An underground cable trench safety inspection and emergency disposal robot and an inspection and risk identification method thereof

Abstract

This invention discloses a robot for safety inspection and emergency response in underground cable trenches, along with its inspection and risk identification method. The robot includes a vehicle motion module, a multi-degree-of-freedom reconnaissance and scanning mechanism, a multi-modal risk detector module, a positioning and autonomous navigation planning module, an edge computing decision-making module, and an emergency execution module. The multi-degree-of-freedom reconnaissance and scanning mechanism achieves multi-height, multi-angle, three-dimensional coverage scanning of the narrow cable trench interior space through vertical lifting, lateral insertion, and rotation scanning. The multi-modal risk detector module collects multi-source information in a single operation. The edge computing decision-making module employs a multi-source fusion risk identification mechanism with master-discrimination, slave-discrimination, and enhanced credibility to comprehensively assess risks and generate corresponding emergency response decisions. This invention enables autonomous inspection and accurate multi-risk identification in underground cable trench environments with low illumination and strong electromagnetic interference, improving the operational safety and intelligent maintenance level of cable trenches.

Description

Technical Field

[0001] This invention relates to the field of intelligent inspection and safety monitoring technology for power facilities, specifically to a robot for safety inspection and emergency response in underground cable trenches and its inspection and risk identification method, and particularly to risk identification and emergency response technology based on multimodal sensor information fusion in complex cable trench environments. Background Technology

[0002] Underground cable trenches are a crucial infrastructure component of power systems, and their safe and stable operation is essential for ensuring power transmission. The internal environment of cable trenches is complex, typically characterized by enclosed spaces, limited ventilation, and harsh conditions such as humidity, high dust levels, and strong electromagnetic interference. Problems such as aging cable joints, insulation damage, and animal intrusion can easily lead to localized overheating, fires, and even explosions, posing a serious threat to the safe and stable operation of the power system.

[0003] Currently, the inspection and monitoring of cable trenches mainly rely on the following methods:

[0004] Manual periodic inspections: Inspectors must enter the harsh environment of trenches and use handheld detection devices (such as infrared thermometers) to conduct inspections. This method has inherent drawbacks such as low efficiency, incomplete coverage, and strong subjectivity, and poses a great threat to the personal safety of inspection personnel in the event of an emergency.

[0005] Fixed monitoring systems: Cameras, temperature sensors, and gas sensors are permanently installed inside the trench. While this system achieves a certain degree of continuous monitoring, its monitoring range is fixed, with numerous blind spots, making it impossible to perform flexible and three-dimensional scanning of the entire trench area. More importantly, it typically only has data acquisition and reporting functions, lacking on-site emergency response capabilities. It cannot intervene promptly after discovering potential hazards, resulting in a management disconnect of "inspection without control."

[0006] Existing inspection robots: Most of the robot technologies already in use are limited to data acquisition based on a single vision or thermal imaging sensor, lacking a multi-source information fusion and judgment mechanism, resulting in high false alarm and false negative rates. Structurally, existing robots often lack sufficient degrees of freedom of movement and cannot penetrate narrow spaces to perform omnidirectional scanning of cables; functionally, their inspection and identification processes are severely disconnected from emergency response processes such as firefighting, and the robots themselves cannot perform firefighting operations, failing to build an integrated intelligent closed-loop control system of "perception-decision-execution".

[0007] Therefore, there is an urgent need for an integrated intelligent equipment that can replace manual labor, perform comprehensive, accurate, and autonomous inspections in complex cable trench environments, and provide efficient on-site emergency response when a hazard is detected. Summary of the Invention

[0008] To address the problems of existing underground cable trench inspection methods, such as high reliance on manual labor, insufficient inspection coverage, low accuracy in risk identification, and lack of on-site emergency response capabilities, this invention aims to provide an underground cable trench safety inspection and emergency response robot and its inspection and risk identification method. In the narrow, low-light, and electromagnetically interference-prone environment of underground cable trenches, the robot achieves autonomous inspection and accurate identification of multiple risks. Through the collaborative design of structure and algorithms, it improves the perception and identification reliability of various types of safety hazards within the cable trench. Upon detecting anomalies or risks, it can automatically execute corresponding emergency response operations, thereby constructing an intelligent closed-loop control system of "inspection—identification—decision-execution—feedback," enhancing the safety and intelligence level of cable trench operation and maintenance.

[0009] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0010] A robot for safety inspection and emergency response of underground cable trenches includes:

[0011] The vehicle motion module is used to drive the robot to move along the channel in the cable trench;

[0012] The multi-degree-of-freedom reconnaissance and scanning mechanism includes a vertical lifting mechanism, a lateral insertion mechanism, and a rotating scanning mechanism. The three together form a composite motion chain of "vertical lifting - lateral insertion - rotating scanning", which is used to achieve three-dimensional coverage scanning at multiple heights, multiple lateral positions, and multiple angles in narrow cable trench spaces.

[0013] The multimodal risk detector module is an integrated sensing module, including a video image unit, a thermal infrared image unit, a catalytic combustion combustible gas sensing unit, an electrochemical sensing unit, and a UHF electromagnetic detection unit. During the inspection, it performs one-time multi-source raw data acquisition on detectable objects in the cable trench and sends the multi-source raw sensing data to the edge computing decision module.

[0014] The positioning and autonomous navigation planning module is used to generate inspection paths or walking instructions based on the positioning results, and drive the vehicle motion module to complete autonomous inspection walking in the cable trench according to the preset inspection task parameters during the inspection process.

[0015] The edge computing decision module is configured with a multi-source follower-judgment fusion algorithm to perform "master judgment - follower judgment - credibility enhancement" data fusion on multimodal sensor data, generate risk judgment results for the cable trench environment, and generate corresponding emergency response strategies and action instructions accordingly.

[0016] The emergency execution module includes an ultrasonic generator and a communication interface unit, which is used to perform corresponding emergency response operations based on the risk assessment results.

[0017] Preferably, the edge computing decision module is configured to perform feature extraction and anomaly screening on the received multi-source raw sensing data, perform multi-source fusion decision-making on data that exceeds a preset threshold or is determined to have abnormal features by image recognition, generate risk assessment results for the cable trench environment, and output corresponding emergency response decision results and control commands accordingly.

