A virtual error signal calculation method under a multi-primary sound source condition
By deploying observation and temporary microphones under multiple primary sound source conditions and utilizing autocorrelation and crosscorrelation processing, the problem of accurately estimating virtual error signals was solved, achieving better noise reduction results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SOUTHEAST UNIV
- Filing Date
- 2024-01-16
- Publication Date
- 2026-05-19
AI Technical Summary
Under conditions with multiple primary sound sources, existing technologies struggle to accurately estimate virtual error signals, especially in locations where error microphones cannot be placed to achieve effective noise reduction.
By deploying observation microphones and temporary microphones, and using autocorrelation and crosscorrelation processing to calculate the observation path sequence, the virtual error signal of the location to be denoised is estimated.
A relatively accurate estimation of the virtual error signal was achieved under multiple primary sound source conditions, thus improving the noise reduction effect.
Smart Images
Figure CN117894291B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of active noise reduction technology, specifically a method for calculating virtual error signals under multiple primary sound source conditions. Background Technology
[0002] Environmental noise pollution has attracted widespread attention worldwide, especially in developing countries with rapid economic growth and accelerating urbanization, where noise pollution is a more prominent problem. Long-term exposure to high-noise environments can cause serious harm to people's physical and mental health, while even general noise interference can disrupt people's normal work and life. Therefore, many scholars have conducted research on noise control.
[0003] Traditional noise control addresses noise from three aspects: the noise source, the noise propagation path, and the noise receiver. Key technologies include sound absorption, sound insulation, and silencers. Its noise control mechanism relies on the interaction between noise waves and acoustic materials or structures to reduce noise, thus falling under the category of passive noise control. However, when the noise signal is low-frequency, passive noise control has limitations in terms of cost, deployment difficulty, and effectiveness. Therefore, the development of active noise control has filled this gap and has become a major research direction in noise control in recent years. Active noise control primarily utilizes the superposition of sound waves with equal amplitude and opposite phase, causing them to cancel each other out and achieve noise reduction.
[0004] In some applications, it's unsuitable to place the error microphone at the desired noise reduction location, such as in the headrest system of a smart chair. The error microphone cannot be placed directly at the ear; it must be placed near the ear. Virtual sensor algorithms were developed to address this situation. These algorithms can achieve maximum noise reduction at the desired location where the error microphone cannot be placed, and the remote microphone method is one of the more commonly used approaches.
[0005] In practical applications, how to better estimate the error signal of the location to be denoised using the observation microphone is a major limitation of this type of virtual sensor algorithm. Summary of the Invention
[0006] To address the shortcomings mentioned in the background section, the present invention aims to provide a method for calculating virtual error signals under multiple primary sound source conditions, thereby providing a more accurate estimation of virtual error signals under such conditions.
[0007] Firstly, the objective of this invention can be achieved through the following technical solution: a method for calculating virtual error signals under multiple primary sound source conditions, the method comprising the following steps:
[0008] Based on the number of primary sound sources, determine the number of observation microphones, determine the location to be noise-reduced, place observation microphones near the location to be noise-reduced, and place temporary microphones at the location to be noise-reduced.
[0009] The primary sound field signals emitted by multiple primary sound sources are received by the observation microphone and the temporary microphone respectively, and the first observation microphone signal and the temporary microphone signal are output respectively.
[0010] Autocorrelation processing is performed on the first observation microphone signal to obtain the autocorrelation sequence of the first observation microphone signal. Cross-correlation processing is performed on the first observation microphone signal and the temporary microphone signal to obtain the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal.
[0011] The observation path sequence of the observation microphone to the temporary microphone is calculated using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal.
[0012] After removing the temporary microphone, under the condition of multiple primary sound sources, the observation microphone receives the primary sound field signal to obtain the second microphone signal. The virtual error signal of the position to be denoised is calculated by using the second observation microphone signal and the observation path sequence of the observation microphone to the temporary microphone.
[0013] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of determining the number of observation microphones based on the number of primary sound sources: determining the number of observation microphones M based on the number of primary sound sources N, such that M≥N.
[0014] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the first observed microphone signal and the temporary microphone signal are labeled as:
[0015] The first observation microphone signal mon=[mon1(n),mon2(n),…,mon m (n),…,mon M [(n)] and the temporary microphone signal err(n);
[0016] Where n represents the discrete time point of the signal, mon m (n) represents the signal vector received by the m-th observation microphone, and M represents the number of observation microphones.
