Nothing Special   »   [go: up one dir, main page]

CN108900943A - A kind of scene adaptive active denoising method and earphone - Google Patents

A kind of scene adaptive active denoising method and earphone Download PDF

Info

Publication number
CN108900943A
CN108900943A CN201810818230.7A CN201810818230A CN108900943A CN 108900943 A CN108900943 A CN 108900943A CN 201810818230 A CN201810818230 A CN 201810818230A CN 108900943 A CN108900943 A CN 108900943A
Authority
CN
China
Prior art keywords
noise
filter
signal
secondary channel
noise reduction
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.)
Granted
Application number
CN201810818230.7A
Other languages
Chinese (zh)
Other versions
CN108900943B (en
Inventor
陈龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan Changhong Electric Co Ltd
Original Assignee
Sichuan Changhong Electric Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Sichuan Changhong Electric Co Ltd filed Critical Sichuan Changhong Electric Co Ltd
Priority to CN201810818230.7A priority Critical patent/CN108900943B/en
Publication of CN108900943A publication Critical patent/CN108900943A/en
Application granted granted Critical
Publication of CN108900943B publication Critical patent/CN108900943B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R1/00Details of transducers, loudspeakers or microphones
    • H04R1/10Earpieces; Attachments therefor ; Earphones; Monophonic headphones
    • H04R1/1091Details not provided for in groups H04R1/1008 - H04R1/1083
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R2460/00Details of hearing devices, i.e. of ear- or headphones covered by H04R1/10 or H04R5/033 but not provided for in any of their subgroups, or of hearing aids covered by H04R25/00 but not provided for in any of its subgroups
    • H04R2460/01Hearing devices using active noise cancellation

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Signal Processing (AREA)
  • Soundproofing, Sound Blocking, And Sound Damping (AREA)
  • Circuit For Audible Band Transducer (AREA)

Abstract

The invention discloses a kind of scene adaptive active denoising methods, including:It picks up ambient noise and is converted into primary noise electric signal;Noise Identification module determines noise type;EQ filter parameter is selected to be loaded into active noise reduction module from filter bank;Active noise reduction module carries out adaptive-filtering to primary noise electric signal, and generates the de-noising electric signal of excitation loudspeaker;After loudspeaker receives de-noising electric signal, de-noising sound wave is generated, and the audio signal after secondary counteracting is issued, the de-noising sound wave is identical as the primary noise electric signal spectrum distribution, acoustic pressure size is identical and opposite in phase.A kind of scene adaptive active noise reduction earphone, including microphone, loudspeaker and active noise reduction module.Noise cancelling headphone can adjust EQ filter parameter in FXLMS noise reduction algorithm, offset the noise that passive noise reduction fails to remove according to the difference of noise energy concentrated area in noise scenarios, export the influence of audio to earphone so as to improve the external noise of scene locating for human ear.