[0018] Preferably, the vertical lifting mechanism of the multi-degree-of-freedom reconnaissance scanning mechanism includes a lifting guide rail, a linear drive assembly, and a torque feedback structure for adjusting the reconnaissance height; the lateral insertion mechanism is located at the end of the lifting mechanism and includes a horizontal telescopic assembly, a guide slider, and an attitude holding structure for allowing the detection unit to penetrate deep into the cable trench; the rotating scanning mechanism includes a rotating drive assembly, an angle feedback unit, and a vibration-damping support structure for performing three-dimensional coverage scanning of the inside of the cable trench.

[0019] Preferably, the vertical lifting mechanism, the horizontal insertion mechanism, and the rotary scanning mechanism are configured to work together in a predetermined linkage sequence, so that the rotary scanning action is performed only when the horizontal insertion mechanism is in the extended state and after the vertical lifting mechanism has completed height positioning, thereby forming a staged, multi-degree-of-freedom linkage scanning method suitable for narrow cable trench spaces.

[0020] Preferably, the emergency execution module includes an ultrasonic generator and a communication interface unit; the ultrasonic generator is installed on the upper end of the multi-degree-of-freedom reconnaissance and scanning mechanism and adopts a directional emission structure, used to drive away small animals in the cable trench during the inspection process; the communication interface unit is configured to receive emergency response decision instructions generated by the edge computing decision module and send the risk analysis results and decision information to the monitoring center.

[0021] Preferably, the ultrasonic generator is coaxially and dynamically mounted with the multi-degree-of-freedom reconnaissance and scanning mechanism, so that the ultrasonic emission direction changes with the rotation direction of the scanning platform, thereby achieving directional removal of cable trenches in different directional areas.

[0022] This invention also provides a method for safety inspection and risk identification of underground cable trenches, which utilizes the aforementioned underground cable trench safety inspection and emergency response robot, and includes the following steps:

[0023] (1) Positioning and initial deployment: Based on the preset inspection task parameters, the robot enters the preset cable trench inspection area in an autonomous inspection mode, and the positioning and autonomous navigation planning module performs fusion positioning based on visual information, inertial measurement data and anti-metal interference laser point cloud to obtain the robot's real-time position information; the initial control state of the inspection task is established according to the real-time position information to provide a position reference for the motion control and multimodal data acquisition of the subsequent reconnaissance scanning mechanism.

[0024] (2) Multi-stage linkage inspection: The vertical lifting mechanism (2), the horizontal insertion mechanism (3), and the multi-degree-of-freedom reconnaissance and scanning mechanism (1) are controlled to perform the three-stage linkage action of "vertical lifting - horizontal insertion - rotation scanning" in sequence. This enables the multi-degree-of-freedom reconnaissance and scanning mechanism to complete the switching of scanning postures at different heights, different horizontal positions, and different angles in the cable trench, forming a multi-condition scanning motion process that covers the internal space of the cable trench;

[0025] (3) Multimodal data acquisition: During the multi-stage linkage inspection process, video images, infrared thermal images, electromagnetic discharge signals and combustible gas and toxic gas concentration sensing data are simultaneously acquired by multimodal sensors set on the multi-degree-of-freedom reconnaissance and scanning mechanism to form an original multimodal sensing dataset characterizing the operating environment of the cable trench.

[0026] (4) Risk fusion identification: The multimodal sensing data is fused and processed. The edge computing decision module is used to output the initial risk results of the main judgment sensor corresponding to different risk types according to the main judgment-secondary judgment fusion mechanism. The results are combined with the detection results of the secondary judgment sensor for weighted correction to generate risk discrimination results that characterize the operating status of the cable trench and its corresponding risk level.

[0027] (5) Disposal decision generation: Based on the risk identification results, the edge computing decision module is called to analyze different risk types and their levels according to the preset risk-disposal mapping rules, and automatically generate corresponding emergency disposal decision results or control parameters to indicate the selection, execution order and control method of subsequent emergency disposal actions;

[0028] (6) Emergency response operation: Based on the emergency response decision results or control parameters, control the emergency execution module to perform corresponding emergency response actions, including selectively performing animal removal, danger warning or information reporting operations based on the decision results;

[0029] (7) Information reporting and data synchronization: The inspection location information, risk assessment results, emergency response execution records and multimodal perception data obtained during the inspection process are uploaded to the monitoring center or back-end system through the communication module to realize remote synchronization and storage of inspection data and response results;

[0030] (8) Feedback and task closed loop: After the emergency response action is completed, the robot is controlled to update the current task status according to the execution status feedback of the emergency execution module and the task completion mark, and automatically executes the task end, exits the current cable trench area or switches to the next inspection task, so as to realize the closed loop control of the inspection and emergency response process.

[0031] The multi-source, follow-up judgment fusion algorithm is a three-layer classification data fusion framework of "master judgment - follow-up judgment - credibility enhancement". The master judgment unit generates an initial risk result based on the master judgment sensor, and the follow-up judgment unit performs weighted correction on it. Specifically, it includes the following steps:

[0032] S1. Extract features from the raw sensing data collected by the multimodal risk detector to form a multimodal feature set, which includes: combustible gas concentration features. Characteristics of toxic gas concentrations Infrared thermal imager heat source characteristics Video image features and UHF electromagnetic characteristics ;

[0033] S2, for different risk types Based on the corresponding primary judgment features Calculate the risk intensity of the primary judgment The calculation formula is:

[0034] ;

[0035] in For the Sigmoid function;

[0036] When the risk type is combustible gas risk hour, ;

[0037] When the risk type is toxic gas risk hour, ;

[0038] When the risk type is animal invasion risk hour, ;

[0039] When the risk type is cable high temperature risk hour, ;

[0040] When the risk type is cable damage risk hour, ;

[0041] S3. Invoke the sub-judgment confirmation mechanism based on the risk type and construct the sub-judgment consistency function. or time continuity reliability Specifically, it includes:

[0042] S3-1. Combustible gas assessment: Time continuity: For combustible gas risk... The reliability of time continuity is calculated using the following formula:

[0043] ;

[0044] In the formula: F represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type F at time τ. The risk intensity threshold corresponding to risk type F;

[0045] S3-2, Judgment of Toxic Gases: Time Continuity: For the risk of toxic gases The reliability of time continuity is calculated using the following formula:

[0046] ;

[0047] In the formula: T represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type T at time τ. The risk intensity threshold corresponding to risk type T;

[0048] S3-3, Animal Intrusion Judgment: Video Confirmation: Regarding the risk of animal intrusion. Using video image features as the criterion, the calculation formula is as follows:

[0049] ;