[0017] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: performing autocorrelation processing on the first observation microphone signal to obtain R. i (k)=E[mon i (n)mon i *(nk)], where E[] represents the expected value calculation, * represents the conjugate, k represents a certain moment in the time series, and R i (k) represents the autocorrelation sequence of the i-th observed microphone signal. Cross-correlation processing is performed on the first observed microphone signal and the temporary microphone signal to obtain r. i (k)=E{mon i (n)err i * (nk)},r i (k) represents the cross-correlation sequence of the i-th observed microphone signal and the temporary microphone signal, i = 1, 2, ..., M.
[0018] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of calculating the observation path sequence of the observation microphone to the temporary microphone using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal:
[0019] Based on the autocorrelation sequence R of the i-th observed microphone signal i (k) and the cross-correlation sequence r of the i-th observed microphone signal and the temporary microphone signal i (k) Calculate the observation path sequence of the observation microphone to the temporary microphone, using the following formula:
[0020] o i (k)=R i -1 (k)r i (k)
[0021] Among them, o i (k) represents the observation path sequence of the i-th observed microphone signal to the temporary microphone signal, R i -1 (k) represents the autocorrelation sequence R. i The inverse of (k), i = 1, 2, ..., M.
[0022] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the observation path sequence o of the i-th observed microphone signal to the temporary microphone signal. i The length of (k) is L, where k = 0, 1, ..., L-1.
[0023] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of calculating the virtual error signal of the location to be denoised using the observed microphone signal and the observed path sequence of the temporary microphone:
[0024] Using the second observation microphone signal MON=[MON1(n),MON2(n),…,MONm (n),…,MON M [n] and the observation path sequence of the i-th observed microphone signal to the temporary microphone signal o i (k) Calculate the virtual error signal at the position to be denoised at this time. in This represents convolution.
[0025] The beneficial effects of this invention are:
[0026] This invention determines the number of observation microphones based on the number of primary sound sources. To model the observation path, observation microphones are placed near the noise reduction location, and temporary microphones are placed at the noise reduction location. The observation microphones and temporary microphones receive the primary sound field signals when multiple primary sound sources are operating, obtaining the first observation microphone signal and the temporary microphone signal. Autocorrelation processing is performed on the first observation microphone signal, and cross-correlation processing is performed on the first observation microphone signal and the temporary microphone signal. Based on the results of the autocorrelation and cross-correlation processing, the observation path sequence of the observation microphones to the temporary microphones is calculated. The temporary microphones are then removed. Under the condition of multiple primary sound sources operating, the observation microphones receive the primary sound field signals to obtain the second microphone signal. The second observation microphone signal and the observation path sequence are used to calculate the virtual error signal at the noise reduction location at this time. This invention can provide a relatively accurate estimate of the virtual error signal under conditions of multiple primary sound sources. Attached Figure Description
[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0028] Figure 1 This is a schematic diagram of the method flow of the present invention;
[0029] Figure 2 This is a schematic diagram of the system model in Embodiment 1 of the present invention;
[0030] Figure 3 This is a comparison chart of the error estimate and the actual error at the location to be denoised in Embodiment 1 of the present invention. Detailed Implementation
[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0032] Example 1:
[0033] The following is a description of the relevant terms used in the embodiments of this application:
[0034] Primary sound source: A noise field to be controlled is called a primary sound field, its source is called a primary sound source, and the noise produced is called primary noise or primary sound wave. The artificially generated "anti" noise used to cancel the primary noise is called secondary noise or secondary sound wave, and the resulting sound field is called a secondary sound field. The actuator that generates secondary noise is called a secondary actuator; if the actuator is a sound source, it is called a secondary sound source; if it is a force source, it is called a secondary force source.
[0035] A microphone (MIC) is a device that converts sound into electrical signals. It is one of the most widely used electroacoustic devices in an audio system. Its function is to convert speech signals into electrical signals, which are then sent to a mixing console or amplifier and played back through loudspeakers. In other words, the microphone is used to pick up sound in an audio system; it is the first component of the entire system, and its performance has a significant impact on the overall sound system.
[0036] Cross-correlation and autocorrelation: Cross-correlation and autocorrelation respectively represent the degree of correlation between two time series and between the values of the same time series at any two different times. That is, the cross-correlation function describes the degree of correlation between the values of random signals x(t) and y(t) at any two different times t1 and t2, and the autocorrelation function describes the degree of correlation between the values of random signal x(t) at any two different times t1 and t2.
[0037] The autocorrelation function describes the degree of correlation between the values of a random signal x(t) at any two different times t1 and t2. The cross-correlation function provides a criterion for judging whether two signals are correlated in the frequency domain. It links the cross spectrum of the signals between two measurement points with their respective autospectral spectra. It can be used to determine the extent to which the output signal comes from the input signal and is very effective in correcting errors caused by noise sources in the measurement.
