Far-field sound pickup optimization method based on directional module

Abstract

The present invention discloses a far-field sound pickup optimization method based on a directional module, and specifically relates to the technical field of sound pickup optimization. By acquiring MEMS capacitance fluctuation information in a microphone array, electromagnetic interference information in an operating environment of the microphone array, and self-oscillation information in a feedback loop, a directional beam reconfiguration delay impact model is constructed according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, and a directional beam reconfiguration delay impact index is generated. Potential operating risks of the microphone array are captured in a timely and effective manner, and the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference in the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration is quantified. A particle swarm algorithm is used to intelligently adjust the gain and phase of the microphone unit according to the degree of delay impact, so that the beamforming process is more accurate and the performance loss caused by delay is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of sound pickup optimization, and more specifically, to a far-field sound pickup optimization method based on a directional module. Background Art

[0002] The far-field sound pickup system based on the directional module integrates two intelligent microphone arrays and has a 15-meter far-field sound pickup distance in the outdoor environment, effectively increasing the sound pickup coverage of the monitoring equipment by 2 times; it provides zero-delay reverberation and noise suppression in the indoor environment, effectively improving the best quality of speech reception; the voice recognition performance is enhanced, greatly improving the efficiency of AI human-computer interaction communication; it also supports unidirectional and omnidirectional microphone data output, high signal-to-noise ratio performance microphone and enhanced SNR mode (through configuration); unified hardware DMA signal output, with high performance and minimal acoustic variation. The module adopts directional differential array microphone (DMA) technology, with a miniature, high efficiency, high AOP, and low power consumption design. Its digital output interface PDM is more adaptable to electrical noise environments than analog signals, and provides stronger error tolerance when signal quality decreases. SAM is the world's first real-time hardware DMA silicon microphone module, and unidirectional or omnidirectional output data can be set for each PDM cycle. The module also supports configuring different time delays between the two microphones to achieve a variety of directional sound pickup modes.

[0003] However, there are certain challenges in the long-term operation of the system. The MEMS capacitors in the microphone array may be affected by environmental factors (such as temperature, humidity, electromagnetic interference, etc.), causing fluctuations in the capacitance characteristics. This fluctuation may affect the accuracy of signal acquisition, especially in complex electromagnetic environments, where electromagnetic interference may further aggravate the self-oscillation phenomenon of the microphone array feedback loop. Self-oscillation can cause system instability, making the gain and phase calculation process untimely, resulting in delays in the reconfiguration of the directional beam. This delay will affect the system's response speed when adjusting the beam direction in real time, thereby reducing the accuracy of the sound pickup effect, especially in rapidly changing audio environments, which may lead to errors in locating and tracking the sound source. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a far-field sound pickup optimization method based on a directional module to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The far-field sound pickup optimization method based on the directional module includes the following steps:

[0007] Step S1, obtaining MEMS capacitance fluctuation information in the microphone array, calculating a MEMS capacitance fluctuation coefficient according to the MEMS capacitance fluctuation information, and evaluating the delay effect of the MEMS capacitance fluctuation in the microphone array on the reconfiguration of the directional beam;

[0008] Step S2, obtaining electromagnetic interference information of the microphone array operating environment, calculating an electromagnetic interference coefficient according to the electromagnetic interference information, and evaluating the delay effect of the electromagnetic interference of the operating environment on the reconfiguration of the directional beam;

[0009] Step S3, obtaining self-oscillation information of the microphone array feedback loop, calculating the self-oscillation coefficient according to the self-oscillation information, and evaluating the delay effect of the self-oscillation of the feedback loop on the reconfiguration of the directional beam;

[0010] Step S4, constructing a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, generating a directional beam reconfiguration delay impact index, and evaluating the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration;

[0011] Step S5, using a particle swarm algorithm to optimize and output the gain and phase of the microphone unit according to the degree of delay influence of the directional beam reconfiguration.