[0050] in: For the Sigmoid function; Indicating the risk of animal invasion At that time, the intensity of animal invasion features obtained from video image processing; ; This represents the sensitivity adjustment coefficient corresponding to the risk of animal invasion, used to adjust the degree of influence of the relative threshold deviation of video image features on the output result;

[0051] S3-4. Cable High Temperature Assessment: UHF + Video Dual Assessment: For cable high temperature risk The calculation formula for UHF follower and video follower is as follows:

[0052] ;

[0053] ;

[0054] And then they are fused together, the fusion formula is:

[0055] ;

[0056] In the formula: The cable represents the high-temperature risk characteristic value of the cable calculated based on the UHF signal at time t. This represents the characteristic value of cable high temperature risk calculated based on video images at time t. It is the comprehensive high temperature risk intensity obtained by merging the UHF judgment result and the video judgment result according to weights, which represents the judgment consistency function value of the cable high temperature risk at time t;

[0057] These are the UHF normalization parameters or judgment threshold parameters corresponding to the high temperature risk of cables, used to adjust the mapping scale of UHF feature values; The video displayed indicates the threshold for determining the high temperature risk of the cable. This represents the weighting coefficient of the UHF decision result in the fusion process; This represents the weighting coefficient of the video from the judgment result in the fusion process; This indicates the sensitivity adjustment coefficient of the video judgment corresponding to the high temperature risk of the cable;

[0058] S3-5. Cable Damage Assessment: For cable damage risk assessment... Confirmation is made using at least one of UHF, video, or infrared detection methods, and the calculation formula is as follows:

[0059] ;

[0060] In the formula: Indicates time The consistency function value of the judgment on cable damage risk is used to characterize the degree of comprehensive confirmation of cable damage risk based on UHF signal, video image and infrared thermal image judgment information; (·) The function definition represents the maximum value function, which is used to select the largest value among multiple judgment results as the confirmation result of cable damage risk; Indicates time The following is a feature value for judging cable damage risk based on UHF electromagnetic detection; Indicates time The following are the cable damage risk assessment feature values ​​obtained from video image analysis; Indicates time The following are the characteristic values ​​for judging cable damage risk obtained from infrared thermal imaging analysis;

[0061] S4. Primary Judge - Secondary Judge Fusion Correction: Generate fusion risk intensity according to the risk type using the following formula. For gas risks, the reliability of time continuity is used; for other risks, the consistency function of judgment is used.

[0062] ;

[0063] ;

[0064] ;

[0065] In the formula: time The fusion risk intensity of combustible gas risk is used to characterize the comprehensive risk level of combustible gas risk after primary judgment and time continuity correction; Indicates time The fusion risk intensity of toxic gas risk is used to characterize the comprehensive risk level of toxic gas risk after correction by the main judgment and time continuity. time The primary risk intensity for assessing the risk of flammable gas; time Reliability of the temporal continuity of flammable gas risk assessment; time The primary risk assessment intensity of the risk of exposure to toxic gases; time The reliability of the temporal continuity of the risk of exposure to toxic gases; Indicates time The fusion risk intensity of the k-th risk category is used to characterize the comprehensive risk level of the corresponding risk after the consistency correction between the primary and secondary judgments; A represents animal invasion risk; H represents cable high temperature risk; D represents cable damage risk.

[0066] S5, Credibility Enhancement / Stability Confirmation: When the intensity of fusion risk is within a continuous statistical window Internal satisfaction At that time, the corresponding risk is confirmed to be valid; The preset continuous statistical window corresponds to the number of sampling periods or the length of the time window; Indicates time The fusion risk intensity of the k-th type of risk is used to characterize the comprehensive risk level of this type of risk after modification by the primary and secondary judgments; This represents the stable confirmation threshold corresponding to the k-th type of risk, used to determine whether the intensity of the fusion risk meets the conditions for risk establishment.

[0067] S6. Output: Output the risk type, risk level and its corresponding fusion risk intensity, and send the risk assessment results to the monitoring center or use them to generate emergency response decision instructions.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] 1. Realize deep three-dimensional coverage scanning inside the cable trench: The present invention adopts a three-stage reconnaissance structure of "vertical lifting - horizontal insertion - rotation scanning", which can break through the narrow space limitation of the trench and enable the multi-modal risk detector module (4) to penetrate into the central axis of the cable trench to perform 360-degree full-domain scanning, effectively eliminating blind spots in the side walls and gaps, and realizing three-dimensional perception and accurate modeling of the internal space of the trench.

[0070] 2. Achieving intelligent risk identification through multi-source fusion: This invention adopts a multi-modal risk detection and "master judgment - slave judgment - credibility enhancement" fusion identification mechanism, which significantly reduces false alarms and false negatives of a single sensor and improves the accuracy and stability of multi-risk identification;

[0071] 3. It can comprehensively identify multiple types of risks such as animal invasion, cable damage, abnormal discharge, high temperature, and gas accumulation, and adapt to the complex operating environment of underground cable trenches;

[0072] 4. Achieve closed-loop control of the entire emergency response process: Integrate risk identification, decision generation and emergency response into a unified system to achieve intelligent closed-loop control of "inspection-identification-decision-execution-feedback", thereby improving the automation and intelligence level of cable trench operation and maintenance;

[0073] 5. Effectively reduces the frequency of manual inspections and lowers the safety risks for maintenance personnel entering high-risk environments. Attached Figure Description

[0074] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0075] Figure 1 This is an overall structural diagram of an underground cable trench safety inspection and emergency response robot according to an embodiment of the present invention;

[0076] Figure 2 This is a full-process logic diagram of robot autonomous inspection and emergency response in an embodiment of the present invention;

[0077] Figure 3 This is a schematic diagram of signal cross-validation for a multi-source follower-discrimination fusion algorithm according to an embodiment of the present invention;

[0078] In the diagram: 1. Multi-degree-of-freedom reconnaissance and scanning mechanism; 2. Vertical lifting mechanism; 3. Lateral insertion mechanism; 4. Multimodal risk detector module; 5. Edge computing decision module; 6. Ultrasonic generator. Detailed Implementation

[0079] To make the features and advantages of this patent more apparent and understandable, specific embodiments are provided below for detailed explanation:

[0080] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0081] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0082] Example 1:

[0083] like Figure 1 As shown, this embodiment provides a robot for safety inspection and emergency response of underground cable trenches, including a vehicle motion module, a multi-degree-of-freedom reconnaissance and scanning mechanism 1, a multi-modal risk detector module 4, an edge computing decision module 5, a positioning and autonomous navigation planning module, and an emergency execution module.

