respectivelyAnd L f ；
The second filter-X least mean square algorithm module 32 uses the narrowband residual noise component y separated by the second linear predictive filtering subsystem 4 LP (n) as an error output and for updating coefficients of the narrowband controller 31; the coefficient update formula of the narrowband controller 31 is:
wherein mu f The updating step length of the narrow-band controller 31 is a positive value;for narrowband reference signal x f (n) output of the secondary channel estimation model in the second filtered-X least mean square algorithm module 32;
The narrowband secondary sound source obtained by the narrowband secondary sound source synthesis subsystem 3 is
The second linear prediction filtering subsystem 4 comprises a second delay element 41 and a second linear prediction filter 42, the second delay element 41 and the second linear prediction filter 42 are connected in series, and the order of the second delay element 41 is D 2 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the second linear prediction filter 42 are respectivelyAnd L 2 The coefficients are updated by using a least mean square algorithm, and the formula is updatedThe method comprises the following steps:
h 2,j (n+1)＝h 2,j (n)+μ 2 e LP (n)e(n-D 2 -j)
wherein mu 2 Taking a positive value for the update step of the second linear prediction filter 42; e, e LP (n) a wideband residual noise component separated for the second linear prediction subsystem 4; e (n) is the residual noise provided by the error microphone;
the narrowband residual noise component and the wideband residual noise component separated from the residual noise by the second linear prediction filtering subsystem 4 are respectively:
e LP (n)＝e(n)-y LP (n)
broadband residual noise component e separated by the second linear predictive filtering subsystem 4 LP (n) as a desired input to the auxiliary filtering subsystem 5.
The auxiliary filtering subsystem 5 comprises a linear filter 51 and a least mean square algorithm module 52;
the coefficients and lengths of the linear filter 51 are respectivelyAnd L 3 The coefficient of the linear filter 51 is updated by using a least mean square algorithm module 52, and the update formula is:
h 3,j (n+1)＝h 3,j (n)+μ 3 e h (n)x w (n-j)
Wherein mu 3 The update step length of the linear filter 51 is a positive value; e, e h (n) is a wideband residual noise component related to the additive noise in the auxiliary noise and the target signal separated by the auxiliary filtering subsystem 5;
the wideband residual noise component related to the wideband reference signal separated by the auxiliary filtering subsystem 5 is
Wideband residual noise component e related to additive noise in auxiliary noise and target signal separated by auxiliary filtering subsystem 5 h (n) as a desired input to the secondary channel on-line recognition subsystem 6.
The secondary channel online identification subsystem 6 comprises: a secondary channel on-line recognition module 61 and an auxiliary noise adjustment module 62;
the secondary channel on-line recognition module 61 includes a secondary channel estimation modelThe secondary channel on-line identification module 61 uses e h (n) is the auxiliary noise v generated by the Gaussian white noise v (n) after passing through the auxiliary noise adjusting module 62 as the desired input 0 (n) as a reference input, estimating a time-varying secondary channel in real time using a least mean square algorithm;
secondary channel estimation model of secondary channel on-line identification module 61Coefficients and lengths of +.>And->The coefficient update formula is as follows:
e s (n)＝e h (n)-y s (n)
wherein mu s The updating step length of the secondary channel estimation model is taken as a positive value; y is s (n) is the output of the secondary channel estimation model of the secondary channel on-line recognition module 61;
auxiliary noise v 0 (n) is:
v 0 (n)＝v(n)G(n)
G(n)＝max{G N (n),G B (n)}
G N (n)＝abs[y LP (n-1)]
G B (n)＝abs[y h (n-1)]
wherein G (n) is a gain adjustment factor of the auxiliary noise adjustment module 62; g N (n) and G B (n) is y respectively LP (n-1) and y h (n-1) a low pass filtered output; abs []Calculating for absolute value; v (n) is zero in mean and variance isAdditive white gaussian noise of (c).
The synthesized secondary sound source is:
y(n)＝y w (n)+y f (n)-v 0 (n)
further, the target noise p (n) and the signal y of the secondary sound source y (n) provided by the secondary speaker after passing through the actual secondary channel S (z) p The difference between (n) is the residual noise, i.e. e (n) =p (n) -y p (n) interference cancellation in the acoustic space, and active noise control is achieved.
