Sound signal and environment signal fused pipe gallery risk detection method and device

Abstract

The invention provides a sound signal and environment signal fused pipe gallery risk detection method and device, and relates to the technical field of pipe gallery safety monitoring, and the method comprises the steps: carrying out the risk judgment of each detection point based on a sound signal and an environment signal, and positioning a target detection point with a risk event in a plurality of detection points; controlling a target acoustic sensor and a target environment sensor of the target detection point to acquire a first acoustic signal and a first environment signal based on the same signal acquisition frequency; and evaluating a risk type and a hazard level of the risk event on the target detection point based on the first sound signal and the first environment signal. The risk in the pipe gallery is preliminarily judged on the basis of at least one of the sound signals and the environment signals, the risk position is quickly positioned, then the risk type and the hazard level of the risk event in the pipe gallery are evaluated by fusing the sound signals and the environment signals, multi-source intelligent decision making is achieved, and the risk assessment efficiency is improved. And the accuracy, comprehensiveness and emergency response speed of pipe gallery risk identification are improved.

Description

Technical Field

[0001] This disclosure relates to the field of utility tunnel safety monitoring technology, and in particular to a method and device for risk detection of utility tunnels that integrates acoustic signals and environmental signals. Background Technology

[0002] With the continuous development of urban infrastructure, underground utility tunnels play a vital role in water supply, power supply, gas supply, and communication. However, due to the complexity of the pipelines within these tunnels and the presence of various potential risks, such as pipeline leaks, equipment failures, and fires, traditional monitoring systems often rely on manual judgment. This results in incomplete and inaccurate monitoring information, slow response times, and difficulty in providing precise risk location and real-time emergency response. Summary of the Invention

[0003] This disclosure aims to at least partially address one of the technical problems in the related art.

[0004] Therefore, the first aspect of this disclosure proposes a method for risk detection in utility tunnels that fuses acoustic signals and environmental signals. The utility tunnel includes multiple detection points, and the method includes:

[0005] Acquire the acoustic signal and environmental signal collected at each detection point;

[0006] A first discriminator is used to determine the risk of each detection point based on the acoustic signal to obtain a first risk detection result. A second discriminator is used to determine the risk of each detection point based on the environmental signal to obtain a second risk detection result. The first discriminator and the second discriminator are obtained through adversarial training.

[0007] Among the multiple detection points, the target detection point where a risk event exists is located. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event.

[0008] The target acoustic sensor and the target environment sensor at the target detection point are controlled to acquire a first acoustic signal and a first environmental signal based on the same signal acquisition frequency;

[0009] The risk type and hazard level of the risk event at the target detection point are assessed based on the first acoustic signal and the first environmental signal.

[0010] In some embodiments of this disclosure, the step of assessing the risk type and hazard level of a risk event at the target detection point based on the first acoustic signal and the first environmental signal includes: assessing the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal using any of the following methods: multivariate statistical analysis, machine learning algorithm, or weighted evaluation algorithm.

[0011] In some embodiments of this disclosure, the method further includes: using a time-series prediction model to predict the impact range and rate of change of risk events at the target detection point based on the first acoustic signal; wherein the time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the changing trends of the corresponding risk events.

[0012] In some embodiments of this disclosure, the method further includes: displaying the location of the target detection point on a terminal display interface, and displaying and alarming the risk type and hazard level of the risk event at the target detection point via a pop-up window.

[0013] A second aspect of this disclosure provides a risk detection device for utility tunnels that fuses acoustic signals and environmental signals. The utility tunnel includes multiple detection points, and the device includes:

[0014] The first acquisition module is used to acquire the acoustic signal and environmental signal collected at each detection point;

[0015] The preliminary judgment module is used to use a first discriminator to perform risk judgment on each of the detection points based on the sound signal to obtain a first risk detection result, and to use a second discriminator to perform risk judgment on each of the detection points based on the environmental signal to obtain a second risk detection result; wherein, the first discriminator and the second discriminator are obtained through adversarial training;

[0016] The positioning module is used to locate the target detection point where a risk event exists among the plurality of detection points. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event.

