Distributed acoustic data acquisition and processing system

Abstract

The invention provides a distributed acoustic data acquisition and processing system, which belongs to the field of signal acquisition and processing, and comprises a power supply module, an external interface module, a data stream receiving module, a network module, a KVM switching module, an operation processing module, a data storage module and an equipment interaction module, the external interface module is used for receiving various input acoustic data and communicating with upper computer equipment in a debugging process; the data stream receiving module is responsible for data conversion and release of acoustic data; the network module is responsible for data exchange among the modules in the device; the operation processing module is used for identifying a noise target and issuing decision suggestions; the data storage module stores original data, operation results and decision information; and the equipment interaction module realizes information interaction with external equipment. And acoustic data are collected in multiple scenes, and target identification is carried out.

Description

Technical Field

[0001] The present invention belongs to the field of signal acquisition and processing, and particularly relates to a distributed acoustic data acquisition and processing system. Background Art

[0002] With the increasing demand for precise detection and environmental perception in various fields, acoustic detection, as an efficient and non-invasive technical means, has become increasingly important and shown broad application prospects in resource exploration, ecological monitoring, security protection, etc. The acoustic data collected by acoustic detection can include underwater acoustic data, environmental acoustic data, mechanical acoustic data, etc. However, acoustic data is often huge and complex in composition, and the recognition of noise signals in acoustic data is also very important in the process of acoustic detection.

[0003] Traditional acoustic data acquisition and processing devices, including sonar stations, data processing computers, etc., usually do not have the ability to analyze and process large-scale concurrent data, and it is difficult to quickly identify and respond to noise targets in acoustic data. This affects the ability of external control devices to make decisions and interact information based on noise information, resulting in low processing and execution efficiency of the collected acoustic data. Summary of the Invention

[0004] In order to improve the deficiencies of traditional acoustic data acquisition and processing devices, the present invention provides a distributed acoustic data acquisition and processing system.

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

[0006] A distributed acoustic data acquisition and processing system specifically includes:

[0007] An external interface module for collecting acoustic data from external acoustic sensors or acoustic devices;

[0008] A device interaction module for information interaction with external devices;

[0009] A data stream receiving module for receiving the acoustic data transmitted from the external interface module and converting the acoustic data into the same format;

[0010] An operation processing module adopting a distributed centerless node architecture, including multiple micro operation units. The multiple micro operation units are used to process the acoustic data after format conversion by the data stream receiving module and the external device information transmitted by the device interaction module, process the acoustic data through an acoustic target recognition method to obtain noise information, perform information interaction with external device information according to the built-in timestamp information and the noise information, classify the noise information by marking, and issue decision-making suggestion information according to the marked noise information.

[0011] Preferably, the external interface module specifically includes a debugging communication interface, an acoustic data interface, and a data buffer pool. The debugging communication interface is used for data communication with a host computer during offline debugging. Specifically, it includes four USB interfaces, a standard Ethernet interface, and an HDMI interface. The USB interfaces and the HDMI interface can be connected to a keyboard, a mouse, a USB flash drive, and a high-definition display, and any micro computing unit in the operation processing module can be selected through a built-in KVM switching module to display its system interface. The host computer device can access any micro computing unit in the operation processing module or any data storage unit in the data storage module through the Ethernet interface to achieve data export and operation debugging.

[0012] The acoustic data interface collects acoustic data from external acoustic sensors or acoustic devices, including a standard input / output audio port, a standard Ethernet port, and an RS485 serial port, and is used to receive various acoustic data information input from the outside. Multiple interfaces can perform parallel synchronous data collection.

[0013] The data buffer pool is used to temporarily store the data received by the acoustic data interface and buffer the collected high-speed audio data stream.

[0014] Preferably, it further includes a network module for constructing a micro local area network to achieve data interaction between modules through multiple network ports. The debugging communication interface, the operation processing module, and the device interaction module in the external interface module perform two-way communication with the network module and receive the information transmitted by the data stream receiving module.

[0015] Preferably, the operation processing module adopts a distributed architecture without a central node. The working mode is specifically as follows: when entering the working state, any micro computing unit is selected for task management and resource allocation, and resources in multiple micro computing units are allocated proportionally in a non-uniform manner according to task requirements to jointly participate in the task; when a task is issued by the device interaction module or new data is transmitted by the data stream receiving module, nodes with idle resources are preferentially called to form a node network for operation processing.

