A service automatic target reporting bomb point sound source positioning method and system

Abstract

The application provides a service automatic target reporting bomb point sound source positioning method and system, and general standard application layer network transmission protocol is used as data transmission basis, distributed sound source positioning unit networking is realized, and bomb point positioning processing is completed. The processing capacity of the processing unit is fully utilized, the steps of large data volume, real-time processing, high synchronization requirement are processed in advance, only the data processing results are transmitted, the mainstream general standard application layer network protocol is used to realize distributed networking, hardware platforms and various devices are supported, and the existing wireless communication network infrastructure and general technology basis are utilized. Frame calculation is performed, frame characteristics are selected as the basis of selection results, and the method is simple, stable, and effective data is ensured on the basis of moderate data volume. The whole processing scheme has good mobility and scalability, processing units are added according to needs, and the accuracy of processing results is ensured.

Description

Technical Field

[0001] This invention relates to the field of sound source localization technology, and in particular to a method and system for locating the sound source of an explosion point in an automatic target reporting service. Background Technology

[0002] To ensure safety, the firing range is closed during various target practice sessions. These training and closures often last for a considerable period, making it difficult to promptly assess the hit rates. Conversely, it is crucial to obtain immediate information about the results after each round of target practice.

[0003] Target location has been widely applied in military, aerospace, and firing range applications. Therefore, numerous organizations have invested significant human, material, and financial resources in developing various location technologies and equipment, such as optical detection, infrared detection, ground radar, reconnaissance aircraft, early warning aircraft, and unmanned aerial vehicles (UAVs). These detection technologies and systems can be categorized by their detection methods, utilizing sound, light, and electromagnetic approaches. Among these, acoustic detection technology demonstrates unique advantages: 1) Less affected by weather conditions. Outdoor testing grounds encounter various adverse weather conditions, while acoustic detection technology can operate in darkness, rain, snow, and fog, achieving continuous and normal detection around the clock, effectively supplementing existing detection technologies. 2) Acoustic detection is not limited by line of sight. Topography has a relatively weak impact on acoustic detection technology, allowing its application in complex terrains such as hills and jungles where visibility is limited. Furthermore, it is unaffected by smoke and dust from explosions. 3) Wide detection range. The attenuation coefficient and directionality are relatively weak for infrasound and ultrasound, enabling the detection of targets over long distances and large areas.

[0004] Currently, acoustic detection technologies utilizing wireless sensor networks have emerged, primarily applied in areas such as gunshot localization, bullet impact point localization, and explosion sound localization. These sensor-based acoustic detection technologies, used in specific environments, employ the construction of private wireless networks and the transmission of data using proprietary communication protocols. Summary of the Invention

[0005] In view of the above problems, the present invention is proposed to provide a method and system for locating the sound source of an explosion point in an automatic target reporting service, which overcomes or at least partially solves the above problems.

[0006] According to one aspect of the present invention, a method for locating the sound source of an explosion point in an automatic target reporting service is provided, the sound source location method comprising:

[0007] Step 100: Place the two arrays in appropriate locations, establish a stable wireless network connection, set the corresponding coordinates and origins, and measure the distance between the origins of the two arrays;

[0008] Step 200: For one of the two arrays, acquire the audio signal, perform analog-to-digital conversion, and store the resulting digitized signal data for use in subsequent steps;

[0009] Step 300: Process the acquired data into frames, as the data frames have some overlap;

[0010] Step 400: Detect whether there are acoustic events in the current frame. If not, process the next frame. If there are, estimate the time delay of each channel in the current frame.

[0011] Step 500: Calculate the orientation of the sound source relative to the array based on the estimated time delay;

[0012] Step 600: Calculate the characteristic value of the obtained azimuth, based on the power of the highest signal amplitude, as follows:

[0013] e j =1-P mm / P max

[0014] In the formula, ej represents the feature value of the j-th frame, and Pmax and Pmm are the power of the highest signal amplitude in the current frame and the average power of the current frame excluding the highest signal amplitude, respectively.

[0015] Step 700: Transmit the calculated azimuth and feature values ​​to the fusion processing computer via a wireless network;

[0016] Step 800: Simultaneously perform the processing steps 200 to 700 on another array;

[0017] Step 900: Receive the processing results from the two arrays on the fusion processing computer, select the effective azimuth calculation result based on the eigenvalues, and calculate the sound source location according to the triangle relationship; the specific selection method is as follows:

[0018]

[0019] In the formula, Ti is the sum of the eigenvalues ​​of the i-th result, and ej,i represents the eigenvalues ​​corresponding to the i-th result and the j-th time. The sum of the eigenvalues ​​of each result is calculated, and the result with the largest sum of eigenvalues ​​is selected as the valid result.

[0020] Step 1000: Output the calculated sound source location result.

[0021] Optionally, step 100 may include the following specific steps:

[0022] Place two independent arrays in the base position;

[0023] A stable wireless network connection is established between the two independent arrays and the computer for data fusion processing. Depending on the site conditions, 5G, 4G, and WiFi networks provided by communication service providers are used for connection, which facilitates subsequent steps to use the wireless network to transmit processing results, fuse data results, and solve the specific location of the sound source.

[0024] Set the corresponding coordinates and origin for the array;

[0025] Finally, the distance between the origins of the two arrays was measured.

[0026] Optionally, in step 200, audio signals are acquired from either of the two arrays.

[0027] The audio signal acquired by the microphone is an analog signal, which needs to be converted from analog to digital, that is, the analog signal is quantized and stored in digital form;

[0028] In order to measure the sound source, the array has multiple microphones arranged in a certain geometry. The digitized signal needs to be saved according to the corresponding microphone channel and a unified time reference for use in subsequent steps.

[0029] Optionally, step 300 specifically includes: the sampled data will continue to arrive over time, and the data is processed by grouping, i.e., framing; framing generally takes 2n data points as one frame, and the value of n is adjusted according to the sampling frequency.

[0030] Optionally, step 400 specifically includes: detecting whether there is an acoustic event in the currently processed frame; if not, processing the next frame; if so, estimating the time delay of each channel for the current frame.

[0031] The presence of acoustic events and the estimation of channel delays in the current frame are determined by utilizing changes in received signal power. The specific processing method is as follows:

[0032] Estimate the mean power of the current frame signal.