Description

A kind of scene adaptive active denoising method and earphone
Technical field
The present invention relates to noise cancelling headphone fields, are a kind of scene adaptive active denoising method and earphone specifically.
Background technique
With people's traffic trip increasingly frequently, ambient noise has seriously affected the quality of trip.Traditional noise reduction Means mainly have sound insulation, material absorption etc., but are limited to the factors such as arrangement space, material property and cost, and conventional method is to height Frequency noise remove effect is preferable, but not satisfactory to low-frequency noise effect.Therefore, active noise reduction start from civil aviaton's military field by Gradually enter into popular life.It is different from traditional noise reduction means, active noise controlling (ANC) be by the principle of sound wave interference cancellation, Original noise is offset using secondary sound source sounding to realize that noise is eliminated.Active noise reduction can be according to environmental change adjust automatically Noise reduction strategies, and it is capable of the noise of the processing special frequency channel of selectivity, to be obviously improved noise reduction quality.Currently, actively The most famous control algolithm that noise cancelling headphone uses is the Filtered-X LMS algorithm (FXLMS) proposed by Widrow.The algorithm characteristic is A filter identical with secondary channel transmission characteristic is placed in reference signal channel to carry out LMS algorithm weight update, with solution Certainly introduce secondary channel bring systematic jitters problem.But based on the noise cancelling headphone of FXLMS algorithm design, in use process The problems such as it is slow that there are convergence rates, only good to narrow-band noise effect, and undesirable to broadband noise control effect, therefore very much Preferable noise reduction effect is unable to get under scene.
Summary of the invention
The purpose of the present invention is to provide a kind of scene adaptive active denoising method and earphones, for solving the prior art It is middle slow using FXLMS algorithm progress active noise reduction convergence rate, and to the problem of broadband noise control effect difference.
The present invention is solved the above problems by following technical proposals:
A kind of scene adaptive active denoising method, including:
Step S1:Microphone picks up ambient noise and is converted into primary noise electric signal;
Step S2:Noise Identification module extracts noise energy collection Mid Frequency and frequency distribution according to primary noise electric signal Characteristic, and matched with preset noise pattern in noise pattern library, determine noise type;
Step S3:EQ filter parameter is selected to be loaded into active noise reduction module, the EQ filter ginseng from filter bank Number includes centre frequency f, bandwidth Q and amount of gain GAIN;
Step S4:Active noise reduction module carries out adaptive-filtering to primary noise electric signal, and generates excitation loudspeaker De-noising electric signal;
Step S5:After loudspeaker receives de-noising electric signal, de-noising sound wave is generated, and by the audio signal after secondary counteracting It issues, the de-noising sound wave is identical as the primary noise electric signal spectrum distribution, acoustic pressure size is identical and opposite in phase.
Principle:
Microphone is converted into electric signal after being picked up to ambient noise at earphone;In Noise Identification module, extraction is made an uproar Acoustical signal energy collection Mid Frequency and frequency distribution feature, and the noise characteristic extracted is compared in default noise pattern library It is right, it obtains and the immediate noise pattern classification of this scene;According to the type of noise, corresponding EQ is selected in noise pattern library Filter parameter (centre frequency (f), bandwidth (Q) and amount of gain (GAIN)), and be loaded into active noise reduction module EQ filter In wave device, active noise module loading EQ filter parameter, active noise reduction module is on the basis of Filtered-X LMS algorithm (FXLMS) On, EQ filter is introduced, is by Noise Identification module, to noise signal energy collection Mid Frequency, frequency due to EQ filter parameter After rate distribution characteristics is analyzed, be compared in noise pattern library after determine, therefore, the introducing of EQ filter enhances Processing and generation de-noising electric signal to noise collection Mid Frequency, and loudspeaker is motivated to generate de-noising sound wave and noise cancellation, and The audio signal after secondary counteracting is issued by loudspeaker.Therefore, either narrow-band noise or broadband noise, can lead to It crosses Noise Identification module and determines type, and then load corresponding EQ filter parameter, obtain the noise reduction capability under the scene greatly Amplitude is promoted, and the noise for entering human ear can be offset according to scene, improves locating noise circumstance, allows user in noisy environment In can also obtain lossless, high-fidelity audio.
Further, the step S4 is specifically included:
Step S41:EQ filter in active noise reduction module loads the EQ filter parameter;
Step S42:Secondary channel is introduced, the secondary channel includes sequentially connected inverse filter Inv_S (z), adaptive Filter W (z) and secondary channel system S (z) are answered, the inverse filter Inv_S (z) is for offsetting secondary channel interference;
Step S43:Signal is generated by noise generator to train secondary channel, to obtain secondary channel transmission function Identification result
Step S44:By the secondary channel transmission function result of the identificationIt substitutes into inverse filter Inv_S (z), then defeated Enter sef-adapting filter W (z) and carries out the processing of LMS adaptive-filtering;
Step S45:Error-filter in active noise reduction module generates the de-noising electric signal of excitation loudspeaker.
Principle:
For better noise reduction, active noise reduction module includes main channel and secondary channel, and signal X (n) is respectively by main logical Error originated from input filter after sound wave counteracting is carried out after road P (z) and secondary channel, signal X (n) signal after main channel P (z) is D (n), signal X (n) are signal after preposition inverse filter Inv_S (z) on secondary channelSignalThrough Crossing sef-adapting filter W (z) afterwards is signal y (n), and signal y (n) exports de-noising electric signal s by secondary channel system S (z) (n), signal e (n) the error originated from input sensor after de-noising electric signal s (n) and signal d (n) sound wave are offset;SignalBy secondary Grade channel transfer function identification resultIt is afterwards signalSignalSignal q (n) after EQ filter with LMS is carried out in signal e1 (n) input adaptive filter W (z) of the signal e (n) after another EQ filter adaptively to filter Wave.
Further, the step S43 is specifically included:In initialization, pumping signal is generated by external control circuit, it should Signal by secondary channel and recognizes it, then picks up the signal by error pick-up, uses after sending circuit system back to LMS algorithm calculates the transmission function of secondary channel, to obtain the estimated value of the channel transfer function S (n)
Principle:It first has to exclusion introducing secondary channel to interfere to noise reduction bring, therefore first to obtain secondary channel transmitting The estimated value of functionSecondary channel interference is offset in the noise reduction module using preposition inverse filter Inv_S (z).Estimate EvaluationBe by external circuit generate pumping signal, after secondary channel send back to circuit system with LMS algorithm into What row was calculated.
Further, the step S44 is specifically included:
A, the signal e (n) that error pick-up receives is solved, specially:
The input signal X (n) of active noise reduction module can reach error through main channel P (z) and two paths of secondary channel and pass Sensor, secondary channel output are expressed as:If assuming at this time adaptive It answers filter W (z) weight coefficient to remain unchanged in a bit of time L, enables r (n)=x (n) * s (n), can be obtainedThe signal e (n) that error pick-up receives at this time is main channel P (z) and the sum of the output signal of secondary channel, i.e.,:
E (n)=d (n)+s (n)=d (n)+rT(n) W (n), wherein d (n) is signal X (n) after main channel P (z) Output signal, W (n) are the transmission function of sef-adapting filter;
B, the best weight coefficient of sef-adapting filter is solved, specially:
Objective function of the least-mean-square-error criterion as signal e (n) is selected, is enabled P=E [d (n) r (n)], R=E [r (n) rT (n)],