[0017] The second acquisition module is used to control the target acoustic sensor and the target environment sensor of the target detection point to acquire the first acoustic signal and the first environmental signal based on the same signal acquisition frequency.

[0018] The detection module is used to assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal.

[0019] In some embodiments of this disclosure, the detection module is specifically used to: assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal using any of the following methods: multivariate statistical analysis, machine learning algorithm, weighted evaluation algorithm.

[0020] In some embodiments of this disclosure, a prediction module is also included; wherein the prediction module is used to: use a time-series prediction model to predict the influence range and rate of change of risk events at the target detection point based on the first acoustic signal; wherein the time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the changing trends of corresponding risk events.

[0021] In some embodiments of this disclosure, a display module is also included; wherein the display module is used to: display the location of the target detection point on the terminal display interface, and display and alarm the risk type and hazard level of the risk event at the target detection point via pop-up window.

[0022] A third aspect of this disclosure provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0023] The memory stores computer-executed instructions;

[0024] The processor executes computer execution instructions stored in the memory to implement the method described in the first aspect above.

[0025] A fourth aspect of this disclosure provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method described in the first aspect above.

[0026] The utility tunnel risk detection method provided in this disclosure, which integrates acoustic and environmental signals, makes a preliminary judgment on the risks in the utility tunnel based on at least one of the acoustic and environmental signals, quickly locates the risk location, and then assesses the risk type and hazard level of the risk event in the utility tunnel by integrating the acoustic and environmental signals, thereby realizing multi-source intelligent decision-making and improving the accuracy, comprehensiveness and emergency response speed of utility tunnel risk identification.

[0027] Additional aspects and advantages of this disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this disclosure. Attached Figure Description

[0028] The above and / or additional aspects and advantages of this disclosure will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, in which:

[0029] Figure 1This is a schematic flowchart illustrating a method for detecting risks in utility tunnels by fusing acoustic and environmental signals, as provided in an embodiment of this disclosure.

[0030] Figure 2 This is a schematic diagram of a pipe gallery risk detection device that fuses acoustic signals and environmental signals, provided in an embodiment of this disclosure. Detailed Implementation

[0031] Embodiments of this disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting this disclosure.

[0032] Specifically, the following describes a method and apparatus for detecting risks in utility tunnels by fusing acoustic signals and environmental signals, according to embodiments of the present disclosure, with reference to the accompanying drawings.

[0033] Figure 1 This is a schematic flowchart illustrating a method for risk detection in utility tunnels that fuses acoustic and environmental signals, as provided in an embodiment of this disclosure. The utility tunnel to be monitored has multiple detection points. For example... Figure 1 As shown, the method for detecting risks in utility tunnels by fusing acoustic signals and environmental signals may include the following steps:

[0034] Step 101: Acquire the acoustic signal and environmental signal collected at each detection point.

[0035] In some embodiments of this disclosure, acoustic sensors and environmental sensors can be installed at each detection point to collect acoustic signals and environmental signals at each detection point, respectively. The acoustic signals include characteristics such as intensity and frequency variation, while the environmental signals can include various parameters related to the pipe gallery environment, such as temperature, humidity, and gas concentration.

[0036] In another implementation, a single sensor, such as an acoustic sensor or an environmental sensor, can be placed at the detection point.

[0037] Step 102: Using a first discriminator, risk assessment is performed on each detection point based on the acoustic signal to obtain a first risk detection result. Using a second discriminator, risk assessment is performed on each detection point based on the environmental signal to obtain a second risk detection result. The first and second discriminators are obtained through adversarial training.

[0038] In one implementation, an acoustic sensor network and an environmental sensor network can be deployed in the utility tunnel space, introducing an independent monitoring mode for parallel sensing and monitoring to achieve efficient risk monitoring. This targeted deployment strategy allows the acoustic sensor network to focus on risk detection through acoustic signals, while the environmental sensor network focuses on risk detection through environmental parameters (such as temperature, humidity, and gas concentration). This clearly defined deployment method can fully leverage the advantages of both discriminators, improving the efficiency and accuracy of front-end risk detection. The first discriminator is deployed in the acoustic sensor network, and the second discriminator is deployed in the environmental sensor network. The lightweight first and second discriminators are connected to multiple network topology ports.