[0084] The edge computing decision module 5 is configured to execute a multi-source follower-judgment fusion algorithm. The multi-source follower-judgment fusion algorithm cross-checks and enhances the credibility of video images, infrared thermal images, electromagnetic discharge signals and the concentrations of combustible and toxic gases collected by the multi-modal risk detector module 4, generates a cable trench risk type determination result, and generates corresponding emergency response strategies and action instructions accordingly.

[0085] The emergency execution module automatically executes a series of actions, including but not limited to ultrasonic expulsion and reporting decision information to the monitoring center, based on the emergency response strategy and action instructions.

[0086] The vehicle motion module, multi-degree-of-freedom reconnaissance and scanning mechanism 1, multimodal risk detector module 4, edge computing decision module 5, positioning and autonomous navigation planning module, and emergency execution module work together to make the robot a closed-loop emergency control system of "inspection-identification-decision-execution-feedback".

[0087] The vehicle motion module provides a highly mobile, highly passable, and highly stable mobile platform for the robot in narrow, uneven, and slippery underground cable trench environments. The vehicle motion module adopts a large four-wheeled run-flat tire drive chassis, which is well known to those skilled in the art. Each wheel set is equipped with existing wide-section anti-skid rubber tires with a V-shaped tread pattern to enhance drainage and grip performance on wet cable trench surfaces. The tire cavity is equipped with an existing air pressure sensor connected to a micro air pump, which can monitor tire pressure changes in real time and perform dynamic inflation and deflation adjustments to achieve adaptive tire pressure regulation. Each wheelset uses a waterproof hub motor for direct drive and is equipped with an independent full-time differential lock control system. This system automatically distributes drive torque by real-time monitoring and calculation of wheel speed, torque, and slip ratio. When single-wheel slippage or insufficient traction is detected, the differential lock closed-loop control is immediately triggered to increase the traction of the drive wheels, enabling automatic escaping and stable forward movement. The four wheelsets are connected to the active hydraulic suspension system via an independent servo speed control structure. This structure allows independent control of the speed and steering angle of each wheelset, enabling omnidirectional movement and crab-like lateral movement. The active hydraulic suspension system automatically adjusts ground clearance based on terrain sensor feedback, maintaining vehicle stability and center of gravity balance on uneven, slippery, or obstacle-filled trench bottoms. This enables high mobility and stable operation of the robot in narrow, wet, sloping, and gravelly cable trench environments, significantly improving the robot's mobility and mission reliability in complex conditions. It also allows for omnidirectional steering, crab-like lateral movement, and dynamic ground clearance adjustment in narrow, uneven trenches.

[0088] In this embodiment, the large four-wheel explosion-proof tire drive chassis, wide-section anti-skid rubber tires, air pressure sensor, micro air pump, waterproof hub motor, full-time differential lock control system, servo speed regulation structure, and active hydraulic suspension system all adopt existing products or structures known to those skilled in the art. The connection or control method between them also adopts existing connection or control method known to those skilled in the art. The vehicle motion module and the edge computing decision module 5 are connected using existing control methods known to those skilled in the art.

[0089] The multi-degree-of-freedom reconnaissance and scanning mechanism forms a composite motion chain of "vertical lifting-lateral insertion-rotation scanning" through a vertical lifting mechanism 2, a lateral insertion mechanism 3, and a multi-degree-of-freedom rotary scanning platform. The vertical lifting mechanism is driven by a combination of multi-stage sleeve-type telescopic columns and electric push rods, and has displacement feedback control and speed adaptive adjustment functions, enabling precise lifting and positioning and anti-collision control at different groove heights. The lateral insertion mechanism 3 is a multi-stage nested structure, which achieves lateral insertion force path control through ball screws and servo drives to achieve graded extension. The multi-degree-of-freedom rotary scanning platform is directly driven by a servo motor through a harmonic reducer. A magnetoelectric absolute encoder is installed coaxially on the main shaft, and an infrared distance sensor array is arranged on the edge of the main shaft to detect the rotation safety threshold. When the obstacle threshold is triggered, a real-time braking command is executed, enabling 360-degree full-area rotation scanning.

[0090] The vertical lifting mechanism 2, the horizontal insertion mechanism 3, the multi-degree-of-freedom rotating scanning platform, the multi-stage sleeve telescopic column, the electric push rod, the servo motor, the harmonic reducer, the spindle, the magnetoelectric absolute encoder, the infrared distance sensor array, and the edge computing decision module 5 in this embodiment all adopt existing products or structures known to those skilled in the art, and the connection or control methods between them also adopt existing connection or control methods known to those skilled in the art.

[0091] The multimodal risk detector module 4 is an integrated heterogeneous sensor unit, internally integrating a video image unit, a thermal infrared image unit, a catalytic combustion combustible gas sensing unit, an electrochemical sensing unit, and a UHF electromagnetic detection unit. These sensing units collaboratively acquire video images, infrared thermal images, electromagnetic discharge signals, and the concentrations of combustible and toxic gases within the cable trench during the rotation of the multi-degree-of-freedom reconnaissance and scanning mechanism. The synchronous acquisition of different types of signals in time and space enables the comparison of thermal anomalies, chemical changes, and deformation characteristics on the same cross-section, providing reliable multi-source data support for subsequent intelligent risk identification.

[0092] The video image unit, thermal infrared image unit, catalytic combustion combustible gas sensing unit, electrochemical sensing unit, and UHF electromagnetic detection unit in this embodiment all adopt existing products or structures known to those skilled in the art. Their connection with the edge computing decision module 5 also adopts existing connection methods known to those skilled in the art. They are installed on the vehicle motion module in a manner known to those skilled in the art to realize the corresponding functions known to those skilled in the art.

[0093] In this embodiment, the robot is equipped with a positioning and autonomous navigation planning module, which provides walking control commands to the vehicle motion module during the inspection of underground cable trenches.

[0094] Specifically, at the start of the inspection task, the localization and autonomous navigation planning module generates corresponding inspection walking instructions based on the robot's current position information and preset inspection task parameters, driving the robot to complete autonomous walking along the cable trench channel. The inspection walking process does not require real-time manual control; the robot can continuously move forward in the trench according to the preset inspection task, and enter the next inspection stage after reaching the designated inspection position or completing the current inspection task.