[0016] Preferably, the operation processing module processes the acoustic data through an acoustic target recognition method to obtain noise information, including: performing preprocessing on the acoustic data such as denoising, pre-emphasis, framing, and windowing, and gradually performing fast Fourier transform, Mel filtering, logarithmic compression, discrete cosine transform, first-order and second-order differences, feature splicing, and model recognition on the preprocessed acoustic data, and outputting the recognized noise information.

[0017] Preferably, the model recognition specifically uses the SA-Res2Net model; by introducing a grouped convolution structure to replace the residual structure in the original Res2Net model, and introducing a self-attention pooling module SAP and a SoftMax activation function after multiple SA-module modules, the SA-Res2Net model is obtained.

[0018] Preferably, the data stream receiving module is used to receive the acoustic data transmitted from the external interface module and convert the acoustic data into the same format. Specifically, it receives the data collected by different acoustic data interfaces from the data buffer pool, defines the data collected by different acoustic data interfaces as different topics, unifies the sampling rate, number of channels, sampling width, and sampling bit depth of the data collected by different acoustic data interfaces, converts them into the same audio format for storage and processing, and transmits the processed data to the operation processing module.

[0019] Preferably, the data stream receiving module and the network module are designed with dual-module redundancy. The data stream receiving module includes a data stream receiving module I and a data stream receiving module II; the network module includes a network module I and a network module II. The data stream receiving module I, the data stream receiving module II, the network module I, and the network module II receive data simultaneously, preferentially organize and publish the input data of the data stream receiving module I and the network module I, and periodically send docking signals; if the docking signals are not received within a predetermined time, the data stream receiving module I and the network module I fail, and the data stream receiving module II and the network module II are switched to publish data.

[0020] Preferably, it further includes a data storage module. The data storage module includes multiple data storage units and is designed to be pluggable, and is used to store the acoustic data after format conversion received by the data stream receiving module, the external device information transmitted by the device interaction module, and the noise information and decision information output by the operation processing module.

[0021] Preferably, the device interaction module is used for information interaction with external devices. The interaction information specifically includes: navigation information, task information, operation control information, and command and control information; data exchange is performed through a network port, and decision-making suggestions or operation control instructions are provided to external devices by combining the task information and command and control information received from the device interaction module.

[0022] The distributed acoustic data acquisition and processing system provided by the present invention has the following beneficial effects:

[0023] The present invention obtains acoustic data through an external interface module, and after unified format processing by a data stream receiving module, it provides the data to an operation processing module, facilitating subsequent modules to process data at high speed and accurately. The operation processing module adopts a distributed architecture without a central node, and uses multiple micro-operation units to process acoustic data, enabling online processing of large-scale acoustic data information, identifying noise targets in the acoustic data, and making a quick response. The device interaction module conducts information interaction on the recognition result of the target with external devices, assisting external devices to execute tasks efficiently, and is applicable to auxiliary decision-making in various scenarios. Brief Description of the Drawings

[0024] In order to more clearly illustrate the embodiments of the present invention and their design schemes, the accompanying drawings required for this embodiment will be briefly introduced below. The accompanying drawings in the following description are only partial embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0025] Figure 1 It is a schematic diagram of the modules of a distributed acoustic data acquisition and processing system of the present invention.

[0026] Figure 2 It is an external interface diagram of the device of the present invention.

[0027] Figure 3 It is a flowchart of debugging and deployment in the offline stage in an embodiment of the present invention.

[0028] Figure 4 It is a flowchart of the operation stage of the data receiving module in an embodiment of the present invention.

[0029] Figure 5 It is a flowchart of the operation stage of the operation processing module and the data storage module in an embodiment of the present invention.

[0030] Figure 6 It is a schematic diagram of the steps of an acoustic target recognition method in an embodiment of the present invention.

[0031] Figure 7 It is an architecture diagram of an acoustic target recognition model adopted in an embodiment of the present invention.

[0032] Figure 8 It is a schematic diagram of the process of information utilization of the device interaction module in an embodiment of the present invention. Detailed Embodiments

[0033] In order to enable those skilled in the art to better understand the technical solutions of the present invention and implement them, the present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and cannot be used to limit the protection scope of the present invention.

[0034] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by terms such as "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "axial", "radial", "circumferential", etc. are based on the orientation or positional relationships shown in the drawings. These are only for the convenience of describing the technical solutions of the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention.

[0035] In addition, terms such as "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In the description of the present invention, it should be noted that unless otherwise clearly specified or limited, the terms "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In the description of the present invention, unless otherwise stated, the meaning of "plurality" is two or more, which will not be elaborated here.