[0012] In a preferred embodiment, by acquiring MEMS capacitance fluctuation information in the microphone array, analyzing the MEMS capacitance fluctuation in the microphone array, and acquiring the MEMS capacitance fluctuation coefficient, the degree of MEMS capacitance fluctuation in the microphone array is measured;

[0013] The logic for obtaining the MEMS capacitance fluctuation coefficient is as follows:

[0014] The MEMS capacitance value of each microphone unit in the microphone array is obtained within a preset time period T, and the instantaneous MEMS capacitance deviation of each microphone unit is calculated, which is expressed as follows: ,in represents the instantaneous MEMS capacitance deviation at time t, represents the MEMS capacitance value at time t, represents the expected MEMS capacitance value; calculate the mean instantaneous MEMS capacitance deviation, the expression is as follows ,in represents the mean value of the instantaneous MEMS capacitance deviation, t={1,2,...,T}, T is a positive integer; calculate the standard deviation of the instantaneous MEMS capacitance deviation, the expression is as follows ,in Indicates the standard deviation of the instantaneous MEMS capacitance deviation; calculates the MEMS capacitance fluctuation coefficient , expressed as follows ,in represents the standard deviation of the instantaneous MEMS capacitance of the i-th microphone unit, represents the instantaneous MEMS capacitance deviation mean of the i-th microphone unit.

[0015] In a preferred embodiment, the electromagnetic interference level of the microphone array operating environment is measured by acquiring electromagnetic interference information of the microphone array operating environment, analyzing the electromagnetic interference situation of the microphone array operating environment, and acquiring the electromagnetic interference coefficient;

[0016] The logic for obtaining the electromagnetic interference coefficient is as follows:

[0017] The electromagnetic interference time domain signal of the microphone array operating environment is collected through the electromagnetic interference monitoring sensor ,in As a time variable, the electromagnetic interference time domain signal is converted into the electromagnetic interference frequency domain signal to calculate the frequency The spectral power density under ,in Indicates frequency The spectral power density under represents the sampling time interval, Represents a complex unit; calculates the degree of electromagnetic interference, the expression is as follows ,in Indicates the degree of electromagnetic interference. represents the lower limit of the operating frequency of the microphone array, Indicates the upper limit of the operating frequency of the microphone array; calculates the electromagnetic interference coefficient , the expression is as follows ,in Indicates the preset reference electromagnetic interference level.

[0018] In a preferred embodiment, the self-oscillation degree of the microphone array feedback loop is measured by acquiring the self-oscillation information of the microphone array feedback loop, analyzing the self-oscillation condition of the microphone array feedback loop, and acquiring the self-oscillation coefficient;

[0019] The logic for obtaining the self-oscillation coefficient is as follows:

[0020] Get the input signal and output signal of the microphone array to calculate the signal distortion coefficient , the expression is as follows ,in represents the input signal at the qth moment, Represents the output signal at the qth moment, and the value range of q is ;

[0021] Perform Fourier transform on the input and output signals of the microphone array and calculate the frequency response , the expression is as follows ,in represents the Fourier transform of the output signal, Represents the Fourier transform of the input signal; calculates the amplitude distortion coefficients , the expression is as follows ,in is the frequency response calculated at sampling point n, Represents the frequency response modulus, and the calculation expression is as follows ,in express The real part of express The imaginary part of is the expected frequency response, Represents the expected frequency response modulus, n={1,2,...,N}, N is a positive integer; calculate the phase distortion coefficient , the expression is as follows ,in is the phase response calculated at sampling point n, expressed as follows , is the expected phase response; calculate the self-oscillation coefficient , expressed as follows .

[0022] In a preferred embodiment, a directional beam reconfiguration delay impact model is constructed based on the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-excited oscillation coefficient to generate a directional beam reconfiguration delay impact index. The model is based on the following formula , where represent the preset proportional coefficients of the MEMS capacitance fluctuation coefficient, electromagnetic interference coefficient, and self-oscillation coefficient, respectively, and Both are greater than 0.

[0023] In a preferred embodiment, the directional beam reconfiguration delay impact index is compared with a preset directional beam reconfiguration delay impact index threshold to evaluate the impact of MEMS capacitance fluctuations in the microphone array, electromagnetic interference in the operating environment, and self-oscillation of the feedback loop on the delay of the directional beam reconfiguration, as follows:

[0024] If the directional beam reconfiguration delay impact index is greater than the directional beam reconfiguration delay impact index threshold, a delay impact deterioration signal is generated;

[0025] If the directional beam reconfiguration delay impact index is less than or equal to the directional beam reconfiguration delay impact index threshold, an operation stable signal is generated.

