A Transmission Line Inspection Method Combining Meteorological Data with Dynamic Confidence Weighting

Abstract

The present invention relates to the technical field of data processing, and particularly to a power transmission line inspection method combining meteorological data with dynamic confidence weighting. By acquiring multi-source meteorological data and performing spatio-temporal reference alignment processing to generate a standardized data set, spatial overlay analysis is carried out with the power grid GIS topological data to construct a meteorological and equipment association topological map to identify threatened equipment; based on a dynamic confidence scoring table, data source weights are assigned, and an optimized meteorological prediction result is generated through weighted fusion and the data interface layer is updated in reverse; in combination with on-site measured data and user correction records, the geographical fence matching rules and weight parameters are dynamically adjusted to achieve closed-loop calibration. The present invention realizes the precise association between meteorological data and the spatial risk of power grid equipment, improves the reliability of extreme weather prediction through multi-source data fusion and adaptive optimization, optimizes the efficiency of emergency resource allocation, reduces the risk of inspection plans being interfered by meteorology, and promotes intelligent decision-making in power operation and maintenance services.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular, to a method for inspecting transmission lines by combining meteorological data with dynamic confidence weighting. Background Art

[0002] In the field of power production operation management and control, the existing technology mainly relies on mobile operation for transmission line inspection, mobile operation for substation operation and maintenance, and operation and maintenance management and control systems to achieve basic operation management, but there are significant defects in the separation of meteorological data and business scenarios. The specific manifestations are as follows: When planning operation and maintenance tasks, real-time meteorological parameters and future weather trend forecasts are not integrated, resulting in inspection and maintenance plans being easily interrupted by sudden bad weather, and it is neither possible to dynamically adjust the operation time sequence through meteorological relevance nor to trigger equipment hidden danger warnings in a timely manner; for extreme meteorological events such as heavy rain and strong wind, there is a lack of a linkage analysis mechanism with power grid geographical space data, and it is impossible to accurately locate substations, transmission lines and maintenance operation points within the scope affected by meteorology, resulting in blindness in the allocation of emergency resources and risk disposal; in addition, the traditional system does not integrate the cross-verification function of multi-source meteorological data (such as numerical weather prediction, wildfire monitoring, icing warning), and the transfer of meteorological warning work orders relies on manual distribution and lacks closed-loop management, resulting in the ineffective release of the decision-making support value of meteorological information in the power operation and maintenance business chain. Summary of the Invention

[0003] Aiming at the deficiencies of the existing technology, the present invention provides a method for inspecting transmission lines by combining meteorological data with dynamic confidence weighting, which solves the technical problems that inspection and maintenance plans are easily interfered by bad weather, the allocation of emergency resources is blind under extreme meteorological events, and the transfer efficiency of multi-source meteorological warning work orders is low due to the lack of deep integration between meteorological data and power operation and maintenance business scenarios.

[0004] To solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0005] The present invention provides a method for inspecting transmission lines by combining meteorological data with dynamic confidence weighting, including:

[0006] Obtain multi-source meteorological data, where the multi-source meteorological data includes numerical weather prediction data, real-time monitoring data and prediction data, and perform spatio-temporal reference alignment processing on the multi-source meteorological data through a protocol parser to generate a standardized meteorological data set. The spatio-temporal reference alignment processing includes unifying the timestamps of different data sources to minute-level accuracy and eliminating spatial reference differences through geographic coordinate interpolation;

[0007] Perform a spatial overlay analysis on the standardized meteorological data set and the power grid GIS topology data, generate a dynamic buffer based on the meteorological influence radius, identify transmission equipment threatened by meteorology through spatial relationship calculation, and construct a meteorology and equipment association topology map. The power grid GIS topology data includes the geographical fences of transmission line corridors and maintenance operation points;

[0008] According to the risk level mapping relationship in the meteorology and equipment association topology map, trigger an event-driven engine to generate a warning work order through a predefined meteorological event threshold. The warning work order encapsulates the identifiers of threatened transmission equipment and disposal suggestions, and is pushed to the operation and maintenance control system through a wireless communication network for closed-loop management of work order issuance, execution, and feedback;

[0009] For the multi-source meteorological prediction data in the same geographical area in the meteorological data interface layer, generate a dynamic confidence score table based on historical error statistical analysis, allocate weights to each data source according to the confidence score table, generate an optimized meteorological prediction result through weighted fusion, and reverse-update the optimized meteorological prediction result to the meteorological data interface layer to overwrite the original single-source data;

[0010] Based on the on-site measured meteorological data and user correction records, calculate the regional deviation between the optimized meteorological prediction result and the measured value, and dynamically adjust the geographical fence matching rule in the spatial overlay analysis and the dynamic weight allocation parameter in the weighted fusion according to the regional deviation to form a closed-loop calibration of meteorological data and equipment risk analysis.

[0011] Furthermore, for the transmission line inspection method combining meteorological data and dynamic confidence weighting in the present invention, the spatio-temporal reference alignment processing of the multi-source meteorological data includes:

[0012] Extract meteorological elements of heterogeneous data sources through a protocol parser, unify the timestamps of the multi-source meteorological data to minute-level accuracy, and eliminate spatial reference differences through a geographic coordinate interpolation algorithm to generate a spatio-temporally continuous standardized meteorological data set;

[0013] The power grid GIS topology data includes the geographical fences of transmission line corridors and maintenance operation points. The spatial overlay analysis is based on the spatial relationship matching between the standardized meteorological data set and the geographical fences, and determines the threat radius of the substation through the calculation of the wind circle range associated with the wind speed.

[0014] Furthermore, for the transmission line inspection method combining meteorological data and dynamic confidence weighting in the present invention, the construction of the meteorology and equipment association topology map includes:

[0015] Divide the dynamic buffer radius according to the meteorological types of heavy rain and strong wind. Based on the threat radius output by the spatial overlay analysis, determine the threatened transmission lines and substations through spatial relationship calculation, and associate the maintenance tasks in the grid GIS topology data to generate a risk level mapping table;

[0016] The early warning work order encapsulates the device identifier, risk type, and disposal suggestions, and is pushed to the mobile operation terminal through the wireless communication network. The wireless communication network uses narrowband Internet of Things technology to transmit the work order data.

[0017] Furthermore, for the transmission line inspection method combining meteorological data and dynamic confidence weighting of the present invention, the generation of the dynamic confidence score table based on historical error statistical analysis includes:

[0018] Based on the historical database of the standardized meteorological data set, statistically analyze the prediction errors of each data source in heavy rain and strong wind scenarios, and use the sliding window statistical method to generate a dynamic confidence score table;

[0019] The dynamic weight assignment is based on the dynamic confidence score table, assigns initial weights according to meteorological types, and performs weight degradation on abnormal forecast values deviating from the population mean through box plot analysis.

[0020] Furthermore, for the transmission line inspection method combining meteorological data and dynamic confidence weighting of the present invention, it further includes:

[0021] Perform linear weighted calculation on the multi-source forecast data according to the dynamic weight assignment parameters, and introduce a time decay factor to adjust the weight ratio for short-term heavy rainfall scenarios. The time decay factor is configured such that the forecast weight ratio for the next 6 hours is 70%, and it decreases to 30% after 24 hours;

[0022] The optimized meteorological prediction result is written back to the meteorological data interface layer through the wireless communication protocol, covering the original single-source data, and providing updated meteorological input for the spatial overlay analysis.

[0023] Furthermore, for the transmission line inspection method combining meteorological data and dynamic confidence weighting of the present invention, the dynamic adjustment of the geographical fence matching rule in the spatial overlay analysis and the dynamic weight assignment parameters in the weighted fusion according to the regional deviation includes:

[0024] Based on the regional deviation between the optimized meteorological prediction result and the measured meteorological data, for heavy rain scenarios, iteratively optimize the buffer radius parameter through the sliding window statistical method;

[0025] The adjustment of the spatio-temporal resolution parameter includes increasing the time granularity of the standardized meteorological data set according to the meteorological mutation scenario mark in the user correction record.