[0095] The positioning and autonomous navigation planning module uses existing products or structures well-known to those skilled in the art, solely to support the robot's autonomous inspection and movement within cable trenches, without involving specific path planning algorithms or navigation strategy limitations. The emergency execution module includes an ultrasonic generator 6, a communication interface unit, and an audible and visual alarm. The ultrasonic generator 6 and the audible and visual alarm are installed on the upper part of the multi-degree-of-freedom rotating scanning platform using existing methods well-known to those skilled in the art. Their power supply and control cables are routed through the internal wiring of the multi-degree-of-freedom rotating scanning platform, via the cable management channels inside the lateral insertion mechanism 3 and the vertical lifting mechanism 2, ultimately connecting to the main control system and power supply inside the robot body. This installation method ensures that the ultrasonic generator 6 rotates with the scanning platform, achieving directional emission and accurately covering the specific area pointed to by the scanning mechanism, used to drive away small animals. The communication interface unit is used to realize information interaction between the robot and the monitoring center or backend system. Specifically, the communication interface unit is configured to receive emergency response decision commands from the edge computing decision module and, during the execution of emergency response operations, collect and record the execution status of the emergency response. The communication interface unit is further used to send the risk assessment results, emergency response instructions, response execution status, and response completion information generated during the inspection process to the monitoring center or backend system, so as to achieve remote synchronization of inspection and emergency response information. After the emergency response operation is completed, the communication interface unit returns a response completion feedback signal to the edge computing decision module to indicate that the current emergency response process has ended, thereby supporting the status update and process switching of the robot's subsequent inspection tasks.

[0096] The ultrasonic generator 6, the audible and visual alarm, the communication interface unit, the monitoring center, the main control system and power supply inside the robot body, and the edge computing decision module 5 in this embodiment are existing products or structures well known to those skilled in the art, and the connection or control methods between them are existing connection and control methods well known to those skilled in the art.

[0097] Example 2:

[0098] This embodiment provides a method for autonomous inspection and emergency response of underground cable trenches using the aforementioned robot, and its full process logic diagram is as follows: Figure 2 As shown, the specific steps include:

[0099] (1) Positioning and initial deployment: Based on the preset inspection task parameters, the robot enters the preset cable trench inspection area in an autonomous inspection mode, and the positioning and autonomous navigation planning module performs fusion positioning based on visual information and inertial measurement data to obtain the robot's real-time position information; the initial control state of the inspection task is established according to the real-time position information to provide a position reference for the motion control and multimodal data acquisition of the subsequent reconnaissance scanning mechanism.

[0100] (2) Multi-stage linkage inspection: control the vertical lifting mechanism 2, the horizontal insertion mechanism 3 and the multi-degree-of-freedom reconnaissance scanning mechanism (1) to perform the three-stage linkage action of "vertical lifting - horizontal insertion - rotation scanning" in sequence. This enables the multi-degree-of-freedom reconnaissance scanning mechanism to complete the scanning posture switching of different heights, different horizontal positions and different angles in the cable trench, forming a multi-condition scanning motion process covering the internal space of the cable trench;

[0101] (3) Multimodal data acquisition: During the multi-stage linkage scanning motion, video images, infrared thermal images, electromagnetic discharge signals and combustible gas and toxic gas concentration sensing data are simultaneously acquired by multimodal sensors set on the multi-degree-of-freedom reconnaissance scanning mechanism to form an original multimodal sensing dataset characterizing the operating environment of the cable trench.

[0102] (4) Risk fusion identification: The multimodal sensing data is fused and processed. The edge computing decision module is used to output the initial risk results of the main judgment sensor corresponding to different risk types according to the main judgment-secondary judgment fusion mechanism. The results are combined with the detection results of the secondary judgment sensor for weighted correction to generate risk discrimination results that characterize the operating status of the cable trench and its corresponding risk level.

[0103] (5) Disposal decision generation: Based on the risk assessment results, the edge computing decision module is called to analyze different risk types and their levels according to the preset risk-disposal mapping rules, and automatically generate corresponding emergency disposal decision results or control parameters to indicate the selection, execution order and control method of subsequent emergency disposal actions.

[0104] (6) Emergency response operation: Based on the emergency response decision results or control parameters, control the emergency execution module to perform the corresponding emergency response actions, including selectively performing animal removal, danger warning or information reporting operations based on the decision results. Danger warning is completed through an audible and visual alarm.

[0105] (7) Information reporting and data synchronization: The inspection location information, risk assessment results, emergency response execution records and multimodal perception data obtained during the inspection process are uploaded to the monitoring center or back-end system through the communication module to realize remote synchronization and storage of inspection data and response results.

[0106] (8) Feedback and Task Closed Loop: After completing the emergency response action, based on the execution status feedback of the emergency execution module and the task completion flag, the robot is controlled to update the current task status and automatically execute the task completion, exit the current cable trench area, or switch to the next inspection task, thereby realizing closed-loop control of the inspection and emergency response process. The robot in this embodiment has the ability to autonomously navigate and locate in the narrow, low-light, and strong electromagnetic interference underground cable trench environment. It can achieve three-dimensional full-coverage scanning of the trench interior space through the composite motion of "vertical lifting-lateral insertion-rotation scanning". It can realize comprehensive risk identification of animal invasion risk, cable structure damage risk, flammable gas accumulation, toxic gas accumulation, and high temperature risk type, and perform corresponding emergency response actions to form a complete "inspection-identification-decision-execution-feedback" intelligent closed-loop control.

[0107] Example 3:

[0108] In this embodiment of the invention, the multi-source follower fusion algorithm adopts a three-level data fusion framework of "master judgment - follower judgment - credibility enhancement" to reliably identify various risks such as high temperature of cable, cable damage, accumulation of combustible gas, accumulation of toxic gas and animal invasion under complex working conditions of cable trench (low light, strong electromagnetic interference, humid and narrow space, etc.).

[0109] The specific identification is as follows:

[0110] S1. Extract features from the raw sensing data collected by the multimodal risk detector to form a multimodal feature set, which includes: combustible gas concentration features. Characteristics of toxic gas concentrations Infrared thermal imager heat source characteristics Video image features and UHF electromagnetic characteristics .

[0111] S2, for different risk types Based on the corresponding primary judgment features Calculate the risk intensity of the primary judgment The calculation formula is:

[0112] ;

[0113] in For the Sigmoid function;

[0114] When the risk type is combustible gas risk hour, ;

[0115] When the risk type is toxic gas risk hour, ;

[0116] When the risk type is animal invasion risk hour, ;

[0117] When the risk type is cable high temperature risk hour, ;

[0118] When the risk type is cable damage risk hour, .