[0038] like Figure 1 As shown, a method for calculating virtual error signals under multiple primary sound sources is presented, the method comprising the following steps:
[0039] Based on the number of primary sound sources, determine the number of observation microphones, determine the location to be noise-reduced, place observation microphones near the location to be noise-reduced, and place temporary microphones at the location to be noise-reduced.
[0040] The process of determining the number of observation microphones based on the number of primary sound sources is as follows: Based on the number of primary sound sources N, determine the number of observation microphones M such that M≥N.
[0041] The primary sound field signals emitted by multiple primary sound sources are received by the observation microphone and the temporary microphone respectively, and the first observation microphone signal and the temporary microphone signal are output respectively.
[0042] Furthermore, the first observation microphone signal and the temporary microphone signal are labeled as follows:
[0043] The first observation microphone signal mon=[mon1(n),mon2(n),…,mon m (n),…,mon M [(n)] and the temporary microphone signal err(n);
[0044] Where n represents the discrete time point of the signal, mon m (n) represents the signal vector received by the m-th observation microphone, and M represents the number of observation microphones.
[0045] Autocorrelation processing is performed on the first observation microphone signal to obtain the autocorrelation sequence of the first observation microphone signal. Cross-correlation processing is performed on the first observation microphone signal and the temporary microphone signal to obtain the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal.
[0046] Among them, autocorrelation processing is performed on the signal from the first observation microphone to obtain R. i (k)=E[mon i (n)mon i * (nk)], where E[] represents the expected value calculation, * represents the conjugate, k represents a certain moment in the time series, and R i (k) represents the autocorrelation sequence of the i-th observed microphone signal. Cross-correlation processing is performed on the first observed microphone signal and the temporary microphone signal to obtain r. i (k)=E{mon i (n)err i * (nk)},r i (k) represents the cross-correlation sequence of the i-th observed microphone signal and the temporary microphone signal, i = 1, 2, ..., M.
[0047] The observation path sequence of the observation microphone to the temporary microphone is calculated using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal.
[0048] The process of calculating the observation path sequence of the observation microphone to the temporary microphone using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal:
[0049] Based on the autocorrelation sequence R of the i-th observed microphone signal i (k) and the cross-correlation sequence r of the i-th observed microphone signal and the temporary microphone signal i (k) Calculate the observation path sequence of the observation microphone to the temporary microphone, using the following formula:
[0050] o i (k)=R i -1 (k)r i (k)
[0051] Among them, o i (k) represents the observation path sequence of the i-th observed microphone signal to the temporary microphone signal, R i -1 (k) represents the autocorrelation sequence R. i The inverse of (k), i = 1, 2, ..., M.
[0052] Wherein, the observation path sequence o of the i-th observed microphone signal to the temporary microphone signal i The length of (k) is L, where k = 0, 1, ..., L-1.
[0053] After removing the temporary microphone, under the condition of multiple primary sound sources, the observation microphone receives the primary sound field signal to obtain the second microphone signal. The virtual error signal of the position to be denoised is calculated by using the second observation microphone signal and the observation path sequence of the observation microphone to the temporary microphone.
[0054] The process of calculating the virtual error signal of the location to be denoised using the signal from the second observation microphone and the observation path sequence of the observation microphone to the temporary microphone:
[0055] Using the second observation microphone signal MON=[MON1(n),MON2(n),…,MON m (n),…,MON M [n] and the observation path sequence of the i-th observed microphone signal to the temporary microphone signal o i (k) Calculate the virtual error signal at the position to be denoised at this time. in This represents convolution.
[0056] Specifically, the present invention will be further illustrated below through embodiments:
[0057] Step S1: The number of primary sound sources N = 5, which are 5-channel Gaussian white noise ref = [ref1(n), ref2(n), ..., ref5(n)], with a frequency of 0–4 kHz. Each channel of noise is independent. Temporary microphones are placed at the location to be denoised. Based on the number of primary sound sources, the number of observation microphones M is determined to be 5, and they are placed around the temporary microphones. The system model is as follows. Figure 2 As shown.
[0058] Step S2: Use the PULSE sound signal measuring instrument to record the virtual error signal err(n) received by the temporary microphone and the primary sound field signal mon=[mon1(n),mon2(n),…,mon5(n)] received by the observation microphone.
