Multifunctional inspection device detection system based on distribution network mobile operation terminal

Abstract

The invention provides a multifunctional inspection device detection system based on a distribution network mobile operation terminal, and relates to the technical field of electric power detection, and the multifunctional inspection device detection system comprises a portable main terminal of a dynamic collaborative architecture, a wearable terminal for augmented reality interaction and a distributed heterogeneous sensing cluster, which form a closed-loop data chain through an electric power dedicated low-delay wireless communication network; the portable main terminal is integrated with a heterogeneous computing unit with adaptive computing power distribution, can perform real-time fusion analysis on multi-dimensional sensing data, and outputs a visual result containing fault location and confidence; the wearable terminal superposes fault information to a real scene in a three-dimensional marking form through a virtual-real fusion positioning technology, and supports eye movement and voice collaborative interaction; multi-module collaborative acquisition of the distributed heterogeneous sensing cluster is combined with a nanosecond timestamp synchronization mechanism, so that the limitation of traditional single parameter detection is broken through, rich and synchronous basic data is provided for subsequent analysis, and the comprehensiveness of routing inspection is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power detection, and in particular to a multifunctional inspection device detection system based on a distribution network mobile operation terminal. Background Art

[0002] With the rapid development of the power industry, the scale of distribution networks continues to expand, and their safe and stable operation is crucial to ensuring power supply. On-site inspections of distribution network equipment are a key measure to ensure normal equipment operation. Currently, it is necessary to test multiple items such as temperature, partial discharge, and abnormal vibration of distribution network equipment, and conduct a comprehensive assessment of the equipment status. However, existing technologies have numerous drawbacks. For one thing, the inspection equipment currently available on the market is limited in functionality, making it impossible to comprehensively and systematically complete all tests on a single instrument. This requires inspectors to carry multiple instruments during their work, making operations cumbersome and inefficient. Furthermore, after each test, instruments must be manually replaced, and the results must be manually recorded. This is not only time-consuming and labor-intensive, but also prone to errors. Furthermore, to ensure the safety of inspectors, constant attention must be paid to parameters such as the electric field and Sf6 concentration. However, existing inspection methods struggle to effectively address these safety factors, posing certain safety risks and making it difficult to adapt to the transition to intelligent operation and inspection models. Summary of the Invention

[0003] In view of the deficiencies in the prior art, the present invention provides a multifunctional inspection device detection system based on a distribution network mobile operation terminal.

[0004] A multifunctional inspection device detection system based on a distribution network mobile operation terminal includes a portable main terminal with a dynamic collaborative architecture, a wearable terminal with augmented reality interaction, and a distributed heterogeneous sensor cluster. The three form a closed-loop data link through a dedicated low-latency wireless communication network for power generation. The portable master terminal integrates a heterogeneous computing unit with adaptive computing power allocation, which can perform real-time fusion analysis of multi-dimensional sensor data and output visual results including fault location and confidence level; The wearable terminal uses virtual-reality fusion positioning technology to overlay fault information on the real scene in the form of three-dimensional markers, supporting collaborative interaction between eye movement and voice; The distributed heterogeneous sensing cluster includes a photothermal coaxial acquisition module, an acoustic-vibration synchronous detection module, and a field-gas joint measurement module. Each module achieves nanosecond-level synchronization through timestamp alignment and has local feature extraction and data compression capabilities.

[0005] Preferably, the photothermal coaxial acquisition module includes a coaxially arranged visible light component and an infrared component, and realizes exposure timing synchronization through a programmable logic unit, and satisfies:

[0006] in, is the corrected exposure time, is the equipment type coefficient; in this embodiment, the overhead line k1=1.2, the switchgear k1=0.8; is the base exposure time, is the standard light intensity, is the real-time light intensity; it should be noted that, in this embodiment, the reference exposure time is (100ms) and the standard light intensity is (1000lux); The temperature field data output by the infrared component and the image data output by the visible light component are spatially aligned through ORB feature point matching, and the matching error is ≤5% of the field of view angle. The photothermal coaxial acquisition module transmits the processed photothermal data to the portable main terminal, providing basic data for its fusion analysis.