[0026] Furthermore, for the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, it further includes that a weather feedback module collects on-site temperature, humidity and rainfall data in real time and transmits them to the meteorological data interface layer through narrowband Internet of Things;

[0027] The weather feedback module deployed on the mobile operation terminal collects on-site temperature, humidity and rainfall data in real time, and the on-site data is transmitted to the meteorological data interface layer through narrowband Internet of Things;

[0028] The user correction record includes manual adjustment information on the risk level in the early warning work order, and the regional deviation calculation generates calibration parameters based on the grid difference analysis of the on-site data and the optimized meteorological prediction result.

[0029] Furthermore, for the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, the historical database stores the forecast data and corresponding actual data in the meteorological data interface layer in the past 30 days, and statistically calculates the hit rate and mean square error according to rainstorm and strong wind meteorological types;

[0030] The confidence score table is generated based on the historical data and is linked with the on-site measured data collected by the weather feedback module, and the initial weight ratio in the dynamic weight allocation is corrected in reverse through deviation analysis.

[0031] Furthermore, for the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, the wireless communication protocol uses narrowband Internet of Things technology to transmit the optimized meteorological prediction result to the mobile operation terminal, and the narrowband Internet of Things supports low-power wide-area communication;

[0032] The time decay factor is configured such that the predicted weight in the weighted fusion accounts for 70% in the next 6 hours and linearly decreases to 30% after 24 hours, and the decreasing logic is associated with the meteorological mutation scenario marker.

[0033] Furthermore, for the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, it further includes:

[0034] According to the measured rainfall distribution in the rainstorm scenario, dynamically expand the buffer coverage range of the transmission line corridor in the meteorological and equipment association topology map;

[0035] The adjustment of the spatio-temporal resolution parameter triggers the update of the sampling frequency of the meteorological data interface layer, and synchronously updates the dynamic weight allocation parameter based on the regional deviation between the optimized meteorological prediction result and the measured value, forming a closed-loop calibration mechanism for data fusion and equipment risk analysis.

[0036] Advantages of the present invention;

[0037] Through the spatio-temporal benchmark alignment of multi-source meteorological data and the spatial overlay analysis of the power grid GIS, the present invention constructs a topological map of the association between meteorology and equipment to achieve precise positioning of meteorological risks, and combines dynamic confidence scoring and weight allocation to optimize the reliability of prediction results; based on the closed-loop feedback calibration mechanism of on-site measured data and user correction records, dynamically adjusts the geographical fence rules and fusion weight parameters to enhance the adaptability of the model to regional meteorological mutations; through the closed-loop management of warning work orders driven by narrowband Internet of Things, dynamically associates meteorological warnings with equipment status and maintenance resources, significantly improves the accuracy of emergency resource allocation and the timeliness of work order execution under extreme weather, effectively reduces the inspection interruption rate and equipment failure risk caused by meteorological factors, and realizes the full-process intelligent decision-making of power operation and maintenance services driven by meteorological data. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the drawings.

[0039] Figure 1 It is a flowchart of a transmission line inspection method combining meteorological data and dynamic confidence weighting provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the specific embodiments and corresponding drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. The technical solutions provided by each embodiment of the present invention are described in detail below in conjunction with the drawings. To better understand the objectives of the present invention, the present invention is further described in detail below.

[0041] Please refer to Figure 1 , the present invention provides a transmission line inspection method combining meteorological data and dynamic confidence weighting, including:

[0042] Step S101, obtaining multi-source meteorological data in real time through the meteorological data interface layer, where the multi-source meteorological data includes numerical forecast data, real-time monitoring data, and prediction data, and performing spatio-temporal benchmark alignment processing on the multi-source meteorological data through a protocol parser to generate a standardized meteorological data set. The spatio-temporal benchmark alignment processing includes unifying the timestamps of different data sources to minute-level accuracy and eliminating spatial benchmark differences through geographic coordinate interpolation;

[0043] Step S102: Perform a spatial overlay analysis on the standardized meteorological data set and the power grid GIS topology data, generate a dynamic buffer based on the meteorological influence radius, identify transmission equipment threatened by meteorology through spatial relationship calculation, and construct a meteorology and equipment association topology map. The power grid GIS topology data includes the geographical fence of the transmission line corridor and the maintenance operation points.

[0044] Step S103: According to the risk level mapping relationship in the meteorology and equipment association topology map, trigger the event-driven engine to generate a warning work order by predefined meteorological event thresholds. The warning work order encapsulates the identification of the threatened transmission equipment and the disposal suggestions, and is pushed to the operation and maintenance management and control system through the wireless communication network for closed-loop management of work order issuance, execution, and feedback.

[0045] Step S104: For the multi-source meteorological prediction data in the same geographical area in the meteorological data interface layer, generate a dynamic confidence score table based on historical error statistical analysis, allocate weights to each data source according to the confidence score table, generate an optimized meteorological prediction result through weighted fusion, and update the optimized meteorological prediction result back to the meteorological data interface layer to overwrite the original single-source data.

[0046] Step S105: Calculate the regional deviation between the optimized meteorological prediction result and the measured value based on the on-site measured meteorological data and the user correction record, and dynamically adjust the geographical fence matching rule in the spatial overlay analysis and the dynamic weight allocation parameter in the weighted fusion according to the regional deviation to form a closed-loop calibration of meteorological data and equipment risk analysis.

[0047] The present invention provides a transmission line inspection method combining meteorological data and dynamic confidence weighting. The specific implementation process of its technical solution is as follows:

[0048] First, access heterogeneous data sources such as the numerical weather prediction center, the wildfire monitoring platform, and the disaster reduction institute in real time through the meteorological data interface layer to obtain multi-source meteorological data including temperature, rainfall, and wind speed. Aiming at the timestamp differences of different data sources, a protocol parser is used to extract the time tags of meteorological elements, and the timestamps are unified to the minute-level accuracy through GPS time synchronization technology. At the same time, based on the geographical coordinate interpolation algorithm, data with inconsistent spatial references are processed for grid alignment. For example, the longitude and latitude coordinates of the meteorological station are matched with the coordinate system of the power grid GIS to eliminate spatial position deviations and generate a spatio-temporally continuous standardized meteorological data set. This data set provides a unified time-space reference for subsequent analysis and avoids analysis errors caused by data source differences.

[0049] When performing spatial overlay analysis on the standardized meteorological dataset and the power grid GIS topology data, the system calls the geographic fence information of the transmission line corridor and the coordinates of the maintenance operation points, and dynamically generates a buffer zone in combination with the meteorological influence radius. For example, in the case of strong wind scenarios, the wind circle range is calculated based on real-time wind speed data, and substations and transmission lines within the coverage of the wind circle are identified through a spatial relationship matching algorithm. At the same time, based on the rainfall intensity distribution predicted by heavy rain, a circular buffer zone centered on the transmission equipment is generated to determine the list of equipment that may be threatened by waterlogging or debris flow. Through spatial overlay analysis, a meteorological and equipment association topology map is constructed, which records the spatial positions of threatened equipment, meteorological risk types, and associated maintenance tasks in the form of structured data.

[0050] When the risk level in the meteorological and equipment association topology map reaches the predefined threshold, the event-driven engine is triggered. For example, in a heavy rain scenario where the predicted rainfall in a certain area exceeds 50 mm / 24 h, the system automatically generates a warning work order, which encapsulates the unique identifier of the threatened equipment, the risk type (such as "risk of insulator flashover caused by heavy rainfall"), and the disposal suggestions (such as "prioritize inspections of pole numbers T-103 to T-115"). The generated work order is transmitted to the operation and maintenance control system through the narrowband Internet of Things. The system automatically allocates maintenance resources according to the priority of the work order and receives on-site feedback information after the work order is completed, forming a closed-loop management process of issuance, execution, and feedback to ensure the timeliness of risk disposal.