[0036] Embodiment

[0037] The present invention provides a distributed acoustic data acquisition and processing system. Taking underwater acoustic data acquisition and processing as an example, as Figure 1 shown, it specifically includes: a power supply module, an external interface module, a data stream receiving module, a network module, a KVM switching module, an arithmetic processing module, a data storage module, and a device interaction module. Among them, the external interface module, the data stream receiving module, the arithmetic processing module, the data storage module, and the device interaction module are all connected to the network module.

[0038] The power supply module is used to supply power to all modules to ensure the stable operation of the system.

[0039] The external interface module includes a debugging communication interface, an acoustic data interface, and a data buffer pool. Among them, the debugging communication interface is used to check the status of the micro arithmetic unit in the arithmetic processing module through the KVM switching module and communicate with the host computer during off-line debugging to realize the deployment and debugging of the system. The acoustic data interface is used to collect acoustic data from external acoustic sensors or acoustic devices and provide multiple data interfaces to achieve parallel synchronous data acquisition. The data buffer pool is used to temporarily store the data received by the acoustic data interface and buffer the collected high-speed audio data stream. Using the external interface module has the advantages of small volume, high integration, convenient installation and deployment, and support for multiple acoustic data interfaces, etc., while improving the versatility.

[0040] The data stream receiving module includes data stream receiving module 1, data stream receiving module 2, network module 1 and network module 2, which are used to receive the data collected from the external interface module, unify the sampling rate, number of channels, sampling width and sampling bit depth of the data, and convert it into an audio format convenient for storage and processing. Subsequently, the acoustic data is published using the MQTT protocol, and the data collected by different acoustic data interfaces is defined as different topics for other modules to subscribe and use. Among them, data stream receiving module 1 and data stream receiving module 2 correspond to data stream receiving module Ⅰ and data stream receiving module Ⅱ respectively; network module 1 and network module 2 correspond to network module Ⅰ and network module Ⅱ respectively.

[0041] The network module is used to manage the data transmission between each module, build a micro local area network inside the system, and realize the data exchange between each module inside the system through several network ports.

[0042] The operation processing module is composed of multiple micro operation units and adopts a distributed architecture without a central node. All micro operation units are equal nodes, and each node can independently perform calculations and storage, and exchange data with other nodes. When entering the working state, a micro operation unit is randomly selected for task management and resource allocation, and the resources in several micro operation units are proportionally allocated in a non-uniform manner according to the task requirements to jointly participate in the task; when tasks are issued from other modules, nodes with more idle resources are preferentially called to form a node network for operation processing.

[0043] In the operation processing module, the acoustic target recognition method is used to obtain the noise target information. Through the timestamp information and noise target information built in the operation processing module, it is matched with the water navigation information and ship operation control information collected by the device interaction module to achieve target annotation. The target annotation information is combined with the task information obtained from the device interaction module and task analysis is carried out to obtain decision-making suggestions. Among them, the process of the acoustic target recognition method is as Figure 6As shown below, specifically: (1) Denoise the acoustic data by filtering the signal through a high-pass filter. (2) Pre-emphasis is performed on the signal after noise filtering to enhance the high-frequency part of the signal and compensate for possible high-frequency losses during signal transmission. (3) Preprocessing of framing and windowing, where the pre-emphasized signal is cut into frames, each frame having a length of 25 ms and a frame shift of 10 ms, and at the same time, a Hamming window function is applied to each frame for windowing. (4) Gradually perform a fast Fourier transform on the preprocessed acoustic data and use Mel filtering. Mel filtering converts the signal from the linear frequency axis (Hertz) to the Mel frequency axis that simulates the human ear's auditory perception, and decomposes the spectrum through a set of triangular filters to extract the characteristics of the speech signal. (5) Logarithmic compression, where the signal spectrum is convolved with each Mel filter, the energy of each Mel frequency channel is calculated, and a logarithmic transformation is performed on these energy values to introduce non-linear components and compress the data. (6) First-order and second-order differences. Since MFCC mainly represents the static information of the spectral characteristics of the audio signal, first-order differences and second-order differences are performed on the extracted MFCC features of the signal. The first-order difference can extract the trend of the audio signal changing over time, representing the dynamic changes of the signal; the second-order difference further describes the acceleration of the signal change and can capture higher-order dynamic information. (7) The extracted features are spliced and input into the network structure SA-Res2Net for model recognition, and finally, the noise target information recognized by the model is published.

