A method for realizing directional propagation of sound source using adaptive following filter
In the sound source directional propagation technology through the adaptive follow filter, the adaptive filter generates opposite signals in a specific direction, which solves the problem that the sound source directional propagation is affected by environmental noise in the prior art, and realizes effective suppression of sound and noise-resistant properties.
Patent Information
- Application Number
- CN202111609825.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-24
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2041-12-24
AI Technical Summary
The existing directed propagation technology of sound source is greatly affected by external environmental noise, lacks effective solutions, and it is difficult to effectively suppress sound propagation in specific directions.
Adaptive follow filter is used to measure the frequency response characteristics of the speakers through the training stage and use an adaptive filter to generate a sound signal with the same size and opposite phase to the sound source speaker in the specified direction in the working stage, thereby realizing sound cancellation and suppression.
It realizes significant reduction in sound in a specific direction, improves noise resistance, simplifies the structure, does not require a sound pickup microphone, only additional speakers are required, and the impact on external environmental noise is reduced.
Smart Images

Figure CN114363768B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of sound source processing, and in particular relates to a method for realizing directional propagation of a sound source by using an adaptive follower filter. Background Art
[0002] In some situations, it is desirable for the loudspeaker to radiate in a specific direction only, and not in other directions, or the sound in other directions is significantly weakened, to ensure privacy or reduce noise interference to the surroundings.
[0003] The current sound source directional propagation technology mostly uses a pickup microphone and is greatly affected by external environmental noise. Summary of the invention
[0004] An embodiment of the present invention provides a method for realizing directional propagation of a sound source by using an adaptive following filter, aiming to solve the problem of lack of solution for directional propagation of a sound source in the prior art. The present invention adopts an adaptive following algorithm to enable a suppression speaker to generate a sound signal with the same magnitude and opposite phase as the sound source speaker in a specified direction, thereby achieving cancellation and suppression of the sound in the specified direction.
[0005] The embodiment of the present invention is implemented as follows:
[0006] A method for achieving directional propagation of a sound source by using an adaptive follower filter, wherein a first speaker src_spk (a speaker for which directional propagation of the sound source is required) and a second speaker ns_spk (a speaker for achieving directional propagation) are used. The second speaker ns_spk is arranged in the radiation area of the first speaker src_spk. The radiation area of the second speaker ns_spk offsets the effective area of the sound of the first speaker src_spk, thereby achieving the effect of directional propagation of the sound source.
[0007] A method for realizing directional propagation of a sound source by using an adaptive following filter includes two stages: a training stage and a working stage. The specific steps are as follows:
[0008] S1. During the training phase, a microphone is placed at the location trn_mic in the area in the propagation direction that needs to be suppressed:
[0009] S101. The test equipment measures the first frequency response of the first speaker src_spk (the speaker of the sound source to be directional propagated) to the regional site (trn_mic) in the propagation direction to be suppressed, so as to simulate the first real frequency response characteristic (Wo) of its air propagation;
[0010] S102. The test equipment measures the second frequency response of the second speaker ns_spk (a speaker for directional propagation) to the regional site (trn_mic) in the propagation direction to be suppressed, so as to simulate the second real frequency response characteristic (Wn) of its air propagation;
[0011] Wherein, a first frequency response filter Woa is provided at the first frequency response position to simulate the corresponding first real frequency response Wo; a second frequency response filter Wna is provided at the second frequency response position to simulate the corresponding second real frequency response Wn; both the first frequency response filter and the second frequency response filter are IIR type filters;
[0012] S2. During the working phase, it is not necessary to place a microphone at the regional location trn_mic in the propagation direction that needs to be suppressed:
[0013] S201. Obtaining the desired signal d(n): The input signal S0(n) of the first speaker src_spk (the speaker that needs to propagate the sound source in a directional manner) is passed through the first frequency response filter Woa to obtain the desired signal d(n);
[0014] S202. Obtain an estimate d′(n) of the desired signal: the output of the input signal S0(n) after passing through the adaptive filter Won and the second frequency response filter Wna of the second speaker (ns_spk, a speaker that realizes directional propagation), completing the estimation of the desired signal d(n); wherein the values of Woa and Wna here have been obtained in step S1 and are known values;
[0015] S203. Obtain the feedback signal er(n) after filtering during the suppression process. The feedback signal er(n) during the suppression process is specifically implemented by the following formula 1:
[0016] er(n)=d(n)-d′(n)
[0017] =Woa H (n)·S0(n)-Won H (n)·S0(n)·Wna (Formula 1);
[0018] Among them: In formula 1, Woa H 、S0、Won H , Wna, the four items in bold represent vectors; H, represents conjugate transpose;
[0019] S204. Update and iterate the filter coefficients. Here, the normalized NLMS algorithm is used for update and iteration to avoid the problem that when S0(n) is large, the gradient amplification is caused, resulting in unstable oscillation in the convergence process, and when S0(n) is small, the algorithm converges slowly.