[0051] For multi-source meteorological prediction data in the same geographical area, the system statistically analyzes the error distributions of each data source based on the historical database. For example, compare the mean square errors of the heavy rain prediction data and the actual rainfall of the meteorological station and Xiangji Technology in the past 30 days, and use the sliding window statistical method to dynamically generate a confidence score table. Initial weights are assigned to different data sources according to the scoring results, and abnormal forecast values deviating from the population mean are identified through box plot analysis, and their weights are dynamically downgraded. Subsequently, linear weighted fusion is performed on the multi-source data to generate an optimized meteorological prediction result. This result is written back to the meteorological data interface layer through a wireless communication protocol, covering the original single-source data, providing updated input for subsequent spatial overlay analysis, and forming a data quality improvement loop.

[0052] Based on the on-site measured meteorological data (such as temperature and humidity) collected by the mobile operation terminal and the manual correction records of the user for the warning work orders, the system calculates the regional deviation between the optimized prediction result and the measured value. For example, a deviation matrix of each geographical unit is generated through grid-based difference analysis to identify the areas with large prediction errors. According to the deviation results, the geographical fence matching rules in the spatial overlay analysis are dynamically adjusted. For example, the coverage radius of the transmission line buffer zone in the heavy rain scenario is expanded; at the same time, the dynamic weight allocation parameters in the weighted fusion model are updated to reduce the weight ratio of the data sources with large errors. The above adjustments are synchronized to the meteorological data interface layer and the operation and inspection control system through the narrowband Internet of Things, realizing the closed-loop calibration of meteorological data and equipment risk analysis, and continuously improving the regional adaptability of the model.

[0053] Each step is closely related through the data flow and feedback mechanism: the standardized meteorological data set provides input for the spatial overlay analysis, the generated topological map drives the automatic triggering of the warning work order, the optimized prediction data updates the interface layer in reverse to improve the subsequent analysis accuracy, and the measured data and user feedback further calibrate the model parameters, forming a complete technical closed-loop from data collection to decision optimization.

[0054] Specifically, in the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the spatio-temporal reference alignment processing of the multi-source meteorological data includes:

[0055] The meteorological elements of the heterogeneous data sources are extracted through a protocol parser, the timestamps of the multi-source meteorological data are unified to the minute-level accuracy, and the spatial reference differences are eliminated through the geographic coordinate interpolation algorithm to generate a spatio-temporally continuous standardized meteorological data set;

[0056] The power grid GIS topological data includes the geographical fences of the transmission line corridors and the maintenance operation points. The spatial overlay analysis is based on the spatial relationship matching between the standardized meteorological data set and the geographical fences, and the threat radius of the substation is determined through the calculation of the wind circle range associated with the wind speed.

[0057] In the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the specific implementation processes of the spatio-temporal reference alignment processing and the spatial overlay analysis are as follows:

[0058] The protocol parser interfaces with the heterogeneous data interfaces of meteorological observatories, disaster reduction institutes, and wildfire monitoring platforms, and extracts core meteorological elements such as temperature, rainfall, and wind speed according to the data formats of different data sources (such as JSON, GRIB2). For example, the parser identifies the timestamp field in the meteorological observatory data, synchronizes the GPS clock with the time tag of the disaster reduction institute data, unifies the time reference to the minute-level accuracy, and eliminates the time-axis misalignment problem caused by the difference in data collection frequencies. At the same time, the geographical coordinate interpolation algorithm is used to grid the data with inconsistent spatial references. For example, the WGS84 coordinate system used by the meteorological observatory is aligned with the CGCS2000 coordinate system of the power grid GIS through the bilinear interpolation algorithm to generate a spatio-temporally continuous standardized meteorological dataset, providing a unified data input reference for subsequent analysis.

[0059] When performing spatial overlay analysis on the standardized meteorological dataset and the power grid GIS topological data, the system calls the geographical fence coordinates of the transmission line corridor and the information of the maintenance operation points, and generates a dynamic buffer in combination with the spatial distribution characteristics of the meteorological elements. For example, for the strong wind scenario, based on the real-time wind speed data in the standardized dataset, the wind speed attenuation model centered on the substation is calculated, and the fan-shaped wind circle range associated with the wind speed level is dynamically generated. The transmission towers and line sections within the coverage of the wind circle are identified through the spatial relationship matching algorithm. For the heavy rain scenario, a radial buffer centered on the power transmission equipment is generated based on the rainfall intensity prediction data, and the spatial boundary of the water accumulation risk area is corrected in combination with the terrain elevation data.

[0060] The above spatial overlay analysis further determines the threat radius of the substation through the wind speed correlation model. Specifically, a wind speed-equipment failure rate correlation model is established based on the historical meteorological disaster case library. When the real-time wind speed exceeds the preset threshold, the threat radius is dynamically calculated according to the equipment type (such as insulators, conductors) and service life. For example, for the old transmission lines that have been in service for more than 15 years, when the wind speed reaches level 8, the threat radius expands to 1.5 times the reference value. The affected equipment is marked through the spatial topology relationship matching algorithm, and the threat radius data is written into the meteorological and equipment association topology map, providing a spatial analysis basis for subsequent risk level mapping.

[0061] In the above processing process, the standardized dataset generated by the spatio-temporal reference alignment provides calibrated input data for the spatial overlay analysis, while the threat radius data output by the spatial analysis directly drives the risk level assessment and the generation of early warning work orders, forming a technical closed-loop from data standardization to spatial decision-making. The collaborative application of the geographical coordinate interpolation algorithm and the wind speed correlation model solves the problem of equipment positioning deviation caused by inconsistent spatial references of multi-source data, and improves the accuracy of determining the meteorological influence range.

[0062] Specifically, in the transmission line inspection method combining meteorological data with dynamic confidence weighting of the present invention, the construction of the meteorological and equipment associated topological map includes:

[0063] Divide the dynamic buffer radius according to meteorological types such as heavy rain and strong wind. Based on the threat radius output by the spatial overlay analysis, determine the threatened transmission lines and substations through spatial relationship calculation, and associate the maintenance tasks in the grid GIS topological data to generate a risk level mapping table;

[0064] The warning work order encapsulates the device identifier, risk type, and disposal suggestions, and is pushed to the mobile operation terminal through the wireless communication network. The wireless communication network uses narrowband Internet of Things technology to transmit the work order data.

[0065] In the transmission line inspection method combining meteorological data with dynamic confidence weighting of the present invention, the specific implementation process of constructing the meteorological and equipment associated topological map and generating the warning work order is as follows: According to the characteristic differences of heavy rain and strong wind meteorological types, the system dynamically divides the buffer generation rules. For the heavy rain scenario, based on the historical water accumulation point distribution data and the predicted value of the real-time rainfall intensity, a radial buffer is generated with the transmission line as the center. The buffer radius expands non-linearly as the rainfall magnitude increases. For example, when the predicted hourly rainfall exceeds 30 mm, the buffer radius expands to 2 times the reference value; for the strong wind scenario, based on the wind speed monitoring data and the tower structure strength parameters, a fluid mechanics model is used to calculate the wind pressure influence range, and a fan-shaped buffer associated with the wind direction is generated. The buffer radius is positively correlated with the square value of the wind speed.

[0066] Based on the threat radius data output by the spatial overlay analysis, determine the threatened equipment through spatial topological relationship calculation. The system calls the tower coordinates, line orientation, and substation location data in the grid GIS, and performs a spatial intersection operation with the meteorological buffer to identify the equipment in the overlapping area. For example, when the heavy rain buffer covers a certain section of the transmission line, associate the historical defect records (such as the aging degree of the insulator) and maintenance plan of the line to generate a risk level mapping table. The table records the equipment code, risk coefficient (such as high risk, medium risk), recommended disposal measures (such as completing insulation detection within 72 hours), and associated maintenance team information, forming a dynamic matching relationship between equipment risks and operation and maintenance resources.