[0044] As Figure 7As shown, during the processing of the acoustic target recognition method, the model adopted is SA-Res2Net, which is composed of multiple self-attention residual blocks SA-module. By introducing a grouped convolution structure to replace the traditional residual structure, it can better extract the multi-scale features of the acoustic target signal. Specifically, the self-attention residual block SA-module builds hierarchical residual connections in a single residual block, groups and convolves the extracted acoustic data feature map into multiple sub-feature maps. Each sub-feature map (except the first one) sequentially performs local feature extraction through a convolutional layer with a 3×3 convolution kernel and adds it to the output of the previous feature map. Then, the sub-feature maps in all channels pass through a convolutional layer with a 1×1 convolution kernel to achieve feature fusion. After fusion, it passes through the SE-Block module (Squeeze-and-Excitation Block) with an attention mechanism added. This module can achieve multi-scale feature representation and adaptive weighting of feature channels, enhancing the model's feature learning ability and discrimination ability. Specifically, the SE-Block module assigns weights to the feature channels through a global pooling layer and several fully connected layers, highlighting the acoustic target signal features and improving the model's recognition performance. Finally, the output is generated through the stacking of multiple SA-modules to generate feature information. The feature information realizes the classification and recognition of the target through the self-attention pooling module SAP and the SoftMax activation function, and finally outputs the result recognized by the model.

[0045] The data storage module is composed of multiple data storage units and is used to store and archive the original acoustic data unified by the data stream receiving module, the noise target information and decision information output by the operation processing module, the task information, navigation information, operation control information, and command and control information published by the device interaction module, etc.

[0046] The device interaction module is used to interact with external devices, and the interaction information includes: navigation information, task information, operation control information, command and control information, etc.

[0047] As Figure 2 shown, it is the appearance interface diagram of the corresponding device of the present invention system, mainly including debugging communication interfaces 101, 102, 103, and 104, acoustic data interfaces 201, 202, and 203, device interaction module interface 301, power module interface and indicator lights 401 and 402, data storage module operation indicator light 501, KVM switch module switch button and indicator light group 601 and 602, and ventilation opening 701.

[0048] In the debugging communication interface, the USB interfaces 101, 102, 103, and 104 are used for system installation using a USB flash drive or connecting devices such as keyboards and mice to facilitate user operations; the standard Ethernet port 105 is used for remote access and debugging tests; the HDMI port 106 is used to transmit the system interface in a visual form to an external display.

[0049] In the acoustic data interface, the standard audio input / output port 201 is used to collect acoustic data transmitted in the form of an audio interface; the standard Ethernet port 202 is used to collect acoustic data transmitted in the form of an Ethernet port; the RS485 serial port 203 is used to collect acoustic data transmitted in the form of a serial port.

[0050] In the device interaction module interface, the standard Ethernet port 301 is used to input or output navigation information, task information, operation control information, command and control information, etc. to external devices to achieve information interaction.

[0051] In the power module interface and indicator lights, the power interface 401 is used for power supply; the power-on indicator light 402 is off when not powered and on when the power supply is normal.

[0052] The operation indicator light 501 of the data storage module will be on when the data storage module is operating normally, off when not operating, and flashing when there is a fault.

[0053] A group of KVM switch module switch buttons 601 are connected to the internal KVM switch module and correspond to each micro-operation unit in the arithmetic processing module. Pressing the button will switch to the corresponding numbered micro-operation unit to view its current status and can be displayed in the visual system interface output by the HDMI port 106; the indicator light 602 corresponding to the KVM switch module switch button 601 is used to display the status of the micro-operation unit, on for normal operation, off for not connected, and flashing for a fault. The ventilation port 701 is used for heat dissipation during operation to avoid damage to functions caused by excessive internal temperature.

[0054] As Figure 3 shown, it is the flowchart of the debugging and deployment in the offline stage of the present invention. The display module, including a display, a keyboard, a mouse, etc., is connected to the KVM switch module through the USB interfaces 101, 102, 103, 104 and the HDMI interface 106, and can access any micro-operation unit through the KVM switch module button 601 to complete initialization operations such as system installation, network connection, environment configuration, and software deployment. The host computer is connected to the network module through the standard Ethernet port 105 and can access any micro-operation unit in the arithmetic processing module and any data storage unit in the data storage module to perform program initialization or debugging operations such as software installation, algorithm debugging, and test simulation.