[0039] In the construction of the first discriminator, acoustic signals are used as input, and risky events and non-risky events are used as the discrimination targets. Adversarial training is performed, and through adversarial learning between the generator and the discriminator, the discriminator can more accurately identify risky and non-risky events. Using a generator-discriminator architecture as the training architecture, the generator continuously generates fake acoustic signal data, and the discriminator distinguishes between real and fake acoustic signals, ultimately training a discriminator capable of accurately identifying risky events based on acoustic signals. After training, the trained discriminator is split into two parts to form the first discriminator.

[0040] In constructing the second discriminator, environmental signals are used as input, and risky events and non-risky events are used as the discrimination targets. Adversarial training is performed, and through adversarial learning between the generator and the discriminator, the discriminator can more accurately identify risky and non-risky events. Using a generator-discriminator architecture as the training architecture, the generator continuously generates false environmental signal data, and the discriminator distinguishes between real and false environmental signals, ultimately training a discriminator capable of accurately identifying risky events based on environmental signals. After training, the trained discriminator architecture is split into two parts to form the second discriminator.

[0041] In one implementation, a supervisory decision-maker can be embedded in the utility tunnel risk management system. This involves combining an acoustic sensor network with a first discriminator and deploying it in an independent monitoring mode. Simultaneously, an environmental sensor network and a second discriminator are also deployed in independent monitoring mode, forming the monitoring and management node within the supervisory decision-maker. As the core decision-making unit of the entire risk monitoring system, the embedded deployment of the supervisory decision-maker ensures the efficiency and real-time nature of the decision-making process. Embedded deployment signifies deep integration between the supervisory decision-maker and the utility tunnel risk management system, enabling direct acquisition of front-end monitoring data for comprehensive analysis and decision-making.

[0042] Step 103: Locate the target detection point where a risk event exists among multiple detection points. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event.

[0043] By using the results of the first and second risk detections, potential risks within the utility tunnel can be preliminarily assessed from multiple dimensions. By fully utilizing the complementarity of acoustic and environmental signals, it is possible to quickly respond to potential risk events in the utility tunnel, rapidly identify target detection points with potential risks, quickly pinpoint risk locations, and improve the efficiency of emergency response.

[0044] Step 104: Control the target acoustic sensor and the target environment sensor at the target detection point to collect the first acoustic signal and the first environmental signal based on the same signal acquisition frequency.

[0045] When at least one of the first and second risk detection results indicates the presence of a potential risk event, the upper-level decision-maker triggers a synchronous detection command. This command simultaneously executes the target acoustic sensor and target environmental sensor at the target detection point, acquiring the first acoustic signal and the first environmental signal at the same signal acquisition frequency. These signals are then transmitted back to the upper-level decision-maker. This synchronous sampling method ensures the temporal consistency between the first acoustic signal and the first environmental signal, providing an accurate data foundation for subsequent comprehensive analysis. Through synchronous sampling, the system can obtain more comprehensive and accurate risk-related information, thereby supporting the precise location and assessment of risks.

[0046] In some embodiments of this disclosure, when only a single sensor is installed at a detection point in the utility tunnel, the detection point with a potential risk event can be used as the first acquisition point. This point is then spatially mapped in another sensor network, and the detection point in the other sensor network closest to the first acquisition point is selected as the second acquisition point. A discriminator is then used to assess the risk of the data from the first and second acquisition points. For example, if a detection point a equipped with an acoustic sensor detects a potential risk event, the location of detection point a is mapped to an environmental sensor network. The detection point closest to detection point a is selected as the second acquisition point from among the detection points equipped with environmental sensors. The first and second acquisition points are then used as target detection points to collect acoustic and environmental signals respectively, and risk assessments are performed on each. Subsequent acquisitions of the first acoustic signal and the first environmental signal at the same frequency are also based on the first and second acquisition points. This method of determining the location based on the nearest sensor position ensures that the collected data is spatially correlated with the first acquisition point, thus providing an accurate reference for subsequent comprehensive analysis. This series of operations ensures the synchronized collaboration of all sensors within the utility tunnel. From spatial location mapping to synchronized data acquisition, it enhances the system's ability to accurately identify and locate risks in complex utility tunnel environments. Through precise spatial perception and data synchronization, the system can obtain comprehensive environmental and acoustic data in a shorter time, providing reliable support for the safety management of the utility tunnel.