[0007] Preferably, the acoustic-vibration synchronous detection module comprises a 32-element ultrasonic array and a three-axis vibration sensor, and beam focusing is achieved by the following formula: P(x,y,z) = Σn=1 32 A n ×S n (x,y,z)×exp (jφ n ) Where P(x,y,z) is the sound pressure at the focal point, A n is the weighting coefficient of the nth array element (0.5-1.0), S n (x,y,z) is the spatial response function of the nth array element, φ n is the phase delay; The time domain signal collected by the vibration sensor is Fourier transformed to obtain a frequency domain spectrum, and the 10Hz-1kHz frequency band characteristics are extracted and compared with the normal spectrum model for difference analysis. A difference of ≥3dB is determined to be abnormal; the sound and vibration synchronization detection module transmits the detected sound and vibration abnormality information to the portable main terminal.

[0008] Preferably, the field-gas joint detection module calculates the safety distance through a three-dimensional electric field attenuation model: D = k² × E -0 . 8 ×√S Where D is the safety distance, k2 is the environmental correction factor, E is the electric field strength, and S is the surface area of ​​the equipment; When the SF6 concentration C≥1ppm, the leakage source is located by the diffusion model: C (x,y,z,t)=Q×t -0 . 5 ×exp[-(x 2 +y 2 +z2 ) / (4Dt)], where Q is the leakage amount, D is the diffusion coefficient, and t is the detection time; the field-gas joint detection module transmits the safety distance information and the gas leakage information to the portable main terminal.

[0009] Preferably, the adaptive computing power allocation of the portable master terminal satisfies: α = (0.6×N1 + 0.4×N2) / (N1 + N2) Among them, α is the computing power ratio of the neural network accelerator, N1 is the number of image data frames, and N2 is the amount of sensor data; When the real-time computing power load is ≥80%, priority scheduling is started, the fault alarm data processing delay is ≤100ms, and the general data processing delay is ≤500ms; the portable master terminal performs fusion analysis based on the received light, heat, acoustic vibration, and field-gas data.

[0010] Preferably, the multi-dimensional sensor data fusion analysis includes: Feature layer fusion: SIFT feature vector V1 is extracted from the photothermal image, and Mel coefficient vector V2 is extracted from the acoustic vibration signal. These are fused into V=w1V1+w2V2 through the weight matrix W=[w1,w2], where w1+w2=1. If the fault type is temperature abnormality, w1=0.6. Decision-making layer fusion: The evidence synthesis formula m (A) = [Σmᵢ(Aᵢ)] / (1-Σmᵢ(Aᵢ) mⱼ(Aⱼ)) is used to output a fault confidence level ≥ 0.85. The fusion analysis results are sent from the portable main terminal to the wearable terminal.

[0011] Preferably, the virtual-reality fusion positioning of the wearable terminal is achieved by perspective projection: [u,v,1]ᵀ = K×[R|t]×[X,Y,Z,1]ᵀ Where K is the camera intrinsic parameter matrix, [R|t] is the pose transformation matrix, (X, Y, Z) is the three-dimensional coordinate of the fault, and (u, v) is the screen coordinate; The three-dimensional mark superposition deviation is ≤3% of the field of view, and it supports multi-terminal synchronous display with a synchronization delay of ≤50ms; the wearable terminal displays the three-dimensional mark based on the fusion analysis results sent by the portable main terminal.

[0012] Preferably, the power-specific wireless communication network adopts an encryption mechanism: Identity Authentication: , where S is the authentication ciphertext, S0 is the plaintext, and E is the SM2 encryption function; Each frame contains a checksum H=SHA256 (D), where D is the data frame content. A checksum failure triggers retransmission. The communication network ensures secure data transmission between the portable master terminal, wearable terminal, and distributed heterogeneous sensor cluster.

[0013] Preferably, the local preprocessing of the distributed heterogeneous sensor cluster includes: Vibration signal denoising: Use wavelet threshold λ=σ×√(2lnN), where σ is the noise standard deviation and N is the signal length; The amount of data after feature extraction is compressed to less than 30% of the original data, and synchronization is achieved through timestamp alignment, with a time deviation of ≤100ms; the preprocessed data is transmitted to the portable main terminal by the distributed heterogeneous sensor cluster.