[0033]

[0034] Where Em represents the mean of the signal in the m-th frame, N is the total number of samples in the current frame, and Ai is the amplitude of the i-th signal in the frame;

[0035] Estimate the variance of the current frame signal

[0036]

[0037] in, Let represent the variance of the signal in the m-th frame, Em represent the mean of the signal in the m-th frame, N represent the total number of samples in the current frame, and Ai represent the amplitude of the i-th signal in the frame.

[0038] Determine whether there is an acoustic event based on the mean and variance of the current frame signal;

[0039] The system collects and calculates the mean Eg and variance of the ambient background noise and the noise level when a signal is present. If the current frame's E m ≤E g , This indicates that there are no acoustic events in the current frame;

[0040] Estimate the delay of each channel in the current frame and use the generalized correlation method to estimate the time delay. Specifically, the method is to calculate the cross power spectrum between the two signals and perform whitening weighting on the calculated power spectrum to suppress the power of noise and sharpen the peak value in the time domain.

[0041] The processed cross-power spectrum is subjected to inverse Fourier transform to obtain the generalized cross-correlation function, and the time delay is estimated based on the peak value of the cross-correlation function.

[0042] Let the sound signals collected by the two microphones be x1(t) and x2(t), the sound source signal be s(t), D be the time delay between the two array elements, and n1(t) and n2(t) be additive noise;

[0043] Assuming the sound source signal s(t) and noise n1(t) and n2(t) are uncorrelated, normally stationary random processes with variance 1 and mean 0; then the two signals x1(t) and x2(t) are expressed as:

[0044]

[0045] Let α = 1, the cross-correlation function of the signals x1(t) and x2(t) received by the microphone is:

[0046]

[0047]

[0048] In the formula, since the sound source signal s(t) and noise n1(t) and n2(t) are uncorrelated, E(s·n) and E(n1·n2) are both 0;

[0049] Rss(τ-D) reaches its maximum value when the cross-correlation function of x1(t) and x2(t) reaches its maximum value;

[0050] Rss(τ-D)≦Rss(0), and the time delay D is the τ that reaches the maximum value.

[0051] For discrete signals with finite observation time, the cross-correlation estimate is:

[0052]

[0053] Where 2M+1 is the length of the correlation function, and N is the length of the entire data, requiring M>|D|;

[0054] Since n represents the number of sampling points, which is n times the sampling period Ts and can only be an integer, it affects the accuracy of the time delay estimation, resulting in an error of ±0.5Ts in the time delay estimation.

[0055] The Fourier transform of the cross-correlation function is the cross-power spectrum function. Adding a window before the correlator is a generalized cross-correlation algorithm. This property is used to suppress noise frequency components in the frequency domain, and then the result is converted to the time domain for analysis and processing.

[0056] Estimating time delay using the cross-correlation method is essentially about finding the maximum value of the function; the point corresponding to the maximum value is the estimated time delay.

[0057] Optionally, step 500 involves calculating the azimuth of the sound source relative to the array based on the estimated time delay; the specific processing method is as follows:

[0058] Step 501: Select an appropriate microphone array geometry:

[0059] Generally, detection arrays are classified into linear arrays, planar arrays, and three-dimensional arrays. In a linear array, the array elements are arranged in a straight line, and positioning is achieved within a half-plane bounded by the line. Linear arrays are simple in model, but their orientation and positioning accuracy is low when the array elements are close to the line. Planar arrays are arrays where the array elements are in the same plane. Planar arrays can detect half a spatial region bounded by the plane. Their azimuth, elevation, and distance are all affected by the position of the array elements and the effective speed of sound, and their elevation detection accuracy is not high. Three-dimensional arrays, composed of multiple sensors, can detect and locate the entire space, greatly improving the elevation detection accuracy, thereby improving the positioning accuracy of three-dimensional coordinates. The appropriate microphone array shape can be selected according to different needs. At the same time, it is necessary to ensure that the spacing between the array elements is much smaller than the distance from the target sound source to the array, so as to ensure that the premise of approximating a plane wave when the sound wave propagates to the microphone in the calculation holds true.

[0060] Step 502: Determine the spacing or coordinate positions between array elements;

[0061] The positions of the array elements are represented using a three-dimensional Cartesian coordinate system; the coordinate values ​​of each element in the coordinate system are determined based on the selected origin, coordinate axis directions, and spacing between elements.

[0062] Step 503: Calculate the direction of the sound source, including the azimuth and elevation angles;

[0063] Each microphone has a certain spacing, and there will be a time delay from the target sound source to different array elements; by combining the spacing, time delay and sound speed, the azimuth and elevation angles of the target sound source are determined.

[0064] The microphone array has N+1 elements, M0, M1, ..., MN. The time delays of the target sound source signal arriving at M0 and M1, M2, ..., MN are τ01, τ02, ..., τ0N, respectively, and the corresponding sound path differences are r01, r02, ..., r0N, respectively. The following formula can be used to calculate them:

[0065] τ 0i =t i -t0 (where i = 1, 2, ..., N)

[0066] r 0i =c·τ 0i (where c is the speed of sound, and i = 1, 2, ..., N)

[0067] Based on the time delay estimation method in step 400, τ01, τ02, ..., τ0N are obtained, and r01, r02, ..., r0N are further calculated. If the distance between the target sound source and M0 is r0, the following formula is obtained based on solid geometry:

[0068]

[0069] In the formula, x, y, z are the position coordinates of the target sound source, and xi, yi, zi (i = 0, 1, 2, ..., N) are the position coordinates of the array element microphones M0, M1, ..., MN in the array. Solving the equations simultaneously yields x, y, z, and r0. Because the distance calculation has a large deviation, r0 is only for reference. The azimuth of the target sound source, including the azimuth and elevation angles, is calculated using x, y, and z.

[0070]

[0071] Optionally, step 600 calculates the characteristic value of the azimuth; specifically, it uses the relative value of the signal power as the characteristic value, and the calculation method is as follows:

[0072] Step 601: Calculate the power of the highest signal amplitude in the current frame:

[0073]

[0074] Where Pmax represents the power of the highest amplitude value Amax in the m-th frame of the signal;

[0075] Step 602: Calculate the average power of the current frame excluding the highest signal amplitude:

[0076]

[0077] Where Pmm represents the average power of the current frame excluding the highest signal amplitude, N is the total number of samples in the current frame, and Ai is the amplitude value of the i-th signal in the frame, where the value of i does not include the maximum signal amplitude max.