[0119] S3. Invoke the decision confirmation mechanism based on the risk type and construct the decision consistency function. or time continuity reliability Specifically, it includes:

[0120] S3-1. Combustible gas assessment: Time continuity: For combustible gas risk... The reliability of time continuity is calculated using the following formula:

[0121] ;

[0122] In the formula: F represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type F at time τ. The risk intensity threshold corresponding to risk type F;

[0123] S3-2, Judgment of Toxic Gases: Time Continuity: For the risk of toxic gases The reliability of time continuity is calculated using the following formula:

[0124]

[0125] In the formula: T represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type T at time τ. The risk intensity threshold corresponding to risk type T;

[0126] S3-3, Animal Intrusion Judgment: Video Confirmation: Regarding the risk of animal intrusion. Using video image features as the criterion, the calculation formula is as follows:

[0127] ;

[0128] in: For the Sigmoid function; Indicating the risk of animal invasion At that time, the intensity of animal invasion features obtained from video image processing; ; This represents the sensitivity adjustment coefficient corresponding to the risk of animal invasion, used to adjust the degree of influence of the relative threshold deviation of video image features on the output result;

[0129] S3-4. Cable High Temperature Assessment: UHF + Video Dual Assessment: For cable high temperature risk The calculation formula for UHF follower and video follower is as follows:

[0130] ;

[0131] ;

[0132] And then they are fused together, the fusion formula is:

[0133] ;

[0134] In the formula: The cable represents the high-temperature risk characteristic value of the cable calculated based on the UHF signal at time t. This represents the characteristic value of cable high temperature risk calculated based on video images at time t. It is the comprehensive high temperature risk intensity obtained by merging the UHF judgment result and the video judgment result according to weights, which represents the judgment consistency function value of the cable high temperature risk at time t;

[0135] These are the UHF normalization parameters or judgment threshold parameters corresponding to the high temperature risk of cables, used to adjust the mapping scale of UHF feature values; The video displayed indicates the threshold for determining the high temperature risk of the cable. This represents the weighting coefficient of the UHF decision result in the fusion process; This represents the weighting coefficient of the video from the judgment result in the fusion process; This indicates the sensitivity adjustment coefficient of the video judgment corresponding to the high temperature risk of the cable;

[0136] S3-5. Cable Damage Assessment: For cable damage risk assessment... Confirmation is made using at least one of UHF, video, or infrared detection methods, and the calculation formula is as follows:

[0137] ;

[0138] In the formula: Indicates time The consistency function value of the judgment on cable damage risk is used to characterize the degree of comprehensive confirmation of cable damage risk based on UHF signal, video image and infrared thermal image judgment information; ( The function definition represents the maximum value function, which is used to select the largest value from multiple judgment results as the confirmation result of cable damage risk; Indicates time The following is a feature value for judging cable damage risk based on UHF electromagnetic detection; Indicates time The following are the cable damage risk assessment feature values ​​obtained from video image analysis; Indicates time The following are the characteristic values ​​for judging cable damage risk obtained from infrared thermal imaging analysis;

[0139] S4. Primary Judge - Secondary Judge Fusion Correction: Generate fusion risk intensity according to the risk type using the following formula. For gas risks, the reliability of time continuity is used; for other risks, the consistency function of judgment is used.

[0140] ;

[0141] ;

[0142] ;

[0143] In the formula: time The fusion risk intensity of combustible gas risk is used to characterize the comprehensive risk level of combustible gas risk after primary judgment and time continuity correction; Indicates time The fusion risk intensity of toxic gas risk is used to characterize the comprehensive risk level of toxic gas risk after correction by the main judgment and time continuity. time The primary risk intensity for assessing the risk of flammable gas; time Reliability of the temporal continuity of flammable gas risk assessment; time The primary risk assessment intensity of the risk of exposure to toxic gases; time The reliability of the temporal continuity of the risk of exposure to toxic gases; Indicates time The fusion risk intensity of the k-th risk category is used to characterize the comprehensive risk level of the corresponding risk after the consistency correction between the primary and secondary judgments; A represents animal invasion risk; H represents cable high temperature risk; D represents cable damage risk.

[0144] S5, Credibility Enhancement / Stability Confirmation: When the intensity of fusion risk is within a continuous statistical window Internal satisfaction At that time, the corresponding risk is confirmed to be valid; The preset continuous statistical window corresponds to the number of sampling periods or the length of the time window; Indicates time The fusion risk intensity of the k-th type of risk is used to characterize the comprehensive risk level of this type of risk after modification by the primary and secondary judgments; This represents the stable confirmation threshold corresponding to the k-th type of risk, used to determine whether the intensity of the fusion risk meets the conditions for risk establishment.

[0145] S6. Output: Output the risk type, risk level and its corresponding fusion risk intensity, and send the risk assessment results to the monitoring center or use them to generate emergency response decision instructions.

[0146] The thresholds and time window parameters involved in the above algorithm are pre-calibrated according to the actual application scenario. In this embodiment, a continuous statistical window is preset, corresponding to the number of sampling periods or the length of the time window. Use 5 to 10 sampling periods, set the gas risk primary judgment threshold to 10% to 20% of the corresponding sensor's full scale, and set the fusion risk intensity confirmation threshold to 0.6 to 0.8. Specific values ​​can be determined through ROC curve optimization using multiple sets of on-site measured data.

[0147] In this invention, to improve the stability and reliability of risk identification results, the edge computing decision module introduces a time continuity detection mechanism for certain risk types. This time continuity detection refers to: within a preset time window, multiple sampling and statistical analyses of the detection results for the same risk are performed; the corresponding risk is confirmed only when the risk determination result meets the conditions for establishment in multiple consecutive or majority sampling periods. This method effectively suppresses misjudgments caused by instantaneous sensor fluctuations, environmental interference, or occasional anomalies, preventing a single detection result from directly triggering risk confirmation, thereby improving the stability of risk identification in complex underground cable trench environments.

[0148] In this invention, the various threshold parameters involved are preset or manually configured according to the actual cable material, operating environment and application scenario. The thresholds can be adjusted according to different cable types and field conditions to adapt to the inspection needs of different cable trench environments.