Embodiment two:
the embodiment provides a robust feedforward type wide and narrow band mixed active noise control method, which is realized based on the robust feedforward type wide and narrow band mixed active noise control system, and comprises the following steps:
step one: setting system parameters
The first linear prediction filter 12, the wide-band controller 21, the narrow-band controller 31, the second linear prediction filter 42, the linear filter 51, and the secondary channel estimation model are respectively setLength and step length of (a); the orders of the first delay links 11 and 41 are respectively set; setting a forgetting factor of the auxiliary noise adjustment module 62; the first linear prediction filter 12, the wideband controller 21, and the narrow band are respectively set Band controller 31, second linear prediction filter 42, linear filter 51, secondary channel estimation model->And the initial values of the gain adjustment factors G (n) of the auxiliary noise adjustment module 62 are all zero; setting auxiliary noise v (n);
step two: synthesizing wideband reference signals and narrowband reference signals
At time n, reference signal x obtained with reference microphone r (n) separating the wideband reference signal and the narrowband reference signal after passing through the first linear predictive filtering subsystem 1, respectively; the separated broadband reference signals are respectively provided for the broadband secondary sound source synthesis subsystem 2 and the auxiliary filtering subsystem 5; the separated narrowband reference signal is provided to a narrowband secondary sound source synthesis subsystem 3;
step three: at time n, first, the broadband reference signal gets the broadband secondary sound source y via the broadband secondary sound source synthesis subsystem 2 w (n) obtaining a narrowband secondary sound source y from the narrowband reference signal via the narrowband secondary sound source synthesis subsystem 3 f (n); second, the auxiliary noise v is obtained by the auxiliary noise adjustment module 62 0 (n) and then y w (n) and y f (n) superposing and synthesizing to obtain a secondary sound source y (n); finally, the residual noise e (n) is separated by the second linear prediction filtering subsystem 4 to obtain a narrow-band residual noise component y LP (n) and wideband residual noise component e LP (n)；
Step four: at time n, wideband residual noise component e LP (n) obtaining wideband residual noise components e related to additive noise in the auxiliary noise and the target signal respectively after passing through the auxiliary filtering subsystem 5 h (n) wideband residual noise component y related to wideband reference signal h (n)；e h (n) as a desired input to the secondary channel online recognition module 61;
step five: update control system
According to the reference signal x r (n) and wideband reference signal x w (n) calculating coefficients of the updated first linear prediction filter 12 at time n+1;
according to the broadband reference signal x w (n) and y h (n) calculating coefficients of the updated wideband controller 21 at time n+1;
from narrowband reference signal x f (n) and narrowband residual noise y LP (n) calculating coefficients of the updated narrowband controller 31 at time n+1;
based on the residual noise e (n) and the wideband residual noise component e LP (n) calculating coefficients of the updated second linear prediction filter 42 at time n+1;
according to the broadband reference signal x w (n) and e h (n) calculating coefficients of the update linear filter 51 at time n+1;
according to the auxiliary noise v 0 (n) and e h (n) computing and updating the secondary channel estimation model in the secondary channel on-line recognition module 61A coefficient at time n+1;
from the narrowband residual noise component y LP (n) and y h (n) calculating a gain adjustment factor at time n+1 for updating the auxiliary noise adjustment module 62;
step six: and returning to the second step, repeating the second step to the fifth step until the system converges and reaches a steady state, and realizing active noise control.