[0078] Step 603: Calculate the feature values ​​of the current frame's calculation result:

[0079] e j =1-P mm / P max

[0080] Where ej represents the feature value of the j-th frame, and Pmax and Pmm are the results calculated in steps 601 and 602, respectively.

[0081] Optionally, in step 700, the calculated azimuth and feature values ​​are transmitted to the fusion processing computer via a wireless network; the TCP / IP protocol is used for transmission to ensure that the calculation results are transmitted stably and reliably to the fusion processing computer.

[0082] Optionally, in step 900, the processing results sent by the two arrays are received separately for the fusion processing calculation, the azimuth calculation result is selected according to the eigenvalues, and the sound source position is calculated according to the triangle relationship, as follows:

[0083] Step 901: Receive the result data sent by the two arrays and select the valid result from them;

[0084] An angle calculated by an array often has multiple values, which need to be filtered based on the eigenvalues;

[0085] Suppose a matrix provides n possible angle values, each occurring K1, K2, ..., Kn times; each result has a corresponding eigenvalue, calculated as described in step 6; corresponding to the result value, there are eigenvalue vectors Q1, Q2, ..., Qn, with:

[0086]

[0087]

[0088] …

[0089]

[0090] In the formula, Qi represents the feature vector corresponding to the i-th result, and ej,i represents the feature value corresponding to the i-th result and the j-th iteration; the feature values ​​of each result are summed, and the result with the largest sum of feature values ​​is selected.

[0091]

[0092] Where Ti is the sum of the eigenvalues ​​of the i-th result, and ej,i represents the eigenvalue of the i-th result at the j-th time; the eigenvalues ​​of each result are summed, and the result with the largest sum of eigenvalues ​​is selected as the valid result;

[0093] Select the result corresponding to the largest Ti as the valid result for subsequent position calculation;

[0094] The above results show the effective outcome of selecting another matrix;

[0095] Step 902: Based on the two effective angle results obtained from the measurement, construct a triangle and use the trigonometric relationship to calculate the location of the sound source;

[0096] To calculate the horizontal distance based on the azimuth angle, a trigonometric relationship is first established: connecting the geometric centers of the two arrays forms one side of a triangle; connecting the two arrays and the sound source respectively forms another triangle; given that the distance between the two arrays is l, and the measured azimuth angles of the sound source are α and β, and the length of the perpendicular line drawn from the sound source S point to the side formed by the arrays is y, with the foot of the perpendicular at a distance x from array 1, we have:

[0097] tanα=y / x

[0098] tanβ=y / (lx)

[0099] Get the values ​​of x and y:

[0100] x = tanβ·l / (tanα + tanβ)

[0101] y=tanβ·tanα·l / (tanα+tanβ)

[0102] The horizontal distances R1 and R2 from the sound source to array 1 and array 2 are further calculated as follows:

[0103]

[0104]

[0105] Then, based on the elevation angles measured by arrays 1 and 2 Calculate height

[0106]

[0107]

[0108] Optionally, step 1000: outputting the calculated sound source location result specifically includes:

[0109] Establish a coordinate system as needed and give the coordinate location of the sound source;

[0110] Give the azimuth, distance, and elevation angle of the two base arrays respectively;

[0111] The specific location of the sound source needs to be given according to the requirements of the final result.

[0112] The present invention also provides a sound source localization system for explosion point detection in service of automatic target reporting, the system comprising:

[0113] A microphone array module assembles microphones into a geometric shape with a certain spacing to pick up sound signals from the environment.

[0114] The analog-to-digital conversion module (ADC) is used because the sound signal acquired by the microphone array is an analog signal. The sound signal picked up by the microphone array needs to be converted from analog to digital (A / D).

[0115] The data framing module processes the obtained digital signal into frames, ensuring that there is some overlap between the frames.

[0116] The delay estimation and angle calculation module estimates the delay of the sound signal reaching different array elements and solves the direction of the sound source based on the geometry of the array and the spacing between the array elements.

[0117] The eigenvalue calculation module calculates the eigenvalue of the corresponding data frame while obtaining the sound source direction result;

[0118] The processing result sending module transmits the processing results and feature values ​​via a wireless network;

[0119] The processing result receiving module receives the processing results and feature values ​​via a wireless network.

[0120] The result selection module selects the valid processing result based on the feature values ​​sent by the array and its processor.

[0121] The sound source location calculation module constructs a triangular geometric relationship based on the selected valid results and the distance between the two arrays to calculate the specific sound source location.

[0122] The output module provides the sound source location as needed, enabling sound source localization.

[0123] This invention provides a method and system for locating the sound source of explosion points in an automatic target reporting service. It uses a general standard application layer network transmission protocol as the foundation for data transmission, effectively selects data processing results, and combines a mature and stable sound source location method to achieve distributed sound source location unit networking, completing the explosion point location processing. First, this method fully utilizes the processing capacity of the processing units, processing steps with large data volumes, real-time processing, and high synchronization requirements first, transmitting only the data processing results. This satisfies the requirements of real-time and synchronous data processing while reducing data transmission volume and lowering network transmission pressure. Second, it utilizes mainstream general standard application layer network protocols to achieve distributed networking, facilitating support from various hardware platforms and devices, and leveraging existing wireless communication network infrastructure and general technical foundations. Third, it performs calculations frame by frame, selecting frame features as the basis for selection results, which is simple and stable, ensuring sufficient effective data with a moderate data volume. Fourth, the entire processing scheme has good mobility and scalability; it can be deployed, moved, and additional processing units can be added as needed to ensure the accuracy of the processing results.

[0124] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0125] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0126] Figure 1 This is a flowchart of the explosion sound source localization method for automatic target reporting in this invention;

[0127] Figure 2 This is a schematic diagram of the explosion point sound source localization system module for automatic target reporting of the present invention;

[0128] Figure 3 This is a schematic diagram (top view / azimuth angle) of the planar method for solving the position of the array and the sound source using triangulation.