[0047] Step 105: Assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal.

[0048] In one implementation, the risk type and hazard level of a risk event at the target detection point can be assessed based on the first acoustic signal and the first environmental signal via a heterogeneous decision channel in the upper-level decision-maker. Through this fusion, the system can combine data from different sensors (acoustic and environmental sensors) for comprehensive analysis, improving the accuracy of risk event judgment in complex environments. In some embodiments of this disclosure, the risk type and hazard level of a risk event at the target detection point can be assessed based on the first acoustic signal and the first environmental signal using any of the following methods: multivariate statistical analysis, machine learning algorithms, or weighted evaluation algorithms to assess the comprehensive impact of the risk event. For example, if the acoustic signal detects a leak sound while the environmental sensor detects a change in gas concentration, the system will combine this information to determine the risk type, such as leak, rupture, or abnormal temperature, and assess its potential hazard level. Finally, second positioning information is generated, providing more accurate data support for subsequent positioning decisions.

[0049] Supervised training of the heterogeneous decision-making channel is used as a monitoring and management node in the higher-level decision-maker. Supervised training is a machine learning method that enables the heterogeneous decision-making channel to accurately identify and process sensor information from different sources by learning from labeled data.

[0050] Monitoring and management nodes are cascaded with decision-making management nodes to form a higher-level decision-maker. Cascading is a system integration method that connects nodes with different functions sequentially to form a complete decision-making system. In this structure, the monitoring and management nodes are responsible for initial risk monitoring and data collection, while the decision-making management nodes conduct in-depth analysis and make decisions based on this data. Through cascading, the higher-level decision-maker can not only efficiently analyze data received from acoustic and environmental sensor networks, but also make intelligent early warning and response decisions based on the integrated information, thereby improving the level of intelligence in utility tunnel management.

[0051] This construction approach not only ensures the rapid acquisition and accurate analysis of risk information, but also enhances the flexibility and intelligence of decision-making through integration and cascading, enabling the utility tunnel management system to provide more reliable risk monitoring and management capabilities in actual operation.

[0052] Simultaneously, the upper-level decision-maker establishes a communication mechanism with the front-end deployed acoustic and environmental sensor networks. This communication mechanism is crucial for the coordinated operation of the entire system. Through this communication, the front-end acoustic and environmental sensor networks can transmit the acquired first acoustic and environmental signals to the upper-level decision-maker in real time. Simultaneously, the upper-level decision-maker can also send instructions to the front-end sensor networks as needed, such as triggering more detailed detection procedures or adjusting monitoring parameters. This two-way communication mechanism ensures information flow and collaborative operation between the various parts of the system, improving the overall intelligence level of the risk monitoring system.

[0053] Optionally, in some embodiments of this disclosure, the location of the target detection point can be displayed on the terminal display interface of the utility tunnel risk management system, and pop-up windows can be displayed and alarms can be triggered for the risk type and hazard level of the risk event at the target detection point. Alternatively, risk simulation and reconstruction can be performed based on the risk type and hazard level of the risk event at the target detection point. The goal of this process is to recreate the detailed spatial distribution when the risk event occurs through simulation technology, thereby accurately depicting the scope of the risk impact. Risk simulation and reconstruction takes into account the pipeline geometry, the layout of the sensor network, and the impact of risk characteristics on the surrounding environment, such as the speed and direction of leak diffusion, and the propagation path of temperature changes. This reconstruction process can help managers more accurately understand the scope of the risk's impact and provide a scientific basis for subsequent emergency response.