Embodiment III: verification of simulated noise and simulated secondary channel conditions
The reference signal consists of three frequency components and additive noise: the normalized angular frequencies of the three frequency components are ω 1 ＝0.1π、ω 2 =0.2pi and ω 3 =0.3pi, the corresponding discrete fourier coefficients are a respectively 1 ＝2.0、b 1 ＝-1.0、a 2 ＝1.0、b 2 ＝-0.5、a 3 ＝0.5、b 3 =0.1; the additive noise is white gaussian noise with a mean of zero and a variance of 0.25. Additive white gaussian noise v in target noise p (n) p The variance of (n) is 0.1. The actual primary channel P (z) adopts an FIR model, and the length and the cut-off frequency of the actual primary channel P (z) are respectively 41 and 0.4 pi; the actual secondary channel S (z) adopts an FIR model, which is imitatedTime-varying, true secondary channel, the length and cut-off frequency of the former half of the model are 21 and 0.4 pi, respectively, and the length and cut-off frequency of the latter half of the model are 11 and 0.4 pi, respectively; secondary channel estimation modelA length of 31; the auxiliary white gaussian noise v (n) has a mean value of zero and a variance of 1.0. The broadband controller adopts an FIR model, and the length of the broadband controller is 51; the narrowband controller uses an FIR model, with a length of 21. The updating step sizes of the narrowband controller, the wideband controller and the secondary online identification module are respectively 0.001, 0.006 and 0.0008. The order of the first delay link is 5; the first linear prediction filter uses an FIR model with a length of 91 and an update step of 0.0002. The order of the second delay element is 32; the second linear prediction filter uses an FIR model with a length of 91 and an update step of 0.0002. The auxiliary filter adopts an FIR model, the length of the auxiliary filter is 61, and the update step length of the auxiliary filter is 0.003. The independent operation times are 100 times; the length of the simulated sampling point is 30000.
FIG. 2 (a) is a graph showing the dynamic variation of the mean square residual error in the third embodiment; fig. 2 (b) is a graph showing the dynamic change of the secondary channel estimation mean square error of the third embodiment. As shown in fig. 2 (a) and 2 (b), after the system reaches a steady state, the steady state values of the system mean square residual errors of the first half and the second half are 0.1041 and 0.1047 respectively, which are close to the variance of the additive white gaussian noise in the target noise, so that the system has good wide-narrow band mixed noise suppression performance; fig. 2 (b) shows that the system of the invention not only can effectively cope with abrupt changes of the secondary channel, but also has good speed and accuracy of online identification of the secondary channel. Further, the system of the invention overcomes the problem of frequency offset without obtaining the frequency value of the prior narrowband reference component by using a non-acoustic sensor.
Embodiment four: verification of actual target noise with actual secondary channel
The actual noise is from the noise of the large-scale cutting machine, and is the non-stationary characteristic of the simulated target noise, the target noise is divided into two parts, and the rotating speed corresponding to the first half partThe target noise includes a narrowband component and a wideband component at 1400rpm with a corresponding rotational speed of 1600rpm for the second half. The actual primary path P (z) employs a linear FIR model with a length and cut-off frequency of 61 and 0.45 pi, respectively. The target noise is the output of the actual reference signal after passing through the actual primary channel P (z). The actual secondary channel model employs a true IIR model widely used in the art (s.m. kuo and d.r. morgan, active Noise Control Systems-Algorithms and DSP Implementation, new York: wiley, 1996); secondary channel estimation model Length 32; the auxiliary white gaussian noise v (n) has a mean value of zero and a variance of 0.1. The broadband controller adopts an FIR model, and the length of the broadband controller is 51; the narrowband controller uses an FIR model, with a length of 21. The updating step sizes of the narrowband controller, the wideband controller and the secondary on-line identification module are respectively 0.35, 0.35 and 0.07. The order of the first delay link is 21; the first linear prediction filter uses an FIR model, with a length of 21 and an update step of 0.5. The order of the second delay link is 21; the second linear prediction filter uses an FIR model, with a length of 21 and an update step of 0.5. The auxiliary filter uses an FIR model with a length of 41 and an update step of 0.5. The independent operation times are 100 times; the actual sampling length is 30000.
FIG. 3 is a dynamic variation curve of target noise and residual noise of the fourth embodiment; when the system reaches a steady state, the noise reduction of the first half system is 12.15dB, and the noise reduction of the second half system is 17.33dB, which shows that the system has good wide-narrow-band mixed noise suppression performance under the conditions of actual non-steady target noise and actual secondary channels. Further, the system of the invention overcomes the problem of frequency offset without obtaining the frequency value of the prior narrowband reference component by using a non-acoustic sensor.
The effectiveness and the practicability of the robust feedforward wide-narrow-band hybrid active noise control system and the robust feedforward wide-narrow-band hybrid active noise control method provided by the invention are jointly verified from two conditions of simulation and experiment respectively, and the actual application of an active noise control technology is further promoted.