[0129] Figure 4 This is a side view (level / tilt angle) schematic diagram of the method for solving the position of the array and the sound source using triangulation in this invention. Detailed Implementation

[0130] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0131] The terms "comprising" and "having," and any variations thereof, in the specification, embodiments, claims, and drawings of this invention are intended to cover non-exclusive inclusion, such as including a series of steps or units.

[0132] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0133] like Figure 1 As shown in the figure, this is a flowchart illustrating the steps of the explosion sound source localization method for automatic target reporting according to the present invention. Detailed explanation follows:

[0134] Step 100: Set up the microphone array and start its processor. Place the two arrays in appropriate locations, establish a stable wireless network connection, set the corresponding coordinates and origin, and measure the distance between the origins of the two arrays.

[0135] Step 200: Acquire audio signals, perform analog-to-digital conversion, and store the resulting digital signal data; process the acquired data into frames, with some overlap between the frames.

[0136] Step 300: Detect whether there are acoustic events in the current frame. If not, process the next frame. If there are, estimate the time delay of each channel for the current frame.

[0137] The array and its processor were used to collect and calculate the ambient background noise and the mean value E when there was a signal, respectively. g With variance If the current frame's E m <E g , This indicates that there are no acoustic events in the current frame;

[0138] The specific method for estimating the average signal power of the current frame is as follows:

[0139]

[0140] Among them, E m Let A represent the mean of the signal in the m-th frame, N be the total number of samples in the current frame, and A be the mean of the signal in the m-th frame. i Let be the amplitude value of the i-th signal in the frame;

[0141] The specific method for estimating the variance of the signal power in the current frame is as follows:

[0142]

[0143] in, E represents the variance of the signal in the m-th frame. m Let A represent the mean of the signal in the m-th frame, N be the total number of samples in the current frame, and A be the mean of the signal in the m-th frame. i Let be the amplitude value of the i-th signal in the frame.

[0144] For frames containing acoustic events, the time delay is estimated using the generalized correlation method. Specifically, this involves calculating the cross-power spectrum between the two signals and applying whitening weighting to the calculated power spectrum to suppress noise power, thereby sharpening the peak value in the time domain and improving the signal-to-noise ratio. The processed cross-power spectrum is then subjected to an inverse Fourier transform to obtain the generalized cross-correlation function, and the time delay is estimated based on the peak value of this cross-correlation function.

[0145] Step 400: Based on the estimated time delay, calculate the orientation of the sound source relative to the array, as follows:

[0146] Each microphone has a certain distance between it, and there will be a time delay from the target sound source to different array elements; by combining the distance, time delay and sound speed, the azimuth and elevation angles of the target sound source can be determined.

[0147] The microphone array has N+1 elements, namely M0, M1, ..., M N The target sound source signal arrives at M0 and M1, M2, ..., M N The time delays are respectively τ 01 τ 02 、…τ 0N The corresponding path differences are r 01 r 02 ...r 0N The following calculation formula applies:

[0148] τ 0i =t i -t0 (where i = 1, 2, ..., N)

[0149] r 0i =c·τ 0i (where c is the speed of sound, and i = 1, 2, ..., N)

[0150] Based on the method for estimating time delay mentioned above, we can obtain τ. 01 τ 02 、…τ 0N Further calculations yielded r 01 r 02 ...r 0N If the distance from the target sound source to M0 is r0, the following formula can be obtained based on solid geometry:

[0151]

[0152] In the formula, x, y, and z are the position coordinates of the target sound source, x i y i , z i (i = 0, 1, 2, ..., N) represent the microphones M0, M1, ..., Mn in the array. N The position coordinates are obtained; the equations are solved simultaneously to obtain x, y, z, and r0; because the distance calculation has a large deviation, r0 is only for reference; the target sound source's azimuth, including azimuth and elevation angles, are calculated using x, y, and z.

[0153]

[0154] Step 500: Calculate the characteristic value of the obtained azimuth, based on the power of the highest signal amplitude, as follows:

[0155] e j =1-P mm / P max

[0156] In the formula, e j P represents the feature value of the current frame (frame j). max and P mm These are the power of the highest signal amplitude in the current frame and the average power of the current frame excluding the highest signal amplitude, respectively.

[0157] Step 600: Transmit the calculated azimuth value and feature value to the fusion processing computer via a wireless network; perform the above operation on another array and its processor simultaneously, and transmit the calculated result and feature value to the fusion processing computer for processing.

[0158] Step 700: Receive the processing results sent by the two arrays on the fusion processing computer respectively; select the effective azimuth calculation result according to the eigenvalues, and calculate the sound source position according to the triangle relationship;

[0159] The specific selection method is as follows:

[0160]

[0161] In the formula T i e is the sum of the eigenvalues ​​of the i-th result. j,i This represents the eigenvalue corresponding to the i-th result and the j-th iteration. The eigenvalues ​​of each result are summed, and the result with the largest sum of eigenvalues ​​is selected as the valid result.

[0162] Calculate the horizontal distance based on the azimuth angle, such as Figure 3As shown. To facilitate distance calculation, a trigonometric relationship is first established: connecting the geometric centers of the two arrays forms one side of a triangle; connecting the two arrays and the sound source respectively forms another triangle. Given the distance between the two arrays is l, and the measured azimuth angles of the sound source are α and β respectively, and a perpendicular line of length y is drawn from the sound source S to the side formed by the arrays, with the foot of the perpendicular at a distance x from array 1, we have:

[0163] tanα=y / x

[0164] tanβ=y / (lx)

[0165] Get the values ​​of x and y:

[0166] x = tanβ·l / (tanα + tanβ)

[0167] y=tanβ·tanα·l / (tanα+tanβ)

[0168] Furthermore, the horizontal distances R1 and R2 from array 1 and array 2 to the sound source can be obtained as follows:

[0169]

[0170]

[0171] Then, based on the elevation angles measured by arrays 1 and 2 Calculate the height, such as Figure 4 As shown

[0172]

[0173]

[0174] The H1 and H2 obtained here may be different and can be used as a reference.

[0175] Step 800: Output the calculated sound source location result.