[0067] When generating a warning work order, the system extracts a list of high-priority devices from the risk level mapping table, encapsulating the unique device identification code, risk type code, and standardized disposal instructions. For example, for the risk of conductor galloping caused by strong winds, the device positioning coordinates, predicted galloping amplitude, and inspection specifications for tightening fittings are embedded in the work order; for the risk of tower foundation scouring caused by heavy rain, the foundation settlement detection process and emergency reinforcement plan are added. The encapsulated work order data is transmitted to the mobile operation terminal through narrowband Internet of Things. Utilizing the wide coverage feature of NB-IoT technology, it maintains stable transmission of work order data in weak signal areas such as mountains and the wild. After receiving, the terminal automatically triggers the functions of patrol navigation path planning and operation process reminder.

[0068] In the above process, the dynamic buffer division rule provides the input of the meteorological influence range for spatial overlay analysis. The calculation result of the spatial topological relationship drives the generation of the risk level mapping table, and the structured data of the mapping table supports the accurate encapsulation and directional push of the warning work order. The application of narrowband Internet of Things technology converts the meteorological analysis result into an executable on-site operation instruction, forming a technical closed-loop from risk identification to disposal execution, solving the problems of low efficiency and large information attenuation in traditional manual work order dispatching. The combination of meteorological type differential processing and equipment status correlation analysis improves the matching accuracy of risk warning and operation and maintenance decision-making.

[0069] Specifically, in the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the generation of the dynamic confidence score table based on historical error statistical analysis includes:

[0070] Based on the historical database of the standardized meteorological data set, the prediction errors of each data source in the heavy rain and strong wind scenarios are statistically analyzed. The window period of the sliding window statistical method is 7 days, and the hit rate and mean square error of each data source within the window are updated daily. The hit rate is defined as the proportion of data with prediction errors within the range of ±2°C or ±5mm.

[0071] The dynamic weight allocation is based on the dynamic confidence score table, initial weights are assigned according to meteorological types, and weight degradation is performed on abnormal forecast values that deviate from the population mean through box plot analysis.

[0072] In the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the specific implementation process of confidence evaluation and dynamic weight allocation is as follows:

[0073] Based on the historical meteorological data in the standardized meteorological dataset, the system stores the prediction records and corresponding actual data of each data source classified by meteorological types such as heavy rain and strong wind. For example, the heavy rain prediction data of the meteorological station, the disaster reduction institute, and the third-party meteorological service provider in the past 30 days are extracted and compared hour by hour with the monitored actual rainfall values, and the absolute error distributions of each data source in the 24-hour, 48-hour, and 72-hour prediction periods are statistically analyzed. The sliding window statistical method is used to dynamically calculate the error indicators. The window period is set to 7 days, and the hit rate (such as the proportion of data within the prediction error of ±2°C) and the mean square error of each data source within the window are updated daily to generate a dynamically updated confidence score table. In the score table, priority labels are assigned to each data source according to the meteorological type.

[0074] The dynamic weight allocation module assigns initial weights to data sources in different meteorological scenarios according to the confidence score table. For example, in the heavy rain scenario, if the historical hit rate of the meteorological station data is 85% and the disaster reduction rate is 72%, the initial weights are set to 55% and 30% respectively, and the remaining weights are assigned to other data sources. For the multi-source prediction data received in real time, the box plot analysis method is used to identify outliers that deviate from the population mean. For example, when the predicted rainfall value of a data source exceeds 1.5 times the interquartile range of the predicted values of other data sources in the same region, it is determined as an abnormal forecast, and the weight degradation mechanism is triggered to temporarily adjust the real-time weight of this data source to 50% of the initial value until its predicted value returns to the normal fluctuation range.

[0075] The updated results of the above confidence score table and dynamic weight allocation parameters act on the weighted fusion process of multi-source meteorological data in real time. For example, in the short-term heavy rainfall scenario, the meteorological station data obtains a higher weight due to its lower recent errors, and its prediction results dominate the fusion calculation; when the third-party data source has an abnormally high predicted value, the box plot analysis automatically reduces its weight proportion to avoid the interference of outliers on the fusion result. The optimized meteorological prediction results are written back to the meteorological data interface layer in reverse, replacing the original single-source data, providing quality-calibrated inputs for subsequent spatial overlay analysis, and at the same time updating the prediction records in the historical database, forming a data quality closed-loop from historical error analysis to real-time weight adjustment.

[0076] The synergistic effect of confidence evaluation and dynamic weight allocation is reflected in that: historical error statistics provide a benchmark basis for weight allocation, real-time anomaly detection dynamically corrects weight parameters, and the optimized prediction results continuously iterate the historical database. The three form a technical closed-loop of self-optimization of data quality. This design solves the problem of unstable prediction results caused by the accuracy fluctuation of multi-source meteorological data and improves the reliability of risk early warning in extreme weather scenarios.

[0077] Specifically, the method for inspecting transmission lines by combining meteorological data with dynamic confidence weighting described in the present invention further includes:

[0078] Perform linear weighted calculation on the multi-source forecast data according to the dynamic weight allocation parameters, and introduce a time decay factor to adjust the weight ratio for the short-term heavy rainfall scenario. The time decay factor is configured such that the predicted weight ratio for the next 6 hours is 70%, and it decreases to 30% after 24 hours.

[0079] The optimized meteorological prediction results are written back to the meteorological data interface layer through a wireless communication protocol, covering the original single-source data, and providing updated meteorological inputs for the spatial overlay analysis.

[0080] In the transmission line inspection method combining meteorological data and dynamic confidence weighting according to the present invention, the specific implementation process of weighted fusion and data writing back is as follows: perform linear weighted calculation on multi-source meteorological forecast data based on dynamic weight allocation parameters, where the weight parameters are derived from the historical confidence score table and real-time anomaly detection results. For example, in the heavy rain scenario, the data from the meteorological station is assigned a weight of 55% due to low recent errors, the weight of the third-party data source is reduced to 15% due to anomalies detected by the box plot, and the remaining weight is assigned to other data sources. The predicted values of multi-source rainfall are weighted and summed according to the proportion to generate a regionalized meteorological prediction result. For the short-term heavy rainfall scenario, introduce a time decay factor to adjust the weight distribution, set the weight ratio of the predicted data for the next 6 hours to 70%, and the weight of the predicted data after 24 hours linearly decreases to 30% to strengthen the decision-making influence of the recent predicted data.

[0081] The optimized meteorological prediction results are written back to the meteorological data interface layer through narrowband Internet of Things technology, replacing the original single-source data. For example, the fused predicted value of rainfall for 72 hours covers the original prediction data of the meteorological station, and at the same time updates the data timestamp and spatial identifier to form the latest version of the standardized meteorological data set. The updated data set is used as an input to re-perform the spatial overlay analysis, driving the real-time refresh of the meteorological and equipment association topology map, ensuring that risk warning work orders are generated based on the latest meteorological information.

[0082] During the data writing back process, the wireless communication protocol uses narrowband Internet of Things technology with low-power wide-area coverage to maintain data transmission stability in weak signal areas along the transmission line. The write-back operation triggers the version management mechanism of the meteorological data interface layer, retains historical version data for deviation backtracking analysis, and at the same time provides an iterative optimization data basis for the dynamic weight allocation module. The spatial overlay analysis module calls the updated meteorological data to recalculate the buffer range and risk level of the threatened equipment, forming a positive feedback link from data fusion to decision optimization.

[0083] The synergistic effect of dynamic weights and time decay factors effectively avoids prediction biases caused by anomalies in a single data source, while the data write-back mechanism enables the dynamic replacement of original data with the results of multi-source fusion. The continuous optimization of meteorological prediction results improves the accuracy of overlay analysis, thereby enhancing the effectiveness of warning work order handling and forming a closed-loop for the mutual improvement of meteorological data quality and business decision-making accuracy.

[0084] Specifically, in the transmission line inspection method combining meteorological data and dynamic confidence weighting of the present invention, the dynamic adjustment of the geographical fence matching rule in the spatial overlay analysis and the dynamic weight allocation parameter in the weighted fusion according to the regional deviation includes:

[0085] Based on the regional deviation between the optimized meteorological prediction result and the measured meteorological data, for the rainstorm scenario, the buffer radius parameter is iteratively optimized by the sliding window statistical method.