[0054] By providing a more intuitive display on the terminal, managers can view the location of risks within the utility tunnel in real time and obtain detailed alarm information. The system visually presents the spatial distribution and characteristics of risk areas through pop-up windows, helping staff quickly identify and respond to potential safety threats. Alarm content can include information such as risk type, affected area, and risk level, enabling utility tunnel managers to take appropriate countermeasures promptly. It not only accurately locates risk events within the utility tunnel but also helps managers quickly understand the actual impact range of risks through simulation and visualization, thereby achieving rapid response and handling. This process significantly improves the intelligence level of the utility tunnel management system, ensuring real-time risk monitoring and effective prevention.

[0055] In some embodiments of this disclosure, the monitoring and management node in the upper-level decision-maker may include a homogeneous decision-making channel in addition to heterogeneous decision-making channels. Homogeneous decision-making channels focus on analyzing information provided by the same type of sensors. This parallel integration approach can fully utilize the advantages of multi-source information, improving the accuracy and reliability of decision-making. The construction of the second decision-making management node enables the system to comprehensively analyze risk information from different monitoring modes, thereby more comprehensively assessing the risk status within the utility tunnel.

[0056] The same-source decision channel uses a time-series prediction model to predict the impact range and rate of change of risk events at the target detection point based on the first acoustic signal. The time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the corresponding trends of risk events.

[0057] By implementing the embodiments of this disclosure, a preliminary judgment of the risks in the utility tunnel can be made based on at least one of acoustic signals and environmental signals, the risk location can be quickly located, and then the risk type and hazard level of the risk event in the utility tunnel can be assessed by fusing acoustic signals and environmental signals, so as to realize multi-source intelligent decision-making and improve the accuracy, comprehensiveness and emergency response speed of utility tunnel risk identification.

[0058] Figure 2 This is a schematic diagram of a pipe gallery risk detection device that fuses acoustic signals and environmental signals, provided as an embodiment of this disclosure. Figure 2 As shown, the tunnel risk detection device that fuses acoustic signals and environmental signals may include: a first acquisition module 201, a preliminary judgment module 202, a positioning module 203, a second acquisition module 204, and a detection module 205.

[0059] The first acquisition module 201 is used to acquire the acoustic signal and environmental signal collected at each detection point.

[0060] The preliminary judgment module 202 is used to use a first discriminator to perform risk judgment on each of the detection points based on the acoustic signal to obtain a first risk detection result, and to use a second discriminator to perform risk judgment on each of the detection points based on the environmental signal to obtain a second risk detection result. The first discriminator and the second discriminator are respectively obtained through adversarial training.

[0061] The positioning module 203 is used to locate the target detection point where a risk event exists among the plurality of detection points. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event.

[0062] The second acquisition module 204 is used to control the target acoustic sensor and the target environment sensor of the target detection point to acquire the first acoustic signal and the first environmental signal based on the same signal acquisition frequency.

[0063] The detection module 205 is used to assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal.

[0064] In some embodiments of this disclosure, the detection module 205 is specifically used to: assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal using any of the following methods: multivariate statistical analysis, machine learning algorithm, weighted evaluation algorithm.

[0065] In some embodiments of this disclosure, such as Figure 2 Based on the illustrated embodiment, the utility tunnel risk detection device may further include a prediction module. The prediction module is used to: predict the impact range and rate of change of a risk event at the target detection point using a time-series prediction model based on the first acoustic signal; wherein the time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the changing trends of corresponding risk events.

[0066] In some embodiments of this disclosure, such as Figure 2 Based on the illustrated embodiment, the utility tunnel risk detection device may further include a display module. The display module is used to: display the location of the target detection point on a terminal display interface, and display and alarm the risk type and hazard level of the risk event at the target detection point via pop-up windows.

[0067] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0068] To implement the above embodiments, this disclosure also proposes an electronic device, including: a processor and a memory communicatively connected to the processor; the memory stores computer execution instructions; the processor executes the computer execution instructions stored in the memory to implement the method provided in the foregoing embodiments.