[0176] like Figure 2As shown in the figure, this invention provides an automatic target reporting system for locating the sound source of explosions. It comprises 10 modules: a microphone array module, an analog-to-digital conversion module, a data framing module, a time delay estimation and angle calculation module, a feature value calculation module, a processing result sending module, a processing result receiving module, a result selection module, a sound source location calculation module, and a result output module. The microphone array module, analog-to-digital conversion module, data framing module, time delay estimation and angle calculation module, feature value calculation module, and processing result sending module are located at the "array and its processor" end. The microphone array module and analog-to-digital conversion module require independent hardware, while the processing result sending module requires the support of a wireless network card (wireless communication module). In this invention, two sets of "arrays and their processors" are required, with a certain distance between them and physical independence. The processing result receiving module, result selection module, sound source location calculation module, and result output module are located at the "fusion processing computer" end. The processing result receiving module requires the support of a wireless network card (wireless communication module). In this invention, the "fusion processing computer" is logically independent of the two sets of "arrays and their processors."

[0177] The specific processing functions of each module are as follows:

[0178] A microphone array module assembles microphones into a geometric shape with a certain spacing to pick up sound signals from the environment.

[0179] The analog-to-digital conversion (ADC) module is used because the sound signals acquired by the microphone array are analog signals. To facilitate digital processing, the sound signals picked up by the microphone array need to be converted from analog to digital (A / D). The ADC module performs the analog-to-digital conversion of the sound signals.

[0180] The data framing module processes the obtained digital signal into frames, ensuring that there is some overlap between the frames.

[0181] The time delay estimation and angle calculation module estimates the time delay (time difference) for the acoustic signal to reach different array elements (array elements), and solves the direction of the sound source based on the geometry of the array and the spacing between the array elements.

[0182] The eigenvalue calculation module calculates the eigenvalues ​​of the corresponding data frame while obtaining the sound source direction result.

[0183] The processing result sending module transmits the processing results and feature values ​​via a wireless network.

[0184] The processing result receiving module receives the processing results and feature values ​​via a wireless network.

[0185] The result selection module selects the valid processing result based on the feature values ​​sent by the array and its processor.

[0186] The sound source location calculation module calculates the specific sound source location by constructing a triangular geometric relationship based on the selected valid results and the distance between the two arrays.

[0187] The output module provides the sound source location as needed, enabling sound source localization.

[0188] The following details each step involved in the technical solution of this invention:

[0189] Set up the microphone array and start its processor.

[0190] Eight identical high-sensitivity broadband microphones were used, with four microphones in each array, forming a microphone array. The array geometry was as follows: the four microphones were located at the origin and on the x-axis and y-axis, respectively; east of the x-axis was considered positive, and north of the y-axis was considered positive. In this example, the focus was on the location of the ground detonation point, so the experimental site was approximated as a plane, and height was not calculated. Array 1 was located on the west side (left), array 2 was located on the east side (right), and the target (expected detonation point) was located north (above) of the line connecting the two arrays. Each of the two arrays, Array 1 and Array 2, establishes its own coordinate origin. The coordinates of each element in Array 1 are: (0, 0, 0), (1.00, 0, 0), (2.00, 0, 0), (0, 1.00, 0); the coordinates of each element in Array 2 are: (-2.00, 0), (-1.00, 0, 0), (0, 0, 0), (0, 1.00, 0).

[0191] The origins of the two microphone arrays are located on the extension of each other's x-axis, 60 meters apart;

[0192] Set up a wireless router to create a wireless local area network. Set the IP addresses (Internet Protocol Addresses) of the two microphone arrays as 192.168.20.11 (array 1) and 192.168.20.12 (array 2). Set up a fusion processing computer located at the midpoint between the two microphone arrays, with the IP address 192.168.20.100.

[0193] The audio signal is acquired, analog-to-digital conversion is performed, and the resulting digitized signal data is stored. The acquired data is processed by dividing it into frames, and the data frames have a certain degree of overlap.

[0194] The sound signals from four microphones were acquired in parallel, using a 16K sampling rate and 16-bit precision to obtain digital signals; each frame of data was 512 bytes long, with a repetition length of 256 bytes, meaning that half of the data overlapped.

[0195] Detect whether there are acoustic events in the current frame being processed. If not, process the next frame; if so, estimate the delay of each channel for the current frame.

[0196] Estimate the average signal power of the current frame:

[0197]

[0198] Among them, E m Let A represent the mean of the signal in the m-th frame, N be the total number of samples in the current frame, and A be the mean of the signal in the m-th frame. i Let be the amplitude value of the i-th signal in the frame;

[0199] Estimate the variance of the signal power in the current frame:

[0200]

[0201] in, E represents the variance of the signal in the m-th frame. m A represents the mean of the signal in the m-th frame, and N is the length of the current frame. i Let be the amplitude value of the i-th signal in the frame.

[0202] Before actually locating the sound source, the mean value E of the ambient background noise was collected and calculated using the array and its processor. g With variance If the current frame's E m <E g , This indicates that there are no acoustic events in the current frame.

[0203] For frames containing acoustic events, the time delay is estimated using the generalized correlation method. Specifically, this involves calculating the cross-power spectrum between the two signals and applying whitening weighting to the calculated power spectrum to suppress noise power, thereby sharpening the peak value in the time domain and improving the signal-to-noise ratio. The processed cross-power spectrum is then subjected to an inverse Fourier transform to obtain the generalized cross-correlation function, and the time delay is estimated based on the peak value of this cross-correlation function.

[0204] Based on the estimated time delay, the direction of the sound source relative to the array is calculated as follows:

[0205] Each microphone has a certain distance between it, and there will be a time delay from the target sound source to different array elements; by combining the distance, time delay and sound speed, the azimuth and elevation angles of the target sound source can be determined.