[0020] S205. When er(n) reaches the expected sufficiently small value after multiple iterations, the iteration ends; at this time, the sound played by src_spk received at the position and direction of trn_mic is significantly weakened; the regional site (trn_mic) in the propagation direction that needs to be suppressed is suppressed, achieving the effect of directional propagation of the sound source.
[0021] In the present invention, when in the training stage, the button key in the circuit is closed to connect the microphone at the regional site (trn_mic) in the propagation direction that needs to be suppressed; when in the working stage, the button key is opened, indicating that no microphone is installed at the regional site (trn_mic) in the propagation direction that needs to be suppressed.
[0022] In the present invention, the adaptive filter Won is connected to the first speaker and the second speaker to form a lateral filter structure; the signal er(n) is the feedback signal after filtering; when er(n) reaches the expected sufficiently small value after multiple iterations, the iteration ends; at this time, at the position and direction of the original mic, the sound played by the received src_spk is significantly weakened, thereby achieving the suppression of the regional site (trn_mic) in the propagation direction that needs to be suppressed, thereby achieving the effect of directional propagation of the sound source.
[0023] As a preferred implementation, the update iteration of the filter coefficients in step S204, the specific update algorithm is as shown in Formula 2:
[0024] Won(n+1)=Won(n)+μ(n)·S0(n)·er(n) (Formula 2)
[0025] in:
[0026] μ(n)=μ / S0 2 (n) (Formula 3)
[0027] The μ in Formula 3 is the iterative basic step size factor.
[0028] As a preferred implementation, the iterative basic step factor μ=0.05.
[0029] The adaptive filter Won is a data-dependent FIR type adaptive filter.
[0030] Compared with the prior art, the present invention has the following beneficial effects:
[0031] 1. The input signal of the adaptive filter Won in the present invention is the pure signal S0(n) at the input end of the first speaker src_spk (the speaker that needs to propagate the sound source in a directionally manner). The filter coefficient will be generated and updated according to a certain criterion (NLMS criterion, i.e., minimum variance) following S0(n), and has nothing to do with the surrounding environmental noise; therefore, the algorithm has good anti-noise characteristics.
[0032] 2. The application of the method of the present invention has a simple structure, and only requires the addition of corresponding speakers, without the need for a pickup microphone. The adaptive following algorithm is only related to the input signal of the sound source speaker, and is less affected by external environmental noise. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 Schematic diagram of the relative position relationship of loudspeakers for realizing directional propagation of sound sources by the adaptive follower filter provided in Embodiment 1 of the present invention;
[0034] Figure 2 It is a schematic diagram of the application structure of the method for realizing directional propagation of sound source by using the adaptive following filter provided in Example 1 of the present invention. DETAILED DESCRIPTION
[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0036] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it may be directly on the other element or there may be an intermediate element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be an intermediate element at the same time.
[0037] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0038] At present, due to the lack of solutions to the problem of directional propagation of sound sources in the prior art, the present invention adopts an adaptive following algorithm to enable the suppression speaker to generate a sound signal that is equal in size and opposite in phase to the sound source speaker in a specified direction, thereby canceling and suppressing the sound in the specified direction; and provides a method for realizing directional propagation of sound sources with an adaptive following filter.
[0039] Example 1
[0040] Embodiment 1 of the present invention provides a method for realizing directional propagation of a sound source by an adaptive following filter, wherein the relative position relationship between the first loudspeaker (src_spk, the loudspeaker for the sound source to be directional propagated) and the second loudspeaker (ns_spk, the loudspeaker for realizing directional propagation) is as follows: Figure 1 As shown, the second speaker is arranged in the radiation area of the first speaker. The radiation area of the second speaker (i.e. the area formed by the thin dashed line) offsets the effective area of the sound of the first speaker, thereby achieving the effect of directional propagation of the sound source.
