Monitoring system for monitoring running state of drop-type lightning arrester on line

Abstract

The invention, which relates to the technical field of power grid safety, discloses a monitoring system for on-line monitoring of the operation state of a drop-out lightning arrester, comprising: a data acquisition and processing module for generating high-quality state characteristic data; the intelligent state monitoring module adopts a multi-algorithm fusion model to perform real-time evaluation of the operation state of the lightning arrester, accurate fault identification and positioning and future state trend pre-judgment; the grading early warning module is used for dynamically generating grading early warning information based on the state monitoring result and pushing the grading early warning information to a related operation and maintenance terminal through multi-channel cooperation; the remote operation and maintenance module is used for carrying out fault emergency remote disposal, equipment core parameter remote configuration and operation and maintenance whole-process closed-loop management; and the data storage tracing module is used for storing total monitoring data, early warning information and operation and maintenance records. According to the invention, through fusion of electrical, mechanical and environmental multi-dimensional parameter acquisition and integrated processing, and through multi-algorithm fusion of an improved CNN-LSTM hybrid model and the like, accurate fault identification, positioning and trend pre-judgment are realized.

Description

Technical Field

[0001] This invention relates to the field of power grid safety technology, specifically to an online monitoring system for the operating status of drop-out surge arresters. Background Technology

[0002] As a critical protection device in power distribution networks, the operating status of surge arresters directly affects the safe and stable power supply of the grid. Currently, the monitoring of the operating status of surge arresters in power distribution networks relies heavily on regular manual inspections, which presents several technical challenges: First, manual inspections are inefficient and costly, making it difficult to cover equipment in remote areas and complex terrains, and are easily affected by subjective factors, leading to missed faults. Second, existing monitoring methods mostly focus on single electrical parameters, lacking a comprehensive consideration of mechanical structural conditions and environmental factors, and thus failing to fully reflect the true operating status of the equipment. Third, fault identification relies heavily on simple threshold judgments, lacking intelligent algorithm support, making it difficult to accurately identify early hidden faults and predict the trend of deterioration. Fourth, fault handling response is delayed, lacking remote emergency handling capabilities, and serious faults can easily escalate into power grid accidents. Fifth, monitoring data is stored in a scattered manner, with poor traceability, failing to provide effective data support for the full life cycle management of the equipment.

[0003] Therefore, there is an urgent need for an online monitoring system that integrates multi-dimensional data acquisition, intelligent status monitoring, rapid early warning, remote operation and maintenance, and data traceability to solve the problems of incomplete monitoring, inaccurate identification, and untimely response in existing technologies, and to ensure the safe and reliable operation of the power distribution network. Summary of the Invention

[0004] To address the aforementioned technical issues, an online monitoring system for the operational status of drop-out surge arresters is provided. This technical solution resolves the problems described above.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] An online monitoring system for the operational status of drop-out surge arresters includes:

[0007] Data acquisition and processing module, intelligent status monitoring module, hierarchical early warning module, remote operation and maintenance module, and data storage and traceability module;

[0008] The data acquisition and processing module is used to collect the electrical operating parameters, mechanical structure parameters and environmental impact parameters of the surge arrester in real time, and generate high-quality status feature data through integrated processing of noise reduction, standardization and feature extraction.

[0009] The intelligent status monitoring module is electrically connected to the data acquisition and processing module, and adopts a multi-algorithm fusion model to perform real-time evaluation of the arrester's operating status, accurate fault identification and location, and prediction of future status trends.

[0010] The hierarchical early warning module is electrically connected with the intelligent state monitoring module, generates hierarchical early warning information dynamically based on the state monitoring result, and pushes the information to related operation and maintenance terminals through multiple channels;

[0011] The remote operation and maintenance module is electrically connected with the hierarchical early warning module, and performs emergency remote disposal of faults, remote configuration of core parameters of equipment, and closed-loop management of the whole operation and maintenance process.

[0012] The data storage and traceability module is electrically connected with the intelligent state monitoring module and the hierarchical early warning module, and stores and traces the whole amount of monitoring data, early warning information and operation and maintenance records.

[0013] Preferably, the data acquisition and processing module comprises:

[0014] The three-phase voltage of the lightning arrester, the leakage current, the vibration parameter of the drop mechanism, the environmental temperature and humidity, and the contamination degree of the insulator are synchronously acquired through high-precision Hall sensors, vibration sensors, temperature and humidity sensors and contamination degree sensors.

[0015] First, the wavelet packet transform denoising algorithm is used to remove electromagnetic interference and environmental noise, and then Z-score standardization processing is performed to eliminate the dimensional difference of the parameters and extract the peak value, mean value and variance characteristics of the parameters to generate standardized state characteristic data.

[0016] Preferably, the intelligent state monitoring module comprises:

[0017] An improved CNN-LSTM hybrid model is constructed, wherein the CNN part extracts the spatial dimension characteristics of the state characteristic data through 3 convolution layers and 2 maximum pooling layers, the LSTM part captures the time sequence correlation characteristics of the data through 2 hidden layers, and the output features of the two parts are fused through a full connection layer to realize real-time evaluation of the running state.

[0018] The model is trained by common fault history data and simulation fault samples covering leakage current exceeding, insulation aging and mechanical jamming, and the network hyperparameters are optimized by the grid search method.

[0019] Preferably, the intelligent state monitoring module further comprises:

[0020] The fault precise positioning unit establishes a spatial positioning model based on the sensor deployment coordinates, locates the fault part by calculating the correlation weight of the fault characteristic parameter and the sensor acquisition data, and identifies the key component fault by combining the Euclidean distance algorithm, wherein the formula for fault positioning is:

[0021]

[0022] In the formula, is the fault characteristic parameter, is the sensor acquisition data, for positioning error;

[0023] The state trend prediction unit establishes an LSTM-ARIMA fusion prediction model based on an attention mechanism, and the attention mechanism automatically assigns weights to the state data of the past 7 days, wherein the prediction model formula is:

[0024]

[0025] wherein, is a fusion weight, and the value range is 0.6-0.8, is a future state prediction value,

[0026] is the state data of the past 7 days, is a long short-term memory model, is an autoregressive integrated moving average model.

[0027] Preferably, the intelligent state monitoring module further comprises:

[0028] The fault grading unit establishes a three-level fault division standard based on the influence range of the fault on the power grid operation, the processing urgency and the state evaluation score, including: slight fault (does not affect normal power supply, does not need to stop, and can be tracked regularly), general fault (affects the stability of operation, and needs to be handled on site within 72 hours) and serious fault (endangers the safety of the power grid, may cause tripping, and needs to be stopped immediately for disposal);

[0029] The health state quantification unit constructs a hierarchical structure by using the analytic hierarchy process, invites 5-8 power operation and maintenance experts to assign weights to the core indicators of voltage, leakage current and vibration parameters, and calculates the health state index by weighted summation.

[0030] Preferably, the construction of the hierarchical structure by using the analytic hierarchy process specifically comprises:

[0031] The calculation formula of the health state index is:

[0032]

[0033] wherein, is the weight of each indicator, is the normalized value of each indicator, is the health state index;

[0034] The value range of the health state index is 0-1, and the health state index ≥0.85 is set as a normal state, 0.7-0.85 is set as an attention state, and <0.7 is set as an abnormal state.

[0035] Preferably, the grading early warning module comprises:

[0036] The early warning threshold dynamic adaptation unit divides scenarios based on the operating life of the lightning arrester and the severity of the environment, and automatically adjusts the early warning threshold in each scenario through a machine learning algorithm.

[0037] The early warning information structured generation unit integrates the device unique number, accurate fault location, fault type, HSI specific value, future 72h trend prediction result and targeted treatment scheme to generate a standardized early warning script.

[0038] The multi-channel collaborative push unit pushes early warning information through four ways: operation and maintenance APP push, SMS notification, Web monitoring platform pop-up window and on-site sound and light alarm, and additionally triggers phone voice notification for serious faults.

[0039] Preferably, the remote operation and maintenance module comprises:

[0040] The emergency fault handling unit automatically sends a remote opening command to the high-voltage control switch associated with the lightning arrester through a 5G private network when a serious fault (HSI < 0.5 and fault type is insulation breakdown, mechanism shedding, etc.) is detected, and the command triggers the switch action after encrypted transmission, cutting off the connection between the faulty device and the power grid.

[0041] The remote parameter configuration unit allows operation and maintenance personnel to remotely adjust data collection frequency, various fault early warning thresholds and prediction model fusion weights through a Web background visual interface, and the parameters are synchronized to the device terminal in real time after adjustment, adapting to the operation needs of different regional power grids.

[0042] The operation and maintenance work order closed loop unit automatically generates an electronic operation and maintenance work order based on early warning information, including fault details, processing time limit and responsibility person classification information, and the work order status is updated in real time, and the device operation and maintenance archives are also updated synchronously.

[0043] Preferably, the data storage and traceability module comprises:

[0044] An InfluxDB+Redis distributed time series database architecture is adopted, InfluxDB is used to store full-amount raw monitoring data, standardized state feature data, fault identification results, early warning information and operation and maintenance records, and Redis is used to cache high-frequency query data (including near 30-day state data and unprocessed early warning information);

[0045] The full-cycle traceability unit supports multi-dimensional combined query of historical data according to device number, time range (accurate to minutes), data type (raw data / monitoring results) and fault type, and can automatically generate device state change trend curve, fault occurrence frequency statistical report and operation and maintenance efficiency analysis report, providing data support for lightning arrester full life cycle management.

[0046] Preferably, it also includes an anti-interference and adaptive calibration unit, the hardware resists electromagnetic interference through shielding ground and power filtering, and the software removes abnormal values through data denoising, and the sensor acquisition channel is automatically calibrated every month.

[0047] Compared with the prior art, the present application has the beneficial effects that:

[0048] The present application proposes to fuse electrical, mechanical and environmental multi-dimensional parameter acquisition and integrated processing, to realize precise fault identification, positioning and 72h trend prediction through an improved CNN-LSTM hybrid model and a multi-algorithm fusion scheme such as Euclidean distance positioning, to improve response and disposal efficiency by matching dynamic early warning threshold, multi-channel push mechanism and 5G remote emergency tripping, work order closed-loop management, to provide reliable data support for the whole life cycle management of lightning arresters by combining InfluxDB+Redis distributed storage, full-dimensional traceability and anti-interference and adaptive calibration design, and to break through the limitations of traditional monitoring. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 The system framework of the present application is shown in the following. DETAILED DESCRIPTION

[0050] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be thought of by those skilled in the art.

[0051] Referring to Figure 1 The online monitoring drop-type lightning arrester operating state monitoring system shown includes:

[0052] a data acquisition and processing module, an intelligent state monitoring module, a hierarchical early warning module, a remote operation and maintenance module, and a data storage and traceability module;

[0053] The data acquisition and processing module is used to acquire lightning arrester electrical operating parameters, mechanical structure parameters and environmental influence parameters in real time, and to generate high-quality state feature data through integrated processing of denoising, standardization and feature extraction;

[0054] The intelligent state monitoring module is electrically connected with the data acquisition and processing module, adopts a multi-algorithm fusion model, and performs real-time evaluation of lightning arrester operating state, precise fault identification and positioning, and future state trend prediction;

[0055] The hierarchical early warning module is electrically connected with the intelligent state monitoring module, dynamically generates hierarchical early warning information based on the state monitoring result, and pushes the information to related operation and maintenance terminals through multi-channel cooperation;

[0056] The remote operation and maintenance module is electrically connected with the hierarchical early warning module, and performs emergency remote disposal of faults, remote configuration of device core parameters, and operation and maintenance full-process closed-loop management.

[0057] The data storage and tracing module is electrically connected with the intelligent state monitoring module and the hierarchical early warning module, and performs storage and multi-dimensional tracing of full-volume monitoring data, early warning information and operation and maintenance records.

[0058] The data acquisition and processing module comprises:

[0059] The three-phase voltage of the lightning arrester, the leakage current, the vibration parameter of the drop mechanism, the environmental temperature and humidity, and the contamination degree of the insulator are synchronously acquired through the high-precision Hall sensor, the vibration sensor, the temperature and humidity sensor, and the contamination sensor.

[0060] First, the wavelet packet transform denoising algorithm is used to remove electromagnetic interference and environmental noise, and then Z-score standardization processing is performed to eliminate the dimensional differences of the parameters, extract the peak value, mean value and variance characteristics of the parameters, and generate standardized state characteristic data.

[0061] An improved CNN-LSTM hybrid model is introduced, which can not only extract the spatial dimension features of the state characteristic data, but also capture the time sequence correlation characteristics, significantly improving the accuracy of lightning arrester fault recognition and trend prediction. At the same time, the network hyperparameters are optimized using the grid search method, so that the adaptability and accuracy of the model are enhanced.

[0062] The intelligent state monitoring module comprises:

[0063] An improved CNN-LSTM hybrid model is constructed, wherein the CNN part extracts the spatial dimension features of the state characteristic data through 3 convolution layers and 2 max pooling layers, and the LSTM part captures the time sequence correlation characteristics of the data through 2 hidden layers. The output features of the two are fused through a fully connected layer to realize real-time evaluation of the running state.

[0064] The model is trained using common fault historical data and simulated fault samples covering leakage current exceeding, insulation aging and mechanical jamming, and the network hyperparameters are optimized using the grid search method.

[0065] The spatial positioning model combined with the Euclidean distance algorithm realizes accurate fault positioning, greatly improves the accuracy of positioning, and can quickly identify key component failures. In addition, the LSTM-ARIMA fusion prediction model based on the attention mechanism provides more flexible and dynamic prediction capabilities for future state trend prediction, enabling the system to provide early warning of potential risks.

[0066] The intelligent state monitoring module further comprises:

[0067] The fault precise positioning unit establishes a spatial positioning model based on sensor deployment coordinates, locates the fault position by calculating the correlation weight of the fault characteristic parameter and the sensor collected data, and identifies the key component fault in combination with the Euclidean distance algorithm, wherein the formula for fault positioning is:

[0068]

[0069] In the formula, is a fault characteristic parameter, is sensor collected data, is a positioning error;

[0070] The state trend prediction unit establishes an LSTM-ARIMA fusion prediction model based on an attention mechanism, and the attention mechanism automatically assigns the weight of the state data in the past 7 days, wherein the formula of the prediction model is:

[0071]

[0072] In the formula, is a fusion weight, and the value range is 0.6-0.8, is a future state prediction value,

[0073] is state data in the past 7 days, is a long short-term memory model, is an autoregressive integrated moving average model.

[0074] The intelligent state monitoring module further comprises:

[0075] The fault grading unit establishes three-level fault division standards based on the influence range of the fault on the power grid operation, the processing urgency and the state evaluation score, including: slight fault (does not affect normal power supply, does not need to stop, and can be tracked regularly), general fault (affects the stability of operation, and needs to be handled on site within 72 hours) and serious fault (endangers the safety of the power grid, may cause tripping, and needs to be stopped immediately for disposal);

[0076] The health state quantification unit constructs a hierarchical structure by using the analytic hierarchy process, invites 5-8 power operation and maintenance experts to assign weights to the core indicators of voltage, leakage current and vibration parameters, and calculates the health state index by weighted summation.

[0077] The construction of the hierarchical structure by using the analytic hierarchy process specifically comprises:

[0078] The calculation formula of the health state index is:

[0079]

[0080] In the formula, a weight of each index, a normalized value of each index, a health status index;

[0081] The health status index ranges from 0 to 1, and is set to be normal when the health status index is greater than or equal to 0.85, to be attention when the health status index is between 0.7 and 0.85, and to be abnormal when the health status index is less than 0.7.

[0082] The analytic hierarchy process is used to quantify the health status index, so that the fault classification is more scientific and detailed. At the same time, combined with the expert weighted value, the health status evaluation is more accurate, and the limitations of the traditional single index evaluation method are avoided.

[0083] The grading early warning module comprises:

[0084] The early warning threshold dynamic adaptation unit divides the scene based on the operating life of the arrester and the severity of the environment, and automatically adjusts the early warning threshold under each scene through a machine learning algorithm;

[0085] The early warning information structured generation unit integrates the device unique number, the fault accurate position, the fault type, the HSI specific value, the future 72h trend prediction result and the targeted treatment scheme to generate a standardized early warning script;

[0086] The multi-channel collaborative pushing unit pushes the early warning information through four ways of operation and maintenance APP pushing, SMS notification, Web monitoring platform pop-up window and on-site sound and light alarm, and additionally triggers a telephone voice notification for serious faults.

[0087] The remote operation and maintenance module comprises:

[0088] The emergency fault handling unit automatically sends a remote opening command to the high-voltage control switch associated with the arrester through a 5G private network when a serious fault (HSI < 0.5 and fault type is insulation breakdown, mechanism shedding, etc.) is monitored, and the command is transmitted after encryption to trigger the switch action to disconnect the fault device from the power grid;

[0089] The remote parameter configuration unit allows operation and maintenance personnel to remotely adjust the data acquisition frequency, each type of fault early warning threshold and the prediction model fusion weight through the Web background visual interface, and the parameters are adjusted in real time and synchronized to the device terminal to adapt to the operation requirements of different regional power grids;

[0090] The operation and maintenance work order closed loop unit automatically generates an electronic operation and maintenance work order according to the early warning information, including fault details, processing time limit and responsibility person classification information, and the work order state is updated in real time, and the device operation and maintenance archives are updated at the same time.

[0091] The data storage and traceability module comprises:

[0092] Adopt InfluxDB+Redis distributed time series database architecture, InfluxDB is used for storing full-amount original monitoring data, standardized state feature data, fault identification result, early warning information and operation and maintenance record, Redis is used for caching high-frequency query data (including nearly 30 days of state data and unprocessed early warning information);

[0093] Full cycle tracing unit supports multi-dimensional combined query of historical data according to equipment number, time range (accurate to minute), data type (original data / monitoring result) and fault type, can automatically generate device state change trend curve, fault occurrence frequency statistical report and operation and maintenance efficiency analysis report, and provides data support for the whole life cycle management of lightning arresters.

[0094] It also includes anti-interference and adaptive calibration unit, the hardware resists electromagnetic interference through shielding ground and power supply filtering, and the software removes abnormal values through data noise reduction, and the sensor acquisition channel is automatically calibrated every month.

[0095] In summary, the advantages of the application are:

[0096] By synchronously collecting electrical, mechanical and environmental three types of core parameters through multiple types of high-precision sensors, combined with integrated data processing technology, the limitations of traditional single parameter monitoring are broken through, the operation state characteristics of lightning arresters are comprehensively captured, and the fault misjudgment and omission caused by parameter loss are avoided;

[0097] An improved CNN-LSTM hybrid model and a multi-algorithm fusion strategy are adopted to accurately extract data space and time sequence features, the fault identification accuracy is high, combined with the Euclidean distance positioning algorithm and the three-level fault division standard, the fault position is accurately positioned and the severity is graded, the early hidden fault is captured in advance, and time is saved for operation and maintenance disposal;

[0098] Through the attention mechanism LSTM-ARIMA fusion prediction model, the device state deterioration trend is predicted in advance for 24h-72h, combined with the quantitative evaluation of the health state index, the traditional post-maintenance is changed into pre-warning and preventive maintenance, and the fault shutdown risk is reduced.

[0099] Relying on 5G private network to realize remote emergency tripping of serious fault, the instruction response is fast, the fault equipment can be quickly cut off to avoid fault expansion; multi-channel cooperative early warning mechanism ensures that early warning information is timely reached, cooperates with operation and maintenance work order closed loop management, and greatly improves the fault disposal efficiency;

[0100] Adopt InfluxDB+Redis distributed storage architecture to realize long-term storage of full-amount data and multi-dimensional tracing; automatically generate state trend curve, fault statistical report and the like to provide data support for equipment selection, maintenance plan making and life evaluation, and reduce operation and maintenance cost;

[0101] Through the anti-interference design of the combination of the hardware shielding ground, power filter and software data noise reduction, and the abnormal value elimination, and the monthly automatic calibration function, the accuracy and stability of the monitoring data are ensured in the complex power grid electromagnetic environment and the outdoor severe conditions, and reliable data basis is provided for the intelligent monitoring.

[0102] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection required by the present application is defined by the appended claims and their equivalents.

Claims

1. An online monitoring system for the operating status of drop-out surge arresters, characterized in that, include: Data acquisition and processing module, intelligent status monitoring module, hierarchical early warning module, remote operation and maintenance module, and data storage and traceability module; The data acquisition and processing module is used to collect the electrical operating parameters, mechanical structure parameters and environmental impact parameters of the surge arrester in real time, and generate high-quality status feature data through integrated processing of noise reduction, standardization and feature extraction. The intelligent status monitoring module is electrically connected to the data acquisition and processing module, and adopts a multi-algorithm fusion model to perform real-time evaluation of the arrester's operating status, accurate fault identification and location, and prediction of future status trends. The hierarchical early warning module is electrically connected to the intelligent status monitoring module, dynamically generates hierarchical early warning information based on status monitoring results, and pushes it to relevant operation and maintenance terminals through multiple channels. The remote operation and maintenance module is electrically connected to the hierarchical early warning module to perform emergency remote fault handling, remote configuration of core equipment parameters, and closed-loop management of the entire operation and maintenance process. The data storage and traceability module is electrically connected to the intelligent status monitoring module and the hierarchical early warning module to store and trace all monitoring data, early warning information and operation and maintenance records in multiple dimensions.

2. The online monitoring system for the operating status of drop-out surge arresters according to claim 1, characterized in that, The data acquisition and processing module includes: High-precision Hall sensors, vibration sensors, temperature and humidity sensors, and pollution sensors are used to simultaneously collect the three-phase voltage of the surge arrester, leakage current, vibration parameters of the drop-out mechanism, ambient temperature and humidity, and pollution level of the insulator. First, wavelet packet transform denoising algorithm is used to specifically remove electromagnetic interference and environmental noise. Then, Z-score standardization is used to eliminate the difference in parameter dimensions and extract the peak, mean and variance features of the parameters to generate standardized state feature data.

3. The online monitoring system for the operating status of drop-out surge arresters according to claim 2, characterized in that, The intelligent status monitoring module includes: An improved CNN-LSTM hybrid model is constructed, in which the CNN part extracts the spatial dimension features of the state feature data through 3 convolutional layers and 2 max pooling layers, and the LSTM part captures the temporal correlation characteristics of the data through 2 hidden layers. The output features of the two are fused through a fully connected layer to achieve real-time evaluation of the running state. The model was trained using historical fault data and simulated fault samples covering common faults such as excessive leakage current, insulation aging, and mechanical jamming. The network hyperparameters were then optimized using a grid search method.

4. The online monitoring system for the operating status of drop-out surge arresters according to claim 3, characterized in that, The intelligent status monitoring module also includes: The fault location unit establishes a spatial location model based on the sensor deployment coordinates. By calculating the correlation weights between fault characteristic parameters and the data collected by each sensor, and combining this with an Euclidean distance algorithm, it locates the fault location and identifies faults in key components. The fault location formula is as follows:

5. In the formula, These are fault characteristic parameters. To collect data for the sensor, This is the positioning error; The state trend prediction unit establishes an LSTM-ARIMA fusion prediction model based on an attention mechanism. The attention mechanism automatically assigns weights to the state data from the past 7 days. The prediction model formula is:

6. In the formula, The fusion weights range from 0.6 to 0.

8. For future state predictions, This is the status data for the past 7 days. For the long short-term memory model, It is an autoregressive integral moving average model.

7. The online monitoring system for the operating status of drop-out surge arresters according to claim 4, characterized in that, The intelligent status monitoring module also includes: The fault classification unit establishes a three-level fault classification standard based on the scope of the fault's impact on power grid operation, the urgency of handling it, and the state assessment score, including: minor faults, general faults, and severe faults. The health status quantification unit uses the analytic hierarchy process (AHP) to construct a hierarchical structure and calculates the health status index through weighted summation.

8. The online monitoring system for the operating status of drop-out surge arresters according to claim 5, characterized in that, The specific method of constructing a hierarchical structure using the analytic hierarchy process is as follows: include: The formula for calculating the health status index is as follows:

9. In the formula, The weights of each indicator, These are the normalized values ​​for each indicator. This is a health status index.

10. The online monitoring system for the operating status of drop-out surge arresters according to claim 6, characterized in that, The tiered early warning module includes: The early warning threshold dynamic adaptation unit divides scenarios based on the lightning arrester's service life and the severity of the environment, and automatically adjusts the early warning threshold for each scenario through machine learning algorithms. The structured early warning information generation unit integrates the equipment's unique number, precise location of the fault, fault type, specific HSI value, 72-hour trend prediction results, and targeted handling solutions to generate standardized early warning text. The multi-channel collaborative push unit simultaneously pushes early warning information through four methods: operation and maintenance APP push, SMS notification, Web monitoring platform pop-up, and on-site sound and light alarm. Serious faults will additionally trigger telephone voice notification.

11. The online monitoring system for the operating status of drop-out surge arresters according to claim 7, characterized in that, The remote operation and maintenance module includes: When a serious fault is detected, the emergency fault handling unit automatically sends a remote tripping command to the high-voltage control switch associated with the surge arrester via the 5G private network. After the command is transmitted in encrypted form, it triggers the switch to disconnect the faulty equipment from the power grid. The remote parameter configuration unit allows maintenance personnel to remotely adjust the data acquisition frequency, early warning thresholds for various types of faults, and fusion weights of the prediction model through a web-based visual interface. After parameter adjustment, the data is synchronized to the device terminal in real time to adapt to the operational needs of different regional power grids. The maintenance work order closed-loop unit automatically generates electronic maintenance work orders based on early warning information. These work orders include fault details, processing time limits, and responsible person information. The work order status is updated in real time, and the equipment maintenance records are updated simultaneously.

12. The online monitoring system for the operating status of drop-out surge arresters according to claim 8, characterized in that, The data storage traceability module includes: The system adopts an InfluxDB+Redis distributed time-series database architecture. InfluxDB is used to store all raw monitoring data, standardized status feature data, fault identification results, early warning information and operation and maintenance records, while Redis is used to cache frequently queried data. The full-cycle traceability unit supports multi-dimensional queries of historical data by device number, time range, data type, and fault type. It can automatically generate equipment status change trend curves, fault frequency statistics reports, and operation and maintenance efficiency analysis reports, providing data support for the full life cycle management of surge arresters.

13. The online monitoring system for the operating status of drop-out surge arresters according to claim 9, characterized in that, It also includes an anti-interference and adaptive calibration unit. The hardware resists electromagnetic interference through shielding grounding and power filtering, while the software removes outliers through data noise reduction and automatically calibrates the sensor acquisition channels every month.

Images