[0206] The microphone array has N+1 elements, namely M0, M1, ..., M N The target sound source signal arrives at M0 and M1, M2, ..., M N The time delays are respectively τ 01 τ 02 、…τ 0N The corresponding path differences are r 01 r 02 ...r 0N The following calculation formula applies:

[0207] τ 0i =t i -t0 (where i = 1, 2, ..., N)

[0208] r 0i =c·τ 0i (where c is the speed of sound, and i = 1, 2, ..., N)

[0209] Based on the method for estimating time delay mentioned above, we can obtain τ. 01 τ 02 、…τ 0N Further calculations yielded r 01 r 02 ...r 0N If the distance from the target sound source to M0 is r0, the following formula can be obtained based on solid geometry:

[0210]

[0211] In the formula, x, y, and z are the position coordinates of the target sound source, x i y i , z i (i = 0, 1, 2, ..., N) represent the microphones M0, M1, ..., Mn in the array. N The position coordinates are obtained; the equations are solved simultaneously to obtain x, y, z, and r0; because the distance calculation has a large deviation, r0 is only for reference; the azimuth of the target sound source is calculated using x, y, and z. Since it is approximated as a plane here, i.e., the elevation angle is preset to 0, only the azimuth angle needs to be calculated:

[0212]

[0213] Calculate the characteristic values ​​of the orientation.

[0214] Calculate the power of the highest signal amplitude in the current frame:

[0215]

[0216] Among them, P max A represents the highest amplitude value A in the m-th frame of the signal. max The power.

[0217] Calculate the average power of the current frame excluding the highest signal amplitude:

[0218]

[0219] Among them, P mm This represents the average power of the current frame excluding the highest signal amplitude, where N is the total number of samples in the current frame, and A is the average power. iLet be the amplitude value of the i-th signal in the frame, where the value of i does not include the maximum amplitude of the highest signal, max.

[0220] Calculate the feature values ​​of the current frame's calculation result:

[0221] e j =1-P mm / P max

[0222] Among them, e j P represents the feature value of the current frame (frame j). max and P mm These are the results obtained from the first two calculations, respectively.

[0223] The table below shows the calculated sound source direction results and their corresponding eigenvalues ​​(partial).

[0224]

[0225]

[0226] The calculated azimuth and feature values ​​are transmitted to the fusion processing computer via a wireless network; the fusion processing computer then receives the processing results sent by the two arrays.

[0227] This invention uses the TCP / IP protocol for transmission to ensure that the calculation results are transmitted stably and reliably to the computer for fusion processing. The pseudocode of the Phython class is given below for specific explanation.

[0228] Array processor end (transmitter end):

[0229]

[0230] Fusion processing computer end (receiving end):

[0231]

[0232] Step 7: Select the effective azimuth calculation result based on the feature value, and calculate the sound source location according to the triangle relationship;

[0233] The specific selection method is as follows:

[0234]

[0235] In the formula T i e is the sum of the eigenvalues ​​of the i-th result. j,i This represents the eigenvalue corresponding to the i-th result and the j-th iteration. The eigenvalues ​​of each result are summed, and the result with the largest sum of eigenvalues ​​is selected as the valid result.

[0236] Calculate T based on the results in the table above.i The results are shown in the table below.

[0237]

[0238] 60.75° is taken as a valid result.

[0239] Calculate the horizontal distance based on the azimuth angle, such as Figure 3 As shown. To facilitate distance calculation, a trigonometric relationship is first established: connecting the geometric centers of the two arrays forms one side of a triangle; connecting the two arrays and the sound source respectively forms another triangle. Given the distance between the two arrays is l, and the measured azimuth angles of the sound source are α and β respectively, and a perpendicular line of length y is drawn from the sound source S to the side formed by the arrays, with the foot of the perpendicular at a distance x from array 1, we have:

[0240] tanα=y / x

[0241] tanβ=y / (lx)

[0242] Get the values ​​of x and y:

[0243] x = tanβ·l / (tanα + tanβ)

[0244] y=tanβ·tanα·l / (tanα+tanβ)

[0245] Furthermore, the horizontal distances R1 and R2 from array 1 and array 2 to the sound source can be obtained as follows:

[0246]

[0247]

[0248] In this example, the selected angle yields the following valid results:

[0249] project Numerical value (unit: °) α 60.75 β 57.78

[0250] The calculated result is as follows:

[0251] project Numerical value (unit: meters) x 28.23 y 50.41 <![CDATA[R1]]> 57.78 <![CDATA[R2]]> 59.59

[0252] Output the calculated sound source location results.

[0253] In this example, the location of the sound source relative to the fusion computer is required. Based on the results obtained above, the location of the sound source relative to the fusion computer is calculated. Final output:

[0254] The sound source is located at a distance of 50.44 meters, 2.01° west of due north, from the fusion computer. This result can also be displayed as a radar chart.

[0255] The actual sound source location was set to due north, 50 meters away. The results above show that it can be well applied to tasks involving determining or detecting the location of explosion points.

[0256] Beneficial Effects: This invention provides a system and method for locating the sound source of explosion points in an automated target reporting service. This method uses a common standard application layer network transmission protocol as the data transmission foundation, effectively selects data processing results, and combines mature and stable sound source localization methods to achieve distributed sound source localization unit networking, completing the explosion point localization process. First, this method fully utilizes the processing capacity of the processing units, processing steps with large data volumes, real-time processing, and high synchronization requirements first, transmitting only the data processing results. This satisfies the requirements of real-time and synchronous data processing while reducing data transmission volume and lowering network transmission pressure. Second, the distributed networking achieved using mainstream common standard application layer network protocols facilitates support from various hardware platforms and devices, and leverages existing wireless communication network infrastructure and common technical foundations. Third, frame-based calculations and selection of frame features as the basis for selection results are simple and stable, ensuring sufficient effective data with a moderate data volume. Fourth, the entire processing scheme has good mobility and scalability, allowing for deployment, relocation, and addition of processing units as needed, ensuring the accuracy of the processing results.