[0041] Wherein: src_spk is the first loudspeaker, specifically the loudspeaker that needs to propagate the sound source in a directional manner;
[0042] ns_spk is a second speaker, specifically a speaker for realizing directional propagation;
[0043] trn_mic, is the regional location in the propagation direction that needs to be suppressed, and is used to place a microphone that obtains the relevant frequency response; during the training phase, the microphone is connected to collect information, and during the working phase, the microphone is disconnected;
[0044] adv_filter is an adaptive follower filter.
[0045] DSP processing system is a digital signal processor system.
[0046] A method for realizing directional propagation of a sound source by using an adaptive following filter includes two stages: a training stage and a working stage. The specific steps are as follows:
[0047] S1. During the training phase, connect the microphones at the regional sites trn_mic in the propagation direction that needs to be suppressed:
[0048] S101. The test equipment measures the first frequency response of the first speaker src_spk (the speaker of the sound source to be directional propagated) to the regional site (trn_mic) in the propagation direction to be suppressed, so as to simulate the first real frequency response characteristic (Wo) of its air propagation;
[0049] S102. The test equipment measures the second frequency response of the second speaker ns_spk (a speaker for directional propagation) to the regional site (trn_mic) in the propagation direction to be suppressed, so as to simulate the second real frequency response characteristic (Wn) of its air propagation;
[0050] Wherein, a first frequency response filter Woa is provided at the first frequency response position to simulate the corresponding first real frequency response Wo; a second frequency response filter Wna is provided at the second frequency response position to simulate the corresponding second real frequency response Wn; both the first frequency response filter and the second frequency response filter are IIR type filters;
[0051] S2. During the working phase, disconnect the microphone at the regional site trn_mic in the propagation direction that needs to be suppressed:
[0052] S201. Obtaining the desired signal d(n): The input signal s0(n) of the first speaker src_spk (the speaker that needs to propagate the sound source in a directional manner) is passed through the first frequency response filter Woa to obtain the desired signal d(n);
[0053] S202. Obtain an estimate d′(n) of the desired signal: the output of the input signal S0(n) after passing through the adaptive filter Won and the second frequency response filter Wna of the second speaker (ns_spk, a speaker that realizes directional propagation), completing the estimation of the desired signal d(n); wherein the values of Woa and Wna here have been obtained in step S1 and are known values;
[0054] S203. Obtain the feedback signal er(n) after filtering during the suppression process. The feedback signal er(n) during the suppression process is specifically implemented by the following formula 1:
[0055] er(n)=d(n)-d′(n)
[0056] =Woa H (n)·S0(n)-Won H (n)·S0(n)·Wna (Formula 1);
[0057] Among them: In formula 1, Woa H 、S0、Won H , Wna, the four items in bold represent vectors; H, represents conjugate transpose;
[0058] S204. Update and iterate the filter coefficients. Here, the normalized NLMS algorithm is used for update and iteration to avoid the problem that when S0(n) is large, the gradient amplification is caused, resulting in unstable oscillation in the convergence process, and when s0(n) is small, the algorithm converges slowly;
[0059] S205. When er(n) reaches the expected sufficiently small value after multiple iterations, the iteration ends; at this time, the sound played by src_spk received at the position and direction of trn_mic is significantly weakened; the regional site (trn_mic) in the propagation direction that needs to be suppressed is suppressed, achieving the effect of directional propagation of the sound source.
[0060] like Figure 2 The box where the "training phase" is located is a circuit connection diagram for obtaining the relevant filter in the training phase; when in the training phase, Figure 2 The middle button key is closed; in the working stage, the button is open, indicating that trn_mic has no microphone installed.
[0061] at the same time, Figure 2 The adaptive filter Won in the middle connects the first speaker and the second speaker to form a lateral filter structure; the signal er(n) is the feedback signal after filtering; when er(n) reaches the expected sufficiently small value after multiple iterations, the iteration ends; at this time, at the position and direction of the original mic, the sound played by the received src_spk is significantly weakened, realizing the suppression of the regional site (trn_mic) in the propagation direction that needs to be suppressed, and achieving the effect of directional propagation of the sound source.