[0014] Preferably, it also includes an equipment health prediction module, which is in communication with the portable main terminal and receives historical fault data and real-time monitoring data output by the portable main terminal; The equipment health prediction module uses an LSTM neural network model, taking the equipment operating parameters of the past six months, including temperature fluctuations, vibration spectrum characteristics, and changes in electric field strength, as input. It automatically learns the parameter correlations during model training and outputs a 30-day equipment health score. The LSTM neural network model includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer matches the dimension of the operating parameters. The hidden layer has three layers, each containing 64 neurons. The output layer uses a Sigmoid activation function to map the health score. When the health score falls below the set threshold, the device health prediction module generates an early warning message and sends it to the portable main terminal, which then pushes it to the wearable terminal for display. The early warning message includes a health trend curve and a possible fault type prompt.

[0015] The present invention provides a multifunctional inspection device detection system based on a distribution network mobile operation terminal. It has the following beneficial effects: 1. By building a dynamic, collaborative, closed-loop data chain, the existing inspection system's independent module operation and data fragmentation issues are resolved. The portable master terminal, wearable terminal, and distributed heterogeneous sensor cluster form a highly efficient, interconnected whole. The distributed heterogeneous sensor cluster's multi-module collaborative data collection, combined with a nanosecond-level timestamp synchronization mechanism, breaks through the limitations of traditional single-parameter detection and enables comprehensive capture of multi-dimensional parameters of distribution network equipment, including light, heat, acoustics, vibration, and field-gas. This provides rich and synchronized basic data for subsequent analysis, significantly improving the comprehensiveness of inspections. At the data processing and analysis level, the portable master terminal's adaptive computing power allocation mechanism dynamically adjusts the computing power ratio based on data type, avoiding resource waste or processing delays caused by fixed computing power allocation and ensuring rapid response to fault alarm data. The multi-dimensional sensor data fusion analysis technology fully taps the value of various data types through the hierarchical integration of feature layers and decision layers, solving the one-sidedness of single-data judgments, significantly improving the accuracy and confidence of fault judgments, and making fault location more precise. For on-site inspection personnel, the wearable terminal's virtual-reality fusion positioning technology overlays fault information on the real scene in the form of three-dimensional markers. Combined with eye movement and voice interaction, it eliminates the inconvenience of traditional handheld terminal operation, allowing inspection personnel to obtain fault information intuitively and conveniently, improving on-site judgment efficiency. At the same time, the field-gas joint detection module's safety distance calculation and gas leak location functions provide inspection personnel with real-time safety guidance, effectively reducing safety risks during the inspection process. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] like Figure 1 The present invention proposes a multifunctional inspection device detection system based on a distribution network mobile operation terminal, which includes a portable main terminal with a dynamic collaborative architecture, a wearable terminal with augmented reality interaction, and a distributed heterogeneous sensor cluster. The three form a closed-loop data link through a power-specific low-latency wireless communication network. The portable master terminal integrates a heterogeneous computing unit with adaptive computing power allocation, which can perform real-time fusion analysis of multi-dimensional sensor data and output visual results including fault location and confidence level; The wearable terminal uses virtual-reality fusion positioning technology to overlay fault information on the real scene in the form of three-dimensional markers, supporting collaborative interaction between eye movement and voice; The distributed heterogeneous sensing cluster includes a photothermal coaxial acquisition module, an acoustic-vibration synchronous detection module, and a field-gas joint measurement module. Each module achieves nanosecond-level synchronization through timestamp alignment and has local feature extraction and data compression capabilities.

[0019] It should be noted that the existing problem of independent operation of various components of the inspection system and poor data flow, which leads to low overall inspection efficiency and information fragmentation, is solved. By building a closed-loop data chain and collaborative architecture, efficient interaction and comprehensive analysis of data from each module can be achieved, improving the integrity and intelligence of distribution network inspection.