[0055] AsFigure 4 and Figure 5 As shown in Figure 5 , the specific process during the system operation is as follows: The data stream receiving module monitors acoustic data from three different types of acoustic data interfaces, namely the standard input / output audio port 201, the standard Ethernet port 202, and the RS485 serial port. When data is received from any interface, it is transmitted into the queue of the data stream receiving module for buffering and the audio format is unified in sequence. The unified acoustic data is in the wav format which is easy to process and store, with a set sampling rate of 25,600 Hz, 1 channel, a sampling width of 2 bytes, and a bit depth of 16 bits. The unified acoustic data is published using the MQTT protocol. The MQTT protocol is a message passing protocol based on the publish / subscribe model. The topic rawdata / audio is used to publish the data received by the standard audio input / output port 201, the topic rawdata / internet is used to publish the acoustic data input by the network port 202, and the topic rawdata / serial is used to publish the acoustic data input by the serial port 203. Subsequently, the above three topics are published to the network module for other modules to subscribe and use.

[0056] The device interaction module conducts data interaction with external devices through the standard Ethernet port 301. The interaction information includes navigation information provided by external navigation devices, task information and command and control information issued from the upper computer, operation control information for devices such as motors and steering gears, etc. The device interaction module also publishes the obtained data to the internal local area network using the MQTT protocol. At the same time, after receiving the decision-making suggestions output by the operation and processing module, it can also publish decision-making control information to external devices.

[0057] The operation and processing module subscribes to the unified acoustic data to receive it, and at the same time subscribes to the navigation information, task information, operation control information, and command and control information published by the device interaction module. Next, by using the acoustic target recognition method for the unified acoustic data, noise target information can be obtained. Corresponding the noise target information with the navigation information and operation control information in the device interaction module and the timestamp information built into the operation and processing module can obtain target annotation information. Integrating the target annotation information with the task information published by the device interaction module and conducting task analysis can yield decision-making suggestions. Combining the decision-making suggestions with the command and control information realizes the release of decisions.

[0058] The data storage module is used to store and back up the data during the operation phase. The stored data includes the original acoustic data unified by the data stream receiving module, the noise target information and decision-making information output by the operation and processing module, the task information, navigation information, operation control information, and command and control information published by the device interaction module, etc.

[0059] such as Figure 8As shown in the figure, it is a schematic flowchart of the utilization of interactive module information. The device interactive module encapsulates the obtained navigation information, operation control information, and timestamp information when obtaining these information into frames. The above information is timestamp-compared with the noise target information output by the acoustic target recognition method according to the timestamp information to form target annotation information covering the above content. The task information and command and control information obtained from external devices by the device interactive module are used to perform task analysis in combination with the target annotation information. If the result of the task analysis is output to a manned vehicle such as a vehicle or a ship, the result will be published in the form of a decision-making suggestion for manual reference; if it is output to an autonomous control device such as an unmanned vehicle or an unmanned underwater vehicle, the result will be directly sent to the operation control device for execution.

[0060] Those skilled in the art should understand that the embodiments of the present invention may provide a method, a system, or a computer program product. Therefore, the present invention may be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may be implemented in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0061] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0062] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0063] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0064] It should be pointed out that the specific implementation methods described above can enable those skilled in the art to understand the invention more comprehensively, but do not limit the invention in any way. Therefore, although the invention has been described in detail in this specification and embodiments, those skilled in the art should understand that the invention can still be modified or replaced by equivalents; and all technical solutions and improvements that do not deviate from the spirit and scope of the invention are included in the protection scope of the patent for the invention. Any figure mark in the claims should not be regarded as limiting the claims involved. Any simple change or equivalent replacement of the technical solution that can be obviously obtained by any technician familiar with the field within the technical scope disclosed in the present invention belongs to the protection scope of the present invention.

Claims

1. A distributed acoustic data acquisition and processing system, characterized in that, Specifically include: An external interface module for collecting acoustic data from external acoustic sensors or acoustic devices; A device interaction module for information interaction with external devices; A data stream receiving module for receiving the acoustic data transmitted from the external interface module and converting the acoustic data into the same format; An operation processing module, adopting an architecture of distributed non - central nodes, including multiple micro - operation units. The multiple micro - operation units are used to process the acoustic data after format conversion by the data stream receiving module and the external device information transmitted by the device interaction module, process the acoustic data through an acoustic target recognition method to obtain noise information, conduct information interaction according to the built - in timestamp information and the noise information and the external device information, perform label classification on the noise information, and publish decision - making suggestion information according to the labeled noise information.