[0069] To implement the above embodiments, this disclosure also proposes a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided in the foregoing embodiments.

[0070] To implement the above embodiments, this disclosure also proposes a computer program product, including a computer program that, when executed by a processor, implements the methods provided in the foregoing embodiments.

[0071] In the foregoing descriptions of the embodiments, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0072] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0073] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of this disclosure includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of this disclosure pertain.

[0074] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0075] It should be understood that various parts of this disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0076] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0077] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0078] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present disclosure have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present disclosure.

Claims

1. A method for risk detection in utility tunnels by fusing acoustic signals and environmental signals, characterized in that, The utility tunnel includes multiple detection points, and the method includes the following steps: Acquire the acoustic signal and environmental signal collected at each detection point; A first discriminator is used to determine the risk of each detection point based on the acoustic signal to obtain a first risk detection result. A second discriminator is used to determine the risk of each detection point based on the environmental signal to obtain a second risk detection result. The first discriminator and the second discriminator are obtained through adversarial training. Among the multiple detection points, the target detection point where a risk event exists is located. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event. The target acoustic sensor and the target environment sensor at the target detection point are controlled to acquire a first acoustic signal and a first environmental signal based on the same signal acquisition frequency; The risk type and hazard level of the risk event at the target detection point are assessed based on the first acoustic signal and the first environmental signal.

2. The method according to claim 1, characterized in that, The assessment of the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal includes: Based on the first acoustic signal and the first environmental signal, the risk type and hazard level of the risk event at the target detection point are assessed using any of the following methods: multivariate statistical analysis, machine learning algorithm, or weighted evaluation algorithm.

3. The method according to claim 1, characterized in that, Also includes: Using a time-series prediction model, the impact range and rate of change of risk events at the target detection point are predicted based on the first acoustic signal; The time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the changing trends of corresponding risk events.

4. The method according to claim 1, characterized in that, Also includes: The location of the target detection point is displayed on the terminal display interface, and a pop-up window displays and alarms the risk type and hazard level of the risk event at the target detection point.

5. A pipe gallery risk detection device that fuses acoustic signals and environmental signals, characterized in that, The utility tunnel includes multiple detection points, and the device includes: The first acquisition module is used to acquire the acoustic signal and environmental signal collected at each detection point; The preliminary judgment module is used to use a first discriminator to perform risk judgment on each of the detection points based on the sound signal to obtain a first risk detection result, and to use a second discriminator to perform risk judgment on each of the detection points based on the environmental signal to obtain a second risk detection result; wherein, the first discriminator and the second discriminator are obtained through adversarial training; The positioning module is used to locate the target detection point where a risk event exists among the plurality of detection points. The target detection point is the detection point where the first risk detection result indicates the existence of a potential risk event and / or the second risk detection result indicates the existence of a potential risk event. The second acquisition module is used to control the target acoustic sensor and the target environment sensor of the target detection point to acquire the first acoustic signal and the first environmental signal based on the same signal acquisition frequency. The detection module is used to assess the risk type and hazard level of the risk event at the target detection point based on the first acoustic signal and the first environmental signal.

6. The apparatus according to claim 5, characterized in that, The detection module is specifically used for: Based on the first acoustic signal and the first environmental signal, the risk type and hazard level of the risk event at the target detection point are assessed using any of the following methods: multivariate statistical analysis, machine learning algorithm, or weighted evaluation algorithm.

7. The apparatus according to claim 5, characterized in that, The device further includes a prediction module; wherein the prediction module is used for: Using a time-series prediction model, the impact range and rate of change of risk events at the target detection point are predicted based on the first acoustic signal; The time-series prediction model is constructed based on the historical acoustic signals of the utility tunnel and the changing trends of corresponding risk events.

8. The apparatus according to claim 5, characterized in that, The device further includes a display module; wherein the display module is used for: The location of the target detection point is displayed on the terminal display interface, and a pop-up window displays and alarms the risk type and hazard level of the risk event at the target detection point.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-4.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-4.

Images