Umbrella type self-adaptive sound source positioning device

Abstract

The invention discloses an umbrella-type adaptive sound source positioning device, which belongs to the field of acoustics, and comprises a measurement system, which is of an umbrella-type power connecting rod structure carrying a microphone array, a temperature sensor and a displacement sensor; the positioning system is an integrated unit which takes an FPGA (Field Programmable Gate Array) as a core and is connected with a sampling module, a feedback module and a display module; wherein the sampling module synchronously digitizes a sound signal, a temperature signal and a displacement signal and then transmits the signals to the FPGA; the FPGA executes a generalized cross-correlation TDOA algorithm based on the received signal to obtain sound source azimuth information, and outputs a control signal; the feedback module receives the control signal and drives the power connecting rod structure so as to change the relative spatial position of each microphone in the microphone array; and the display module displays sound source azimuth information in real time. According to the invention, an umbrella type mechanical structure is used, a proper structure can be adaptively adjusted so as to more accurately position a sound source, the real-time performance is good, and the anti-noise performance is strong.

Description

Technical Field

[0001] This invention belongs to the field of acoustics, and in particular relates to an umbrella-type adaptive sound source localization device. Background Technology

[0002] Sound source localization refers to the process of locating signals with unknown directions and distances using sound sensors (microphones). Sound source localization technology determines the spatial location of a sound source by analyzing sound wave signals and is widely used in fields such as intelligent voice interaction, security monitoring, industrial inspection, and medical diagnosis.

[0003] Sound source localization technologies are mainly divided into several categories: time delay estimation (TDOA), beamforming, high resolution spectrum estimation (MUSIC / ESPRIT), and machine learning.

[0004] Time Delay Allocation (TDOA) calculates the time difference between sound waves arriving at different microphones and combines this with array geometry to determine the sound source location. Generalized Cross-Correlation (GCC) algorithms are commonly used to enhance noise immunity. It is simple in principle, has low computational complexity, and low hardware cost. However, it relies on high-precision time synchronization, and its positioning performance is related to the microphone matrix structure, creating a conflict between complexity and accuracy.

[0005] Beamforming methods create directional beams by adjusting the weights of each microphone, with the direction of maximum output power indicating the direction of the sound source. Typical methods include controllable beamforming, which offers strong real-time performance and excellent noise and interference suppression. However, they suffer from high computational complexity, limited low-frequency positioning performance, and high hardware costs.

[0006] High-resolution spectral estimation utilizes the orthogonality between the signal and noise subspaces to locate sound sources by searching for spatial spectral peaks. The most representative algorithm, such as MUSIC, offers high resolution, can distinguish multiple sound sources with similar angles, and has strong noise resistance. However, it suffers from high computational complexity, poor real-time performance, and requires narrowband sound signals; broadband signals require segmented processing.

[0007] Machine learning, on the other hand, transforms the sound source localization problem into a pattern recognition or regression problem through machine learning models, thereby completing the sound source localization. The basic process is as follows: First, acoustic signals from sound sources at different locations are collected using a microphone array. After preprocessing such as denoising and synchronization, time-domain, frequency-domain, or array spatial features are extracted. Then, these features are combined with the real-world location labels of the sound sources to train a classification or regression model (such as SVM or CNN), enabling the model to learn the mapping relationship between features and location. Finally, the trained model is used to analyze new signals and output the sound source's location information. The entire process does not rely on a precise physical model and can adapt to noise, reverberation, and other interference in complex environments through data learning. It does not require a precise physical model, can adapt to complex environments (reverberation, noise), and improves localization accuracy through big data. However, it relies on a large amount of labeled data (high collection costs); its generalization ability for unseen scenarios (such as new types of noise or changes in array layout) is weak; and when high real-time requirements are needed, the computational cost of complex models (such as Transformers) may be too high.

[0008] More microphones generally mean higher accuracy, but an excessive number of microphones increases computational complexity and thus affects real-time sound source localization. Therefore, this device uses five microphones to reduce complexity. Traditional sound source localization devices have a fixed distance between microphones, resulting in significant performance differences under various conditions. They also struggle to distinguish between near and far fields. In summary, sound source localization requires a trade-off between accuracy, cost, and real-time performance. Currently, TDOA (Tone-to-Area Detection and Analysis) solutions still suffer from the following problems: Insufficient real-time performance: Specifically, current devices require data acquisition first, followed by communication and data processing via a host computer; or they use data storage devices to save the recorded data and then process it later. Real-time processing is not possible on-site, resulting in poor real-time performance.

[0009] Poor adaptability: Specifically, the fixed structure of the microphone matrix means that in different application scenarios, due to changes in the characteristics and spatial location of the sound source, the fixed-spacing sound source localization device is prone to poor localization performance and weak adaptability.

[0010] Poor flexibility: Specifically, the sound source localization device is not portable and has insufficient storage capacity. At the same time, its structure is unreasonable and it performs poorly in complex outdoor environments.

[0011] Poor noise resistance: Specifically, when using related algorithms for sound source localization, it is greatly affected by environmental interference, and the algorithm performance needs further optimization and improvement.

[0012] Therefore, the core challenge of existing technologies lies in balancing anti-interference capabilities in complex environments, hardware costs, and real-time performance. Summary of the Invention

[0013] To address the problems mentioned above in the background technology, the sound source localization device of the present invention improves upon existing sound source localization devices by using the TDOA localization algorithm to determine the spatial location of the sound source through analysis of sound wave signals.

[0014] To achieve the above objectives, the present invention provides an umbrella-type adaptive sound source localization device, comprising: Measurement and positioning systems; The measurement system includes: Microphone array module, used to receive sound waves from a sound source and output multiple sound signals; The temperature measurement module is used to acquire the ambient temperature and output a temperature signal; The displacement measurement module is used to acquire the displacement of each microphone in the microphone array relative to the center rod and output the displacement signal; The power linkage module is used to support the microphone array module, temperature measurement module and displacement measurement module, and to change the spatial position of the microphone according to the control signal; The positioning system includes: The sampling module is connected to the microphone array module, temperature measurement module, and displacement measurement module, and is used to synchronously acquire the sound signal, temperature signal, and displacement signal and output digital signal. The FPGA main controller, connected to the sampling module, is used to estimate the time delay based on the digital signal to obtain the time delay difference; and to calculate the sound source location information by executing the TDOA algorithm according to the time delay difference, temperature signal and displacement signal, and at the same time generate a control signal for adjusting the spatial position of the microphone. The feedback module is connected between the FPGA main controller and the power linkage module, and is used to amplify and filter the control signal and drive the power linkage module. The display module is connected to the FPGA main controller and is used to output the sound source location information.

[0015] Preferably, the microphone array module includes at least five microphone units, each microphone unit being disposed on a corresponding link of the power linkage module; The microphone unit is used to receive sound signals emitted by the sound source; The sound signal is processed by a filtering and amplification circuit and then output to the sampling module.

[0016] Preferably, the power linkage module includes a central rod, a rotating rod, a support rod, and an actuator; One end of the rotating rod is hinged to the central rod, and the other end is fixed to the corresponding microphone; One end of the support rod is hinged to the corresponding rotating rod, and the other end is slidably connected to the central rod via a slider. The actuator is connected to the slider and drives the slider to move linearly on the central rod, thereby causing the rotating rod to rotate.

[0017] Preferably, the displacement measurement module is a linear sliding resistive distance sensor, which is mounted on the central rod and measures the linear displacement of the slider; The displacement signal is used to calculate the current spatial coordinates of each microphone.

[0018] Preferably, the temperature measurement module is used to measure the ambient temperature in real time and provide the temperature signal to the FPGA main controller; The FPGA master controller is used to correct the speed of sound in the air based on the temperature signal, so as to compensate for the error caused by the change in the speed of sound in the TDOA algorithm.

[0019] Preferably, the sampling module includes a multi-channel synchronous ADC chip. The input terminal of the ADC chip is connected to the filtering and amplification circuits of the microphone array module, the temperature measurement module, and the displacement measurement module, respectively. The output terminal of the ADC chip is connected to the FPGA main controller through a parallel bus for simultaneous sampling of multiple signals.

[0020] Preferably, the process of the FPGA main controller executing the TDOA algorithm includes: Perform an FFT transformation on every two microphone audio signals to obtain the frequency domain signal; Calculate the cross-power spectrum of the frequency domain signal; The cross-power spectrum was then PHAT-weighted. The weighted cross-power spectrum is subjected to IFFT transformation to obtain the generalized cross-correlation function; The peak position of the generalized cross-correlation function is detected to obtain the time delay between the corresponding microphone pairs; The sound source location is calculated based on the time delay and the current microphone coordinates.

[0021] Preferably, after obtaining the sound source location information, the FPGA master controller further executes the following: Determine whether the sound source is in the near-field model or the far-field model. If it is in the far-field model, only the azimuth and elevation angles are output; if it is in the near-field model, the azimuth, elevation angles, and distance information are output. Next, based on the judgment result, a corresponding control signal is generated and the baseline length of the microphone array is adjusted through the feedback module.

[0022] Preferably, the feedback module includes: A differential amplifier circuit is used to receive the control signals output by the FPGA main controller; A power amplifier, connected to the differential amplifier circuit, is used to amplify the power of the control signal; An LC filter circuit, connected to the output of the power amplifier, is used to filter out high-frequency noise and drive the actuator through the amplified control signal to make real-time adaptive adjustments to the microphone array.

[0023] Preferably, the display module includes an LCD screen, which is connected to the FPGA main controller via a parallel interface; The LCD display screen is used to display information in real time, including: When the far-field model is active, the azimuth and elevation angles of the sound source are displayed; When the near-field model is active, it displays the azimuth, elevation, and distance of the sound source. And the new baseline length and the number of adjustments after each microphone array structure adjustment.

[0024] Compared with the prior art, the present invention has the following advantages and technical effects: This invention uses an integrated device that combines data reception, data processing, and data feedback, reducing post-processing of data and improving system real-time performance.

[0025] This invention uses an "umbrella-like" mechanical structure to achieve variable spacing of the microphone array, thereby improving the ability to distinguish sound source signals under different sound source characteristics and spatial positions, thus adapting to different positioning requirements and improving the versatility of the device. This invention uses an "umbrella-like" mechanical structure to improve the portability and lightweight of the device and reduce the spatial complexity of the microphone matrix system; This invention uses the GCC algorithm for fast time delay estimation, incorporates spatial temperature information for correction, and uses FPGA-based main control to improve anti-interference capabilities, thereby enhancing the overall performance and positioning accuracy of the system. It is suitable for near-field or far-field sound source localization indoors or outdoors.

[0026] The invention has achieved excellent results in actual sound source localization tests. By monitoring the status of external sound signals in real time, the umbrella-shaped mechanical structure can adaptively adjust the appropriate structure, which can more accurately locate the sound source with an accuracy rate of over 90%. It also has good real-time performance and strong noise resistance. Attached Figure Description

[0027] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the overall structure of the device according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the microphone array module according to an embodiment of the present invention; Figure 3This is a schematic diagram illustrating the operation of the displacement measurement module in an embodiment of the present invention. Figure 4 This is a flowchart illustrating the operation of the feedback module in an embodiment of the present invention. Figure 5 This is a flowchart illustrating the distance derivation process in the spatial model positioning solution of this invention. Figure 6 This is a schematic diagram of the time delay estimation process according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the structure during the distance derivation process in the spatial model positioning solution of this invention. Detailed Implementation

[0028] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0029] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0030] like Figure 1 As shown, this embodiment provides an umbrella-type adaptive sound source localization device, including: Measurement and positioning systems; The measurement system includes: Microphone array module, used to receive sound waves from a sound source and output multiple sound signals; The temperature measurement module is used to acquire the ambient temperature and output a temperature signal; The displacement measurement module is used to acquire the displacement of each microphone in the microphone array relative to the center rod and output the displacement signal; The power linkage module supports the microphone array module, temperature measurement module, and displacement measurement module, and changes the spatial position of the microphone according to the control signal; The positioning system includes: The sampling module, connected to the microphone array module, temperature measurement module, and displacement measurement module, is used to synchronously acquire sound signals, temperature signals, and displacement signals and output digital signals. The FPGA main controller, connected to the sampling module, is used to estimate the time delay based on the digital signal and obtain the time delay difference. Based on the time delay difference, temperature signal, and displacement signal, it calculates the sound source location information by executing the TDOA algorithm and generates control signals for adjusting the spatial position of the microphone. The feedback module, connected between the FPGA main controller and the power linkage module, is used to amplify and filter the control signal and drive the power linkage module. The display module, connected to the FPGA main controller, is used to output the sound source location information.

[0031] Specifically, the apparatus in this embodiment includes a measurement system and a positioning system.

[0032] The measurement system is an umbrella-shaped power linkage structure equipped with a multi-microphone matrix and a temperature sensor. The measurement system includes a microphone array module, a temperature measurement module, a displacement measurement module, and a power linkage module; the microphone array module, temperature measurement module, and displacement measurement module are all located on the power linkage module.

[0033] Meanwhile, the microphone array module consists of multiple microphone sound sensors and corresponding filtering and amplification circuits; the temperature measurement module consists of a temperature sensor and a filtering and amplification circuit. The displacement measurement module consists of an infrared distance sensor and a filtering and amplification circuit. The power linkage module includes a basic mechanical metal linkage structure and a power structure. The power structure is a motor (actuator), responsible for providing the power source for the changes in the mechanical structure (the power source can also be manually adjusted).

[0034] The positioning system consists of an FPGA main controller, a feedback module, and a display module.

[0035] The FPGA main controller serves as the control core, connected to the feedback module and the display module. The feedback module, primarily composed of signal conditioning circuitry, is connected to both the FPGA main controller and the power linkage module. The display module is connected to the FPGA main controller. The sampling module mainly consists of an ADC chip and its peripheral basic circuitry.

[0036] Furthermore, the microphone array module includes at least five microphone units, each of which is respectively mounted on a corresponding link of the power linkage module; The microphone unit is used to receive sound signals emitted by a sound source; The audio signal is processed by a filtering and amplification circuit and then output to the sampling module.

[0037] Specifically, the microphone array module consists of five microphone sound sensors and corresponding filtering and amplification circuits. Each microphone sound sensor and its corresponding filtering and amplification circuit constitute a group, for a total of five groups. For example... Figure 2As shown, the five groups are fixed to corresponding umbrella-shaped connecting rods. The four microphones on the outer sides can rotate around the central rod as their corresponding connecting rods move. The microphone module's function is to receive sound signals from the sound source, perform preliminary filtering and amplification, and transmit the data to the FPGA main control for further processing. The microphone (sound sensor) is used to receive sound, and the peripheral circuitry provides filtering and amplification functions. The selected microphone is an electret condenser microphone, model XF-18D. This microphone is omnidirectional with a sensitivity of 55±3dB.

[0038] The umbrella-shaped five-element microphone array implemented here features microphones with variable spacing when receiving signals. This allows the sound source localization system to significantly improve its adaptability and localization performance by flexibly adjusting the spacing between the microphones in the array. On the one hand, the spacing can be dynamically matched according to the frequency characteristics of the sound source (e.g., using a small spacing for high-frequency signals to avoid phase ambiguity, and using a large spacing for low-frequency signals to obtain sufficient time difference), thereby covering a wider frequency range and improving the localization accuracy of sound sources in different frequency bands. On the other hand, it can adapt to different distance scenarios such as near-field and far-field (using a small spacing to capture wavefront details in the near field, and using a large spacing to increase the baseline and improve angular resolution in the far field), and the configuration can be adjusted according to interference such as environmental noise and reverberation (e.g., using a small spacing to reduce multipath effects in complex environments). Ultimately, it achieves universal adaptation from a single scenario to diverse scenarios, reduces the system's dependence on specific application scenarios, and improves deployment flexibility and practical value.

[0039] The overall space of the device is as follows Figure 2 As shown, details of the specific connection structure have been omitted. The power linkage structure mainly consists of two parts: the instrument linkage and the actuator. The instrument linkage includes a central rod, a rotating rod, and a support rod. The rotation trajectory of the rotating rod is similar to the opening and closing trajectory of an umbrella rib, except that the rotating rod can rotate vertically 180 degrees around its center.

[0040] The actuator moves up and down, driving the support rod to move. The support rod drives the rotating rod to move, which in turn drives the microphones arranged on it to move, thus realizing the spatial structure transformation of the microphone matrix.

[0041] The algorithm logic is shown in the figure. Figure 5As shown, after system initialization, the sampling module uses an ADC to synchronously sample audio, temperature, and displacement information. This data is then transmitted to the FPGA main controller. The algorithm set in the FPGA checks if the collected audio signal meets a threshold until a characteristic audio signal appears. The algorithm then further processes the valid signal, estimates the time delay, and combines the temperature and displacement values ​​with a spatial model of the microphone matrix to perform preliminary positioning calculations and display. Further near-field and far-field judgments and optimal spacing determination are then performed, and the microphone structure is adaptively adjusted for more precise positioning and display. The time delay estimation process is as follows: Figure 6 As shown.

[0042] Furthermore, the power linkage module includes a central rod, a rotating rod, a support rod, and an actuator; One end of the rotating rod is hinged to the central rod, and the other end is fixed to the corresponding microphone; One end of the support rod is hinged to the corresponding rotating rod, and the other end is slidably connected to the central rod via a slider. The actuator connects to the slider and drives the slider to move linearly on the central rod, thereby causing the rotating rod to rotate.

[0043] Specifically, the power linkage module includes a linkage structure and an actuator. The former supports the overall microphone array spatial structure of the measurement system, including basic linkages such as the central rod, rotating rod, and support rod; the latter mainly receives the control signals processed by the feedback module, exerts a force on the rotating linkage, and moves it a corresponding distance in the corresponding direction, thereby changing the spatial structure accordingly.

[0044] Furthermore, the latter's actuator components primarily provide the power for the linkage's transformation. The spatial structure of the entire microphone matrix is ​​variable, exhibiting an umbrella-like trajectory during its transformation. The spatial structure is as follows... Figure 2 As shown. The four side connecting rods are rotatable connecting rods (hereinafter referred to as swivel rods), capable of rotating around the central rod in a vertical plane. Simultaneously, a support connecting rod is connected to the center of each swivel rod. The support connecting rods support the swivel rods, and their ends can move vertically up and down along the central rod, thus changing the rotation angle of the swivel rods. The power source for vertical movement can be either a motor actuator or manual pushing.

[0045] Furthermore, the displacement measurement module is a linear sliding resistive distance sensor, which is mounted on the central rod and measures the linear displacement of the slider; The displacement signal is used to calculate the current spatial coordinates of each microphone.

[0046] Specifically, the displacement measurement module consists of a capacitive distance sensor and a filtering and amplification circuit. It is responsible for measuring the displacement information from the bottom of the support rod to the center rod and transmitting the data to the FPGA for further processing.

[0047] The displacement measurement module in this embodiment primarily determines displacement by measuring the displacement of the lower end of the support rod relative to the center, and calculates the spatial position change of the microphone module corresponding to each link through geometric relationships. This device uses a linear sliding resistive distance sensor, the basic principle of which is that a brush contacts a resistive film to change the resistance value. This converts the distance into an electrical signal, which is then acquired by the sampling module. The basic process is as follows: Figure 3 As shown.

[0048] Furthermore, the temperature measurement module is used to measure the ambient temperature in real time and provide the temperature signal to the FPGA main controller; The FPGA master controller is used to correct the speed of sound in the air based on the temperature signal in order to compensate for the error caused by the change in the speed of sound in the TDOA algorithm.

[0049] Specifically, the temperature measurement module consists of a temperature sensor and a filtering and amplification circuit. The temperature sensor measures the air temperature in the current environment and transmits the data to the FPGA for further processing. Since temperature has a significant impact on the speed of sound in air, this module is added to improve positioning accuracy.

[0050] Furthermore, the sampling module includes a multi-channel synchronous ADC chip. The input terminals of the ADC chip are connected to the filtering and amplification circuits of the microphone array module, temperature measurement module, and displacement measurement module, respectively. The output terminals of the ADC chip are connected to the FPGA main controller via a parallel bus for simultaneous sampling of multiple signals.

[0051] Specifically, the sampling module uses the ACM7060C as the core analog-to-digital converter chip to process the measurement results of the sound, temperature, and displacement signals from the measurement module before transmitting them to the FPGA main controller for further processing. The 7606C is an 8-channel, 16-bit converter with a maximum speed of 1MHz, supporting both parallel and serial reading, meeting the requirements for real-time sampling.

[0052] Furthermore, the process of the FPGA main controller executing the TDOA algorithm includes: Perform an FFT transformation on every two microphone audio signals to obtain the frequency domain signal; Calculate the cross-power spectrum of a frequency domain signal; PHAT weighting of the cross-power spectrum; The weighted cross-power spectrum is subjected to IFFT transformation to obtain the generalized cross-correlation function; The peak position of the generalized cross-correlation function is detected to obtain the time delay between the corresponding microphone pairs; The location of the sound source is calculated based on the time delay and the current microphone coordinates.

[0053] Specifically, the TDOA-based localization algorithm mainly consists of two steps: first, calculating the time delay difference between different array elements in the microphone array; second, establishing a mathematical model for sound source localization based on the spatial position and geometric relationship of the array; and finally, solving for the sound source location. The most commonly used algorithm is the GCC algorithm, which has low computational complexity and is easy to implement.

[0054] Taking the i-th and j-th microphones as examples: Step 1: Calculate the signal and FFT of the signal: and ; in, Indicates the i-th signal. Indicates the j-th signal. This represents the frequency domain Fourier transform of the i-th signal. Let represent the frequency domain Fourier transform of the j-th signal.

[0055] Step 2: Calculate the cross-power spectrum ; in, Represents the cross power spectrum. express . conjugate.

[0056] Step 3: Perform IFFT calculation on the cross-power spectrum to obtain... : (2); in, Let ω represent the autocorrelation function, and ω represent the angular frequency.

[0057] Step 4: To improve the robustness of the GCC algorithm in reverberant and noisy environments, a weighted operation is performed on the cross-correlation function to highlight the power spectrum peak: (3); In the formula This is represented as a weighting function. Specifically, this embodiment uses the PHAT function, which acts as a whitening filter, smoothing the cross-power spectrum between signals and highlighting the peak value of the cross-correlation function, thus exhibiting good anti-reverberation capability. However, the anti-reverberation capability of GCC-PHAT is greatly affected by noise, so noise interference must be removed when applying the GCC-PHAT weighting function.

[0058] Step 5: After obtaining GCC, perform testing. Peak value, and then estimate time difference: (4); Here, the argmax operation refers to the operation that maximizes the cross-correlation function. The value of the variable is used as an estimate of the time delay.

[0059] Furthermore, after obtaining the sound source location information, the FPGA master controller performs the following further actions: Determine whether the sound source is in the near-field model or the far-field model. If it is in the far-field model, only the azimuth and elevation angles are output; if it is in the near-field model, the azimuth, elevation angles, and distance information are output. Next, based on the judgment result, a corresponding control signal is generated and the baseline length of the microphone array is adjusted through the feedback module.

[0060] Specifically, in this embodiment, the FPGA main controller collects and analyzes the sound signal, temperature signal, and displacement signal transmitted from the sampling module, obtains the sound source location information through the TDOA method based on GCC inside the FPGA, and then outputs it to the display screen. Based on the positioning, it outputs control signals, which are then applied to the power linkage through the feedback module to control the change of the spatial structure of the microphone array, thereby achieving the effect of adaptive control of the optimal microphone spacing.

[0061] The FPGA selected is the EP4CE10F17C8N chip, which includes 180 user I / O pins, 423,936 bits of embedded RAM, and 46 9-bit embedded hardware multipliers. It boasts abundant resources to meet performance computing requirements.

[0062] The following section uses the near-field model as an example to derive the process of solving the spatial model for location.

[0063] First, the geometric equations of the relevant microphone matrix are listed. After obtaining the relevant time delay difference through time delay estimation, the equations are substituted to obtain the positioning information. For a far-field model, the azimuth and elevation angles are obtained; for a near-field model, the distance is added.

[0064] Let M1, M2, M3, M4, and M5 be five microphones; their corresponding coordinates are as follows: The coordinates of the sound source are: i = 1, 2, ..., 5. .set up , , , , L1, L2, L3, L4, and L5 represent the incident angles and distances of sound source S relative to M1, M2, M3, M4, and M5, respectively. Initially, all five microphones are d away from the origin, with M1 being the microphone located directly above the origin. c is the speed of sound in air. The moment when the sound source emits a sound signal. (i=1, 2…5) represents the time when the i-th microphone is reached. for and The time delay difference between them.

[0065] Distance derivation: like Figure 7 As shown, the M1 microphone and center stick in the lower left corner are examples. The angle at which the connecting rod containing M1 rotates downwards around the center is 0-90°.

[0066] Since M1 is always fixed, the coordinates of M1 are always [missing information]. .

[0067] like Figure 5 As shown, it is easy to obtain at the rotation angle At that time, the coordinates of M2 are .

[0068] According to the geometric structure, we can obtain... and The following relationship exists (Equation 5). Wherein, This is the distance the end of the support rod moves downwards from its initial position (when the rotating rod is horizontal). The angle at which the connecting rod containing M2 rotates downwards around the center is 0-90°.

[0069] (5); And because It can be accurately measured by the displacement measurement module, and the solution can be obtained as follows: ; Thus, the coordinates of M2 are obtained.

[0070] The coordinates of the remaining microphones can be obtained by analogy.

[0071] The solutions for L1, L2, L3, L4, and L5 can be obtained using the basic formula for the distance between two points in space (6).

[0072] (6); because: (7); Therefore, we can solve for: (8); in, (9); (10); (11); in, This is obtained from the aforementioned GCC method.

[0073] Furthermore, the unknown vector U can be obtained by solving equation (12): (12); Finally, the coordinates of the desired location S are obtained. Substituting into equation (6) yields the distance. .

[0074] Azimuth angle derivation: As mentioned above, for and The time delay difference between them, for example for: (13); Combining the overall structural diagram, and using the Law of Cosines, we get: (14); Given d, L2, and L4, the azimuth angle can be solved. Similarly, the remaining azimuth angles can be obtained, and the direction of the sound source can be determined by combining them.

[0075] Furthermore, the feedback module includes: Differential amplifier circuit, used to receive control signals output by FPGA main controller; A power amplifier, connected to a differential amplifier circuit, is used to amplify the power of control signals; The LC filter circuit, connected to the output of the power amplifier, is used to filter out high-frequency noise and drive the actuator through the amplified control signal to make real-time adaptive adjustments to the microphone array.

[0076] Specifically, the feedback module primarily amplifies and filters the output control signal of the FPGA to match the drive signal of the actuator in the power linkage. For example, it amplifies and filters the output control signal of the FPGA to match the drive signal of the electromagnetic actuator. Figure 4 As shown, the circuit includes a differential circuit, a TDF8599, and an LC filter circuit. The TDF8599 amplifies the signal to meet the driving requirements of the electromagnetic actuator.

[0077] Furthermore, the display module includes an LCD screen, which is connected to the FPGA main controller via a parallel interface; The LCD display screen is used to show real-time sound source location information, including direction and distance. This includes: When the far-field model is active, the azimuth and elevation angles of the sound source are displayed; When the near-field model is active, it displays the azimuth, elevation, and distance of the sound source. And the new baseline length and the number of adjustments after each microphone array structure adjustment.

[0078] Specifically, the display module in this embodiment displays sound source localization information in real time, including azimuth and distance information. In the far-field model, the useful information displayed is angle information, including pitch and azimuth angles; in the near-field model, in addition to angle information, distance information is also displayed. The switching between near-field and far-field models is determined by the following formula: (1); in, Let λ be the distance from the sound source to the center of the microphone array, D be the spacing between two adjacent microphone elements, and λ be the minimum wavelength of the sound source signal. When the equation holds true, the sound source is considered to be in the far field; otherwise, it is considered to be in the near field.

[0079] This embodiment proposes a practically usable "umbrella-type" adaptive sound source localization device. The realization of this device has important practical guiding significance for the engineering and lightweighting of the sound source localization field, and can also provide a necessary platform for subsequent research on sound source localization devices.

[0080] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An umbrella adaptive sound source localization device, characterized by, The umbrella type adaptive sound source positioning device comprises a measuring system and a positioning system. The measuring system comprises: a microphone array module for receiving sound waves from a sound source and outputting multiple sound signals; a temperature measuring module for obtaining ambient temperature and outputting a temperature signal; a displacement measuring module for obtaining the displacement of each microphone in the microphone array relative to the central rod and outputting a displacement signal; a power linkage module for supporting the microphone array module, the temperature measuring module, and the displacement measuring module, and changing the spatial position of the microphone according to a control signal; The positioning system comprises: a sampling module connected to the microphone array module, the temperature measuring module, and the displacement measuring module, for synchronously collecting the sound signals, the temperature signal, and the displacement signal and outputting a digital signal; an FPGA master connected to the sampling module, for performing time delay estimation based on the digital signal to obtain a time delay difference, and for calculating sound source position information by executing a TDOA algorithm according to the time delay difference, the temperature signal, and the displacement signal, while generating a control signal for adjusting the spatial position of the microphone; a feedback module connected between the FPGA master and the power linkage module, for amplifying and filtering the control signal and driving the power linkage module; a display module connected to the FPGA master, for outputting the sound source position information.