[0062] As a preferred implementation, the update iteration of the filter coefficients in step S204, the specific update algorithm is as shown in Formula 2:
[0063] Won(n+1)=Won(n)+μ(n)·S0(n)·er(n) (Formula 2)
[0064] in:
[0065] μ(n)=μ / S0 2 (n) (Formula 3)
[0066] The μ in Formula 3 is the iterative basic step size factor.
[0067] As a preferred implementation, the iterative basic step factor μ=0.05.
[0068] The adaptive filter Won is a data-dependent FIR type adaptive filter.
[0069] Compared with the prior art, the present invention has the following beneficial effects:
[0070] The input signal of the adaptive filter Won in the present invention is the pure signal S0(n) at the input end of the first speaker src_spk (the speaker that needs to propagate the sound source in a directionally manner), and its filter coefficient will be generated and updated according to a certain criterion (NLMS criterion, i.e., minimum variance) following S0(n), and has nothing to do with the surrounding environmental noise; therefore, this algorithm has good anti-noise characteristics.
[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for realizing directional propagation of sound source by adaptive following filter, characterized in that: It includes two stages: training stage and working stage. The specific steps are as follows: S1. During the training phase, connect the microphones at the regional sites trn_mic in the propagation direction that needs to be suppressed: S101. The test equipment measures the first frequency response of the first speaker src_spk to the regional site trn_mic in the propagation direction to be suppressed, so as to simulate the first real frequency response characteristic of the air propagation; S102. The test device measures the second frequency response of the second speaker ns_spk to the regional site trn_mic in the propagation direction to be suppressed, so as to simulate the second real frequency response characteristics of the air propagation; Wherein, the first frequency response position is provided with a first frequency response filter Woa to simulate the corresponding first real frequency response Wo; the second frequency response position is provided with a second frequency response filter Wna to simulate the corresponding second real frequency response Wn; S2. During the working phase, disconnect the microphone at the regional site trn_mic in the propagation direction that needs to be suppressed: S201. Obtaining a desired signal d(n): The input signal S0(n) of the first speaker src_spk is passed through the first frequency response filter Woa to obtain a desired signal d(n); S202. Obtaining an estimate of the desired signal d′(n): the output of the input signal S0(n) after passing through the adaptive filter Won and the second frequency response filter Wna of the second speaker ns_spk, completing the estimation of the desired signal d(n); S203. Obtain the feedback signal er(n) after filtering during the suppression process. The feedback signal er(n) during the suppression process is specifically implemented by the following formula 1: er(n)=d(n)-d′(n) = Woa H (n)·S0(n)-Won H (n)·S0(n)·Wna (Eq. 1); Among them: In formula 1, Woa H 、S0、Won H , Wna, the four items in bold represent vectors; H, represents conjugate transpose; S204. Iterative update of filter coefficients, where the normalized NLMS algorithm is used for iterative update; S205. When er(n) reaches the expected sufficiently small value after multiple iterations, the iteration ends; at this time, at the position and direction of trn_mic, the sound played by the first loudspeaker src_spk is significantly weakened; the regional site trn_mic in the propagation direction that needs to be suppressed is suppressed, so as to achieve the effect of directional propagation of the sound source; The update iteration of the filter coefficients described in step S204, the specific update algorithm is as shown in Formula 2: Won(n+1)=Won(n)+μ(n)·S0(n)·er(n) (Formula 2) in: μ(n) = μ / S0 2 (n) (Equation 3) The μ in Formula 3 is the iterative basic step size factor.
2. The method for realizing directional propagation of sound source by using adaptive following filter according to claim 1, characterized in that: The first frequency response filter and the second frequency response filter are both IIR filters.
3. The method for realizing directional propagation of sound source by using adaptive following filter according to claim 1, characterized in that: The adaptive filter Won is connected to the first speaker and the second speaker to form a lateral filter structure.
4. The method for realizing directional propagation of sound source by using adaptive following filter according to claim 3, characterized in that: The iteration basic step factor μ=0.
05.
5. The method for realizing directional propagation of sound source by using adaptive following filter according to claim 1, characterized in that: The adaptive filter Won is a data-dependent FIR type adaptive filter.