(1) when steady input signal X (n), then the Quadratic Function Optimization of weight vector
Since matrix R is that positive definite is symmetrical, show J (n) existence anduniquess minimum value, thus can calculate best weight coefficient arrow Amount, calculates to obtain W0=-R-1P;
(2) it when input signal X (n) is unstable, is solved using recursion, i.e. the output recursion of sef-adapting filter W (z) Relationship is:W (n+1)=W (n)-μ Δ W (n), wherein μ is the convergence step-length that automatic adjusument introduces, using single error signal Estimation of the squared gradient as mean square error gradient:Obtain sef-adapting filter Weight coefficient vector more new formula is:W (n+1)=- 2 μ e (n) r (n) of W (n).
The table of the main channel by loudspeaker, the signal e (n) after the sound wave counteracting after secondary channel is given in step A It is related with sef-adapting filter W (z), main channel P (z) and secondary channel system S (z) up to formula, whereinIn, variable l, the i.e. L of small letter.Signal e (n) can be considered random change Amount further solves the best weight coefficient of sef-adapting filter W (z), and according to input signal, two ways can be taken to be counted Calculate optimum weight vector.
Further, there are two the EQ filter is set, one of EQ filter is for receiving secondary channel transmitting letter Number identification resultOutput signal is simultaneously filtered rear input adaptive filter W (z), another EQ filter is for receiving Error pick-up received signal e (n) is simultaneously filtered rear input adaptive filter W (z).
It can be seen that the signal of different frequency distribution character from second of calculation for calculating optimum weight vector, Its noise energy is also concentrated in different frequency range, thus in order to enhance active noise reduction module to the noise reduction effect of a certain scene, master Dynamic noise reduction module introduces processing of the enhancing of EQ filter two to noise collection Mid Frequency.After increasing EQ filter, sef-adapting filter W (z) bigger weight can be occupied by frequency range being selected in weight coefficient recurrence calculation, finally allow noise reduction of the noise reduction module under the scene Ability is improved significantly.
Further, there are three the EQ filter is set, one of EQ filter is for receiving secondary channel transmitting letter Number identification resultOutput signal is simultaneously filtered rear input adaptive filter W (z), other two EQ filter is used respectively It is filtered in by the output signal of main channel P (z), secondary channel system S (z), filtered signal carries out sound wave counteracting and inputs Error-filter and sef-adapting filter W (z).
EQ filter filtering first is carried out to main channel signal and secondary channel signal before entering error pick-up, this When, the signal d'(n of main channel error originated from input sensor)=d (n) * eq (n), the signal y' of secondary channel error originated from input sensor (n)=s (n) * eq (n), it can be seen that EQ filter, which is equivalent to, has different gains to each frequency range of S (n), and selectes frequency Duan Zengyi is bigger, therefore the d'(n exported) in select frequency range amplitude bigger, and then allow selected frequency range pair into sef-adapting filter The adjusting weight of sef-adapting filter increases, and so that noise energy collection Mid Frequency is occupied bigger weight in weight coefficient update, most Noise reduction capability of the noise reduction module under the scene is allowed to be improved significantly eventually.
A kind of scene adaptive active noise reduction earphone, including microphone and loudspeaker, the microphone are sequentially connected preposition Amplifying circuit, Noise Identification module, active noise reduction module and power amplification circuit, the power amplification circuit and the loudspeaker Connection, wherein:
Microphone, for picking up ambient noise and being converted to primary noise electric signal;
Noise Identification module, for extracting noise immune collection Mid Frequency and frequency distribution from the primary noise electric signal Characteristic, and matched in noise pattern library with preset noise pattern, it determines noise type, is then selected in filter bank It selects corresponding EQ filter parameter and is loaded into active noise reduction module;
Active noise reduction module carries out the processing of the improvement FXLMS adaptive-filtering based on EQ filter to primary noise signal, The de-noising electric signal of excitation loudspeaker is generated, and is exported to power amplification circuit;
Loudspeaker generates de-noising sound wave, the de-noising sound wave and primary noise signal spectrum after receiving de-noising electric signal Distribution, acoustic pressure size are identical, opposite in phase, and loudspeaker exports the audio signal after offset noise.
Active noise reduction module makes sub-loudspeaker generate the de-noising sound wave opposite with main loudspeaker amplitude same phase, reaches Active noise reduction.
Compared with prior art, the present invention having the following advantages that and beneficial effect:
(1) noise cancelling headphone can adjust FXLMS noise reduction algorithm according to the difference of noise energy concentrated area in noise scenarios Middle EQ filter parameter offsets the noise that passive noise reduction fails to remove, so as to improve scene locating for human ear external noise to ear The influence of machine output audio.
(2) present invention improves the FXLMS algorithm of type using EQ filter, can be had according to scene difference, signal Different frequency distribution characteristics, noise energy are also concentrated in different frequency range, in order to enhance active noise reduction module to special scenes Noise reduction effect, the noise reduction module introduces EQ filter to enhance to the processing of noise collection Mid Frequency, can be under many scenes It realizes to broadband noise filter effect.
Detailed description of the invention
Fig. 1 is Headphone structure block diagram of the invention;
Fig. 2 is noise reduction flow chart of the invention;
Fig. 3 is the schematic diagram of the third specific embodiment of the invention;
Fig. 4 is the schematic diagram of specific embodiment in the present invention the 4th.
Specific embodiment
Firstly, being given to the signal being referred to herein before introducing specific embodiments of the present invention and being explained as follows table:
The present invention is described in further detail below with reference to embodiment, embodiments of the present invention are not limited thereto.
Embodiment 1:
In conjunction with shown in attached drawing 1 and Fig. 2, a kind of scene adaptive active denoising method, including:
Step S1:Microphone picks up ambient noise and is converted into primary noise electric signal;
Step S2:Noise Identification module extracts noise energy collection Mid Frequency and frequency distribution according to primary noise electric signal Characteristic, and matched with preset noise pattern in noise pattern library, determine noise type;
Step S3:EQ filter parameter is selected to be loaded into active noise reduction module, the EQ filter ginseng from filter bank Number includes centre frequency f, bandwidth Q and amount of gain GAIN;
Step S4:Active noise reduction module carries out adaptive-filtering to primary noise electric signal, and generates excitation loudspeaker De-noising electric signal;
It specifically includes:Step S41:EQ filter in active noise reduction module loads the EQ filter parameter;
Step S42:Secondary channel is introduced, the secondary channel includes sequentially connected inverse filter Inv_S (z), adaptive Filter W (z) and secondary channel system S (z) are answered, the inverse filter Inv_S (z) is for offsetting secondary channel interference;
Step S43:Signal is generated by noise generator to train secondary channel, to obtain secondary channel transmission function Identification result
Step S44:By the secondary channel transmission function result of the identificationIt substitutes into inverse filter Inv_S (z), then defeated Enter sef-adapting filter W (z) and carries out the processing of LMS adaptive-filtering;
Step S45:Error-filter in active noise reduction module generates the de-noising electric signal of excitation loudspeaker;
Step S5:After loudspeaker receives de-noising electric signal, de-noising sound wave is generated, and by the audio signal after secondary counteracting It issues, the de-noising sound wave is identical as the primary noise electric signal spectrum distribution, acoustic pressure size is identical and opposite in phase.
Principle:
Microphone is converted into electric signal after being picked up to ambient noise at earphone;In Noise Identification module, extraction is made an uproar Acoustical signal energy collection Mid Frequency and frequency distribution feature, and the noise characteristic extracted is compared in default noise pattern library It is right, it obtains and the immediate noise pattern classification of this scene;According to the type of noise, corresponding EQ is selected in noise pattern library Filter parameter (centre frequency (f), bandwidth (Q) and amount of gain (GAIN)), and be loaded into active noise reduction module EQ filter In wave device.Active noise reduction module includes main channel and secondary channel, and signal X (n) passes through main channel P (z) and secondary channel respectively Error originated from input filter after progress sound wave counteracting afterwards, signal X (n) signal after main channel P (z) is d (n), and signal X (n) exists It is signal after preposition inverse filter Inv_S (z) on secondary channelSignalBy sef-adapting filter W It (z) is afterwards signal y (n), signal y (n) is by secondary channel system S (z) output de-noising electric signal s (n), de-noising electric signal s (n) Signal e (n) error originated from input sensor after being offset with signal d (n) sound wave;SignalIt is recognized by secondary channel transmission function As a resultIt is afterwards signalSignalSignal q (n) and signal e (n) after EQ filter pass through another EQ LMS adaptive-filtering is carried out in signal e1 (n) input adaptive filter W (z) after filter.EQ filter, which loads noise, to be known The filter parameter selected in other module is to be concentrated by Noise Identification module to noise signal energy due to EQ filter parameter After frequency range, frequency distribution feature are analyzed, be compared in noise pattern library after determine, therefore, EQ filter draws Enter the processing enhanced to noise collection Mid Frequency and generate de-noising electric signal, and loudspeaker is motivated to generate de-noising sound wave and noise It offsets, and is issued the audio signal after secondary counteracting by loudspeaker.Therefore, either narrow-band noise or broadband noise, Type can be determined by Noise Identification module, and then loads corresponding EQ filter parameter, make the noise reduction capability under the scene It is improved significantly, the noise for entering human ear can be offset according to scene, improves locating noise circumstance, allows user in noise Also lossless, high-fidelity audio can be obtained in heterocycle border.
Embodiment 2:
On the basis of embodiment 1, the step S44 is specifically included:
A, the signal e (n) that error pick-up receives is solved, specially:
The input signal X (n) of active noise reduction module can reach error through main channel P (z) and two paths of secondary channel and pass Sensor, secondary channel output are expressed as:If assuming at this time adaptive It answers filter W (z) weight coefficient to remain unchanged in a bit of time L, enables r (n)=x (n) * s (n), can be obtainedThe signal e (n) that error pick-up receives at this time is main channel P (z) and the sum of the output signal of secondary channel, i.e.,:
E (n)=d (n)+s (n)=d (n)+rT(n) W (n), wherein d (n) is signal X (n) after main channel P (z) Output signal, W (n) are the transmission function of sef-adapting filter;
B, the best weight coefficient of sef-adapting filter is solved, specially:
Objective function of the least-mean-square-error criterion as signal e (n) is selected, is enabled P=E [d (n) r (n)], R=E [r (n) rT (n)],
(1) when steady input signal X (n), then the Quadratic Function Optimization of weight vector
Since matrix R is that positive definite is symmetrical, show J (n) existence anduniquess minimum value, thus can calculate best weight coefficient arrow Amount, calculates to obtain W0=-R-1P;
(2) it when input signal X (n) is unstable, is solved using recursion, i.e. the output recursion of sef-adapting filter W (z) Relationship is:W (n+1)=W (n)-μ Δ W (n), wherein μ is the convergence step-length that automatic adjusument introduces, using single error signal Estimation of the squared gradient as mean square error gradient:Obtain sef-adapting filter Weight coefficient vector more new formula is:W (n+1)=- 2 μ e (n) r (n) of W (n).
The table of the main channel by loudspeaker, the signal e (n) after the sound wave counteracting after secondary channel is given in step A It is related with sef-adapting filter W (z), main channel P (z) and secondary channel system S (z) up to formula.Signal e (n) can be considered with Machine variable further solves the best weight coefficient of sef-adapting filter W (z), and according to input signal, can take two ways into Row calculates optimum weight vector.
Embodiment 3:
On the basis of embodiment 2, in conjunction with shown in attached drawing 3, there are two the EQ filter is set, one of EQ filter For receiving secondary channel transmission function identification resultOutput signal is simultaneously filtered rear input adaptive filter W (z), Another EQ filter is for receiving error pick-up received signal e (n) and being filtered rear input adaptive filter W (z)。
It can be seen that the signal of different frequency distribution character from second of calculation for calculating optimum weight vector, Its noise energy is also concentrated in different frequency range, thus in order to enhance active noise reduction module to the noise reduction effect of a certain scene, master Dynamic noise reduction module introduces processing of the enhancing of EQ filter two to noise collection Mid Frequency.After increasing EQ filter, sef-adapting filter W (z) bigger weight can be occupied by frequency range being selected in weight coefficient recurrence calculation, finally allow noise reduction of the noise reduction module under the scene Ability is improved significantly.
Embodiment 4:
On the basis of embodiment 2, in conjunction with shown in attached drawing 4, there are three the EQ filter is set, one of EQ filter For receiving secondary channel transmission function identification resultOutput signal is simultaneously filtered rear input adaptive filter W (z), Other two EQ filter is respectively used to filter the output signal of main channel P (z), secondary channel system S (z), filtered Signal carries out sound wave and offsets simultaneously error originated from input filter and sef-adapting filter W (z).
EQ filter filtering first is carried out to main channel signal and secondary channel signal before entering error pick-up, this When, the signal d'(n of main channel error originated from input sensor)=d (n) * eq (n), the signal y' of secondary channel error originated from input sensor (n)=s (n) * eq (n) equally makes noise energy collection Mid Frequency occupy bigger weight in weight coefficient update, finally allows described Noise reduction capability of the noise reduction module under the scene is improved significantly.
Embodiment 5:
As shown in connection with fig. 1, a kind of scene adaptive active noise reduction earphone, including microphone and loudspeaker, the microphone It is sequentially connected pre-amplification circuit, Noise Identification module, active noise reduction module and power amplification circuit, the power amplification circuit It is connect with the loudspeaker, wherein:
Microphone, for picking up ambient noise and being converted to primary noise electric signal;
Noise Identification module, for extracting noise immune collection Mid Frequency and frequency distribution from the primary noise electric signal Characteristic, and matched in noise pattern library with preset noise pattern, it determines noise type, is then selected in filter bank It selects corresponding EQ filter parameter and is loaded into active noise reduction module;
Active noise reduction module carries out the processing of the improvement FXLMS adaptive-filtering based on EQ filter to primary noise signal, The de-noising electric signal of excitation loudspeaker is generated, and is exported to power amplification circuit;
Loudspeaker generates de-noising sound wave, the de-noising sound wave and primary noise signal spectrum after receiving de-noising electric signal Distribution, acoustic pressure size are identical, opposite in phase, and loudspeaker exports the audio signal after offset noise.
Active noise reduction module makes sub-loudspeaker generate the de-noising sound wave opposite with main loudspeaker amplitude same phase, reaches Active noise reduction.
Although reference be made herein to invention has been described for explanatory embodiment of the invention, and above-described embodiment is only this hair Bright preferable embodiment, embodiment of the present invention are not limited by the above embodiments, it should be appreciated that those skilled in the art Member can be designed that a lot of other modification and implementations, these modifications and implementations will fall in principle disclosed in the present application Within scope and spirit.