[0086] The adjustment of the spatio-temporal resolution parameter includes increasing the time granularity of the standardized meteorological data set according to the meteorological mutation scenario marker in the user correction record.

[0087] In the transmission line inspection method combining meteorological data and dynamic confidence weighting of the present invention, the specific implementation processes of the geographical fence matching rule and the spatio-temporal resolution adjustment are as follows: Based on the optimized meteorological prediction result and the measured meteorological data reported by the mobile operation terminal, the system performs grid-based regional deviation analysis. For example, the transmission line corridor is divided into 1 km × 1 km geographical units, the absolute deviation between the predicted rainfall and the measured value in each unit is compared, and the average deviation rate in the past 7 days is calculated using the sliding window statistical method. When the deviation rate exceeds 15%, the iterative optimization of the buffer radius parameter is triggered. The optimization process corrects the model based on historical disaster data. For example, in the rainstorm scenario, if the measured rainfall in a certain area is continuously higher than the predicted value, the buffer radius is expanded by 5% per day, and the spatial threat range in the meteorological and equipment association topology map is updated.

[0088] The adjustment of the spatio-temporal resolution parameter is based on the meteorological mutation marker in the user correction record. When the user manually marks a short-term heavy rainfall or sudden strong wind in a certain area through the mobile operation terminal, the system automatically extracts the meteorological data stream in that area and increases the time granularity from the hourly level to the 10-minute level. For example, for the meteorological mutation scenario in the marked area, the time stamp interval of the standardized meteorological data set is compressed from 60 minutes to 10 minutes, and the data sampling frequency of the spatial overlay analysis module is synchronously adjusted to enhance the capture ability of rapid fluctuations in meteorological parameters.

[0089] The optimized buffer radius parameter and time granularity parameter are transmitted back to the meteorological data interface layer via narrowband Internet of Things, driving the real-time update of spatial overlay analysis. The updated meteorological and device association topology map recalculates the list of threatened devices and generates a corrected warning work order for pushing to the terminal. At the same time, the adjusted time granularity data provides high-precision input for the dynamic weight allocation module, optimizing the timeliness of multi-source data fusion. The interaction between user correction marks and the system's automatic optimization results forms a two-way calibration mechanism for meteorological data analysis and business decision-making.

[0090] The dynamic expansion of the geofence parameter and the scenario-based adjustment of the time granularity work together to solve the problem that traditional static rules are difficult to adapt to regional meteorological mutations. The buffer optimization driven by measured data reduces the false alarm risk caused by prediction deviation, and the enhanced time granularity improves the response speed to short-term extreme weather. The two continuously improve the system's adaptability to complex meteorological conditions through closed-loop feedback.

[0091] Specifically, the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention further includes that the weather feedback module collects on-site temperature, humidity and rainfall data in real time and transmits it to the meteorological data interface layer via narrowband Internet of Things;

[0092] The weather feedback module deployed on the mobile operation terminal collects on-site temperature, humidity and rainfall data in real time, and the on-site data is transmitted to the meteorological data interface layer via narrowband Internet of Things;

[0093] The user correction record includes manual adjustment information on the risk level in the warning work order, and the regional deviation calculation generates calibration parameters based on the grid difference analysis of the on-site data and the optimized meteorological prediction result.

[0094] In the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the specific implementation process of weather feedback module data collection and regional deviation calibration is as follows: The mobile operation terminal integrates temperature, humidity and rainfall sensors to collect on-site meteorological data of the transmission line corridor in real time. The sensor data is subjected to format standardization processing by the edge computing node and converted into a structured data format compatible with the meteorological data interface layer. For example, the analog signal is converted into a JSON data packet with a time stamp, and the device geographical location code is attached. The standardized on-site data is transmitted to the meteorological data interface layer via narrowband Internet of Things. The low-power consumption characteristic of NB-IoT is used to realize long-cycle data backhaul in the field environment, and the transmission interval is dynamically adjusted according to the network signal strength, and the lowest supports a data reporting frequency of once every 5 minutes.

[0095] The user correction record module receives the manually adjusted risk level information submitted by the mobile work terminal. For example, when the on-site inspection personnel find that the actual wind speed around a certain tower is lower than the predicted value in the early warning work order, they lower the risk level from "high risk" to "medium risk" through the terminal interface. The correction record includes the equipment code, correction time, and correction basis text. The system extracts the geographical location information in the correction record, correlates and optimizes the predicted value of the corresponding grid in the meteorological prediction result, and performs grid-based difference analysis.

[0096] The regional deviation calculation module matches the on-site measured data with the optimized prediction result according to the 1km×1km grid cells, and calculates the absolute deviation values of temperature, humidity, and rainfall in each cell. For the grid cells with user correction records, a weight coefficient of 3 is assigned to strengthen the deviation impact, and a calibration parameter matrix with weight distribution is generated. The calibration parameters are synchronized to the dynamic weight allocation module through the meteorological data interface layer, triggering real-time adjustment of the multi-source data fusion weights. For example, reducing the initial weight ratio of the data source with a large deviation in the recent period.

[0097] The calibration parameters also act on the iterative optimization of the geographical fence matching rules. For example, when the measured rainfall value in a certain grid cell is continuously higher than the predicted value and there is a user correction record, the coverage radius of the buffer zone of the transmission line in this area is extended, and the spatial threat range in the meteorological and equipment association topology map is updated. The real-time data transmitted by the narrowband Internet of Things and the user correction information form a two-way data stream, supporting the continuous optimization of the meteorological prediction model and business decision rules, and constructing a closed-loop feedback mechanism from on-site perception to system calibration.

[0098] The synergistic effect of on-site data collection and user correction is reflected in that: sensor data provides an objective monitoring basis, and manual correction injects business experience judgment. The two are transformed into quantitative calibration parameters through grid-based difference analysis, driving the adaptive adjustment of the meteorological prediction model and risk warning rules, and enhancing the regional adaptation ability of the system under complex terrain conditions.

[0099] Specifically, for the transmission line inspection method combining meteorological data and dynamic confidence weighting described in the present invention, the historical database stores the forecast data and corresponding actual data in the meteorological data interface layer in the past 30 days, and statistically calculates the hit rate and mean square error according to the meteorological types of heavy rain and strong wind;

[0100] The confidence score table is generated based on the historical data, and is linked with the on-site measured data collected by the weather feedback module, and the initial weight ratio in the dynamic weight allocation is reversely corrected through deviation analysis.

[0101] In the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, the specific implementation process of the linkage between the historical database and the confidence score is as follows: The historical database stores the forecast data and corresponding actual data in the past 30 days in the meteorological data interface layer classified by meteorological types. For the heavy rain scenario, the 24-hour cumulative rainfall prediction values from data sources such as meteorological observatories and disaster reduction institutes are extracted and compared hour by hour with the measured values of meteorological monitoring stations, and the hit rate indicators of each data source are statistically calculated. For example, the proportion of data with a prediction error within ±5 mm; for the strong wind scenario, the mean square error of the wind speed prediction of each data source is statistically calculated, and the window period is set to be updated in a rolling manner for 7 days. The system generates a dynamic confidence score table based on the hit rate and the mean square error. The real-time score values for each meteorological type are recorded in the score table according to the data source identifier, and the score fluctuation trend in the recent 3 days is marked.

[0102] A linkage mechanism is established between the confidence score table and the on-site measured data of the weather feedback module. When the mobile operation terminal reports the measured rainfall data in a certain area, the system extracts the corresponding optimized meteorological prediction result in this area and calculates the grid deviation rate between the predicted value and the measured value. If the deviation rate of the same data source exceeds the threshold within 3 consecutive window periods, a downgrading operation of the dynamic weight distribution parameter is triggered. For example, the historical hit rate of a certain third-party data source in the heavy rain scenario is 78%, but due to the recent increase in the measured deviation rate to 12%, its initial weight is dynamically adjusted from 25% to 15%, and at the same time, the weight proportion of the data source with a lower deviation rate is increased.