[0020] As an optional embodiment, the photothermal coaxial acquisition module includes a coaxially arranged visible light component and an infrared component, and realizes exposure timing synchronization through a programmable logic unit, and satisfies:

[0021] in, is the corrected exposure time, is the equipment type coefficient; in this embodiment, the overhead line k1=1.2, the switchgear k1=0.8; is the base exposure time, is the standard light intensity, is the real-time light intensity; it should be noted that, in this embodiment, the reference exposure time is (100ms) and the standard light intensity is (1000lux); The temperature field data output by the infrared component and the image data output by the visible light component are spatially aligned through ORB feature point matching, with a matching error of ≤5% of the field of view angle. The coaxial photothermal acquisition module transmits the processed photothermal data to the portable main terminal, providing the basis for its fusion analysis. This addresses the existing issues of asynchronous photothermal acquisition and spatial misalignment, which lead to large deviations between temperature anomalies and the actual device position. Through temporal synchronization and spatial alignment, the consistency of photothermal data is ensured, the accuracy of fault location is improved, and reliable photothermal parameters are provided for the fusion analysis of the portable main terminal.

[0022] As an optional embodiment, the acoustic-vibration synchronous detection module includes a 32-element ultrasonic array and a three-axis vibration sensor, and beam focusing is achieved by the following formula: P (x,y,z) = Σn=1 32 A n ×S n (x,y,z)×exp (jφ n ) Among them, P(x,y,z) is the sound pressure at the focal point, A n is the weighting coefficient of the nth array element (0.5-1.0), S n (x, y, z) is the spatial response function of the nth array element, φ n is the phase delay; The time domain signal collected by the vibration sensor is Fourier transformed to obtain a frequency domain spectrum, and the 10Hz-1kHz frequency band characteristics are extracted and compared with the normal spectrum model for difference analysis. A difference of ≥3dB is determined to be abnormal; the sound and vibration synchronization detection module transmits the detected sound and vibration abnormality information to the portable main terminal.

[0023] This technology addresses the existing issues of low focusing accuracy and incomplete signal analysis in acoustic and vibration detection, which can lead to missed or misdetected partial discharges and mechanical faults. By improving ultrasonic detection accuracy through beam focusing and combining it with vibration spectrum analysis, it enables comprehensive assessment of equipment acoustic and vibration anomalies, providing accurate clues to acoustic and vibration faults to the portable host terminal.

[0024] As an optional embodiment, the field-gas joint detection module calculates the safety distance using a three-dimensional electric field attenuation model: D = k² × E -0 . 8 ×√S Where D is the safety distance, k2 is the environmental correction factor, E is the electric field strength, and S is the surface area of ​​the equipment; When the SF6 concentration C≥1ppm, the leakage source is located by the diffusion model: C (x,y,z,t)=Q×t -0 . 5 ×exp[-(x 2 +y 2 +z 2 ) / (4Dt)], where Q is the leakage amount, D is the diffusion coefficient, and t is the detection time; the field-gas joint detection module transmits the safety distance information and the gas leakage information to the portable main terminal.

[0025] The invention solves the problems in the existing technology of inaccurate safety distance calculation and fuzzy gas leakage positioning, which threaten the safety of patrol personnel, provides reliable safety guidance for patrol personnel, and provides environmental safety parameters for comprehensive fault judgment of portable main terminals.

[0026] As an optional embodiment, the adaptive computing power allocation of the portable master terminal satisfies: α = (0.6×N1 + 0.4×N2) / (N1 + N2) Among them, α is the computing power ratio of the neural network accelerator, N1 is the number of image data frames, and N2 is the amount of sensor data; When the real-time computing power load is ≥80%, priority scheduling is started, the fault alarm data processing delay is ≤100ms, and the general data processing delay is ≤500ms; the portable master terminal performs fusion analysis based on the received light, heat, acoustic vibration, and field-gas data.

[0027] This solves the existing problem of fixed terminal computing power allocation and high data processing latency, which leads to delayed fault response. Through adaptive computing power allocation and priority scheduling, it ensures rapid processing of critical fault data, improving the processing efficiency and response speed of portable master terminals to various fault information.