Claims (7)

1. a kind of scene adaptive active denoising method, which is characterized in that including:
Step S1:Microphone picks up ambient noise and is converted into primary noise electric signal;
Step S2:Noise Identification module extracts noise energy collection Mid Frequency according to primary noise electric signal and frequency distribution is special Property, and matched with preset noise pattern in noise pattern library, determine noise type;
Step S3:EQ filter parameter is selected to be loaded into active noise reduction module, the EQ filter parameter packet from filter bank Include centre frequency f, bandwidth Q and amount of gain GAIN;
Step S4:Active noise reduction module carries out adaptive-filtering to primary noise electric signal, and generates the de-noising of excitation loudspeaker Electric signal;
Step S5:After loudspeaker receives de-noising electric signal, de-noising sound wave is generated, and the audio signal after secondary counteracting is sent out Out, the de-noising sound wave is identical as the primary noise electric signal spectrum distribution, acoustic pressure size is identical and opposite in phase.
2. a kind of scene adaptive active denoising method according to claim 1, which is characterized in that the step S4 is specific Including:
Step S41:EQ filter in active noise reduction module loads the EQ filter parameter;
Step S42:Secondary channel is introduced, the secondary channel includes sequentially connected inverse filter Inv_S (z), adaptive filter Wave device W (z) and secondary channel system S (z), the inverse filter Inv_S (z) is for offsetting secondary channel interference;
Step S43:Signal is generated to train secondary channel by noise generator, to obtain the identification of secondary channel transmission function As a result
Step S44:By the secondary channel transmission function result of the identificationIt substitutes into inverse filter Inv_S (z), then inputs certainly Adaptive filter W (z) carries out the processing of LMS adaptive-filtering;
Step S45:Error-filter in active noise reduction module generates the de-noising electric signal of excitation loudspeaker.
3. a kind of scene adaptive active denoising method according to claim 2, which is characterized in that the step S43 tool Body includes:In initialization, pumping signal is generated by external control circuit, which by secondary channel and distinguishes it Know, the signal then picked up by error pick-up, send back to after circuit system with LMS algorithm to the transmission function of secondary channel into Row calculates, to obtain the estimated value of the channel transfer function S (n)
4. a kind of scene adaptive active denoising method according to claim 3, which is characterized in that the step S44 tool Body includes:
A, the signal e (n) that error pick-up receives is solved, specially:
The input signal X (n) of active noise reduction module can reach error pick-up through main channel P (z) and two paths of secondary channel, Secondary channel output is expressed as:If assuming adaptive-filtering at this time Device W (z) weight coefficient remains unchanged in a bit of time L, enables r (n)=x (n) * s (n), can be obtainedThe signal e (n) that error pick-up receives at this time is main channel P (z) and the sum of the output signal of secondary channel, i.e. e (n)=d (n)+s (n)=d (n)+rT(n) W (n), wherein d (n) is signal X (n) output signal after main channel P (z), W (n) are the transmission function of sef-adapting filter;
B, the best weight coefficient of sef-adapting filter is solved, specially:
Objective function of the least-mean-square-error criterion as signal e (n) is selected, is enabled P=E [d (n) r (n)], R=E [r (n) rT(n)],
(1) when steady input signal X (n), then the Quadratic Function Optimization of weight vector
J (n)=E [e2(n)]=E [d2 (n)]+2E [d (n) rT(n)]W+WTE[r(n)rT(n)] W=E [d2(n)]+2PTW+WTBy It is that positive definite is symmetrical in matrix R, shows J (n) existence anduniquess minimum value, thus can calculate optimum weight vector, calculate to obtain W0 =-R-1P;
(2) it when input signal X (n) is unstable, is solved using recursion, i.e. the output recurrence relation of sef-adapting filter W (z) For:W (n+1)=W (n)-μ Δ W (n), wherein μ is the convergence step-length that automatic adjusument introduces, using the flat of single error signal Estimation of the square gradient as mean square error gradient:Obtain sef-adapting filter power system Counting vector more new formula is:W (n+1)=- 2 μ e (n) r (n) of W (n).
5. a kind of scene adaptive active denoising method according to claim 4, which is characterized in that the EQ filter is set There are two, one of EQ filter is for receiving secondary channel transmission function identification resultOutput signal is simultaneously filtered Input adaptive filter W (z) afterwards, another EQ filter is for receiving error pick-up received signal e (n) and being filtered Input adaptive filter W (z) after wave.
6. a kind of scene adaptive active denoising method according to claim 4, which is characterized in that the EQ filter is set There are three, one of EQ filter is for receiving secondary channel transmission function identification resultOutput signal is simultaneously filtered Input adaptive filter W (z) afterwards, other two EQ filter are respectively used to main channel P (z), secondary channel system S (z) Output signal filtering, filtered signal carries out sound wave and offsets and error originated from input filter and sef-adapting filter W (z).
7. a kind of scene adaptive active noise reduction earphone, which is characterized in that including microphone and loudspeaker, the microphone is successively Connect pre-amplification circuit, Noise Identification module, active noise reduction module and power amplification circuit, the power amplification circuit and institute Loudspeaker connection is stated, wherein:
Microphone, for picking up ambient noise and being converted to primary noise electric signal;
Noise Identification module, it is special for extracting noise immune collection Mid Frequency and frequency distribution from the primary noise electric signal Property, and matched in noise pattern library with preset noise pattern, it determines noise type, is then selected in filter bank Corresponding EQ filter parameter is loaded into active noise reduction module;
Active noise reduction module carries out the processing of the improvement FXLMS adaptive-filtering based on EQ filter to primary noise signal, generates The de-noising electric signal of loudspeaker is motivated, and is exported to power amplification circuit;
Loudspeaker receives and generates de-noising sound wave after de-noising electric signal, the de-noising sound wave and the distribution of primary noise signal spectrum, Acoustic pressure size is identical, opposite in phase, and loudspeaker exports the audio signal after offset noise.
CN201810818230.7A 2018-07-24 2018-07-24 A kind of scene adaptive active denoising method and earphone Active CN108900943B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810818230.7A CN108900943B (en) 2018-07-24 2018-07-24 A kind of scene adaptive active denoising method and earphone