[0149] This invention is not limited to the identification of the specific risk types mentioned above. Any risk or abnormal state that is identified and confirmed based on the perception data collected by the multimodal risk detector and in accordance with the "primary judgment - secondary judgment - credibility enhancement" fusion judgment process described in this invention is within the scope of application of the technical solution of this invention.

[0150] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A robot for safety inspection and emergency response of underground cable trenches, characterized in that, include: The vehicle motion module is used to drive the robot to move along the channel in the cable trench; The multi-degree-of-freedom reconnaissance scanning mechanism (1) includes a vertical lifting mechanism (2), a lateral insertion mechanism (3) and a rotating scanning mechanism. The three constitute a composite motion chain of "vertical lifting - lateral insertion - rotating scanning" to achieve three-dimensional coverage scanning at multiple heights, multiple lateral positions and multiple angles in a narrow cable trench space. The multimodal risk detector module (4) is an integrated sensing module, including a video image unit, a thermal infrared image unit, a catalytic combustion combustible gas sensing unit, an electrochemical sensing unit, and a UHF electromagnetic detection unit. During the inspection, it performs one-time multi-source raw data collection on detectable objects in the cable trench and sends the multi-source raw sensing data to the edge computing decision module. The positioning and autonomous navigation planning module is used to generate inspection paths or walking instructions based on the positioning results, and drive the vehicle motion module to complete autonomous inspection walking in the cable trench according to the preset inspection task parameters during the inspection process. The edge computing decision module (5) is configured with a multi-source follower-judgment fusion algorithm to perform "master judgment - follower judgment - credibility enhancement" data fusion on multimodal sensor data, and generate risk judgment results for cable trench environment, and generate corresponding emergency response strategies and action instructions accordingly; The emergency execution module includes an ultrasonic generator (6) and a communication interface unit, which is used to perform corresponding emergency response operations based on the risk assessment results.

2. The underground cable trench safety inspection and emergency response robot according to claim 1, characterized in that, The edge computing decision module (5) is configured to perform feature extraction and anomaly screening on the received multi-source raw sensing data, and to perform multi-source fusion decision-making on data that exceeds a preset threshold or is determined to be abnormal by image recognition, so as to generate risk judgment results of the cable trench environment, and output corresponding emergency response decision results and control instructions accordingly.

3. The underground cable trench safety inspection and emergency response robot according to claim 1, characterized in that, The vertical lifting mechanism (2) of the multi-degree-of-freedom reconnaissance scanning mechanism (1) includes a lifting guide rail, a linear drive assembly and a torque feedback structure, used to adjust the reconnaissance height; the lateral insertion mechanism (3) is set at the end of the lifting mechanism, including a horizontal telescopic assembly, a guide slider and an attitude holding structure, used to allow the detection unit to penetrate into the cable trench; the rotating scanning mechanism includes a rotating drive assembly, an angle feedback unit and a vibration-damping support structure, used to perform three-dimensional coverage scanning of the inside of the cable trench.

4. The underground cable trench safety inspection and emergency response robot according to claim 3, characterized in that, The vertical lifting mechanism, the horizontal insertion mechanism, and the rotary scanning mechanism are configured to work together in a predetermined linkage sequence, so that the rotary scanning action is performed only when the horizontal insertion mechanism is in the extended state and after the vertical lifting mechanism has completed height positioning, thereby forming a staged, multi-degree-of-freedom linkage scanning method suitable for narrow cable trench spaces.

5. The underground cable trench safety inspection and emergency response robot according to claim 1, characterized in that, The emergency execution module includes an ultrasonic generator (6) and a communication interface unit; the ultrasonic generator is installed on the upper end of the multi-degree-of-freedom reconnaissance scanning mechanism (1) and adopts a directional emission structure to drive away small animals in the cable trench during the inspection process; the communication interface unit is configured to receive emergency response decision instructions generated by the edge computing decision module and send the risk analysis results and decision information to the monitoring center.

6. The underground cable trench safety inspection and emergency response robot according to claim 5, characterized in that, The ultrasonic generator (6) is coaxially and dynamically mounted with the multi-degree-of-freedom reconnaissance and scanning mechanism (1), so that the ultrasonic emission direction changes with the rotation direction of the scanning platform, thereby realizing the directional removal of different areas of the cable trench.

7. A method for safety inspection and risk identification of underground cable trenches, employing a robot for safety inspection and emergency response of underground cable trenches as described in any one of claims 1-6, characterized in that, Includes the following steps: (1) Positioning and initial deployment: Based on the preset inspection task parameters, the robot enters the preset cable trench inspection area in an autonomous inspection mode, and the positioning and autonomous navigation planning module performs fusion positioning based on visual information, inertial measurement data and anti-metal interference laser point cloud to obtain the robot's real-time position information; the initial control state of the inspection task is established according to the real-time position information to provide a position reference for the motion control and multimodal data acquisition of the subsequent reconnaissance scanning mechanism; (2) Multi-stage linkage inspection: The vertical lifting mechanism (2), the horizontal insertion mechanism (3) and the multi-degree-of-freedom reconnaissance scanning mechanism (1) are controlled to perform the three-stage linkage action of "vertical lifting - horizontal insertion - rotation scanning" in sequence. The multi-degree-of-freedom reconnaissance scanning mechanism can complete the scanning posture switching of different heights, different horizontal positions and different angles in the cable trench, forming a multi-condition scanning motion process covering the internal space of the cable trench; (3) Multimodal data acquisition: During the multi-stage linkage inspection process, video images, infrared thermal images, electromagnetic discharge signals and combustible gas and toxic gas concentration sensing data are simultaneously acquired by multimodal sensors set on the multi-degree-of-freedom reconnaissance and scanning mechanism to form an original multimodal sensing dataset characterizing the operating environment of the cable trench. (4) Risk fusion identification: The multimodal sensing data is fused and processed. The edge computing decision module is used to output the initial risk results of the main judgment sensor corresponding to different risk types according to the main judgment-secondary judgment fusion mechanism. The results are combined with the detection results of the secondary judgment sensor for weighted correction to generate risk discrimination results that characterize the operating status of the cable trench and its corresponding risk level. (5) Disposal decision generation: Based on the risk identification results, the edge computing decision module is called to analyze different risk types and their levels according to the preset risk-disposal mapping rules, and automatically generate corresponding emergency disposal decision results or control parameters to indicate the selection, execution order and control method of subsequent emergency disposal actions; (6) Emergency response operation: Based on the emergency response decision results or control parameters, control the emergency execution module to perform corresponding emergency response actions, including selectively performing animal removal, danger warning or information reporting operations based on the decision results; (7) Information reporting and data synchronization: The inspection location information, risk assessment results, emergency response execution records and multimodal perception data obtained during the inspection process are uploaded to the monitoring center or back-end system through the communication module to realize remote synchronization and storage of inspection data and response results; (8) Feedback and task closed loop: After the emergency response action is completed, the robot is controlled to update the current task status according to the execution status feedback of the emergency execution module and the task completion mark, and automatically executes the task end, exits the current cable trench area or switches to the next inspection task, so as to realize the closed loop control of the inspection and emergency response process.