[0059] Step S3: Set the observation path sequence of the observation microphone to the temporary microphone. i (k) has a length of 1024. Autocorrelation processing is performed on the observed microphone signal to obtain R. i (k)=E[mon i (n)mon i * (nk)], where R i (k) represents the autocorrelation sequence of the i-th observed microphone signal, and r is the result of cross-correlation processing of the observed microphone signal and the temporary microphone signal. i (k)=E[mon i (n)err * [(nk)], where r i (k) represents the cross-correlation sequence of the i-th observed microphone signal and the temporary microphone signal, k = 0, 1, ..., 1023, i = 1, 2, ... 5.
[0060] Step S4: Calculate the observation path sequence of the observed microphone to the temporary microphone using the autocorrelation and cross-correlation processing results. i (k)=R i -1 (k)r i (k), i = 1, 2, ... 5.
[0061] Step S5: Remove the temporary microphone. Under the condition of multiple primary sound sources, use the observed microphone signal and the observed path sequence at this time to estimate the error signal at the location to be denoised. The true error err(n) and the estimated error err_es(n) at the location to be denoised are as follows: Figure 3 As shown. Next, the estimation accuracy is calculated for err(n) and err_es(n). NN represents the data length, and the NR is approximately 12.8 dB, indicating that the method of this invention can achieve relatively accurate error estimation.
[0062] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0063] The foregoing has shown and described the basic principles, main features, and advantages of this disclosure. Those skilled in the art should understand that this disclosure is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of this disclosure. Various changes and modifications can be made to this disclosure without departing from its spirit and scope, and all such changes and modifications fall within the scope of this disclosure as claimed.
Claims
1. A method for calculating virtual error signals under multiple primary sound sources, characterized in that, Includes the following steps: Based on the number of primary sound sources, determine the number of observation microphones, determine the location to be noise-reduced, place observation microphones near the location to be noise-reduced, and place temporary microphones at the location to be noise-reduced. The primary sound field signals emitted by multiple primary sound sources are received by the observation microphone and the temporary microphone respectively, and the first observation microphone signal and the temporary microphone signal are output respectively. Autocorrelation processing is performed on the first observation microphone signal to obtain the autocorrelation sequence of the first observation microphone signal. Cross-correlation processing is performed on the first observation microphone signal and the temporary microphone signal to obtain the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal. The observation path sequence of the observation microphone to the temporary microphone is calculated using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal. The process of calculating the observation path sequence of the observation microphone to the temporary microphone using the autocorrelation sequence of the first observation microphone signal and the cross-correlation sequence of the first observation microphone signal and the temporary microphone signal: According to the Autocorrelation sequence of the observed microphone signal and the Cross-correlation sequence of the observed microphone signal and the temporary microphone signal The observation path sequence of the observation microphone to the temporary microphone is calculated using the following formula: in, Indicates the first The observation path sequence of the observed microphone signal to the temporary microphone signal. Represents the autocorrelation sequence The reverse, ; Remove the temporary microphone. Under the condition of multiple primary sound sources, the observation microphone receives the primary sound field signal to obtain the second observation microphone signal. The virtual error signal of the position to be denoised is calculated by using the second observation microphone signal and the observation path sequence of the observation microphone to the temporary microphone. The process of calculating the virtual error signal of the location to be denoised using the signal from the second observation microphone and the observation path sequence of the observation microphone to the temporary microphone: Using the signal from the second observation microphone and the The observation path sequence of the observed microphone signal to the temporary microphone signal. Calculate the virtual error signal at the location to be denoised. ,in This represents convolution.
2. The method for calculating virtual error signals under multiple primary sound source conditions according to claim 1, characterized in that, The process of determining the number of observation microphones based on the number of primary sound sources: based on the number of primary sound sources... Determine the number of observation microphones ,make .
3. The method for calculating virtual error signals under multiple primary sound source conditions according to claim 1, characterized in that, The first observation microphone signal and the temporary microphone signal are labeled as follows: First observation microphone signal and temporary microphone signal ; in, Represents discrete time points of the signal. Indicates the first The signal vector received by each observation microphone This indicates the number of observed microphones.
4. The method for calculating virtual error signals under multiple primary sound sources according to claim 3, characterized in that, The autocorrelation processing of the first observation microphone signal yields... ,in This indicates the expected calculation. Indicates conjugate. Represents a specific moment in a time series. Indicates the first The autocorrelation sequence of the observed microphone signals is obtained by cross-correlation processing of the first observed microphone signal and the temporary microphone signal. , Indicates the first Cross-correlation sequences of observed microphone signals and temporary microphone signals, .
5. The method for calculating virtual error signals under multiple primary sound sources according to claim 1, characterized in that, The first The observation path sequence of the observed microphone signal to the temporary microphone signal. The length is , .