Some steps in the embodiments of the present invention may be implemented by using software, and the corresponding software program may be stored in a readable storage medium, such as an optical disc or a hard disk.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (9)
1. A robust feedforward type wide-narrow-band mixed active noise control system is characterized in that the active noise control system respectively adopts a reference microphone to collect reference signals, adopts an error microphone to collect residual noise and adopts a secondary loudspeaker to provide a secondary sound source; the actual primary channel in the acoustic space is a channel model of the reference signal propagating to the error microphone; the actual secondary channel in the acoustic space is a model of the channel that propagates the secondary sound source provided by the secondary speaker to the error microphone; the active noise control system comprises a first linear prediction filtering subsystem (1), a broadband secondary sound source synthesis subsystem (2), a narrowband secondary sound source synthesis subsystem (3), a second linear prediction filtering subsystem (4), an auxiliary filtering subsystem (5) and a secondary channel online identification subsystem (6);
The first linear prediction filtering subsystem (1) is respectively connected with the broadband secondary sound source synthesis subsystem (2), the narrowband secondary sound source synthesis subsystem (3) and the auxiliary filtering subsystem (5); the broadband secondary sound source synthesis subsystem (2) is respectively connected with the first linear prediction filtering subsystem (1) and the auxiliary filtering subsystem (5); the narrow-band secondary sound source synthesis subsystem (3) is respectively connected with the first linear prediction filtering subsystem (1) and the second linear prediction filtering subsystem (4); the second linear prediction filtering subsystem (4) is respectively connected with the narrow-band secondary sound source synthesis subsystem (3), the auxiliary filtering subsystem (5) and the secondary channel on-line identification subsystem (6); the auxiliary filtering subsystem (5) is respectively connected with the first linear prediction filtering subsystem (1), the broadband secondary sound source synthesis subsystem (2), the second linear prediction filtering subsystem (4) and the secondary channel on-line identification subsystem (6); the secondary channel online identification subsystem (6) is respectively connected with the second linear prediction filtering subsystem (4) and the auxiliary filtering subsystem (5);
the first linear predictive filtering subsystem (1) is used for synthesizing a broadband reference signal and a narrowband reference signal; the broadband secondary sound source synthesis subsystem (2) is used for synthesizing a broadband secondary sound source; the narrow-band secondary sound source synthesis subsystem (3) is used for synthesizing a narrow-band secondary sound source; the second linear prediction filtering subsystem (4) is used for separating a narrow-band residual noise component and a broadband residual noise component from residual noise; the auxiliary filtering subsystem (5) is used for separating broadband residual noise components related to broadband reference signals and broadband residual noise components related to additive noise in auxiliary noise and target signals from the broadband residual noise components; the secondary channel online identification subsystem (6) is used for estimating a time-varying secondary channel model in real time;
The narrow-band residual noise component separated by the second linear prediction filtering subsystem (4) is respectively used as error output of the narrow-band secondary sound source synthesis subsystem (3) and input of an auxiliary noise adjustment module in the secondary channel online identification subsystem (6); meanwhile, the broadband residual noise component which is separated by the auxiliary filtering subsystem (5) and related to the broadband reference signal is respectively used as an error output of the broadband secondary sound source synthesis subsystem (2) and an auxiliary noise adjustment module input in the secondary channel on-line identification subsystem (6); simultaneously, the broadband residual noise component which is separated by the auxiliary filtering subsystem (5) and related to the auxiliary noise and additive noise in the target signal is used as the expected input of the secondary channel on-line identification subsystem (6); the method can improve the independence among the broadband secondary sound source synthesis subsystem (2), the narrowband secondary sound source synthesis subsystem (3) and the secondary channel online identification subsystem (6), improve the accuracy and the speed of online identification of the secondary channel, improve the dynamic performance of the whole system, reduce the influence of the introduced auxiliary noise on the residual noise, and improve the noise suppression performance of the whole system.