[0103] The dynamic weight adjustment result is updated in real time to the multi-source data weighted fusion module, and the optimized meteorological prediction data is written back to the historical database in reverse, forming a closed loop for iterative optimization of data quality. The historical database synchronously records the weight adjustment log, including the data source identifier, adjustment time and adjustment range that trigger the downgrading, providing a traceability basis for the subsequent update of the confidence score table. The linkage mechanism between the confidence score table and the weight distribution parameter enables the parameter adjustment of the meteorological prediction model to refer to both the long-term historical performance and the short-term real-time deviation, enhancing the adaptability of the model to different meteorological scenarios.

[0104] The classification storage of historical data provides a statistical basis for the confidence score, the real-time deviation analysis injects a dynamic correction basis for the weight distribution, and the weight adjustment result continuously iterates the content of the historical database. The three form a technical closed loop through the data flow and parameter update mechanism, solving the problem that the traditional static weight distribution method is difficult to adapt to the accuracy fluctuation of meteorological data sources and improving the regional accuracy of the multi-source data fusion result.

[0105] Specifically, in the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, the wireless communication protocol uses narrowband Internet of Things technology to transmit the optimized meteorological prediction result to the mobile operation terminal, and the narrowband Internet of Things supports low-power wide-area communication;

[0106] The time decay factor is configured such that the weight of the 6-hour future prediction in the weighted fusion accounts for 70%, and the weight linearly decreases to 30% after 24 hours. The decreasing logic is associated with the meteorological mutation scenario marker.

[0107] In the transmission line inspection method combining meteorological data and dynamic confidence weighting according to the present invention, the specific implementation process of narrowband Internet of Things communication and time decay factor configuration is as follows: The narrowband Internet of Things technology uses an authorized frequency band to achieve low-power transmission of optimized meteorological prediction results, and matches the signal strength change along the transmission line through an adaptive modulation and coding mechanism. For example, in areas with weak signal coverage in mountainous areas, the terminal device switches to the BPSK modulation method to reduce the transmission rate to maintain connection stability, while in plain areas, the QPSK modulation is used to improve data transmission efficiency. The optimized meteorological prediction results are encapsulated into lightweight data packets, appended with timestamps and geographical grid codes, and transmitted to the mobile operation terminal through the narrowband Internet of Things after encryption. After the terminal device parses the data packet, it triggers local warning prompts and navigation path updates.

[0108] The configuration of the time decay factor is dynamically associated with the meteorological mutation scenario marker. When the user modifies the record or the real-time monitoring data marks that there is a short-time heavy rainfall in a certain area, the weighted fusion module activates the time decay factor, sets the weight of the 6-hour future prediction data to account for 70%, and the weight decreases by 10% every 6 hours thereafter, until it drops to 30% after 24 hours. For example, in the case of a rainstorm mutation scenario, the weight of the 3-hour future prediction data provided by the meteorological station is increased to 80%, while the weight of the prediction data after 24 hours drops to 20%, strengthening the decision-making influence of recent data. The decay logic parameters are stored in the rule library of the dynamic weight allocation module, supporting dynamic adjustment of the decreasing gradient according to meteorological types and regional characteristics.

[0109] The optimized data transmitted by the narrowband Internet of Things is synchronously written back to the meteorological data interface layer, triggering real-time updates of the spatial overlay analysis module. The updated meteorological prediction results and time decay factor parameters jointly act on the dynamic weight allocation process, forming a technical closed-loop from data transmission to model iteration. The warning information received by the mobile operation terminal contains time decay marks. For example, the high-risk period in the next 6 hours is highlighted on the interface to assist the inspection personnel in prioritizing the handling of approaching meteorological threats.

[0110] The low-power characteristic of the narrowband Internet of Things and the dynamic configuration of the time decay factor work together to solve the problem of balancing long-distance data transmission and prediction timeliness in the field environment. The meteorological mutation scenario marker triggers targeted adjustment of the weight decay logic, while the on-site data transmitted back by the Internet of Things reversely calibrates the decay parameters. The two improve the spatio-temporal matching accuracy of meteorological prediction and business handling through closed-loop feedback.

[0111] Specifically, the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention further includes:

[0112] Dynamically expanding the buffer coverage range of the transmission line corridor in the meteorological and equipment associated topology map according to the measured rainfall distribution in the rainstorm scenario;

[0113] The adjustment of the spatio-temporal resolution parameter triggers the update of the sampling frequency of the meteorological data interface layer, and synchronously updates the dynamic weight allocation parameter based on the regional deviation between the optimized meteorological prediction result and the measured value, forming a closed-loop calibration mechanism for data fusion and equipment risk analysis.

[0114] In the transmission line inspection method combining meteorological data with dynamic confidence weighting according to the present invention, the specific implementation process of the geographical fence optimization and closed-loop calibration is as follows: Based on the measured rainfall distribution data in the rainstorm scenario, the system performs a grid-based waterlogging risk analysis of the transmission line corridor. The measured data is compared with the optimized meteorological prediction result according to a 1km×1km geographical unit, and the area where the prediction deviation rate exceeds 20% is identified. Combining the historical pole tilt records and geological settlement data, the buffer coverage range is dynamically expanded. For example, in a certain area, the measured rainfall is 15% higher than the predicted value and there are 3 historical landslide records, the buffer radius is expanded from the reference 500 meters to 800 meters, the spatial threat boundary in the meteorological and equipment associated topology map is updated, and the maintenance priority mark of the affected poles is associated.

[0115] The adjustment of the spatio-temporal resolution parameter is triggered by the meteorological mutation mark in the user correction record. When the mobile operation terminal marks a short-term heavy rainfall in a certain area, the sampling frequency of the meteorological data interface layer is increased from once per hour to once every 10 minutes, and the time granularity of the standardized meteorological data set is synchronously refined to the minute level. The high-precision data stream is input into the spatial overlay analysis module, driving the real-time iteration of the dynamic buffer model. At the same time, the calibration parameter is transmitted back to the dynamic weight allocation module through the narrowband Internet of Things, reducing the weight ratio of the data source with a large recent deviation.

[0116] The synchronous update of the calibration parameter and the weight allocation parameter forms a cross-step closed loop. For example, in the rainstorm mutation scenario, the data with improved spatio-temporal resolution identifies that the predicted value of a certain data source continuously deviates from the measured value. The dynamic weight allocation module reduces its weight from 30% to 15%, and at the same time expands the geographical fence range of this area. The updated parameter set is written into the historical database through the meteorological data interface layer, providing data support for the iteration of the subsequent confidence score table, forming a complete feedback link from on-site perception to model optimization.

[0117] The dynamic expansion of the geographical fence and the scenario-based adjustment of the spatio-temporal resolution work together to solve the problem that traditional static models are difficult to adapt to sudden regional meteorological changes. The buffer optimization driven by measured data reduces the risk of missed reports, the increased sampling frequency enhances the response sensitivity to short-term extreme weather, and the parameter synchronization and update mechanism realizes the multi-module linkage of meteorological analysis, weight allocation, and risk warning, improving the overall decision-making accuracy of the system under complex meteorological conditions.

[0118] The specific embodiments of the technical solution of the present invention are as follows:

[0119] Embodiment 1;

[0120] In the application of mobile operation for power transmission inspection, the system calls the predicted rainfall data for the next 72 hours in the standardized meteorological dataset and performs spatial matching with the GIS coordinates of the power transmission line inspection section. Through the interface, the rainfall intensity curve and risk level (low / medium / high) for the next 3 days in a certain section (such as from pole tower number T-103 to T-115) are displayed, and the inspection personnel adjust the inspection priority according to the prediction results. If the predicted rainfall exceeds 50 mm / 24 h, the system automatically marks this section as "high risk" and recommends delaying the inspection or enabling the unmanned aerial vehicle inspection plan.

[0121] Embodiment 1 solves the problem of the execution interruption of traditional inspection tasks caused by sudden weather changes through the dynamic association of meteorological data and inspection plans, improving the timeliness of plan adjustment.