[0257] The above specific embodiments further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for locating the sound source of an explosion point in an automated target reporting service, characterized in that, The sound source localization method includes: Step 100: Place the two arrays in appropriate locations, establish a stable wireless network connection, set the corresponding coordinates and origins, and measure the distance between the origins of the two arrays; Step 200: For one of the two arrays, acquire the audio signal, perform analog-to-digital conversion, and store the resulting digitized signal data for use in subsequent steps; Step 300: Process the acquired data into frames, as the data frames have some overlap; Step 400: Detect whether there are acoustic events in the current frame. If not, process the next frame. If there are, estimate the time delay of each channel in the current frame. Step 500: Calculate the orientation of the sound source relative to the array based on the estimated time delay; Step 600: Calculate the characteristic value of the obtained azimuth, based on the power of the highest signal amplitude, as follows: In the formula, e j Let P represent the feature value of the j-th frame. max and P mm These are the power of the highest signal amplitude in the current frame and the average power of the current frame excluding the highest signal amplitude, respectively. Step 700: Transmit the calculated azimuth and feature values ​​to the fusion processing computer via a wireless network; Step 800: Simultaneously perform the processing steps 200 to 700 on another array; Step 900: Receive the processing results from the two arrays on the fusion processing computer, select the effective azimuth calculation result based on the eigenvalues, and calculate the sound source location according to the triangle relationship; the specific selection method is as follows: In the formula T i e is the sum of the eigenvalues ​​of the i-th result. j,i This represents the eigenvalue corresponding to the i-th result and the j-th iteration. For each result, the eigenvalues ​​are summed, and the result with the largest sum of eigenvalues ​​is selected as the valid result; K i This indicates the number of times the i-th result appears; Step 1000: Output the calculated sound source location result.

2. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... The specific steps of step 100 include: Place two independent arrays in the base position; A stable wireless network connection is established between the two independent arrays and the computer for data fusion processing. Depending on the site conditions, 5G, 4G, and WiFi networks provided by communication service providers are used for connection, which facilitates subsequent steps to use the wireless network to transmit processing results, fuse data results, and solve the specific location of the sound source. Set the corresponding coordinates and origin for the array; Finally, the distance between the origins of the two arrays was measured.

3. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... In step 200, audio signals are acquired for either of the two arrays. The audio signal acquired through the microphone is an analog signal, which needs to be converted from analog to digital, that is, the analog signal is quantized and stored in digital form; In order to measure the sound source, the array has multiple microphones arranged in a certain geometry. The digitized signal needs to be saved according to the corresponding microphone channel and a unified time reference for use in subsequent steps.

4. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... Step 300 specifically includes: the sampled data will continue to arrive over time, and the data will be processed by grouping, i.e., framing; framing generally takes 2n data points as a frame, and the value of n is adjusted according to the sampling frequency.

5. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... Step 400 specifically includes: detecting whether there is an acoustic event in the currently processed frame; if not, processing the next frame; if so, estimating the time delay of each channel for the current frame. The presence of acoustic events and the estimation of channel delays in the current frame are determined by utilizing changes in received signal power. The specific processing method is as follows: Estimate the mean power of the current frame signal. Among them, E m Let A represent the mean of the signal in the m-th frame, and N be the total number of samples in the current frame. i Let be the amplitude value of the i-th signal in the frame; Estimate the variance of the current frame signal in, E represents the variance of the signal in the m-th frame. m Let A represent the mean of the signal in the m-th frame, and N be the total number of samples in the current frame. i Let be the amplitude value of the i-th signal in the frame; Determine whether there is an acoustic event based on the mean and variance of the current frame signal; The system collects and calculates the mean E of ambient background noise and the mean E when there is a signal, respectively. g With variance If the current frame This indicates that there are no acoustic events in the current frame; Estimate the delay of each channel in the current frame and use the generalized correlation method to estimate the time delay. Specifically, the method is to calculate the cross power spectrum between the two signals and perform whitening weighting on the calculated power spectrum to suppress the power of noise and sharpen the peak value in the time domain. The processed cross-power spectrum is subjected to inverse Fourier transform to obtain the generalized cross-correlation function, and the time delay is estimated based on the peak value of the cross-correlation function. Let the sound signals collected by the two microphones be x1(t) and x2(t), the sound source signal be s(t), D be the time delay between the two array elements, and n1(t) and n2(t) be additive noise; Assuming the sound source signal s(t) and noise n1(t) and n2(t) are mutually uncorrelated, normally stationary random processes with variance of 1 and mean of 0, then the two signals x1(t) and x2(t) can be expressed as: Let α=1, the cross-correlation function of the signals x1(t) and x2(t) received by the microphone is: In the formula, since the sound source signal s(t) and noise n1(t) and n2(t) are uncorrelated, E(s·n) and E(n1·n2) are both 0; Rss(τ-D) reaches its maximum value when the cross-correlation function of x1(t) and x2(t) reaches its maximum value; Rss(τ-D)≦Rss(0), and the time delay D is the τ that reaches the maximum value. For discrete signals with finite observation time, the cross-correlation estimate is: Where 2M + 1 is the length of the correlation function, and N is the length of the entire data, requiring M > |D|; Since n represents the number of sampling points, which is n times the sampling period Ts and can only be an integer, it affects the accuracy of the time delay estimation, resulting in an error of ±0.5Ts in the time delay estimation. The Fourier transform of the cross-correlation function is the cross-power spectrum function. Adding a window before the correlator is a generalized cross-correlation algorithm. This property is used to suppress noise frequency components in the frequency domain, and then the result is converted to the time domain for analysis and processing. Estimating time delay using the cross-correlation method is essentially about finding the maximum value of the function. The point corresponding to the maximum value is the estimated time delay.

6. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... Step 500: Based on the estimated time delay, calculate the azimuth of the sound source relative to the array; the specific processing method is as follows: Step 501: Select an appropriate microphone array geometry: Detection arrays are classified into linear arrays, planar arrays, and three-dimensional arrays. In a linear array, the array elements are arranged in a straight line, and positioning is achieved within a half-plane bounded by the line. Linear arrays are simple in design, but their orientation and positioning accuracy is low when the elements are close to a straight line. Planar arrays are arrays where the array elements are in the same plane. Planar arrays can detect half a spatial region bounded by the plane, but their azimuth, elevation, and distance are all affected by the position of the array elements and the effective speed of sound. Moreover, the elevation angle detection accuracy is not high. Three-dimensional arrays, composed of multiple sensors, can detect and locate the entire space, greatly improving the elevation angle detection accuracy and thus improving the positioning accuracy of three-dimensional coordinates. The appropriate microphone array shape can be selected according to different needs. At the same time, it is necessary to ensure that the spacing between the array elements is much smaller than the distance from the target sound source to the array, so as to ensure that the premise of approximating a plane wave when the sound wave propagates to the microphone in the calculation holds true. Step 502: Determine the spacing or coordinate positions between array elements; The positions of the array elements are represented using a three-dimensional Cartesian coordinate system; the coordinate values ​​of each element in the coordinate system are determined based on the selected origin, coordinate axis directions, and spacing between elements. Step 503: Calculate the direction of the sound source, including the azimuth and elevation angles; Each microphone has a certain spacing, and there will be a time delay from the target sound source to different array elements; by combining the spacing, time delay and sound speed, the azimuth and elevation angles of the target sound source are determined. The microphone array has N+1 elements, namely M0, M1, ..., M N The target sound source signal arrives at M0 and M1, M2, ..., M N The time delays are respectively τ 01 τ 02 、…τ 0N The corresponding path differences are r 01 r 02 ...r 0N The following calculation formula applies: Based on the time delay estimation method in step 400, τ is obtained. 01 τ 02 、…τ 0N Further calculations yielded r 01 r 02 ...r 0N If the distance from the target sound source to M0 is r0, the following formula is obtained based on solid geometry: In the formula, x, y, and z are the position coordinates of the target sound source, x i y i , z i (i=0, 1, 2, ..., N) are the array element microphones M0, M1, ..., M in the array. N The position coordinates are obtained; the equations are solved simultaneously to obtain x, y, z, and r0; because the distance calculation has a large deviation, r0 is only for reference; the target sound source's azimuth, including azimuth and elevation angles, are calculated using x, y, and z. 。 7. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... Step 600 calculates the characteristic value of the azimuth; specifically, it uses the relative value of the signal power as the characteristic value, and the calculation method is as follows: Step 601: Calculate the power of the highest signal amplitude in the current frame: Among them, P max A represents the highest amplitude value A in the m-th frame of the signal. max The power; Step 602: Calculate the average power of the current frame excluding the highest signal amplitude: Among them, P mm This represents the average power of the current frame excluding the highest signal amplitude, where N is the total number of samples in the current frame, and A is the average power. i Let i be the amplitude value of the i-th signal in the frame, where the value of i does not include the maximum amplitude of the highest signal. Step 603: Calculate the feature values ​​of the current frame's calculation result: Among them, e j Let P represent the feature value of the j-th frame. max and P mm These are the results calculated in steps 601 and 602, respectively.

8. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... In step 700, the calculated azimuth and feature values ​​are transmitted to the fusion processing computer via a wireless network; the TCP / IP protocol is used for transmission to ensure that the calculation results are transmitted stably and reliably to the fusion processing computer.

9. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... In step 900, the fusion processing computer receives the processing results sent by the two arrays respectively, selects the azimuth calculation result based on the eigenvalues, and calculates the sound source position according to the triangle relationship, as follows: Step 901: Receive the result data sent by the two arrays and select the valid result from them; An angle calculated by an array often has multiple values, which need to be filtered based on the eigenvalues; Suppose a matrix provides n possible angle values, and each value appears K1, K2, ..., K times. n Each result has a corresponding feature value, calculated as described in step 6; corresponding to the result value, there is a feature value vector Q1, Q2, ..., Q n ,have: In the formula Q i Let e ​​represent the eigenvalue vector corresponding to the i-th result. j,i This represents the feature value corresponding to the i-th result and the j-th iteration; Summing the eigenvalues ​​of each result, and selecting the result with the largest sum of eigenvalues; Where T i e is the sum of the eigenvalues ​​of the i-th result. j,i This represents the feature value corresponding to the i-th result and the j-th iteration; Summing the eigenvalues ​​of each result, and selecting the result with the largest sum of eigenvalues ​​as the valid result; Select the largest T i The corresponding results are used as valid results for subsequent position calculations; The above results show the effective outcome of selecting another matrix; Step 902: Based on the two effective angle results obtained from the measurement, construct a triangle and use the trigonometric relationship to calculate the location of the sound source; To calculate the horizontal distance based on the azimuth angle, a trigonometric relationship is first established: connecting the geometric centers of the two arrays forms one side of a triangle; connecting the two arrays and the sound source respectively forms another triangle; the distance between the two arrays is known to be... The measured azimuth angles of the sound source are α and β. The length of the perpendicular line drawn from the sound source S point to the side formed by the array is y, and the distance from the foot of the perpendicular to array 1 is x. Therefore: Get the values ​​of x and y: The horizontal distances R1 and R2 from the sound source to array 1 and array 2 are further calculated as follows: Then, calculate the altitude based on the elevation angles φ1 and φ2 measured by arrays 1 and 2. 。 10. The method for locating the sound source of an explosion point in an automatic target reporting service according to claim 1, characterized in that... Step 1000: Outputting the calculated sound source location result specifically includes: Establish a coordinate system as needed and give the coordinate location of the sound source; Give the azimuth, distance, and elevation angle of the two base arrays respectively; The specific location of the sound source needs to be given according to the requirements of the final result.

11. A sound source localization system for automatically reporting explosion points, the system comprising: A microphone array module assembles microphones into a geometric shape with a certain spacing to pick up sound signals from the environment. The analog-to-digital conversion module is used because the sound signal acquired by the microphone array is an analog signal, and the sound signal picked up by the microphone array needs to be converted from analog to digital. The data framing module processes the obtained digital signal into frames, ensuring that there is some overlap between the frames. The delay estimation and angle calculation module detects whether there are acoustic events in the currently processed frame. If not, it processes the next frame. If so, estimate the time delay of each channel for the current frame; Based on the estimated time delay, the orientation of the sound source relative to the array is calculated; The eigenvalue calculation module calculates the eigenvalue of the corresponding data frame while obtaining the sound source direction result, and calculates the azimuth eigenvalue based on the power of the highest signal amplitude. The processing result sending module transmits the processing results and feature values ​​via a wireless network; The processing result receiving module receives the processing results and feature values ​​via a wireless network. The result selection module selects the valid processing result based on the feature values ​​sent by the array and its processor. The fusion processing computer receives the processing results sent by the two arrays respectively, selects the azimuth calculation result based on the feature values, and calculates the sound source position according to the triangle relationship. The sound source location calculation module constructs a triangular geometric relationship based on the selected valid results and the distance between the two arrays to calculate the specific sound source location. The output module provides the sound source location as needed, enabling sound source localization.

Images