2. A robust feed forward type wide narrowband hybrid active noise control system as claimed in claim 1, characterized in that said first linear predictive filtering subsystem (1) comprises a first delay element (11) and a first linear predictive filter (12), said first delay element (11) and said first linear predictive filter (12) being connected in series, said first delay element (11) having an order D 1 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the first linear prediction filter (12) are respectivelyAnd L 1 The coefficient is updated by using a least mean square algorithm, and the updating formula is as follows:
h 1,j (n+1)＝h 1,j (n)+μ 1 x w (n)x r (n-D 1 -j)
wherein mu 1 Taking a positive value for the update step size of the first linear prediction filter (12); x is x w (n) wideband reference signal, x, separated for the first linear prediction subsystem (1) r (n) a reference signal provided for the reference microphone; n is the moment, n is more than or equal to 0;
the narrowband reference signals and the broadband reference signals synthesized by the first linear prediction filtering subsystem (1) are respectively as follows:
x w (n)＝x r (n)-x f (n)。
3. a robust feed forward wide narrowband hybrid active noise control system as in claim 2, wherein the wideband secondary sound source synthesis subsystem (2) comprises a wideband controller (21) and a first filter-X least mean square algorithm module (22);
The wideband controller (21) employs a linear filter having coefficientsAnd respectively have the length ofAnd L w ；
The first filter-X least mean square algorithm module (22) uses the wideband reference signal X separated by the auxiliary filter subsystem (5) w (n) related wideband residual noise component y h (n) as an error output and for updating coefficients of the wideband controller (21); the coefficient update formula of the broadband controller (21) is as follows:
wherein mu w Taking a positive value for the update step length of the broadband controller (21);for the wideband reference signal x w (n) output of a secondary channel estimation model in said first filter-X least mean square algorithm module (22);
the broadband secondary sound source obtained by the broadband secondary sound source synthesis subsystem (2) is as follows:
4. a robust feed forward wide narrowband hybrid active noise control system as in claim 3, wherein said narrowband secondary sound source synthesis subsystem (3) comprises a narrowband controller (31) and a second filter-X least mean square algorithm module (32);
the narrowband controller (31) adopts a linear filter, and the coefficients and the lengths of the linear filter are respectivelyAnd L f ；
The second filtering-X least mean square algorithm module (32) utilizes the narrow band residual noise component y separated by the second linear prediction filtering subsystem (4) LP (n) as an error output and for updating coefficients of the narrowband controller (31); the coefficient update formula of the narrowband controller (31) is as follows:
wherein mu f Taking a positive value for the update step length of the narrowband controller (31);for the narrowband reference signal x f (n) output of a secondary channel estimation model in said second filter-X least mean square algorithm module (32);
the narrowband secondary sound source obtained by the narrowband secondary sound source synthesis subsystem (3) is as follows:
5. a robust feed forward wide narrowband hybrid active noise control system in accordance with claim 4, wherein said second linear predictive filtering subsystem (4) comprises a second delay element (41) and a second linear predictive filter (42), said second delay element (41) and said second linear predictive filter (42) being connected in series, said second delay element (41) having an order D 2 The method comprises the steps of carrying out a first treatment on the surface of the The coefficients and lengths of the second linear prediction filter (42) are respectivelyAnd L 2 The coefficient is updated by using a least mean square algorithm, and the updating formula is as follows:
h 2,j (n+1)＝h 2,j (n)+μ 2 e LP (n)e(n-D 2 -j)
wherein mu 2 Taking a positive value for an update step of the second linear prediction filter (42); e, e LP (n) a wideband residual noise component separated for the second linear prediction subsystem (4); e (n) residual noise provided for the error microphone;
The narrow-band residual noise component and the wide-band residual noise component separated from the residual noise by the second linear prediction filtering subsystem (4) are respectively:
e LP (n)＝e(n)-y LP (n)
broadband residual noise component e separated by the second linear predictive filtering subsystem (4) LP (n) serving as a desired input to the auxiliary filtering subsystem (5).
6. A robust feed forward wide narrowband hybrid active noise control system in accordance with claim 5, wherein said auxiliary filtering subsystem (5) comprises a linear filter (51) and a least mean square algorithm module (52);
the coefficients and lengths of the linear filters (51) are respectivelyAnd L 3 And updating the coefficients of the linear filter (51) by using the least mean square algorithm module (52), wherein the updating formula is as follows:
h 3,j (n+1)＝h 3,j (n)+μ 3 e h (n)x w (n-j)
wherein mu 3 Taking a positive value for the update step length of the linear filter (51); e, e h (n) a wideband residual noise component related to additive noise in the auxiliary noise and the target signal separated by the auxiliary filtering subsystem (5);
the wideband residual noise component related to the wideband reference signal separated by the auxiliary filtering subsystem (5) is:
broadband residual noise component e related to additive noise in auxiliary noise and target signal separated by auxiliary filtering subsystem (5) h (n) serving as a desired input to the secondary channel on-line recognition subsystem (6).