[0122] Embodiment 2;

[0123] In the power grid single map interface, the rainstorm warning area covers the power transmission line corridor with a red semi-transparent layer, and the system automatically calculates the affected substations (such as substation A) and lines (such as line L-205). Through spatial overlay analysis, a list of affected equipment is generated and associated with maintenance tasks. For example, if the rainstorm buffer covers 3 pole towers of line L-205, the system marks this section of the line as "high risk of foundation scouring" and pushes the reinforcement plan to the maintenance team.

[0124] Embodiment 2 accurately locates the equipment affected by meteorology based on spatial overlay analysis, optimizes the allocation of emergency resources, and reduces the blindness of manual investigation.

[0125] Embodiment 3;

[0126] After the high wind warning work order is issued, the mobile operation terminal receives the work order information and displays the threatened substations (such as substation B) and the list of fittings to be inspected. After the inspection personnel confirm on-site that the actual wind speed is lower than the predicted value, they submit a correction feedback through the terminal, and the system dynamically reduces the priority of subsequent warning work orders in this area. After the work order is completed, the system automatically archives and generates an analysis report, and updates the confidence score table under the high wind scenario.

[0127] Example 3 forms a closed-loop link from early warning issuance to execution optimization through work order feedback and dynamic adjustment of model parameters, improving the matching accuracy between meteorological early warnings and actual business scenarios.

[0128] The technical solution of the present invention realizes the deep integration of meteorological data and power operation and maintenance services through the following steps:

[0129] The present invention accesses heterogeneous data sources such as numerical weather prediction centers, wildfire monitoring platforms, and disaster reduction institutes through a meteorological data interface layer to obtain multi-source meteorological data such as temperature, rainfall, and wind speed. A protocol parser is used to extract meteorological elements from each data source, and the timestamps are unified to minute-level accuracy through GPS time synchronization technology. Based on a geographic coordinate interpolation algorithm, the spatial reference differences are eliminated (such as aligning the WGS84 coordinate system with the CGCS2000 coordinate system of the power grid GIS), generating a spatio-temporally continuous standardized meteorological data set. For example, grid interpolation processing is performed on the real-time monitoring data of meteorological stations and disaster reduction institutes to form 5km×5km grid data covering the transmission line corridor, providing a unified spatio-temporal reference for subsequent analysis. The geographic coordinate interpolation algorithm uses bilinear interpolation to align meteorological data in the WGS84 coordinate system with the CGCS2000 coordinate system of the power grid GIS.

[0130] Perform spatial overlay analysis on the standardized meteorological data set and the power grid GIS topology data, and dynamically generate buffers based on the meteorological influence radius. For the heavy rain scenario, a radial buffer centered on the transmission tower is generated according to the rainfall intensity prediction data, and the boundary of the water accumulation risk area is corrected in combination with the terrain elevation; for the strong wind scenario, the wind circle range of the substation is calculated through a wind speed attenuation model to identify the threatened transmission lines. For example, when the real-time wind speed reaches level 8, the system automatically marks the old lines in service for more than 15 years within the threat radius, constructing a meteorological and equipment association topology map, which records the spatial coordinates of the threatened equipment, the risk type (such as the risk of insulator flashover), and the associated maintenance tasks.

[0131] Based on the historical database, the prediction errors (such as hit rate, mean square error) of each data source in heavy rain and strong wind scenarios are statistically analyzed, and a dynamic confidence score table is generated using the sliding window statistical method. Initial weights are assigned according to the scoring results, and abnormal forecast values are detected in real time through box plot analysis, and the weight parameters are dynamically adjusted. For example, the initial weight of the meteorological station data in the heavy rain scenario is set to 55%. If its predicted rainfall deviates from the mean of other data sources by 1.5 times, the weight is downgraded to 30%. Multiple source data are linearly weighted and fused to generate an optimized meteorological prediction result, and a time decay factor is introduced to strengthen the weight of recent predictions (70% for the next 6 hours, decreasing to 30% after 24 hours). The optimized result is written back to the meteorological data interface layer through narrowband Internet of Things to update the original single-source data.

[0132] The mobile operation terminal collects on-site temperature, humidity and rainfall data in real time through sensors, and transmits them to the meteorological data interface layer via the narrowband Internet of Things. The system calculates and optimizes the grid deviation between the predicted results and the measured values, and dynamically adjusts the geographic fence matching rules (such as expanding the radius of the rainstorm buffer zone) and the weight distribution parameters. For example, if the measured rainfall in a certain area continues to be 5% higher than the predicted value, the trigger buffer radius will be expanded from 500 meters to 800 meters. The early warning work order is automatically generated according to the risk level mapping table, encapsulating the equipment identification and disposal suggestions (such as "priority inspection of towers T-103 to T-115"), and is pushed to the terminal through NB-IoT. After the work order is executed, the on-site feedback data triggers the iteration of the model parameters, forming a closed-loop link from data collection to decision optimization.

[0133] The spatiotemporal resolution parameters are dynamically adjusted according to the meteorological mutation marks in the user's correction records. For example, after marking the short-term heavy rainfall area, the time granularity of the standardized data set is increased from the hour level to the 10-minute level, and the spatial overlay analysis module synchronously updates the sampling frequency. The calibration parameters and weight allocation parameters are updated in conjunction with each other, and the historical database stores the weight adjustment log to support the iterative optimization of the confidence score table. Narrowband Internet of Things technology ensures low-power data transmission in the wild environment. For example, BPSK modulation is used to maintain connection stability in mountainous areas, and data packets are optimized and encrypted before transmission to the terminal, triggering navigation path updates and risk warnings. The narrowband Internet of Things adopts NB-IoT technology, supports BPSK / QPSK adaptive modulation, and switches to BPSK mode when the signal strength is lower than -90dBm to maintain transmission stability.

[0134] The above implementation methods achieve accurate positioning of meteorological risks and efficient dispatch of emergency resources through standardized data alignment, dynamic weight allocation and closed-loop feedback mechanism. For example, in heavy rain scenarios, the system predicts high-risk sections 24 hours in advance, and after the work order is issued, maintenance resources are allocated in a targeted manner to reduce the inspection interruption rate; in high wind scenarios, real-time wind speed calibration is used to reduce the false alarm rate. Multi-source data fusion and adaptive optimization technology significantly improve the protection capabilities of power equipment in extreme weather and the intelligent level of operation and inspection decision-making.

[0135] The present invention systematically solves the problem of integrating meteorological data with power operation and inspection business scenarios through the following technical solutions:

[0136] Unify the timestamps and spatial benchmarks of multi-source meteorological data through spatio-temporal benchmark alignment processing, generate a standardized meteorological dataset, and conduct spatial overlay analysis with the power grid GIS topology data. Dynamically generate buffers based on the meteorological influence radius, identify transmission equipment threatened by rainstorms and strong winds, and construct a meteorological and equipment association topology map. This map associates equipment status with maintenance tasks through a risk level mapping relationship, and predefined meteorological event thresholds trigger the automatic generation of warning work orders to achieve precise matching of meteorological risks and equipment positioning. This mechanism directly maps meteorological prediction results to the power grid geographical space, solving the problem of weak anti-interference ability of traditional inspection plans due to lack of meteorological relevance.

[0137] Generate a dynamic confidence score table based on historical error statistics, dynamically allocate data source weights in combination with real-time anomaly detection, and generate an optimized meteorological prediction result through weighted fusion. Introduce a time decay factor to strengthen the decision-making weight of recent prediction data, and reverse update the optimized result to the meteorological data interface layer through narrowband Internet of Things. Field measured data and user correction records generate calibration parameters through grid difference analysis, dynamically adjust the geographical fence matching rules and weight allocation parameters, and form a closed-loop calibration of data quality and business rules. This technical path solves the blindness problem of emergency resource allocation under extreme meteorological events, and improves decision-making accuracy through multi-source data cross-validation and adaptive optimization.