2. The distributed acoustic data acquisition and processing system according to claim 1, characterized in that The external interface module specifically includes a debugging communication interface, an acoustic data interface, and a data buffer pool. The debugging communication interface is used for data communication with the upper computer during offline debugging, specifically including four USB interfaces (101 - 104), a standard Ethernet interface (105), and an HDMI interface (106). The USB interfaces and the HDMI interface can connect a keyboard, a mouse, a USB flash drive, and a high - definition display, and select any micro - operation unit in the operation processing module through the built - in KVM switching module to display its system interface; The upper computer device accesses any micro - operation unit in the operation processing module or any data storage unit in the data storage module through the Ethernet interface to achieve data export and operation debugging; The acoustic data interface collects acoustic data from external acoustic sensors or acoustic devices, including a standard input / output audio port (201), a standard Ethernet port (202), and an RS485 serial port (203), and is used to receive various external input acoustic data information. Multiple interfaces support parallel synchronous data collection; A data buffer pool for temporarily storing the data received by the acoustic data interface and buffering the collected high - speed audio data stream.

3. A distributed acoustic data acquisition and processing system according to claim 1, wherein It also includes a network module for constructing a micro - local area network to achieve data interaction between modules through multiple network ports. The debugging communication interface, the operation processing module, and the device interaction module in the external interface module communicate bidirectionally with the network module and receive the information transmitted by the data stream receiving module.

4. A distributed acoustic data acquisition and processing system according to claim 1, wherein The operation processing module adopts an architecture of distributed non - central nodes, and its working mode is specifically as follows: When entering the working state, select any micro - operation unit for task management and resource allocation, and allocate resources in multiple micro - operation units proportionally in a non - balanced manner according to task requirements to jointly participate in the task; When there is a task issued by the device interaction module or new data transmitted by the data stream receiving module, preferentially call the nodes with idle resources to form a node network for operation processing.

5. A distributed acoustic data acquisition and processing system according to claim 1, characterized in that, The operation processing module processes the acoustic data through an acoustic target recognition method to obtain noise information, including: performing preprocessing on the acoustic data such as denoising, pre-emphasis, framing, and windowing, and gradually performing fast Fourier transform, Mel filtering, logarithmic compression, discrete cosine transform, first-order and second-order differences, feature splicing, and model recognition on the preprocessed acoustic data, and outputting the recognized noise information.

6. A distributed acoustic data acquisition and processing system according to claim 5, characterized in that, The model recognition specifically uses the SA-Res2Net model; by introducing a grouped convolution structure to replace the residual structure in the original Res2Net model, and introducing a self-attention pooling module SAP and a SoftMax activation function after multiple SA-module modules, the SA-Res2Net model is obtained.

7. A distributed acoustic data acquisition and processing system according to claim 1, wherein The data stream receiving module is used to receive the acoustic data transmitted from the external interface module and convert the acoustic data into the same format. Specifically, it receives the data collected by different acoustic data interfaces from the data buffer pool, defines the data collected by different acoustic data interfaces as different topics, unifies the sampling rate, number of channels, sampling width, and sampling bit depth of the data collected by different acoustic data interfaces, converts them into the same audio format for storage and processing, and transmits the processed data to the operation processing module.

8. The distributed acoustic data acquisition and processing system according to claim 3, wherein, The data stream receiving module and the network module are designed with dual-module redundancy. The data stream receiving module includes a data stream receiving module I and a data stream receiving module II; the network module includes a network module I and a network module II. The data stream receiving module I, the data stream receiving module II, the network module I, and the network module II receive data simultaneously, preferentially organize and publish the input data of the data stream receiving module I and the network module I, and periodically send docking signals; if the docking signals are not received within a predetermined time, the data stream receiving module I and the network module I fail, and switch to the data stream receiving module II and the network module II to publish data.

9. A distributed acoustic data acquisition and processing system according to claim 1, wherein, It further includes a data storage module. The data storage module includes multiple data storage units and adopts a pluggable design, and is used to store the acoustic data after format conversion received by the data stream receiving module, the external device information transmitted by the device interaction module, and the noise information and decision information output by the operation processing module.

10. A distributed acoustic data acquisition and processing system according to claim 1, characterized in that, The device interaction module is used for information interaction with external devices. The interaction information specifically includes: task information, operation control information, and command and control information; data exchange is performed through a network port (301), and decision-making suggestions or operation control instructions are provided to external devices in combination with the task information, operation control information, and command and control information received from the device interaction module.

Images