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810818230.7A CN108900943B (en) 2018-07-24 2018-07-24 A kind of scene adaptive active denoising method and earphone

Publications (2)

Publication Number Publication Date
CN108900943A true CN108900943A (en) 2018-11-27
CN108900943B CN108900943B (en) 2019-11-05

Family

ID=64351774

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810818230.7A Active CN108900943B (en) 2018-07-24 2018-07-24 A kind of scene adaptive active denoising method and earphone

Country Status (1)

Country Link
CN (1) CN108900943B (en)

Cited By (30)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109545181A (en) * 2018-12-13 2019-03-29 四川长虹电器股份有限公司 A kind of adaptive digital active noise reduction framework
CN109714023A (en) * 2018-12-28 2019-05-03 歌尔股份有限公司 Adaptive filter method, sef-adapting filter and noise control system
CN109769060A (en) * 2019-02-02 2019-05-17 吉林大学 A kind of mobile phone active noise reducing device and method
CN110290442A (en) * 2019-07-17 2019-09-27 北京市劳动保护科学研究所 Active noise reduction earphone and its design method
CN110996205A (en) * 2019-11-28 2020-04-10 歌尔股份有限公司 Earphone control method, earphone and readable storage medium
CN111193977A (en) * 2019-12-13 2020-05-22 恒玄科技(上海)股份有限公司 Noise reduction method of earphone, self-adaptive FIR filter, noise removal filter bank and earphone
CN111462723A (en) * 2020-03-31 2020-07-28 上海联影医疗科技有限公司 Active noise reduction method and device
CN111627414A (en) * 2019-02-28 2020-09-04 上海汽车集团股份有限公司 Active denoising method and device and electronic equipment
CN111836165A (en) * 2020-07-10 2020-10-27 深圳市昂思科技有限公司 Compensation method for frequency response curve of electroacoustic device in active noise reduction system
CN111883095A (en) * 2020-04-27 2020-11-03 珠海市杰理科技股份有限公司 Active noise reduction method, device and system and related equipment
CN111970609A (en) * 2020-08-27 2020-11-20 Oppo广东移动通信有限公司 Sound quality adjusting method, sound quality adjusting system, electronic device, earphone, and computer-readable storage medium
CN112037749A (en) * 2020-09-03 2020-12-04 浙江红黑科技有限公司 Fascia rifle of making an uproar falls in initiative
WO2021010011A1 (en) * 2019-07-12 2021-01-21 ソニー株式会社 Information processing device, information processing method, and program
CN112289295A (en) * 2020-06-08 2021-01-29 珠海市杰理科技股份有限公司 Active noise reduction system training method and related equipment
CN112331226A (en) * 2020-09-29 2021-02-05 江苏清微智能科技有限公司 Voice enhancement system and method for active noise reduction system
CN112767908A (en) * 2020-12-29 2021-05-07 安克创新科技股份有限公司 Active noise reduction method based on key sound recognition, electronic equipment and storage medium
CN112951195A (en) * 2021-02-24 2021-06-11 辽宁省视讯技术研究有限公司 Vehicle-mounted active noise reduction dynamic regulation and control method and system
CN113132848A (en) * 2021-04-13 2021-07-16 北京安声科技有限公司 Filter design method and device and in-ear active noise reduction earphone
CN113207065A (en) * 2021-05-10 2021-08-03 杭州兆华电子有限公司 Acoustic calibrator and method based on ANC feedforward topology
CN113242491A (en) * 2021-06-29 2021-08-10 展讯通信(上海)有限公司 Noise reduction processing method, apparatus, device, storage medium, and program
CN113284480A (en) * 2020-12-11 2021-08-20 西安艾科特声学科技有限公司 Noise reduction effect estimation method for active noise control system
CN113365176A (en) * 2020-03-03 2021-09-07 华为技术有限公司 Method and device for realizing active noise elimination and electronic equipment
CN113643682A (en) * 2021-10-13 2021-11-12 展讯通信(上海)有限公司 Noise reduction method, chip module and equipment
CN114242096A (en) * 2021-08-20 2022-03-25 北京士昌鼎科技有限公司 Noise reduction system based on time-frequency domain
TWI764151B (en) * 2019-05-20 2022-05-11 仁寶電腦工業股份有限公司 Method for sound filtering and sound filter
WO2023087565A1 (en) * 2021-11-19 2023-05-25 深圳市韶音科技有限公司 Open acoustic apparatus
US11689845B2 (en) 2021-11-19 2023-06-27 Shenzhen Shokz Co., Ltd. Open acoustic device
CN116439913A (en) * 2023-04-14 2023-07-18 中国人民解放军海军潜艇学院 Marine active hearing protection type earmuff and protection method thereof
CN118098189A (en) * 2024-02-29 2024-05-28 东莞市达源电机技术有限公司 Intelligent motor noise reduction method
CN118411998A (en) * 2024-07-02 2024-07-30 杭州知聊信息技术有限公司 Audio noise processing method and system based on big data