[0028] As an optional embodiment, the multi-dimensional sensor data fusion analysis includes: Feature layer fusion: SIFT feature vector V1 is extracted from the photothermal image, and Mel coefficient vector V2 is extracted from the acoustic vibration signal. These are fused into V=w1V1+w2V2 through the weight matrix W=[w1,w2], where w1+w2=1. If the fault type is temperature abnormality, w1=0.6. Decision-making layer fusion: The evidence synthesis formula m (A) = [Σmᵢ(Aᵢ)] / (1-Σmᵢ(Aᵢ) mⱼ(Aⱼ)) is used to output a fault confidence level ≥ 0.85. The fusion analysis results are sent from the portable main terminal to the wearable terminal.

[0029] This technology addresses the issues of insufficient multi-dimensional data fusion and low reliability of fault diagnosis in existing technologies. By integrating the feature layer and the decision layer, it fully utilizes the data advantages of each module, improves the accuracy and confidence of fault diagnosis, and lays the foundation for wearable terminals to display reliable fault information.

[0030] As an optional embodiment, the virtual-reality fusion positioning of the wearable terminal is achieved by perspective projection: [u,v,1]ᵀ = K×[R|t]×[X,Y,Z,1]ᵀ Where K is the camera intrinsic parameter matrix, [R|t] is the pose transformation matrix, (X, Y, Z) is the three-dimensional coordinate of the fault, and (u, v) is the screen coordinate; The three-dimensional mark superposition deviation is ≤3% of the field of view, and it supports multi-terminal synchronous display with a synchronization delay of ≤50ms; the wearable terminal displays the three-dimensional mark based on the fusion analysis results sent by the portable main terminal.

[0031] This solves the existing problem of non-intuitive fault information display and information asynchrony across multiple terminals, which hinders on-site inspection and collaboration. Through virtual-reality integrated positioning and multi-terminal synchronization, fault information is intuitively presented and shared across the team, improving on-site inspection efficiency and collaboration.

[0032] As an optional embodiment, the power-specific wireless communication network adopts an encryption mechanism: Identity Authentication: , where S is the authentication ciphertext, S0 is the plaintext, and E is the SM2 encryption function; Each frame contains a checksum H=SHA256 (D), where D is the data frame content. A checksum failure triggers retransmission. The communication network ensures secure data transmission between the portable master terminal, wearable terminal, and distributed heterogeneous sensor cluster.

[0033] This solves the existing problem of insecure data transmission and easy loss and tampering in inspection systems. Through encryption authentication and data verification, it ensures the security, integrity and accuracy of data during transmission, ensuring the reliable operation of data interaction throughout the system.

[0034] As an optional embodiment, the local preprocessing of the distributed heterogeneous sensor cluster includes: Vibration signal denoising: Use wavelet threshold λ=σ×√(2lnN), where σ is the noise standard deviation and N is the signal length; The amount of data after feature extraction is compressed to less than 30% of the original data, and synchronization is achieved through timestamp alignment, with a time deviation of ≤100ms; the preprocessed data is transmitted to the portable main terminal by the distributed heterogeneous sensor cluster.

[0035] This solves the existing problem of sensor data being noisy, bulky, and poorly synchronized, which increases the processing burden on terminals. Local preprocessing reduces data noise and volume, improves data synchronization, reduces computing pressure on portable master terminals, and improves overall system efficiency.

[0036] As an optional embodiment, it further includes an equipment health prediction module, which is in communication with the portable main terminal and receives historical fault data and real-time monitoring data output by the portable main terminal; The equipment health prediction module uses an LSTM neural network model, taking the equipment operating parameters of the past six months, including temperature fluctuations, vibration spectrum characteristics, and changes in electric field strength, as input. It automatically learns the parameter correlations during model training and outputs a 30-day equipment health score. The LSTM neural network model includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer matches the dimension of the operating parameters. The hidden layer has three layers, each containing 64 neurons. The output layer uses a Sigmoid activation function to map the health score. When the health score is less than or equal to the set threshold, the device health prediction module generates an early warning message and sends it to the portable main terminal, which is then pushed to the wearable terminal for display. The early warning message includes a health trend curve and a prompt indicating the type of possible fault. It should be noted that the device health score ranges from 0 to 100, and the threshold is set by the staff. The default initial value is 60. This solves the problem that existing technologies can only detect the current status of equipment but cannot predict equipment health trends in advance, making it difficult to respond to sudden failures. By using neural network-based health prediction and leveraging potential correlations between data to predict trends, this technology provides a forward-looking basis for inspection planning and equipment maintenance, improving the reliability of distribution network equipment operation.

[0037] It mainly includes a portable inspection computing terminal, a wearable display terminal and a wireless expandable sensor group. The three are connected wirelessly to form an efficient system architecture for data collection, processing and display. The portable inspection computing terminal is a miniaturized independent device carried by distribution network inspection personnel. It includes a sensor network signal transceiver module, an edge computing module, and a cellular network signal transceiver module connected in sequence, and also includes a data storage module connected to the edge computing module. Sensor network signal transceiver module: It has powerful signal reception and transmission capabilities, can receive wireless data signals sent by wireless scalable sensor groups, interact with edge computing terminals, and send wireless data signals to wearable display terminals to ensure smooth flow of data within the system. Edge computing module: As the core computing unit of the entire system, it undertakes the task of real-time computing and analysis. It can process various types of data collected by the wireless scalable sensor group in real time, determine whether there is a fault, analyze the fault location and abnormal parameter quantity, and give accurate inspection conclusions. At the same time, the module can also generate visual detection results, including images, charts, videos, etc., to facilitate staff to intuitively understand the detection situation. Cellular network signal transceiver module: responsible for establishing a connection with the enterprise intranet to achieve data transmission. Through this module, the system can upload the detection data to the enterprise intranet in a timely manner so that relevant personnel can conduct further analysis and processing. At the same time, it can also receive instructions from the enterprise intranet to achieve remote control and management. Data storage module: It is used to store the data output by the edge computing module locally as a data backup to prevent data loss. It can also ensure the integrity of the data when network transmission fails or other abnormal situations occur, and provide support for subsequent data analysis. The wearable display terminal includes a sensor network signal receiving module and a display module electrically connected to the sensor network signal receiving module, and the sensor network signal receiving module and the display module are encapsulated in the same housing. The sensor network signal receiving module is responsible for receiving the wireless data signal sent by the portable inspection computing terminal and transmitting it to the display module. The display module then displays the received data to the user in an intuitive manner, including information such as fault judgment results, to facilitate user monitoring and guidance of on-site maintenance. The wireless scalable sensor group includes a multi-spectral chip, an ultrasonic acquisition module, a vibration acquisition module, an electric field size and distance acquisition module, and an SF6 and oxygen concentration acquisition module, which are respectively connected to the portable inspection computing terminal; the wireless scalable sensor group also includes a visible light acquisition module and an infrared acquisition module, which are respectively connected to the multi-spectral chip; The visible light acquisition module and infrared acquisition module are encapsulated in the same housing as a single device. They are used for contactless detection of real-time video and infrared images of distribution network equipment. The collected data is transmitted to the multispectral chip via a wired connection. The multispectral chip fuses the visible light and infrared images to obtain physical ID signals and perform temperature detection. This image fusion technology can more comprehensively and accurately reflect the operating status of the equipment. The ultrasonic acquisition module, vibration acquisition module, electric field size and distance acquisition module, SF6 and oxygen concentration acquisition module are separately packaged, that is, they are packaged as independent devices, and there is no electrical connection between the independent devices. These modules are used to detect the ultrasonic information, vibration waveform, electric field size and distance, SF6 and oxygen concentration and other parameters of the distribution network equipment. The data information collected by each sensor module will be sent to the portable inspection and calculation terminal via wireless transmission to realize the synchronous acquisition of multiple parameters of the equipment; First, the wireless scalable sensor group begins operation. Each sensor module collects relevant data from the distribution network equipment. The visible light acquisition module and the infrared acquisition module collect real-time video and infrared images of the equipment, which are then fused and processed by a multispectral chip and transmitted to the portable inspection computing terminal. The ultrasonic acquisition module, vibration acquisition module, electric field size and distance acquisition module, SF6 and oxygen concentration acquisition module also wirelessly transmit their collected data to the portable inspection computing terminal. After receiving the data, the sensor network signal transceiver module of the portable inspection computing terminal transmits it to the edge computing module, which performs real-time analysis on the data. Based on the temperature matrix of the infrared image, it queries whether there are abnormal pixels within the image range to determine whether there is a temperature abnormality fault and marks the location of the temperature abnormality point. It judges the ultrasonic frequency data at different locations to determine whether there is a partial discharge fault and marks the fault location through the ultrasonic array. It compares the spectral characteristics of the vibration waveform and acceleration waveform with the built-in mechanical vibration spectrum characteristic library to determine whether there is a mechanical vibration fault and mark the fault location. By analyzing the strength of the 50Hz power frequency signal after digital filtering, it determines whether the operator maintains a safe distance from the charged object. By analyzing the degree of ionization of the gas in the high-frequency electromagnetic field, it determines whether there is an SF6 leakage fault. The edge computing module sends the fault judgment results obtained through analysis to the wearable display terminal. After the sensor network signal receiving module of the wearable display terminal receives the data, the display module displays the fault judgment results to the user. The user can guide on-site maintenance work based on this information. At the same time, the cellular network signal transceiver module of the portable inspection computing terminal uploads the detection data to the enterprise intranet, realizing remote storage and sharing of data, making it convenient for the enterprise to uniformly manage and analyze the operating status of distribution network equipment.

[0038] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A multifunctional inspection device detection system based on a distribution network mobile operation terminal, characterized in that: It includes a portable main terminal with a dynamic collaborative architecture, a wearable terminal with augmented reality interaction, and a distributed heterogeneous sensor cluster. The three form a closed-loop data link through a low-latency wireless communication network dedicated to power. The portable master terminal integrates a heterogeneous computing unit with adaptive computing power allocation, which can perform real-time fusion analysis of multi-dimensional sensor data and output visual results including fault location and confidence level; The wearable terminal uses virtual-reality fusion positioning technology to overlay fault information on the real scene in the form of three-dimensional markers, supporting collaborative interaction between eye movement and voice; The distributed heterogeneous sensing cluster includes a photothermal coaxial acquisition module, an acoustic-vibration synchronous detection module, and a field-gas joint measurement module. Each module achieves nanosecond-level synchronization through timestamp alignment and has local feature extraction and data compression capabilities.

2. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 1 is characterized in that: The photothermal coaxial acquisition module includes a coaxially arranged visible light component and an infrared component, and realizes exposure timing synchronization through a programmable logic unit, and meets the following requirements: ； in, is the corrected exposure time, is the equipment type coefficient; is the base exposure time, is the standard light intensity, is the real-time light intensity; The temperature field data output by the infrared component and the image data output by the visible light component are spatially aligned through ORB feature point matching, and the matching error is ≤5% of the field of view angle. The photothermal coaxial acquisition module transmits the processed photothermal data to the portable main terminal, providing basic data for its fusion analysis.

3. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 2 is characterized in that: The acoustic-vibration synchronous detection module includes a 32-element ultrasonic array and a three-axis vibration sensor, and beam focusing is achieved through the following formula: P (x,y,z) = Σn=1 32 A n ×S n (x,y,z)×exp (jφ n ) Where P (x,y,z) is the sound pressure at the focal point, A n is the weighting coefficient of the nth array element (0.5-1.0), S n (x,y,z) is the spatial response function of the nth array element, φ n is the phase delay; The time domain signal collected by the vibration sensor is Fourier transformed to obtain a frequency domain spectrum, and the 10Hz-1kHz frequency band characteristics are extracted and compared with the normal spectrum model for difference analysis. A difference of ≥3dB is determined to be abnormal; the sound and vibration synchronization detection module transmits the detected sound and vibration abnormality information to the portable main terminal.

4. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 2 is characterized in that: The field-gas joint measurement module calculates the safety distance using a three-dimensional electric field attenuation model: D = k2×E -0 . 8 ×√S Where D is the safety distance, k2 is the environmental correction factor, E is the electric field strength, and S is the surface area of ​​the equipment; When the SF6 concentration C≥1ppm, the leakage source is located by the diffusion model: C (x,y,z,t)=Q×t -0 . 5 ×exp [-(x 2 +y 2 +z 2 ) / (4Dt)], where Q is the leakage amount, D is the diffusion coefficient, and t is the detection time; the field-gas joint detection module transmits the safety distance information and the gas leakage information to the portable main terminal.

5. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 4 is characterized in that: The adaptive computing power allocation of the portable master terminal satisfies: α = (0.6×N1 + 0.4×N2) / (N1 + N2) Among them, α is the computing power ratio of the neural network accelerator, N1 is the number of image data frames, and N2 is the amount of sensor data; When the real-time computing power load is ≥80%, priority scheduling is started, the fault alarm data processing delay is ≤100ms, and the general data processing delay is ≤500ms; the portable master terminal performs fusion analysis based on the received light, heat, acoustic vibration, and field-gas data.

6. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 4 is characterized in that: The multi-dimensional sensor data fusion analysis includes: Feature layer fusion: SIFT feature vector V1 is extracted from the photothermal image, and Mel coefficient vector V2 is extracted from the acoustic vibration signal. These are fused into V=w1V1+w2V2 through the weight matrix W=[w1,w2], where w1+w2=1. If the fault type is temperature abnormality, w1=0.

6. Decision-making layer fusion: The evidence synthesis formula m (A) = [Σmᵢ(Aᵢ)] / (1-Σmᵢ(Aᵢ) mⱼ(Aⱼ)) is used to output a fault confidence level ≥ 0.

85. The fusion analysis results are sent from the portable main terminal to the wearable terminal.

7. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 6 is characterized in that: The virtual-reality fusion positioning of the wearable terminal is achieved through perspective projection: [u,v,1]ᵀ = K×[R|t]×[X,Y,Z,1]ᵀ Where K is the camera intrinsic parameter matrix, [R|t] is the pose transformation matrix, (X, Y, Z) is the three-dimensional coordinate of the fault, and (u, v) is the screen coordinate; The three-dimensional mark superposition deviation is ≤3% of the field of view, and it supports multi-terminal synchronous display with a synchronization delay of ≤50ms; the wearable terminal displays the three-dimensional mark based on the fusion analysis results sent by the portable main terminal.

8. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 6 is characterized in that: The power-specific wireless communication network adopts an encryption mechanism: Identity Authentication: , where S is the authentication ciphertext, S0 is the plaintext, and E is the SM2 encryption function; Each frame contains a checksum H=SHA256 (D), where D is the data frame content. A checksum failure triggers retransmission. The communication network ensures secure data transmission between the portable master terminal, wearable terminal, and distributed heterogeneous sensor cluster.

9. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 1 is characterized in that: The local preprocessing of the distributed heterogeneous sensor cluster includes: Vibration signal denoising: Use wavelet threshold λ=σ×√(2lnN), where σ is the noise standard deviation and N is the signal length; The amount of data after feature extraction is compressed to less than 30% of the original data, and synchronization is achieved through timestamp alignment, with a time deviation of ≤100ms; the preprocessed data is transmitted to the portable main terminal by the distributed heterogeneous sensor cluster.

10. The multifunctional inspection device detection system based on the distribution network mobile operation terminal according to claim 1, characterized in that: It also includes an equipment health prediction module, which is connected to the portable main terminal for communication and receives historical fault data and real-time monitoring data output by the portable main terminal; The equipment health prediction module uses an LSTM neural network model, taking the equipment operating parameters of the past six months, including temperature fluctuations, vibration spectrum characteristics, and changes in electric field strength, as input. It automatically learns the parameter correlations during model training and outputs a 30-day equipment health score. The LSTM neural network model includes an input layer, a hidden layer, and an output layer. The number of nodes in the input layer matches the dimension of the operating parameters. The hidden layer has three layers, each containing 64 neurons. The output layer uses a Sigmoid activation function to map the health score. When the health score is less than or equal to the set threshold, the equipment health prediction module generates an early warning message and sends it to the portable main terminal, which is then pushed to the wearable terminal for display. The early warning message includes a health trend curve and a prompt for the possible fault type.

Images