[0138] The warning work order encapsulates the threatened equipment identification, risk type and disposal suggestions, and is pushed to the mobile operation terminal through narrowband Internet of Things, supporting real-time data transmission under low-power wide-area communication. The work order execution result and on-site feedback data are transmitted back to the operation and maintenance control system, triggering dynamic update of the risk level and iteration of model parameters. The meteorological mutation scenario marker links the spatio-temporal resolution adjustment and weight decay logic to achieve dynamic sorting of work order priorities and directional allocation of resources. This process converts multi-source warning information into executable closed-loop management instructions, solves the problems of low efficiency of traditional manual work order dispatching and large information attenuation, and improves the circulation and execution efficiency of meteorological warning work orders in complex scenarios.

Claims

1. A transmission line inspection method combining meteorological data with dynamic confidence weighting, characterized in that: include: Acquire multi-source meteorological data, the multi-source meteorological data including numerical forecast data, real-time monitoring data and forecast data, perform spatiotemporal benchmark alignment processing on the multi-source meteorological data through a protocol parser, and generate a standardized meteorological data set, the spatiotemporal benchmark alignment processing including unifying timestamps of different data sources to minute-level accuracy and eliminating spatial benchmark differences through geographic coordinate interpolation; Perform spatial overlay analysis on the standardized meteorological data set and the power grid GIS topology data, generate a dynamic buffer based on the meteorological impact radius, identify the transmission equipment threatened by the meteorological weather through spatial relationship calculation, and construct a meteorological and equipment related topology map. The power grid GIS topology data includes the geographic fences and maintenance operation points of the transmission line corridor; According to the risk level mapping relationship in the meteorological and equipment-related topological map, an event-driven engine is triggered by a predefined meteorological event threshold to generate an early warning work order, which encapsulates the identification of the threatened power transmission equipment and disposal suggestions, and is pushed to the operation and inspection control system through a wireless communication network for closed-loop management of work order issuance, execution and feedback; For multi-source meteorological forecast data of the same geographical area in the meteorological data interface layer, a dynamic confidence score table is generated based on historical error statistical analysis, and weights of each data source are assigned according to the confidence score table. An optimized meteorological forecast result is generated through weighted fusion, and the optimized meteorological forecast result is reversely updated to the meteorological data interface layer to cover the original single-source data; Based on the on-site measured meteorological data and user correction records, the regional deviation between the optimized meteorological forecast result and the measured value is calculated, and the geographic fence matching rules in the spatial overlay analysis and the dynamic weight allocation parameters in the weighted fusion are dynamically adjusted according to the regional deviation to form a closed-loop calibration of meteorological data and equipment risk analysis.

2. The power transmission line inspection method combining meteorological data and dynamic confidence weighting according to claim 1, characterized in that: The multi-source meteorological data performs time-space reference alignment processing including: Extract meteorological elements from heterogeneous data sources through a protocol parser, the protocol parser supports protocol conversion in JSON, GRIB2 and XML formats, extract the timestamp field of meteorological elements through regular expression matching, and call the GPS clock synchronization module to unify the timestamp to minute-level accuracy, unify the timestamps of the multi-source meteorological data to minute-level accuracy, and eliminate spatial benchmark differences through geographic coordinate interpolation algorithms to generate a time-space continuous standardized meteorological data set; The power grid GIS topology data includes the geographic fences and maintenance operation points of the transmission line corridors. The spatial overlay analysis is based on the spatial relationship matching between the standardized meteorological data set and the geographic fences, and the threat radius of the substation is determined by calculating the wind circle range associated with the wind speed.

3. The power transmission line inspection method combining meteorological data and dynamic confidence weighting according to claim 1, characterized in that: The construction of the meteorological and equipment related topological map includes: Divide the dynamic buffer zone radius according to the meteorological types of heavy rain and strong wind, determine the threatened transmission lines and substations through spatial relationship calculation based on the threat radius output by the spatial overlay analysis, and associate the maintenance tasks in the power grid GIS topology data to generate a risk level mapping table; The early warning work order encapsulates equipment identification, risk type and disposal suggestions, and is pushed to the mobile operation terminal via a wireless communication network, which uses narrowband Internet of Things technology to transmit work order data.

4. The power transmission line inspection method combining meteorological data and dynamic confidence weighting according to claim 1, characterized in that: The generating of a dynamic confidence score table based on historical error statistical analysis includes: Based on the historical database of the standardized meteorological data set, the prediction errors of each data source in the heavy rain and strong wind scenes are counted, and a dynamic confidence score table is generated using a sliding window statistical method; The dynamic weight allocation is based on the dynamic confidence score table, allocates initial weights according to meteorological types, and performs weight degradation on abnormal forecast values ​​that deviate from the group mean through box plot analysis.

5. The power transmission line inspection method combining meteorological data with dynamic confidence weighting according to claim 4, characterized in that: Also includes: The multi-source forecast data is linearly weighted according to the dynamic weight allocation parameters, and a time decay factor is introduced to adjust the weight ratio for short-term heavy rainfall scenarios. The time decay factor adopts a linear decreasing logic, with the weight ratio of the next 6 hours being 70%, and the weight ratio decreasing by 10% every 6 hours until the weight ratio drops to 30% after 24 hours. The decreasing logic is dynamically adjusted based on the meteorological mutation scenario marker. When short-term heavy rainfall occurs in the marked area, the weight ratio of the next 3 hours is increased to 80%; The optimized meteorological forecast results are written back to the meteorological data interface layer via a wireless communication protocol, overwriting the original single-source data, and providing updated meteorological input for the spatial overlay analysis.

6. The power transmission line inspection method combining meteorological data and dynamic confidence weighting according to claim 1, characterized in that: The dynamically adjusting the geo-fence matching rules in the spatial overlay analysis and the dynamic weight allocation parameters in the weighted fusion according to the regionalized deviation includes: Based on the regional deviation between the optimized meteorological forecast results and the measured meteorological data, for the rainstorm scene, the buffer radius parameter is iteratively optimized by the sliding window statistical method; The adjustment of the spatiotemporal resolution parameters includes increasing the time granularity of the standardized meteorological data set according to the meteorological mutation scene mark in the user correction record.

7. The power transmission line inspection method combining meteorological data with dynamic confidence weighting according to claim 3, characterized in that: It also includes a weather feedback module that collects on-site temperature, humidity and rainfall data in real time and transmits it to the meteorological data interface layer via the narrowband Internet of Things; The weather feedback module deployed in the mobile operation terminal collects on-site temperature, humidity and rainfall data in real time, and the on-site data is transmitted to the meteorological data interface layer via the narrowband Internet of Things; The user correction record includes manual adjustment information of the risk level in the early warning work order, and the regionalized deviation calculation generates calibration parameters based on the gridded difference analysis between the field data and the optimized meteorological forecast result.

8. The power transmission line inspection method combining meteorological data with dynamic confidence weighting according to claim 4, characterized in that: The historical database stores the forecast data and corresponding actual data of the past 30 days in the meteorological data interface layer, and classifies and calculates the hit rate and mean square error according to the rainstorm and gale meteorological types; The confidence score table is generated based on the historical data and linked with the field measured data collected by the weather feedback module, and the initial weight ratio in the dynamic weight allocation is reversely corrected through deviation analysis.

9. The power transmission line inspection method combining meteorological data with dynamic confidence weighting according to claim 5, characterized in that: The wireless communication protocol uses narrowband Internet of Things technology to transmit the optimized weather forecast results to the mobile operation terminal, and the narrowband Internet of Things supports low-power wide-area communication; The time decay factor is configured such that the prediction weight for the next 6 hours in the weighted fusion accounts for 70%, and the weight decreases linearly to 30% after 24 hours. The decreasing logic is associated with the meteorological mutation scene tag.

10. The power transmission line inspection method combining meteorological data with dynamic confidence weighting according to claim 6, characterized in that: Also includes: Dynamically expand the buffer coverage of the power transmission line corridor in the meteorological and equipment-related topological map according to the measured rainfall distribution of the rainstorm scene; The adjustment of the spatiotemporal resolution parameters triggers the update of the sampling frequency of the meteorological data interface layer, and synchronously updates the dynamic weight allocation parameters based on the regional deviation between the optimized meteorological forecast results and the measured values, forming a closed-loop calibration mechanism for data fusion and equipment risk analysis.

Images