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101833949A (en) * 2010-04-26 2010-09-15 浙江万里学院 Active noise control method for eliminating and reducing noise
US8705783B1 (en) * 2009-10-23 2014-04-22 Advanced Bionics Methods and systems for acoustically controlling a cochlear implant system
CN205248032U (en) * 2015-09-15 2016-05-18 宁波仁栋电气有限公司 Low -noise dry type transformer
CN105788604A (en) * 2016-04-07 2016-07-20 虞安波 FXLMS-based optimized active noise reduction method
US20170178617A1 (en) * 2015-12-17 2017-06-22 Harman Becker Automotive Systems Gmbh Active noise control by adaptive noise filtering

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8705783B1 (en) * 2009-10-23 2014-04-22 Advanced Bionics Methods and systems for acoustically controlling a cochlear implant system
CN101833949A (en) * 2010-04-26 2010-09-15 浙江万里学院 Active noise control method for eliminating and reducing noise
CN205248032U (en) * 2015-09-15 2016-05-18 宁波仁栋电气有限公司 Low -noise dry type transformer
US20170178617A1 (en) * 2015-12-17 2017-06-22 Harman Becker Automotive Systems Gmbh Active noise control by adaptive noise filtering
CN107025910A (en) * 2015-12-17 2017-08-08 哈曼贝克自动系统股份有限公司 Pass through the Active noise control of auto adapted noise filtering
CN105788604A (en) * 2016-04-07 2016-07-20 虞安波 FXLMS-based optimized active noise reduction method