7. A robust feed forward wide narrowband hybrid active noise control system in accordance with claim 6, wherein said secondary channel online identification subsystem (6) comprises: a secondary channel on-line identification module (61) and an auxiliary noise adjustment module (62);
the secondary channel on-line recognition module (61) comprises a secondary channel estimation modelThe secondary channel on-line identification module (61) uses e h (n) is an auxiliary noise v generated by the Gaussian white noise v (n) passing through the auxiliary noise adjusting module (62) and being a desired input 0 (n) as a reference input, estimating a time-varying secondary channel in real time using a least mean square algorithm;
a secondary channel estimation model of the secondary channel on-line identification module (61)Coefficient and length of (a) are respectivelyAnd->The coefficient update formula is as follows:
e s (n)＝e h (n)-y s (n)
wherein mu s The updating step length of the secondary channel estimation model is taken as a positive value; y is s (n) is the output of a secondary channel estimation model of the secondary channel on-line recognition module (61);
the auxiliary noise v 0 (n) is:
v 0 (n)＝v(n)G(n)
G(n)＝max{G N (n),G B (n)}
G N (n)＝abs[y LP (n-1)]
G B (n)＝abs[y h (n-1)]
wherein G (n) is a gain adjustment factor of the auxiliary noise adjustment module (62); g N (n) and G B (n) is y respectively LP (n-1) and y h (n-1) a low pass filtered output; abs []Calculating for absolute value; v (n) is zero in mean and variance isAdditive white gaussian noise of (c).
8. The robust feed forward wide narrowband hybrid active noise control system of claim 7, wherein the synthesized secondary sound source is:
y(n)＝y w (n)+y f (n)-v 0 (n)
and y (n) is output to the secondary speaker, and interference with the target noise is canceled in the acoustic space.
9. An active noise control method, wherein the method is implemented based on the robust feedforward wide narrowband hybrid active noise control system of claim 8, the method comprising:
step one: setting system parameters;
respectively set up the firstA linear prediction filter (12), a wideband controller (21), a narrowband controller (31), a second linear prediction filter (42), a linear filter (51), and a secondary channel estimation modelLength and step length of (a); the steps of the first delay link (11) and the first delay link (41) are respectively set; setting a forgetting factor of the auxiliary noise adjustment module (62); the first linear prediction filter (12), the broadband controller (21), the narrowband controller (31), the second linear prediction filter (42), the linear filter (51) and the secondary channel estimation model are respectively arranged >And the initial values of the gain adjustment factors G (n) of the auxiliary noise adjustment module (62) are all zero; setting auxiliary noise v (n);
step two: synthesizing a wideband reference signal and a narrowband reference signal;
at time n, reference signal x obtained with reference microphone r (n) separating the wideband reference signal and the narrowband reference signal respectively after passing through the first linear predictive filtering subsystem (1); the separated broadband reference signals are respectively provided for a broadband secondary sound source synthesis subsystem (2) and an auxiliary filtering subsystem (5); the separated narrowband reference signal is provided to a narrowband secondary sound source synthesis subsystem (3);
step three: at time n, first, a broadband reference signal is passed through the broadband secondary sound source synthesis subsystem (2) to obtain a broadband secondary sound source y w (n) obtaining a narrowband secondary sound source y by the narrowband reference signal through a narrowband secondary sound source synthesis subsystem (3) f (n); secondly, the auxiliary noise v is obtained by the auxiliary noise adjusting module (62) 0 (n) and then y w (n) and y f (n) superposing and synthesizing to obtain a secondary sound source y (n); finally, the residual noise e (n) is separated by a second linear prediction filtering subsystem (4) to obtain a narrow-band residual noise component y LP (n) and wideband residual noise component e LP (n)；
Step four: at time n, the wideband residual noise Acoustic component e LP (n) obtaining wideband residual noise components e related to additive noise in the auxiliary noise and the target signal respectively after passing through the auxiliary filtering subsystem (5) h (n) wideband residual noise component y related to wideband reference signal h (n)；e h (n) serving as a desired input to the secondary channel online recognition module (61);
step five: updating the control system;