[0026] In a preferred embodiment, when a delay-impact-degraded signal is generated, a particle swarm algorithm is used to optimize the gain and phase of the microphone unit, as follows:

[0027] Step S5-1, initialize the particle swarm and randomly generate gain and phase; the position vector of each particle is ,in represents the gain of the mth microphone unit, represents the phase of the mth microphone unit, m={1,2,...,M}, M is a positive integer; initialize the velocity of each particle, the velocity vector of each particle is ,in represents the speed of the gain of the mth microphone unit, represents the speed of the phase of the mth microphone element;

[0028] Step S5-2, calculate the fitness according to the fitness function, the fitness function is as follows ,in represents fitness, represents the directivity beam reconfiguration delay impact index, represents the directivity beam reconfiguration delay impact index threshold;

[0029] Step S5-3, update the speed and position of the particle. The speed of the particle is updated according to the following formula: ,in Represents the particle The speed of iterations, Represents the particle The speed of iterations, represents the best historical position of the particle, represents the global optimal position, is the learning factor, controlling the particle pair and Dependence, is a random number; the particle position is updated according to the following formula ,in Represents the particle The position of the iteration, Represents the particle The position of the iteration;

[0030] Step S5-4: If the fitness is greater than or equal to the preset fitness threshold, repeat steps S5-2 and S5-3 for iterative update; if the fitness is less than the preset fitness threshold, The iteration is terminated, and the final gain and final phase combination of the microphone unit are output.

[0031] Technical effects and advantages of the present invention:

[0032] 1. The present invention obtains the MEMS capacitance fluctuation information in the microphone array, the electromagnetic interference information of the operating environment of the microphone array, and the self-oscillation information of the feedback loop, and constructs a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, generates a directional beam reconfiguration delay impact index, and timely and effectively captures the potential operating risks of the microphone array, and quantifies the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration, and uses a particle swarm algorithm to intelligently adjust the gain and phase of the microphone unit according to the delay impact, so that the beamforming process is more accurate and the performance loss caused by delay is reduced.

[0033] 2. The present invention introduces a particle swarm algorithm that dynamically optimizes gain and phase, combined with precise evaluation and adjustment of environmental factors, to greatly improve the accuracy and response speed of beam adjustment while ensuring system stability. It can maintain high-precision far-field sound pickup capabilities in complex electromagnetic environments, providing efficient, accurate and stable solutions for multiple application fields such as conferencing, intelligent voice assistants, and security monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0035] Figure 1 is a flow chart of a method according to an embodiment of the present invention;

[0036] Figure 2 Schematic diagram of the smart microphone array of the present invention. DETAILED DESCRIPTION

[0037] 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.

[0038] Example: Figure 1-Figure 2 The present invention provides a far-field sound pickup optimization method based on a directional module, comprising the following steps:

[0039] Step S1, obtaining MEMS capacitance fluctuation information in the microphone array, calculating a MEMS capacitance fluctuation coefficient according to the MEMS capacitance fluctuation information, and evaluating the delay effect of the MEMS capacitance fluctuation in the microphone array on the reconfiguration of the directional beam;

[0040] Step S2, obtaining electromagnetic interference information of the microphone array operating environment, calculating an electromagnetic interference coefficient according to the electromagnetic interference information, and evaluating the delay effect of the electromagnetic interference of the operating environment on the reconfiguration of the directional beam;

[0041] Step S3, obtaining self-oscillation information of the microphone array feedback loop, calculating the self-oscillation coefficient according to the self-oscillation information, and evaluating the delay effect of the self-oscillation of the feedback loop on the reconfiguration of the directional beam;

[0042] Step S4, constructing a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, generating a directional beam reconfiguration delay impact index, and evaluating the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration;

[0043] Step S5, using a particle swarm algorithm to optimize the gain and phase of the microphone unit according to the degree of delay influence of the directional beam reconfiguration;

[0044] Step S1, obtaining MEMS capacitance fluctuation information in the microphone array, calculating a MEMS capacitance fluctuation coefficient according to the MEMS capacitance fluctuation information, and evaluating the delay effect of the MEMS capacitance fluctuation in the microphone array on the reconfiguration of the directional beam;

[0045] The MEMS capacitance fluctuation coefficient is a key indicator used to measure the degree of fluctuation of the MEMS capacitance in the microphone array due to environmental changes (such as temperature, humidity, electromagnetic interference, etc.) during operation. This coefficient reflects the relative size of the MEMS capacitance deviation from its stable value, thereby evaluating the potential impact of these fluctuations on the performance of the pickup system. The fluctuation of the MEMS capacitance may cause the sensitivity and response of the microphone to the sound signal to change, affecting the quality of the pickup. The MEMS capacitance fluctuation coefficient can be used to evaluate the stability of the system in different environments. The larger the MEMS capacitance fluctuation coefficient, the more drastic the MEMS capacitance fluctuation, and the recalculation of gain and phase may become complicated, resulting in an increase in the delay when the beam directivity is reconfigured. A smaller MEMS capacitance fluctuation coefficient indicates that the MEMS capacitance remains relatively stable during the operation of the system and the capacitance changes less. In this way, the gain and phase adjustment process of the microphone array will be simpler and faster, and the delay when the beam directivity is reconfigured will be smaller, thereby improving the response speed and stability of the system;

[0046] Therefore, by acquiring the MEMS capacitance fluctuation information in the microphone array, analyzing the MEMS capacitance fluctuation in the microphone array, and acquiring the MEMS capacitance fluctuation coefficient, the degree of MEMS capacitance fluctuation in the microphone array is measured;

[0047] The logic for obtaining the MEMS capacitance fluctuation coefficient is as follows:

[0048] The MEMS capacitance value of each microphone unit in the microphone array is obtained within a preset time period T, and the instantaneous MEMS capacitance deviation of each microphone unit is calculated, which is expressed as follows: ,in represents the instantaneous MEMS capacitance deviation at time t, represents the MEMS capacitance value at time t, represents the expected MEMS capacitance value; calculate the mean instantaneous MEMS capacitance deviation, the expression is as follows ,in represents the mean value of the instantaneous MEMS capacitance deviation, t={1,2,...,T}, T is a positive integer; calculate the standard deviation of the instantaneous MEMS capacitance deviation, the expression is as follows ,in Indicates the standard deviation of the instantaneous MEMS capacitance deviation; calculates the MEMS capacitance fluctuation coefficient , expressed as follows ,in represents the standard deviation of the instantaneous MEMS capacitance of the i-th microphone unit, represents the mean instantaneous MEMS capacitance deviation of the i-th microphone unit;

[0049] Step S2, obtaining electromagnetic interference information of the microphone array operating environment, calculating an electromagnetic interference coefficient according to the electromagnetic interference information, and evaluating the delay effect of the electromagnetic interference of the operating environment on the reconfiguration of the directional beam;

[0050] The electromagnetic interference coefficient is an indicator used to measure the degree of electromagnetic interference suffered by the microphone array in its operating environment. Electromagnetic interference will affect the signal acquisition and processing process of the microphone array, especially when precise beamforming and directional beam reconfiguration are required. Strong electromagnetic interference may lead to inaccurate gain and phase calculations, thereby affecting the system's response speed and pickup quality. Therefore, calculating and evaluating the electromagnetic interference coefficient is of great significance for optimizing the performance of the system. Electromagnetic interference will affect the gain and phase calculations, resulting in delays in the reconfiguration of directional beams. Based on the electromagnetic interference coefficient, constructing a delay model to evaluate the impact of electromagnetic interference on beam directivity reconfiguration can help the system better adapt to and cope with interference factors in the electromagnetic environment, ensure that beamforming and directional beam reconfiguration can be performed with the lowest possible delay, and improve the system's response speed, stability and accuracy. Through this process, the system can optimize its gain and phase adjustment strategy, reduce mis-pointing and positioning errors, enhance the system's robustness, real-time and self-recovery capabilities, and thus continue to provide high-quality pickup services in a changing environment;

[0051] Therefore, by acquiring the electromagnetic interference information of the microphone array operating environment, analyzing the electromagnetic interference situation of the microphone array operating environment, and obtaining the electromagnetic interference coefficient, the electromagnetic interference degree of the microphone array operating environment is measured;

[0052] The logic for obtaining the electromagnetic interference coefficient is as follows:

[0053] The electromagnetic interference time domain signal of the microphone array operating environment is collected through the electromagnetic interference monitoring sensor ,in It is a time variable, which is used to represent the change of electromagnetic interference signal over time. It converts the electromagnetic interference time domain signal into the electromagnetic interference frequency domain signal to calculate the frequency The spectral power density under ,in Indicates frequency The spectral power density under represents the sampling time interval, represents a plural unit, i.e. ; Calculate the degree of electromagnetic interference, the expression is as follows ,in Indicates the degree of electromagnetic interference. represents the lower limit of the operating frequency of the microphone array, Indicates the upper limit of the operating frequency of the microphone array; calculates the electromagnetic interference coefficient , the expression is as follows ,in Indicates the preset reference electromagnetic interference level;

[0054] It should be noted that the reference electromagnetic interference level can be obtained by actually measuring the response of the microphone array under different electromagnetic interference conditions, calculating the typical value or mean value of the electromagnetic interference when the microphone array is in stable operation, and the technicians in this field can give the exact reference electromagnetic interference level according to the actual situation.

[0055] Step S3, obtaining self-oscillation information of the microphone array feedback loop, calculating the self-oscillation coefficient according to the self-oscillation information, and evaluating the delay effect of the self-oscillation of the feedback loop on the reconfiguration of the directional beam;

[0056] The self-oscillation coefficient is used to measure the impact of the self-oscillation of the microphone array feedback loop on the system performance, especially the delay effect on the reconfiguration of the directional beam. In the microphone array, the self-oscillation of the feedback loop may cause system response time delay, inaccurate beam pointing, and even noise interference. Therefore, evaluating and calculating the self-oscillation coefficient is crucial to optimizing system performance. The larger the self-oscillation coefficient, the stronger the self-oscillation in the feedback loop, which may cause system response delay, thereby affecting the reconfiguration of the directional beam. The smaller the self-oscillation coefficient, the faster the system response speed. The directional beam can be reconfigured relatively quickly without significant delay due to the interference of the self-oscillation. The self-oscillation coefficient can quantify the intensity of the self-oscillation in the feedback loop, thereby evaluating its impact on the performance of the microphone array, especially the delay effect on the reconfiguration of the directional beam. By calculating the self-oscillation coefficient, the design of the microphone array can be optimized to reduce the adverse effects of self-oscillation on system performance. By adjusting the stability of the feedback loop, dynamically adjusting the beam configuration rate, or taking delay compensation measures, the performance and stability of the microphone array can be effectively improved.

[0057] Therefore, by obtaining the self-oscillation information of the microphone array feedback loop, analyzing the self-oscillation situation of the microphone array feedback loop, and obtaining the self-oscillation coefficient, the self-oscillation degree of the microphone array feedback loop is measured;

[0058] The logic for obtaining the self-oscillation coefficient is as follows:

[0059] Get the input signal and output signal of the microphone array to calculate the signal distortion coefficient , the expression is as follows ,in represents the input signal at the qth moment, Represents the output signal at the qth moment, and the value range of q is ;

[0060] It should be noted that the input signal is the original sound signal received by the microphone array, and the output signal is the signal generated after being processed by the microphone array feedback loop;

[0061] Perform Fourier transform on the input and output signals of the microphone array and calculate the frequency response , the expression is as follows ,in represents the Fourier transform of the output signal, Represents the Fourier transform of the input signal; calculates the amplitude distortion coefficients , the expression is as follows ,in is the frequency response calculated at sampling point n, Represents the frequency response modulus, and the calculation expression is as follows ,in express The real part of express The imaginary part of is the expected frequency response, Represents the expected frequency response modulus, n={1,2,...,N}, N is a positive integer; calculate the phase distortion coefficient , the expression is as follows ,in is the phase response calculated at sampling point n, expressed as follows , is the expected phase response; calculate the self-oscillation coefficient , expressed as follows ;

[0062] Step S4, constructing a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, generating a directional beam reconfiguration delay impact index, and evaluating the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration;

[0063] The directional beam reconfiguration delay impact model is constructed based on the MEMS capacitance fluctuation coefficient, electromagnetic interference coefficient, and self-oscillation coefficient, and the directional beam reconfiguration delay impact index is generated. The model is based on the following formula , where represent the preset proportional coefficients of the MEMS capacitance fluctuation coefficient, electromagnetic interference coefficient, and self-oscillation coefficient, respectively, and All are greater than 0;

[0064] It should be noted that before building the directional beam reconfiguration delay impact model, it is necessary to ensure that the MEMS capacitance fluctuation coefficient, electromagnetic interference coefficient, and self-oscillation coefficient are all normalized. Common normalization methods include Min-Max normalization and Z-Score normalization. Set it according to the actual situation. For example, adopt the expert empowerment method, that is, invite experts in related fields to determine the preset proportion coefficients of various indicators through professional opinion surveys and comprehensive evaluations;

[0065] It can be seen from the above calculation expression that the larger the MEMS capacitance fluctuation coefficient, the larger the electromagnetic interference coefficient, and the larger the self-oscillation coefficient, the larger the directional beam reconfiguration delay impact index, indicating that the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop have a deeper impact on the delay of the directional beam reconfiguration; conversely, the smaller the MEMS capacitance fluctuation coefficient, the smaller the electromagnetic interference coefficient, and the smaller the self-oscillation coefficient, the smaller the directional beam reconfiguration delay impact index, indicating that the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop have a shallower impact on the delay of the directional beam reconfiguration;

[0066] The directional beam reconfiguration delay impact index is compared with a preset directional beam reconfiguration delay impact index threshold to evaluate the impact of MEMS capacitance fluctuations in the microphone array, electromagnetic interference in the operating environment, and self-oscillation of the feedback loop on the delay of directional beam reconfiguration, as follows:

[0067] If the directional beam reconfiguration delay impact index is greater than the directional beam reconfiguration delay impact index threshold, it indicates that the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop have a very deep impact on the delay of the directional beam reconfiguration, and a delay impact deterioration signal is generated;

[0068] If the directional beam reconfiguration delay impact index is less than or equal to the directional beam reconfiguration delay impact index threshold, it indicates that the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop have little effect on the delay of the directional beam reconfiguration, and a stable operation signal is generated;

[0069] Step S5, using a particle swarm algorithm to optimize the gain and phase of the microphone unit according to the degree of delay influence of the directional beam reconfiguration;

[0070] When the signal is degraded by the generation delay, the particle swarm algorithm is used to optimize the gain and phase of the microphone unit, as follows:

[0071] Step S5-1, initialize the particle swarm and randomly generate gain and phase; each particle represents a potential gain and phase combination of a microphone unit; the position vector of each particle is ,in represents the gain of the mth microphone unit, represents the phase of the mth microphone unit, m={1,2,...,M}, M is a positive integer; initialize the velocity of each particle, the velocity vector of each particle is ,in represents the speed of the gain of the mth microphone unit, represents the speed of the phase of the mth microphone element;

[0072] Step S5-2, calculate the fitness according to the fitness function, the fitness function is as follows ,in represents fitness, represents the directivity beam reconfiguration delay impact index, represents the delay impact index threshold of directional beam reconfiguration. The fitness function is used to evaluate the delay impact under a given gain and phase configuration. The goal is to minimize the delay of directional beam reconfiguration through particle swarm optimization;

[0073] Step S5-3, update the speed and position of the particle. The speed of the particle is updated according to the following formula: ,in Represents the particle The speed of iterations, Represents the particle The speed of iterations, represents the best historical position of the particle, represents the global optimal position, is the learning factor, controlling the particle pair and Dependence, is a random number; the particle position is updated according to the following formula ,in Represents the particle The position of the iteration, Represents the particle The position of the iteration;

[0074] Step S5-4: If the fitness is greater than or equal to the preset fitness threshold, repeat steps S5-2 and S5-3 for iterative update; if the fitness is less than the preset fitness threshold, Then the iteration is terminated and the final gain and final phase combination of the microphone unit are output, that is, ,in represents the final gain of the mth microphone unit, represents the final phase of the mth microphone unit;

[0075] The present invention obtains MEMS capacitance fluctuation information in a microphone array, electromagnetic interference information in the operating environment of the microphone array, and self-oscillation information of a feedback loop, and constructs a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, and generates a directional beam reconfiguration delay impact index, so as to timely and effectively capture potential operating risks of the microphone array, and quantify the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference in the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration, and adopts a particle swarm algorithm to intelligently adjust the gain and phase of the microphone unit according to the delay impact degree, so as to make the beamforming process more accurate and reduce the performance loss caused by delay.

[0076] The present invention introduces a particle swarm algorithm that dynamically optimizes gain and phase, combined with precise evaluation and adjustment of environmental factors, to greatly improve the accuracy and response speed of beam adjustment while ensuring system stability. It can maintain high-precision far-field sound pickup capabilities in complex electromagnetic environments, providing efficient, accurate and stable solutions for multiple application fields such as conferencing, intelligent voice assistants, and security monitoring.

[0077] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0078] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0079] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A far-field sound pickup optimization method based on a directional module, characterized in that: The steps include: Step S1, obtaining MEMS capacitance fluctuation information in the microphone array, calculating a MEMS capacitance fluctuation coefficient according to the MEMS capacitance fluctuation information, and evaluating the delay effect of the MEMS capacitance fluctuation in the microphone array on the reconfiguration of the directional beam; Step S2, obtaining electromagnetic interference information of the microphone array operating environment, calculating an electromagnetic interference coefficient according to the electromagnetic interference information, and evaluating the delay effect of the electromagnetic interference of the operating environment on the reconfiguration of the directional beam; Step S3, obtaining self-oscillation information of the microphone array feedback loop, calculating the self-oscillation coefficient according to the self-oscillation information, and evaluating the delay effect of the self-oscillation of the feedback loop on the reconfiguration of the directional beam; Step S4, constructing a directional beam reconfiguration delay impact model according to the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, generating a directional beam reconfiguration delay impact index, and evaluating the degree of delay impact of the MEMS capacitance fluctuation in the microphone array, the electromagnetic interference of the operating environment, and the self-oscillation of the feedback loop on the directional beam reconfiguration; Step S5, using a particle swarm algorithm to optimize the gain and phase of the microphone unit according to the degree of delay influence of the directional beam reconfiguration; According to the MEMS capacitance fluctuation coefficient, electromagnetic interference coefficient, and self-oscillation coefficient, a directional beam reconfiguration delay impact model is constructed to generate a directional beam reconfiguration delay impact index Zbxy. The formula based on the model is as follows Wherein, Zjdr, Dcgr, and Zjzd represent the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, respectively; w1, w2, and w3 represent the preset proportional coefficients of the MEMS capacitance fluctuation coefficient, the electromagnetic interference coefficient, and the self-oscillation coefficient, respectively; and w1, w2, and w3 are all greater than 0.

2. The far-field sound pickup optimization method based on the directional module according to claim 1 is characterized in that: By acquiring MEMS capacitance fluctuation information in the microphone array, analyzing the MEMS capacitance fluctuation in the microphone array, and acquiring the MEMS capacitance fluctuation coefficient, the degree of MEMS capacitance fluctuation in the microphone array is measured; The logic for obtaining the MEMS capacitance fluctuation coefficient is as follows: The MEMS capacitance value of each microphone unit in the microphone array is obtained within a preset time period T, and the instantaneous MEMS capacitance deviation of each microphone unit is calculated, which is expressed as follows pc t =|dr t -yqd|, where pc t represents the instantaneous MEMS capacitance deviation at time t, dr t represents the MEMS capacitance value at the tth moment, and yqd represents the expected MEMS capacitance value; the mean value of the instantaneous MEMS capacitance deviation is calculated as follows Where drj represents the mean value of the instantaneous MEMS capacitance deviation, t = {1, 2, ..., T}, T is a positive integer; calculate the standard deviation of the instantaneous MEMS capacitance deviation, the expression is as follows Where zjb represents the standard deviation of the instantaneous MEMS capacitance deviation; calculate the MEMS capacitance fluctuation coefficient Zjdr, expressed as follows Among them zjb i represents the standard deviation of the instantaneous MEMS capacitance of the i-th microphone unit, drj i represents the instantaneous MEMS capacitance deviation mean of the i-th microphone unit.

3. The far-field sound pickup optimization method based on the directional module according to claim 1 is characterized in that: By acquiring electromagnetic interference information of the microphone array operating environment, analyzing the electromagnetic interference situation of the microphone array operating environment, and acquiring the electromagnetic interference coefficient, the electromagnetic interference degree of the microphone array operating environment is measured; The logic for obtaining the electromagnetic interference coefficient is as follows: The electromagnetic interference time domain signal xh(sj) of the microphone array operating environment is collected by the electromagnetic interference monitoring sensor, where sj is the time variable. The electromagnetic interference time domain signal is converted into the electromagnetic interference frequency domain signal to calculate the spectral power density at frequency f, which is expressed as follows Where PSD(f) represents the spectral power density at frequency f, JG represents the sampling time interval, and j represents the complex unit; the electromagnetic interference degree is calculated as follows Where Pinter represents the degree of electromagnetic interference, f low represents the lower limit of the operating frequency of the microphone array, f high Indicates the upper limit of the operating frequency of the microphone array; Calculate the electromagnetic interference coefficient Dcgr, the expression is as follows Wherein Pref represents a preset reference electromagnetic interference level.

4. The far-field sound pickup optimization method based on the directional module according to claim 1 is characterized in that: By acquiring self-oscillation information of the microphone array feedback loop, analyzing the self-oscillation situation of the microphone array feedback loop, and acquiring the self-oscillation coefficient, the degree of self-oscillation of the microphone array feedback loop is measured; The logic for obtaining the self-oscillation coefficient is as follows: Get the input signal and output signal of the microphone array to calculate the signal distortion coefficient Ddist. The expression is as follows Where X(q) represents the input signal at the qth moment, Y(q) represents the output signal at the qth moment, and the value range of q is [0, ZQ], where ZQ represents the total operation time of the microphone array; Perform Fourier transform on the input signal and output signal of the microphone array respectively, and calculate the frequency response PL, which is expressed as follows Where YF represents the Fourier transform of the output signal, and XF represents the Fourier transform of the input signal; the amplitude distortion coefficient Dampl is calculated as follows Among them, PL n is the frequency response calculated at sampling point n, |PL n | represents the frequency response modulus, and the calculation expression is as follows Among them, α[PL n ] indicates PL n The real part, β[PL n ] indicates PL n The imaginary part of PLY is the expected frequency response, |PLY| represents the expected frequency response modulus, n = {1,2,...,N}, N is a positive integer; calculate the phase distortion coefficient Dphase, the expression is as follows Where ∠PL n is the phase response calculated at sampling point n, expressed as follows ∠PLY is the expected phase response; calculate the self-oscillation coefficient Zjzd, expressed as follows 5. The far-field sound pickup optimization method based on the directional module according to claim 1, characterized in that: The directional beam reconfiguration delay impact index is compared with a preset directional beam reconfiguration delay impact index threshold to evaluate the impact of MEMS capacitance fluctuations in the microphone array, electromagnetic interference in the operating environment, and self-oscillation of the feedback loop on the delay of directional beam reconfiguration, as follows: If the directional beam reconfiguration delay impact index is greater than the directional beam reconfiguration delay impact index threshold, a delay impact deterioration signal is generated; If the directional beam reconfiguration delay impact index is less than or equal to the directional beam reconfiguration delay impact index threshold, an operation stable signal is generated.

6. The far-field sound pickup optimization method based on the directional module according to claim 5 is characterized in that: When the signal is degraded by the generation delay, the particle swarm algorithm is used to optimize the gain and phase of the microphone unit, as follows: Step S5-1, initialize the particle swarm and randomly generate gain and phase; the position vector of each particle is wz m =[ZY m ,XW m ], where wz m is the particle position vector, ZY m represents the gain of the mth microphone unit, XW m represents the phase of the mth microphone unit, m={1,2,...,M}, M is a positive integer; initialize the velocity of each particle, the velocity vector of each particle is in represents the speed of the gain of the mth microphone unit, represents the speed of the phase of the mth microphone element; Step S5-2, calculate the fitness according to the fitness function, the fitness function is as follows HS = Zbxy-Zbxy z , where HS represents the fitness, Zbxy represents the directional beam reconfiguration delay impact index, and Zbxy z represents the directivity beam reconfiguration delay impact index threshold; Step S5-3, update the speed and position of the particle. The speed of the particle is updated according to the following formula sd m (d+1)=sd m (d)+c1*r1*(pbest-wz m )+c2*r2*(gbest-wz m ), where wz m is the particle position vector, sd m (d+1) represents the velocity of the particle at the d+1th iteration, sd m (d) represents the speed of the particle in the dth iteration, pbest represents the historical optimal position of the particle, gbest represents the global optimal position, c1 and c2 are learning factors that control the particle's dependence on pbest and gbest, r1 and r2 are random numbers; the particle's position is updated according to the following formula wz m (d+1)=wz m (d)+sd m (d+1), where wz m (d+1) represents the position of the particle at the d+1th iteration, wz m (d) represents the position of the particle at the dth iteration; Step S5-4: if the fitness is greater than or equal to the preset fitness threshold, repeat steps S5-2 and S5-3 for iterative update; if the fitness is less than the preset fitness threshold, terminate the iteration and output the final gain and final phase combination of the microphone unit.

Images