Cited By (43)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109545181A (en) * 2018-12-13 2019-03-29 四川长虹电器股份有限公司 A kind of adaptive digital active noise reduction framework
CN109714023B (en) * 2018-12-28 2023-07-11 歌尔股份有限公司 Adaptive filtering method, adaptive filter and noise control system
CN109714023A (en) * 2018-12-28 2019-05-03 歌尔股份有限公司 Adaptive filter method, sef-adapting filter and noise control system
CN109769060A (en) * 2019-02-02 2019-05-17 吉林大学 A kind of mobile phone active noise reducing device and method
CN111627414B (en) * 2019-02-28 2024-01-30 上海汽车集团股份有限公司 Active denoising method and device and electronic equipment
CN111627414A (en) * 2019-02-28 2020-09-04 上海汽车集团股份有限公司 Active denoising method and device and electronic equipment
TWI764151B (en) * 2019-05-20 2022-05-11 仁寶電腦工業股份有限公司 Method for sound filtering and sound filter
WO2021010011A1 (en) * 2019-07-12 2021-01-21 ソニー株式会社 Information processing device, information processing method, and program
CN110290442A (en) * 2019-07-17 2019-09-27 北京市劳动保护科学研究所 Active noise reduction earphone and its design method
CN110996205A (en) * 2019-11-28 2020-04-10 歌尔股份有限公司 Earphone control method, earphone and readable storage medium
CN111193977A (en) * 2019-12-13 2020-05-22 恒玄科技(上海)股份有限公司 Noise reduction method of earphone, self-adaptive FIR filter, noise removal filter bank and earphone
WO2021175267A1 (en) * 2020-03-03 2021-09-10 华为技术有限公司 Method for implementing active noise cancellation, apparatus, and electronic device
CN113365176A (en) * 2020-03-03 2021-09-07 华为技术有限公司 Method and device for realizing active noise elimination and electronic equipment
CN111462723A (en) * 2020-03-31 2020-07-28 上海联影医疗科技有限公司 Active noise reduction method and device
CN111462723B (en) * 2020-03-31 2023-09-26 上海联影医疗科技股份有限公司 Active noise reduction method and device
CN111883095B (en) * 2020-04-27 2023-09-15 珠海市杰理科技股份有限公司 Active noise reduction method, device, system and related equipment
CN111883095A (en) * 2020-04-27 2020-11-03 珠海市杰理科技股份有限公司 Active noise reduction method, device and system and related equipment
CN112289295A (en) * 2020-06-08 2021-01-29 珠海市杰理科技股份有限公司 Active noise reduction system training method and related equipment
CN112289295B (en) * 2020-06-08 2023-12-26 珠海市杰理科技股份有限公司 Active noise reduction system training method and related equipment
CN111836165A (en) * 2020-07-10 2020-10-27 深圳市昂思科技有限公司 Compensation method for frequency response curve of electroacoustic device in active noise reduction system
CN111970609A (en) * 2020-08-27 2020-11-20 Oppo广东移动通信有限公司 Sound quality adjusting method, sound quality adjusting system, electronic device, earphone, and computer-readable storage medium
CN111970609B (en) * 2020-08-27 2022-07-12 Oppo广东移动通信有限公司 Sound quality adjusting method, sound quality adjusting system and computer readable storage medium
CN112037749A (en) * 2020-09-03 2020-12-04 浙江红黑科技有限公司 Fascia rifle of making an uproar falls in initiative
CN112331226A (en) * 2020-09-29 2021-02-05 江苏清微智能科技有限公司 Voice enhancement system and method for active noise reduction system
CN112331226B (en) * 2020-09-29 2024-04-12 江苏清微智能科技有限公司 Voice enhancement system and method for active noise reduction system
CN113284480A (en) * 2020-12-11 2021-08-20 西安艾科特声学科技有限公司 Noise reduction effect estimation method for active noise control system
CN113284480B (en) * 2020-12-11 2024-03-26 西安艾科特声学科技有限公司 Noise reduction effect estimation method for active noise control system
CN112767908B (en) * 2020-12-29 2024-05-21 安克创新科技股份有限公司 Active noise reduction method based on key voice recognition, electronic equipment and storage medium
CN112767908A (en) * 2020-12-29 2021-05-07 安克创新科技股份有限公司 Active noise reduction method based on key sound recognition, electronic equipment and storage medium
CN112951195A (en) * 2021-02-24 2021-06-11 辽宁省视讯技术研究有限公司 Vehicle-mounted active noise reduction dynamic regulation and control method and system
CN113132848A (en) * 2021-04-13 2021-07-16 北京安声科技有限公司 Filter design method and device and in-ear active noise reduction earphone
CN113207065A (en) * 2021-05-10 2021-08-03 杭州兆华电子有限公司 Acoustic calibrator and method based on ANC feedforward topology
CN113207065B (en) * 2021-05-10 2022-04-12 杭州兆华电子股份有限公司 Acoustic calibrator and method based on ANC feedforward topology
CN113242491A (en) * 2021-06-29 2021-08-10 展讯通信(上海)有限公司 Noise reduction processing method, apparatus, device, storage medium, and program
CN114242096A (en) * 2021-08-20 2022-03-25 北京士昌鼎科技有限公司 Noise reduction system based on time-frequency domain
CN113643682A (en) * 2021-10-13 2021-11-12 展讯通信(上海)有限公司 Noise reduction method, chip module and equipment
WO2023060793A1 (en) * 2021-10-13 2023-04-20 展讯通信(上海)有限公司 Noise reduction method, chip, chip module, and device
US11689845B2 (en) 2021-11-19 2023-06-27 Shenzhen Shokz Co., Ltd. Open acoustic device
WO2023087565A1 (en) * 2021-11-19 2023-05-25 深圳市韶音科技有限公司 Open acoustic apparatus
CN116439913A (en) * 2023-04-14 2023-07-18 中国人民解放军海军潜艇学院 Marine active hearing protection type earmuff and protection method thereof
CN116439913B (en) * 2023-04-14 2024-03-15 中国人民解放军海军潜艇学院 Marine active hearing protection type earmuff and protection method thereof
CN118098189A (en) * 2024-02-29 2024-05-28 东莞市达源电机技术有限公司 Intelligent motor noise reduction method
CN118411998A (en) * 2024-07-02 2024-07-30 杭州知聊信息技术有限公司 Audio noise processing method and system based on big data

Also Published As

Publication number Publication date
CN108900943B (en) 2019-11-05

Similar Documents

Publication Publication Date Title
CN108900943B (en) A kind of scene adaptive active denoising method and earphone
US6917688B2 (en) Adaptive noise cancelling microphone system
US7092529B2 (en) Adaptive control system for noise cancellation
CN103518383B (en) Paired microphone for suppressing noise
US8620650B2 (en) Rejecting noise with paired microphones
US20010028718A1 (en) Null adaptation in multi-microphone directional system
CN105407440B (en) Hearing device comprising an orientation system
US20090268920A1 (en) Cardioid beam with a desired null based acoustic devices, systems and methods
JP5617133B2 (en) Directional output signal generation system and method
US8000482B2 (en) Microphone array processing system for noisy multipath environments
JP2010513987A (en) Near-field vector signal amplification
US20180047410A1 (en) Signal Processor with side-tone noise reduction for a headset
CN105810202B (en) It is a kind of to drop hypoechoic method, apparatus and communication apparatus
CN111935584A (en) Wind noise processing method and device for wireless earphone assembly and earphone
CN111971975B (en) Active noise reduction method, system, electronic equipment and chip
Gil-Cacho et al. Wiener variable step size and gradient spectral variance smoothing for double-talk-robust acoustic echo cancellation and acoustic feedback cancellation
CN113630684A (en) Earphone with active noise reduction function and noise reduction method thereof
WO2022020122A1 (en) Ear-wearable device with active noise cancellation system that uses internal and external microphones
CN108429996A (en) Active noise control method, circuit and relevant device
KR102517939B1 (en) Capturing far-field sound
CN114582312A (en) Active control method and system for anti-interference adaptive road noise in vehicle
EP1305975B1 (en) Adaptive microphone array system with preserving binaural cues
CN113015052A (en) Wearable electronic device with reduced low frequency noise
CN107431869B (en) Hearing device
JPH06292293A (en) Microphone equipment

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
GR01 Patent grant
GR01 Patent grant