according to the reference signal x r (n) and the wideband reference signal x w (n) calculating coefficients at time n+1 for updating the first linear prediction filter (12);
according to the broadband reference signal x w (n) and wideband residual noise component y h (n) calculating coefficients at time n+1 for updating the broadband controller (21);
from the narrowband reference signal x f (n) and narrowband residual noise y LP (n) calculating coefficients at time n+1 for updating the narrowband controller (31);
based on the residual noise e (n) and wideband residual noise component e LP (n) calculating coefficients at time n+1 for updating the second linear prediction filter (42);
according to the broadband reference signal x w (n) and e h (n) calculating coefficients at time n+1 for updating the linear filter (51);
according to the auxiliary noise v 0 (n) and e h (n) computationally updating a secondary channel estimation model in the secondary channel on-line recognition module (61) A coefficient at time n+1;
from the narrowband residual noise component y LP (n) and y h (n) calculating and updating a gain adjustment factor of the auxiliary noise adjustment module (62) at time n+1;
step six: and returning to the second step, and repeating the second step to the fifth step until the system converges and reaches a steady state, thereby realizing active noise control.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310665844.7A CN116721649A (en) | 2023-06-06 | 2023-06-06 | Robust feedforward type wide-narrow-band hybrid active noise control system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310665844.7A CN116721649A (en) | 2023-06-06 | 2023-06-06 | Robust feedforward type wide-narrow-band hybrid active noise control system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116721649A true CN116721649A (en) | 2023-09-08 |
Family
ID=87870826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310665844.7A Pending CN116721649A (en) | 2023-06-06 | 2023-06-06 | Robust feedforward type wide-narrow-band hybrid active noise control system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116721649A (en) |
-
2023
- 2023-06-06 CN CN202310665844.7A patent/CN116721649A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108428445B (en) | Self-adaptive active noise reduction method of error-free microphone | |
| CN113327570B (en) | Narrow-band feedforward type active noise control system and method | |
| CN111986644B (en) | Wide-narrow band hybrid feedforward active noise control system | |
| CA2094984A1 (en) | Noise control system | |
| CN106531145B (en) | Recurrence active noise control method based on M estimator | |
| CN101379872A (en) | Hearing aid with self-adapting feedback inhibition system | |
| CN111627414B (en) | Active denoising method and device and electronic equipment | |
| CN111063331A (en) | Parallel noise cancellation system using harmonic filtering | |
| CN105976806B (en) | Active noise control method based on maximum entropy | |
| CN113851104B (en) | Feedback type active noise control system and method containing secondary channel online identification | |
| Patel et al. | Modified phase-scheduled-command FxLMS algorithm for active sound profiling | |
| Gomathi et al. | Variable step size for improving convergence of FxLMS algorithm | |
| CN109658947A (en) | A kind of active noise controlling method of synchronous modeling and control | |
| CN113593515B (en) | Wide-narrow band hybrid active noise control system for coping with frequency offset | |
| CN116721649A (en) | Robust feedforward type wide-narrow-band hybrid active noise control system and method | |
| CN113299260B (en) | Active noise reduction method based on EMFNL filter on-line modeling secondary channel | |
| Wang et al. | Online secondary path modeling for active sound quality control systems | |
| JPH09511081A (en) | Time domain adaptive control system | |
| CN113096629B (en) | Relative path virtual sensing method for single-channel feedback active noise control system | |
| CN115394311B (en) | Robust narrow-band feedback type active noise control system and method | |
| Chang et al. | Enhanced offline secondary path modeling method for narrowband active noise control system | |
| CN117524182A (en) | Narrow-band feed-forward hybrid active noise control system and method | |
| Sharifi-Tehrani | Design, simulation and implementation of an active sound-noise cancellation system for use in a cockpit intercommunication system | |
| Yang et al. | Hybrid filtered-x adaptive vibration control with internal feedback and online identification | |
| Wu et al. | Multi-reference Adaptive Gain FXLMS Algorithm for Active Noise Control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |