Acoustic-electric composite detection device

Abstract

The invention relates to the technical field of electronic detection, in particular to an acoustoelectric composite detection device which comprises a radar antenna front end, a passive acoustic sensor array, a multi-channel data acquisition module, a frequency synthesis module, a signal processing module, a data processing module and a display control module. The radar antenna front end obtains target space information; the passive sound sensor array receives sound wave vibration signals; the multi-channel data acquisition module is used for synchronously acquiring radar and sound signals; the frequency synthesis module provides time synchronization; the signal processing module extracts target features; the data processing module resolves target information through a data fusion algorithm and performs intelligent identification; and the display control module realizes information visualization and control. According to the invention, through cooperative detection and data fusion of radar and acoustics, the detection probability, the classification precision and the anti-interference capability of low, slow and small targets are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electronic detection, in particular to a sound and electricity composite detection device. BACKGROUND

[0002] In the field of low-altitude security and unmanned aerial vehicle control, the effective detection and identification of low-altitude, slow-speed, and small-size targets is a major technical challenge. Such targets usually have small radar scattering cross-section, slow flight speed, and low flight altitude, making them easily submerged in complex ground clutter and environmental noise, resulting in problems such as large detection blind area, high false alarm rate, and limited identification capability of traditional single detection methods.

[0003] In order to improve the detection capability of low, slow, and small targets, several solutions based on multi-sensor fusion have been proposed in existing technologies. For example, in the patent with the application number CN202411378317.9 and the patent name "A short-range low-altitude photoelectric blind-filling monitoring method, system, device, and terminal", photoelectric sensors are mainly used for blind-filling monitoring, but there are limitations in the effective distance and adaptability to bad weather. In the patent with the application number CN202310154577.7 and the patent name "Radar and video fusion three-dimensional velocity measurement device and method", the focus is on the fusion of optical image and radar velocity information, and the use of target acoustic characteristics is lacking. In the patent with the application number CN202210000758.X and the patent name "A real-time situation awareness system and method based on low-altitude airborne multi-sensor fusion", the platform and application scenario are quite different from the ground fixed defense system. In the patent with the application number CN202410314714.3 and the patent name "Multi-source data fusion method and detection device for low, slow, and small unmanned aerial vehicles", although it involves multi-source data, it does not explicitly disclose the deep coordination mechanism of radar and acoustic array in the hardware architecture and intelligent triggering level.

[0004] In particular, in the patent with the application number CN201610474368.0 and the patent name "Composite detection method and system for low-altitude, slow-speed, and small target flight speed", although acoustic and radio composite detection is used, the technical core focuses on the measurement of a single parameter - speed, the system structure is relatively simple, and the function is limited to basic early warning, which cannot meet the urgent needs of high-precision classification and identification of targets.

[0005] In the field of acoustic detection, existing research such as "Zhao, Shenn, Zhus, Hao, Zhou, Chao, et al. Optimization design of microphone stereo array in acoustic detection of unmanned aerial vehicles [J]. Journal of Electronic Measurement and Instrument, 2025" optimizes the layout of the microphone array to improve the direction-finding performance, but this only solves a local problem in passive acoustic detection and does not involve the system-level coordination, synchronous acquisition, and deep fusion of acoustic and radar, two heterogeneous sensors.

[0006] In summary, the prior art solution depends on complementary homogeneous sensors, or only performs shallow data fusion, or fails to achieve the advantages of radar active precise ranging and acoustic passive characteristic identification in system architecture. Therefore, there is an urgent need in the art for a composite detection device and method that can deeply couple the long-range spatial perception ability of radar and the target characteristic identification ability of acoustics at the levels of hardware, synchronous acquisition, signal processing and data analysis, to achieve a longer detection distance, higher identification accuracy and stronger anti-interference ability for low, slow and small targets. SUMMARY

[0007] The purpose of the present application is to solve the problems existing in the prior art, and provide an electro-acoustic composite detection device, comprising: a radar antenna front end for directional transmission of electromagnetic waves and high-sensitivity reception of target echoes in a preset area to obtain long-range and high-precision target spatial information; a passive acoustic sensor array for passive reception of acoustic vibration signals generated or radiated by the target; a multi-channel data acquisition module connected with the radar antenna front end and the passive acoustic sensor array, for synchronous acquisition and preprocessing of the target spatial information and the acoustic vibration signals to obtain target echo signals and acoustic vibration echo signals; a frequency synthesis module connected with the radar antenna front end and the multi-channel data acquisition module, respectively, for providing local oscillator signals and realizing time synchronization between modules; a signal processing module connected with the multi-channel data acquisition module, for signal processing of the target echo signals and the acoustic vibration echo signals, extracting target features, and obtaining signal processing results; a data processing module connected with the signal processing module, for obtaining target information according to the signal processing results, and intelligently identifying and data fusing the target features through a data fusion algorithm; a display control module connected with the data processing module, for realizing information visualization and control.

[0008] Preferably, the radar antenna front end comprises an antenna array, a TR module array and a feed network: the antenna array is connected with the feed network through the TR module array, and the feed network is connected with the frequency synthesis module to obtain synchronous clock signals, local oscillator signals and control signals; the antenna array is arranged by a plurality of antenna units in a specific pattern, and realizes electronic scanning of electromagnetic beams through beam forming technology; The TR module array is connected to the multi-channel data acquisition module, and is used for transmitting intermediate frequency signals to the multi-channel data acquisition module.

[0009] Preferably, the passive acoustic sensor array is composed of several high-sensitivity microphones, and the passive acoustic sensor array is adaptively optimized according to the detection requirements of dynamic targets to improve the spatial resolution of target direction finding and the anti-interference capability. After the passive acoustic sensor array passively receives acoustic signals, the acoustic signal data is collected through the multi-channel data acquisition module, and the acoustic fingerprint feature extraction, passive detection and target classification are completed through the signal processing module and the data processing module.

[0010] Preferably, the radar antenna front end and the passive acoustic sensor array form a multi-modal collaborative mechanism to support an intelligent triggering mechanism. When the radar antenna front end detects a suspicious target, the passive acoustic sensor array is automatically activated to directionally listen to the airspace where the suspicious target is located, the system false alarm rate is reduced through dual-mode verification, and the invalid working time of the passive acoustic sensor array is reduced to optimize the system power consumption.

[0011] Preferably, the multi-channel data acquisition module is connected to the frequency synthesis module to obtain an accurate time synchronization signal. The multi-channel data acquisition module has several groups of independent and performance-adapted analog front ends and ADC channels, which are used to complete high-speed and high-dynamic-range sampling of radar intermediate frequency signals, and high-precision and synchronous sampling of acoustic audio signals.

[0012] Preferably, electromagnetic shielding and isolation technology is used inside the multi-channel data acquisition module to prevent radar high-frequency signals from interfering with acoustic analog circuits and to ensure the signal-to-noise ratio of acoustic signals.

[0013] Preferably, the data fusion algorithm includes: A space-time alignment algorithm is used to calibrate the space-time of the radar data and the acoustic data in the target features, and to eliminate false signals of a single sensor. An adaptive weighted fusion algorithm is used to dynamically adjust the confidence weight of the two types of data according to the real-time reliability of the radar data and the acoustic data, so as to realize data fusion anti-false alarm. A multi-modal classifier based on deep learning is used to analyze the fused feature data to realize high-precision classification and recognition of targets.

[0014] Preferably, the phase center of the radar antenna front end and the acoustic center of the passive acoustic sensor array are consistent in physical space.

[0015] Preferably, the signal processing module is realized by FPGA, and the radar signal processing chain and the acoustic signal processing chain are run internally and in parallel. The radar signal processing chain comprises pulse compression, moving target detection, CFAR detection and micro-Doppler analysis. The acoustic signal processing chain comprises signal preprocessing, sound source positioning and voiceprint feature extraction.

[0016] Preferably, the display control module provides integrated situation display, system state monitoring, manual control instruction input and data communication interface.

[0017] Compared with the prior art, the present application has the following beneficial effects: The present application obtains the Doppler center frequency range and the normalized half-width through the radar type and the clutter type, obtains the constraint conditions comprising the center frequency point, the sidelobe region, the null region and the main lobe region according to the Doppler center frequency range and the normalized half-width, and obtains the filter information according to the constraint conditions and the global pilot vector matrix, so that the amplitude and the phase of the target filter can be optimized at the same time. The present application realizes the synchronous, high-fidelity acquisition and isolation of the radar intermediate frequency signal and the acoustic audio signal through the cooperation of the multi-channel data acquisition module and the frequency synthesis module, and realizes the real-time processing and feature extraction of the radar and acoustic signals by the signal processing module in parallel, so as to realize the synchronous acquisition and feature-level fusion of the heterogeneous sensor data, and provide high-quality, time-space consistent input for the back-end data fusion. The present application runs the data fusion algorithm containing time-space alignment, adaptive weighted fusion and deep learning classifier through the data processing module, calculates the target information according to the target features and performs intelligent identification, realizes the complementary and collaborative verification of multi-source information, effectively eliminates the false alarm of a single sensor, and significantly improves the target classification accuracy, anti-interference ability and overall reliability of the system in complex environments. BRIEF DESCRIPTION OF DRAWINGS

[0018] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included only to illustrate preferred embodiments and are not to be considered as limiting of the present application.

[0019] Figure 1 Figure 1 is a structural diagram of a sound-electricity composite detection device according to the present application. DETAILED DESCRIPTION

[0020] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application. Obviously, the described embodiments are part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0021] Those skilled in the art can understand that the singular form "a", "an", "the" used herein can include the plural form, unless specifically stated otherwise. It should be further understood that the use of the phrase "comprising" in the specification of the present application means that the features, integers, steps, operations, elements and / or components described exist, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0022] Embodiments Please refer to Figure 1 As shown in the drawings, the acoustic-electric composite detection device provided by the embodiment fully utilizes the radar with high integration and low cost based on the TR chip antenna array design, optimizes and upgrades the detection equipment for low, slow and small targets to a great extent, and each module cooperates to realize the cooperative detection and identification of low, slow and small targets, including: The radar antenna front end is used for directional emission of electromagnetic waves in a preset area and high-sensitivity reception of target echoes to obtain remote and high-precision target spatial information. Specifically, in the embodiment, the radar antenna front end is the active detection unit of the acoustic-electric composite detection device.

[0023] Preferably, the radar antenna front end includes an antenna array, a TR module array and a feed network. The antenna array is connected to the feed network through the TR module array, and the feed network is connected to a frequency synthesis module of the acoustic-electric composite detection device to obtain synchronous clock signals, local oscillator signals and control signals. Specifically, in the embodiment, the frequency synthesis module generates a high-frequency local oscillator signal with high stability and low phase noise, and sends it to the feed network. The feed network adopts a multi-layer board structure design to realize the functions of Wilkinson power divider and phase shifter. Its function is to provide synchronous clock, local oscillator signals and control signals for the TR module array, and to ensure the amplitude and phase consistency of each channel signal, which is the key to realize beam scanning and formation. The antenna array is arranged by a plurality of antenna units in a specific rule, and electronic scanning of electromagnetic beams is realized through beam forming technology. Specifically, in the embodiment, the antenna array adopts a microstrip patch antenna form to constitute a planar array. The array can realize electronic scanning within a range of ±45° in azimuth and elevation by controlling the phase of each unit, and can cover a wide airspace without mechanical rotation. Since all the antenna units radiate electromagnetic waves simultaneously and in phase, they interfere and superimpose in space to form a main beam with extremely high directivity and concentrated energy. This is beam forming. Through phase shifters in the feed network, the phase of the signal distributed to each antenna unit can be accurately controlled. By changing the phase distribution, the pointing direction of the beam can be changed without physically rotating the antenna, thereby realizing electronic scanning and quickly covering a wide preset airspace. The TR module array is connected to the multi-channel data acquisition module of the acoustic-electric composite detection device, and is used for transmitting intermediate frequency signals to the multi-channel data acquisition module of the acoustic-electric composite detection device. Specifically, in the embodiment, the TR module array corresponds to each unit (or unit group) of the antenna array, and is integrated with an independent transmitting / receiving module. Each TR module includes a transmitting channel, and the transmitting channel includes a power amplifier. After the excitation signal from the frequency synthesizing module is amplified to the required power, the signal is radiated out through the antenna. Each TR module includes a receiving channel, and the receiving channel includes a low-noise amplifier (such as a noise figure < 2 dB) and a mixer. The weak echo signal received is amplified and down-converted, and an intermediate frequency signal is output. After the transmission is completed, the TR module immediately switches from the transmitting mode to the receiving mode, and the echo signal reflected by the target is captured by the antenna unit. The signal energy is extremely weak, and is smaller than the transmitted signal by tens of billions of times. The intermediate frequency signals output by all the TR modules are sent into the multi-channel data acquisition module in parallel and synchronously, and the timing (when to transmit, when to switch to receive, and when to sample) of the entire transmission and reception process is strictly controlled by the synchronous clock and trigger signal generated by the frequency synthesizing module. This ensures the consistency of the whole system, and provides an accurate time reference for subsequent data fusion.

[0024] The passive acoustic sensor array is used for passively receiving acoustic vibration signals generated or radiated by a target. Specifically, in the embodiment, the passive acoustic sensor array is a passive detection and feature identification unit of the device. During the working process, the passive acoustic sensor array is arranged near the front end of the radar antenna, and the installation interface is designed in a standardized manner to support quick replacement of different array types.

[0025] Preferably, the passive acoustic sensor array of the sound-electricity composite detection device is composed of several high-sensitivity microphones, and the passive acoustic sensor array of the sound-electricity composite detection device is adaptively optimized according to the detection requirements of the dynamic target to improve the spatial resolution of target direction finding and the anti-interference capability. Specifically, in the embodiment, the passive acoustic sensor array is composed of a plurality of (for example, 16) MEMS microphones arranged in an optimized circular array, and works in a frequency range of 20 Hz to 2 kHz, which can effectively capture the characteristic noise generated by the rotors and motors of the unmanned aerial vehicle. After the passive acoustic sensor array of the sound-electricity composite detection device passively receives the acoustic signal, the sound-electricity composite detection device acquires acoustic signal data through the multi-channel data acquisition module of the sound-electricity composite detection device, and completes the extraction of acoustic fingerprint features, passive detection and target classification through the signal processing module and the data processing module of the sound-electricity composite detection device.

[0026] Preferably, the radar antenna front end and the passive acoustic sensor array of the sound-electricity composite detection device form a multi-modal cooperative mechanism to support an intelligent triggering mechanism. When the radar antenna front end detects a suspicious target, the passive acoustic sensor array of the sound-electricity composite detection device is automatically activated to perform directional acoustic listening on the airspace where the suspicious target of the sound-electricity composite detection device is located, the system false alarm rate is reduced through dual-mode verification, and the invalid working time of the passive acoustic sensor array is reduced to optimize the system power consumption.

[0027] Preferably, the phase center of the radar antenna front end and the acoustic center of the passive acoustic sensor array of the sound-electricity composite detection device are consistent in physical space.

[0028] More preferably, the analog voltage signals produced by all microphones are synchronously sent into the acoustic acquisition channels of the multi-channel data acquisition module. Here, "synchronous" is crucial, and the global clock provided by the frequency synthesis module ensures that the synchronized multi-channel digital signals are processed by the digital beamforming module in the back-end (e.g., FPGA) to assume that the sound source comes from a certain direction and calculate the theoretical time difference of the sound wave from that direction to each microphone. Then, it uses the calculated time difference to compensate for the signals of each channel, so that the signals from that assumed direction are perfectly aligned and in phase on all channels, resulting in significant enhancement, while noise and interference signals from other directions, because their arrival time difference does not match the assumption, will be out of phase or even opposite phase when superimposed, canceling each other out or being weakened. This is equivalent to forming a sharp auditory cone or acoustic lens in the target direction. By dynamically changing these delay values, this auditory cone can be electronically scanned without physically rotating the array to search the entire monitoring airspace. Through scanning, the device can find the beam pointing that maximizes the output signal energy, which is the azimuth and elevation angle estimate of the target. This is acoustic direction finding. Once the beam is locked on the target, the system performs spectral analysis and time-frequency analysis (e.g., extracting mel frequency cepstral coefficients) on the enhanced high-quality audio signal.

[0029] The multi-channel data acquisition module is connected to the radar antenna front end and the passive acoustic sensor array of the acoustic-electric composite detection device to synchronously acquire and preprocess the target spatial information and acoustic wave vibration signals of the acoustic-electric composite detection device, obtaining target echo signals and acoustic wave vibration echo signals. Specifically, in this embodiment, the channel data acquisition module is the sensory digitization center of the device, responsible for synchronously and high-fidelity acquisition of heterogeneous sensor signals.

[0030] Preferably, the multi-channel data acquisition module of the acoustic-electric composite detection device is connected to the frequency synthesis module of the acoustic-electric composite detection device to obtain accurate time synchronization signals. The multi-channel data acquisition module of the acoustic-electric composite detection device has a plurality of independent and performance-adapted analog front ends and ADC channels, which are used to complete high-speed and high-dynamic-range sampling of radar intermediate frequency signals and high-precision and synchronous sampling of acoustic audio signals. Specifically, in the embodiment, the multi-channel data acquisition module includes a radar acquisition channel, which uses an ADC with a 16-bit precision and a maximum sampling rate of 250 MS / s to acquire radar intermediate frequency signals. The multi-channel data acquisition module includes an acoustic acquisition channel, which uses an ADC with a 24-bit precision and a sampling rate of 48 kS / s, and the number of channels is unified with the number of passive acoustic sensor array units, and is responsible for synchronously acquiring signals of all microphones. The radar acquisition channel and the acoustic acquisition channel are controlled by the same global clock and synchronous trigger signal provided by the frequency synthesis module, so as to ensure that the time synchronization error between radar and acoustic data is less than 1 microsecond, and to provide an accurate time and space alignment basis for the back-end data fusion.

[0031] Preferably, electromagnetic shielding and isolation technology is used inside the multi-channel data acquisition module of the acoustic-electric composite detection device to prevent radar high-frequency signals from interfering with acoustic analog circuits and to ensure the signal-to-noise ratio of acoustic signals. Specifically, in the embodiment, the analog front end of the acoustic channel is provided with electromagnetic shielding and filtering circuits to suppress radio frequency interference generated during radar operation.

[0032] More preferably, intermediate frequency analog signals from the radar antenna front end (TR module array) are received, the signals have a high frequency (usually in the MHz to GHz range) and a large dynamic range (because the strengths of echo signals differ greatly), audio analog signals from the passive acoustic sensor array (each microphone) are received, the signals have a low frequency (usually 20 Hz-2 kHz), are extremely weak and susceptible to interference, a global unified and high-stability master clock and a periodic synchronous trigger signal are generated by the frequency synthesis module, and are simultaneously sent to all ADC channels of the radar and the acoustic, the radar channel ADC uses a high-speed and high-sampling-rate ADC (such as 16-bit precision, 250 MS / s) to capture the details of the intermediate frequency signal at a high enough time resolution, so as to accurately analyze the distance and speed information of the target, the acoustic channel ADC uses a high-resolution and high-signal-to-noise-ratio ADC (such as 24-bit precision, 48 kS / s) to capture the subtle features of the acoustic wave at a high enough amplitude resolution, so as to provide high-quality data for subsequent voiceprint recognition, under the command of the synchronous trigger signal, all ADC channels are started at the same time, and the analog signals of the respective channels are converted into digital signals in parallel and independently, the converted radar digital stream (I / Q data) and acoustic digital stream (audio samples) are packaged respectively and attached with time stamps and other information, and are sent in real time and in parallel to the signal processing module (usually FPGA) through a high-speed data bus (such as LVDS, Ethernet).

[0033] A frequency synthesis module is connected with the radar antenna front end and the multi-channel data acquisition module of the acoustic-electric composite detection device respectively, and is used for providing a local oscillator signal and realizing time synchronization among modules. Specifically, in the embodiment, the frequency synthesis module is the frequency heart of the device, adopts a hybrid architecture of DDS+PLL, generates a high-stability, low-phase-noise frequency and clock signal, and specific outputs include: a local oscillator signal provided for the radar antenna front end, a high-speed sampling clock and a periodic synchronization pulse provided for the multi-channel data acquisition module. The module guarantees the time sequence consistency and frequency stability of the whole system. The frequency synthesis module generates a synchronization signal to drive the radar antenna front end to emit electromagnetic waves. The radio frequency signal generated by the frequency synthesis module is distributed to each TR module through a feed network, and is synchronously radiated after power amplification. When receiving, each TR module performs low-noise amplification and down-conversion on the echo signal, and the output intermediate frequency signal is sent to the multi-channel data acquisition module. At the same time, the multi-channel data acquisition module is triggered to synchronously collect radar echoes and acoustic signals.

[0034] More preferably, the frequency synthesis module adopts advanced technologies such as direct digital frequency synthesis or phase-locked loop to generate a high-frequency sinusoidal wave signal with high frequency stability and extremely low phase noise, which is the local oscillator signal. The pure local oscillator signal is accurately distributed to each TR module of the radar antenna front end through the feed network. In the transmitting process: the TR module takes it as a reference to generate and amplify the final electromagnetic wave. In the receiving process: the mixer of the TR module uses it to mix with the received echo signal to generate an intermediate frequency signal. The down-conversion process can be performed because the local oscillator signal provides a stable frequency reference. The frequency synthesis module also derives two key synchronization signals when generating the local oscillator signal: a high-speed sampling clock, which is a frequency-fixed square wave signal and serves as a metronome for ADC sampling, and a synchronous trigger pulse, which is a periodic pulse signal and serves as a starting gun for starting an action (such as a radar transmitting / receiving cycle) of the whole system. These clock and trigger signals are sent to the radar antenna front end (to control the transmitting timing and TR component switching) and the multi-channel data acquisition module (to control the starting sampling time of the ADC) at the same time.

[0035] A signal processing module is connected with the multi-channel data acquisition module of the acoustic-electric composite detection device, and is used for signal processing of target echo signals and acoustic vibration echo signals, extraction of target features, and obtaining of signal processing results. Specifically, in the embodiment, the signal processing module is implemented by a high-performance FPGA, and two real-time processing pipelines run in parallel inside the FPGA. The radar signal processing chain includes pulse compression, moving target detection, CFAR detection, micro-Doppler analysis, etc., and the acoustic signal processing chain includes signal preprocessing, sound source positioning, voiceprint feature extraction, etc. The signal processing module processes the radar signal and the acoustic signal in parallel, and extracts the spatial motion information and the voiceprint feature information of the target respectively.

[0036] Preferably, the signal processing module of the acoustic-electric composite detection device is realized by FPGA, and the radar signal processing chain and the acoustic signal processing chain are run in parallel inside; The radar signal processing chain of the acoustic-electric composite detection device includes pulse compression, moving target detection, CFAR detection, and micro-Doppler analysis; The acoustic signal processing chain of the acoustic-electric composite detection device includes signal preprocessing, sound source positioning, and acoustic feature extraction.

[0037] More preferably, according to the original intermediate frequency digital signal collected by the radar ADC, the signal is compressed into an extremely narrow peak pulse through a pulse compression algorithm, and the compressed signal is output. Using the Doppler effect, the compressed signal is filtered to remove background clutter such as stationary ground objects and buildings through a clutter suppression filter (such as MTI). Then, a series of pulses are subjected to FFT analysis, i.e., moving target display, and MTD data is output. According to the noise level of the surrounding environment, a detection threshold is adaptively set. Only when the signal strength exceeds this threshold is it determined to be a real target point, and a target detection report is output. The target echo sequence confirmed by CFAR detection is subjected to higher-order time-frequency analysis to detect the micro-frequency modulation caused by the parts of the target (such as the rotation of the unmanned aerial vehicle rotor and the swing of the limbs when walking), and the micro-Doppler spectrum or time-frequency diagram is output.

[0038] More preferably, the multi-channel original audio digital signal synchronously collected by the acoustic ADC is subjected to digital filtering (to further filter background noise), pre-emphasis (to enhance high-frequency components), and other preprocessing operations to improve signal quality. The preprocessed multi-channel synchronous audio signal is subjected to calculation and compensation of the time difference of sound waves arriving at different microphones to form one or more electronically scannable "auditory beams". The system scans the space to find the strongest beam pointing direction, and outputs the target azimuth and elevation angle. The enhanced audio signal from a specific direction after beamforming is subjected to spectral analysis to extract feature parameters such as mel-frequency cepstral coefficients, spectral centroids, and zero-crossing rates that can represent the nature of sound, and an acoustic feature vector is output.

[0039] The data processing module is connected to the signal processing module of the acoustic-electric composite detection device and is used to obtain target information according to the signal processing results, intelligently recognize and fuse target features through data fusion algorithms. Specifically, in this embodiment, the data processing module is the intelligent brain of the device and is realized by a multi-core CPU and an embedded AI chip. Through time and space registration, data association, track filtering and prediction, deep learning models, and other methods, the data processing module completes core tasks such as data fusion and target tracking, target recognition and classification. The data processing module performs time and space registration, association, and fusion on the radar and acoustic information, and uses the fused composite features to complete accurate identification and stable tracking of the target.

[0040] Preferably, the acoustic-electric composite detection device data fusion algorithm comprises: a space-time alignment algorithm for space-time calibration according to radar data and acoustic data in acoustic-electric composite detection device target features, and elimination of false signals of a single sensor; an adaptive weighted fusion algorithm for dynamically adjusting the confidence weight of two types of data according to the real-time reliability of acoustic-electric composite detection device radar data and acoustic-electric composite detection device acoustic data, and realizing data fusion anti-false alarm; a multi-modal classifier based on deep learning for analyzing the fused feature data to realize high-precision classification and identification of targets.

[0041] More preferably, the radar features (distance, azimuth, speed) and acoustic features (azimuth, acoustic fingerprint) sent by the signal processing module have inherent space-time differences in the original data observed by the radar and acoustic sensors due to the different physical installation positions of the radar and acoustic sensors. This step unifies all features to the same reference coordinate system and the same time through a preset coordinate transformation formula and timestamp compensation, realizes the preliminary matching of the point seen by the radar and the sound heard by the acoustic sensor in space-time, and in continuous multiple detection periods, the device receives multiple possible target reports. The data association algorithm (such as nearest neighbor, joint probability data association) is responsible for playing "connect four", and in continuous multiple detection periods, the system receives multiple possible target reports. The data association algorithm (such as nearest neighbor, joint probability data association) is responsible for playing "connect four", and judges which target at the current time is the same as the target at the last time, thereby forming the continuous motion trajectory (track) of the target, outputting the preliminary radar-acoustic feature pair and the stable target track. According to the real-time situation, the confidence (weight) of the radar and acoustic data is dynamically evaluated, the intelligent model outputs the target classification and identification results according to the fused features including radar micro-Doppler spectrum and acoustic acoustic fingerprint feature vector, uses Kalman filtering algorithm to smooth the track of the target and predict its possible position at the next time, realizes stable tracking, integrates all information into a complete target resume, and outputs a structured target information list including target ID, latitude and longitude coordinates, motion speed and heading, target type, target model and threat level.

[0042] The display control module is connected with the acoustic-electric composite detection device data processing module, and is used for realizing information visualization and control.

[0043] Preferably, the sound-electricity composite detection device display control module provides comprehensive situation display, system state monitoring, manual control instruction input and data communication interface, specifically, in the embodiment, the display control module provides man-machine interactive interface, mainly completes comprehensive situation display, system state monitoring and control, alarm and data processing, etc., and the final target information is sent to the display control module for visualization, while receiving control instructions, forming a closed loop.

[0044] The embodiment focuses on reliability design, each module only uses mature technology and devices in the industry to ensure the development cycle; the embodiment supports the combined use of multiple radar antenna front ends, and flexibly realizes 90° / 180° / 270° / 360° airspace monitoring in the azimuth dimension; the signal processing module and the data processing module of the embodiment are deployed on a high-performance computing platform in the back end, realizing the software, adaptation and rapid deployment of the device functions; the embodiment uses solidification, digitization and software technology to improve the integration of the equipment and reduce the mutual connection between devices; the modules of the embodiment use standardized design to ensure the universality and interchangeability of the subsystems and components, and focus on fault diagnosis and system detection design, mainly using software and supplemented by hardware, which is convenient for equipment operation and management.

[0045] The above is only the preferred embodiment of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments, and any technical solution falling within the concept of the present application belongs to the protection scope of the present application. It should be noted that for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application are also considered to be within the protection scope of the present application.

Claims

1. A combined acoustic and electrical detection device, characterized in that, include: The radar antenna front end is used to directionally transmit electromagnetic waves within a preset area and receive target echoes in order to obtain long-distance, high-precision target spatial information. A passive acoustic sensor array is used to passively receive acoustic vibration signals generated or radiated by a target. A multi-channel data acquisition module is connected to the radar antenna front end and the passive acoustic sensor array, and is used to synchronously acquire and preprocess the target spatial information and the acoustic vibration signal to obtain the target echo signal and the acoustic vibration echo signal. The frequency synthesis module is connected to the radar antenna front end and the multi-channel data acquisition module respectively, and is used to provide local oscillator signals and realize time synchronization between modules. The signal processing module is connected to the multi-channel data acquisition module and is used to perform signal processing on the target echo signal and the acoustic vibration echo signal, extract target features, and obtain signal processing results. A data processing module, connected to the signal processing module, is used to obtain target information based on the signal processing result, and to intelligently identify and fuse the target features using a data fusion algorithm. The display control module is connected to the data processing module and is used to realize information visualization and control.

2. The acoustic-electric composite detection device according to claim 1, characterized in that, The radar antenna front end includes an antenna array, a TR module array, and a feed network: The antenna array is connected to the feed network via the TR module array, and the feed network is connected to the frequency synthesis module to obtain synchronized clock signals, local oscillator signals, and control signals; The antenna array consists of several antenna elements arranged in a specific pattern, and the electromagnetic beam is electronically scanned using beamforming technology. The TR module array is connected to the multi-channel data acquisition module and is used to transmit intermediate frequency signals to the multi-channel data acquisition module.

3. The acoustic-electric composite detection device according to claim 1, characterized in that, The passive acoustic sensor array consists of several high-sensitivity microphones. The passive acoustic sensor array is adaptively optimized according to the detection requirements of dynamic targets in order to improve the spatial resolution of target direction finding and anti-interference capability. After passively receiving acoustic signals, the passive acoustic sensor array collects acoustic signal data through the multi-channel data acquisition module, and completes voiceprint feature extraction, passive detection, and target classification through the signal processing module and the data processing module.

4. The acoustic-electric composite detection device according to claim 1, characterized in that, The radar antenna front end and the passive acoustic sensor array form a multi-modal collaborative mechanism, supporting an intelligent triggering mechanism: When the radar antenna front end detects a suspicious target, the passive acoustic sensor array is automatically activated to conduct directional acoustic monitoring of the airspace where the suspicious target is located. The system false alarm rate is reduced through dual-modal verification, and the ineffective working time of the passive acoustic sensor array is reduced to optimize system power consumption.

5. The acoustic-electric composite detection device according to claim 1, characterized in that, The multi-channel data acquisition module is connected to the frequency synthesis module to obtain a precise time synchronization signal; The multi-channel data acquisition module has several independent and performance-adaptive analog front-ends and ADC channels, which are used to complete high-speed, high dynamic range sampling of radar intermediate frequency signals, as well as high-precision, synchronous sampling of acoustic audio signals.

6. The acoustic-electric composite detection device according to claim 5, characterized in that, The multi-channel data acquisition module employs electromagnetic shielding and isolation technology to prevent high-frequency radar signals from interfering with the acoustic analog circuits and to ensure the signal-to-noise ratio of the acoustic signals.

7. The acoustic-electric composite detection device according to claim 1, characterized in that, The data fusion algorithm includes: A spatiotemporal alignment algorithm is used to perform spatiotemporal calibration based on radar and acoustic data in the target features, eliminating false alarm signals from a single sensor; An adaptive weighted fusion algorithm is used to dynamically adjust the confidence weights of the two types of data based on the real-time reliability of the radar data and the acoustic data, thereby achieving data fusion to resist false alarms. A deep learning-based multimodal classifier is used to analyze fused feature data to achieve high-precision target classification and recognition.

8. The acoustic-electric composite detection device according to claim 1, characterized in that, The phase center of the radar antenna front end and the acoustic center of the passive acoustic sensor array are physically aligned.

9. The acoustic-electric composite detection device according to claim 1, characterized in that, The signal processing module is implemented by an FPGA, and internally the radar signal processing chain and the acoustic signal processing chain run in parallel. The radar signal processing chain includes pulse compression, moving target detection, CFAR detection, and micro-Doppler analysis; The acoustic signal processing chain includes signal preprocessing, sound source localization, and voiceprint feature extraction.

10. The acoustic-electric composite detection device according to claim 1, characterized in that, The display control module provides comprehensive situation display, system status monitoring, manual control command input, and data communication interface.

Images