2. The umbrella type adaptive sound source positioning device according to claim 1, wherein: the microphone array module comprises at least five microphone units, each of which is arranged on a corresponding linkage of the power linkage module; the microphone units are used for receiving sound signals emitted by a sound source; the sound signals are processed by a filter-amplifier circuit and then output to the sampling module.

3. The umbrella type adaptive sound source positioning device according to claim 1, wherein: the power linkage module comprises a central rod, a rotating rod, a support rod, and an actuator; one end of the rotating rod is hinged to the central rod, and the other end is fixed to a corresponding microphone; one end of the support rod is hinged to a corresponding rotating rod, and the other end is slidingly connected to the central rod through a sliding block; the actuator is connected to the sliding block and drives the sliding block to move linearly on the central rod, thereby rotating the rotating rod.

4. The umbrella type adaptive sound source positioning device according to claim 1, wherein: the displacement measuring module is a linear sliding resistance distance sensor, which is installed on the central rod and measures the linear displacement of the sliding block; the displacement signal is used to calculate the spatial coordinates of the current microphones.

5. The umbrella type adaptive sound source positioning device according to claim 1, wherein: the temperature measuring module is used to measure the ambient temperature in real time and provide the temperature signal to the FPGA master; the FPGA master is used to correct the sound speed in the air according to the temperature signal, so as to compensate for the error caused by the change of sound speed in the TDOA algorithm.

6. The umbrella type adaptive sound source positioning device according to claim 1, wherein: ​ The sampling module comprises a multi-channel synchronous ADC chip, input ends of the ADC chip are connected with filter-amplifier circuits of the microphone array module, the temperature measuring module and the displacement measuring module respectively, and an output end of the ADC chip is connected with the FPGA master control through a parallel bus, for simultaneously sampling multiple signals.

7. The umbrella-shaped adaptive sound source positioning device according to claim 1, wherein, the process of the TDOA algorithm executed by the FPGA master control comprises: performing FFT transformation on every two microphone sound signals to obtain frequency domain signals; calculating cross power spectrum of the frequency domain signals; performing PHAT weighting on the cross power spectrum; performing IFFT transformation on the weighted cross power spectrum to obtain a generalized cross-correlation function; detecting a peak position of the generalized cross-correlation function to obtain a time delay between the corresponding microphone pair; calculating a sound source position according to the time delay and current microphone coordinates.

8. The umbrella-shaped adaptive sound source positioning device according to claim 1, wherein, after obtaining the sound source position information, the FPGA master control further executes: judging whether the sound source is in a near-field model or a far-field model, if in the far-field model, only azimuth and elevation angles are output, and if in the near-field model, azimuth, elevation and distance information are output; then, generating corresponding control signals according to the judgment result and adjusting a baseline length of the microphone array through the feedback module.

9. The umbrella-shaped adaptive sound source positioning device according to claim 1, wherein, the feedback module comprises: a differential amplifier circuit for receiving the control signals output by the FPGA master control; a power amplifier connected with the differential amplifier circuit for amplifying power of the control signals; an LC filter circuit connected with an output end of the power amplifier for filtering high-frequency noise, and driving an actuator through the power-amplified control signals to perform real-time adaptive adjustment on the microphone array.

10. The umbrella-shaped adaptive sound source positioning device according to claim 1, wherein, the display module comprises an LCD display screen connected with the FPGA master control through a parallel interface; the LCD display screen is used for displaying information in real time, including: when the far-field model is effective, displaying azimuth and elevation angles of the sound source; when the near-field model is effective, displaying azimuth, elevation and distance of the sound source; and a new baseline length after each adjustment of the microphone array structure and the adjustment times.

Images