8. The method for safety inspection and risk identification of underground cable trenches according to claim 7, characterized in that, The multi-source adjudication fusion algorithm is a three-layer classification data fusion framework of "master judgment - slave judgment - credibility enhancement". The master judgment unit generates an initial risk result based on the master judgment sensor, and the slave judgment unit performs weighted correction on it. Specifically, it includes the following steps: S1. Extract features from the raw sensing data collected by the multimodal risk detector to form a multimodal feature set, which includes: combustible gas concentration features. Characteristics of toxic gas concentrations Infrared thermal imager heat source characteristics Video image features and UHF electromagnetic characteristics ; S2, for different risk types Based on the corresponding primary judgment features Calculate the risk intensity of the primary judgment The calculation formula is: ； in For the Sigmoid function; When the risk type is combustible gas risk hour, ; When the risk type is toxic gas risk hour, ; When the risk type is animal invasion risk hour, ; When the risk type is cable high temperature risk hour, ; When the risk type is cable breakage risk hour, ; S3. Invoke the sub-judgment confirmation mechanism based on the risk type and construct the sub-judgment consistency function. or time continuity reliability Specifically, it includes: S3-1. Combustible gas assessment: Time continuity: For combustible gas risk... The reliability of time continuity is calculated using the following formula: ； In the formula: F represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type F at time τ. The risk intensity threshold corresponding to risk type F; S3-2, Judgment of Toxic Gases: Time Continuity: For the risk of toxic gases The reliability of time continuity is calculated using the following formula: ； In the formula: T represents the combustible gas risk type identifier, used to characterize the target risk category for the current continuous assessment; t represents the current time or the current sampling period; τ represents the historical sampling time within the statistical window; This indicates the number of sampling periods used for continuous statistics, or the preset time window; I(·) represents an indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. This indicates the intensity of the primary risk corresponding to risk type T at time τ. The risk intensity threshold corresponding to risk type T; S3-3, Animal Intrusion Judgment: Video Confirmation: Regarding the risk of animal intrusion. Using video image features as the criterion, the calculation formula is as follows: ； in: For the Sigmoid function; Indicating the risk of animal invasion At that time, the intensity of animal invasion features obtained from video image processing; ; This represents the sensitivity adjustment coefficient corresponding to the risk of animal invasion, used to adjust the degree of influence of the relative threshold deviation of video image features on the output result; S3-4. Cable High Temperature Assessment: UHF + Video Dual Assessment: For cable high temperature risk The formula for calculating the UHF follower and video follower is as follows: ； ； And then they are fused together, the fusion formula is: ； In the formula: The cable represents the high-temperature risk characteristic value of the cable calculated based on the UHF signal at time t. This represents the characteristic value of cable high temperature risk calculated based on video images at time t. It is the comprehensive high temperature risk intensity obtained by merging the UHF judgment result and the video judgment result according to weights, which represents the judgment consistency function value of the cable high temperature risk at time t; These are the UHF normalization parameters or judgment threshold parameters corresponding to the high temperature risk of cables, used to adjust the mapping scale of UHF feature values; The video displayed indicates the threshold for determining the high temperature risk of the cable. This represents the weighting coefficient of the UHF decision result in the fusion process; This represents the weighting coefficient of the video from the judgment result in the fusion process; This indicates the sensitivity adjustment coefficient of the video judgment corresponding to the high temperature risk of the cable; S3-5. Cable Damage Assessment: For cable damage risk assessment... Confirmation is made using at least one of UHF, video, or infrared detection methods, and the calculation formula is as follows: ； In the formula: Indicates time The consistency function value of the judgment on cable damage risk is used to characterize the degree of comprehensive confirmation of cable damage risk based on UHF signal, video image and infrared thermal image judgment information; ( The function definition represents the maximum value function, which is used to select the largest value from multiple judgment results as the confirmation result of cable damage risk; Indicates time The following is a feature value for judging cable damage risk based on UHF electromagnetic detection; Indicates time The following are the cable damage risk assessment feature values ​​obtained from video image analysis; Indicates time The following are the characteristic values ​​for judging cable damage risk obtained from infrared thermal imaging analysis; S4. Primary Judge - Secondary Judge Fusion Correction: Generate fusion risk intensity according to the risk type using the following formula. For gas risks, the reliability of time continuity is used; for other risks, the consistency function of judgment is used. ； ； ； In the formula: time The fusion risk intensity of combustible gas risk is used to characterize the comprehensive risk level of combustible gas risk after primary judgment and time continuity correction; Indicates time The fusion risk intensity of toxic gas risk is used to characterize the comprehensive risk level of toxic gas risk after correction by the main judgment and time continuity. time The primary risk intensity for assessing the risk of flammable gases; time Reliability of the temporal continuity of flammable gas risk assessment; time The primary risk assessment intensity of the risk of exposure to toxic gases; time The reliability of the temporal continuity of the risk of exposure to toxic gases; Indicates time The fusion risk intensity of the k-th risk category is used to characterize the comprehensive risk level of the corresponding risk after the consistency correction between the primary and secondary judgments; A represents animal invasion risk; H represents cable high temperature risk; D represents cable damage risk. S5, Credibility Enhancement / Stability Confirmation: When the intensity of fusion risk is within a continuous statistical window Internal satisfaction At that time, the corresponding risk is confirmed to be valid; The preset continuous statistical window corresponds to the number of sampling periods or the length of the time window; Indicates time The fusion risk intensity of the k-th type of risk is used to characterize the comprehensive risk level of this type of risk after modification by the primary and secondary judgments; This represents the stable confirmation threshold corresponding to the k-th type of risk, used to determine whether the intensity of the fusion risk meets the conditions for risk establishment. S6. Output: Output the risk type, risk level and its corresponding fusion risk intensity, and send the risk assessment results to the monitoring center or use them to generate emergency response decision instructions.

Images