Power distribution operation weather risk dynamic early warning method and system

Abstract

The application provides a power distribution operation weather risk dynamic early warning method and system. The method comprises the following steps: querying weather forecast data according to a current operation position, time-aligning the weather forecast data and weather detection data to obtain a synchronous weather data set, extracting features and performing weighted fusion on the synchronous weather data set according to a fusion weight coefficient to obtain weather characteristic values, converting the weather characteristic values into initial risk values according to a numerical mapping rule, performing weighted summation on the initial risk values according to a risk weight coefficient to obtain a basic weather risk value, performing scene correction on the basic weather risk value according to power distribution operation data and the current operation position to obtain a comprehensive risk value, and matching the comprehensive risk value with a risk and instruction relationship to obtain a risk grade and an early warning instruction. Through multi-source weather data synchronous fusion, power distribution operation scene correction and hierarchical mapping early warning, the application realizes scene-based early warning of weather risks, so as to improve the safety of power distribution operations.

Description

Technical Field

[0001] This application relates to the field of risk warning technology, and in particular to a dynamic early warning method and system for meteorological risks in power distribution operations. Background Technology

[0002] With the continuous expansion of urban and rural power distribution network construction and operation and maintenance, a large number of live-line work, power outage maintenance and emergency repairs occur in open-air environments. Moreover, the work sites are often scattered in areas with significant differences, such as urban and rural areas, mountainous areas and river valleys. They are highly susceptible to meteorological factors such as strong winds, lightning, rainstorms, high temperature and humidity, icing and low visibility, which can lead to risks such as personal injury, wind-induced discharge of lines, insulation dampness or icing-induced line breakage. Therefore, a meteorological risk management method that can combine on-site micro-meteorological changes and differences in work scenarios to provide dynamic early warning is needed.

[0003] Existing meteorological risk early warning methods for power distribution operations mainly rely on regional weather forecasts, manual visual observation, monitoring at fixed weather stations, and general threshold-based warnings. The basic idea is to determine the operation window based on regional forecasts issued by meteorological departments, or to collect data such as wind speed, rainfall, temperature, and humidity at fixed monitoring points, and then trigger warnings according to a unified threshold. If necessary, operation control is implemented based on general meteorological signals such as blue, yellow, orange, and red. While these methods are simple to implement and easy to manage uniformly, they are essentially static and coarse-grained meteorological control models, failing to truly reflect the coupling relationship between real-time micro-meteorology at the operation site and the actual operation scenario. Summary of the Invention

[0004] In view of this, this application provides a dynamic early warning method and system for meteorological risks in power distribution operations, which improves the accuracy of meteorological risk early warning and the safety of power distribution operations.

[0005] On the one hand, embodiments of this application provide a method for dynamic early warning of meteorological risks in power distribution operations, the method comprising: Based on the received current work location, query meteorological forecast data, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. Based on the preset fusion weight coefficients, feature extraction and weighted fusion are performed on the synchronous meteorological dataset to obtain meteorological feature values ​​for each data type. The meteorological characteristic values ​​are converted into corresponding initial risk values ​​according to the preset numerical mapping rules, and the initial risk values ​​are weighted and summed according to the preset risk weight coefficients to obtain the basic meteorological risk values. Based on the preset power distribution operation data and the current operation location, the basic meteorological risk value is modified according to the scenario to obtain a comprehensive risk value; The comprehensive risk value is matched with the preset risk and instruction relationship to obtain the risk level and early warning instruction.

[0006] In an optional implementation, the step of querying meteorological forecast data based on the received current work location and performing time alignment processing on the meteorological forecast data and the real-time collected meteorological monitoring data to obtain a synchronized meteorological dataset includes: A meteorological observation area is generated based on the current working location and a preset center distance, and meteorological forecast data is queried through a preset meteorological data service interface based on the meteorological observation area. Real-time data collection of instantaneous wind speed, rainfall, ambient temperature, relative humidity, and visibility at the current work location to obtain the first meteorological detection data; A lightning warning area is generated based on the preset warning distance and the meteorological observation area. The location of lightning occurrence and the speed of storm movement within the lightning warning area are collected in real time. The real-time lightning distance value is calculated based on the current working position, the location of lightning occurrence, and the speed of storm movement to obtain the second meteorological detection data. Based on the collection timestamps in each data set, the meteorological forecast data, the first meteorological detection data, and the second meteorological detection data are time-aligned to obtain a synchronized meteorological dataset.

[0007] In an optional implementation, the step of extracting features from the synchronous meteorological dataset and performing weighted fusion based on preset fusion weight coefficients to obtain meteorological feature values ​​for each data type includes: Based on the preset feature types, feature extraction is performed on the synchronous meteorological dataset to obtain wind speed feature set, rainfall feature set and lightning feature set; Based on the wind speed feature set, a first fusion weight coefficient is determined from the preset fusion weight coefficients, and the wind speed feature set is weighted and summed according to the first fusion weight coefficient to generate wind speed feature values. Based on the rainfall feature set, a second fusion weight coefficient is determined from the fusion weight coefficients, and the rainfall feature set is weighted and summed according to the second fusion weight coefficient to generate flood feature values; The probability of lightning occurrence within the lightning warning area is calculated based on the lightning feature set, and lightning prediction feature values ​​are generated.

[0008] In an optional implementation, the step of converting the meteorological characteristic values ​​into corresponding initial risk values ​​according to a preset numerical mapping rule, and then weighting and summing the initial risk values ​​according to a preset risk weighting coefficient to obtain a basic meteorological risk value includes: Based on the wind speed safety threshold and rainfall safety threshold in the preset numerical mapping rules, the ratio of the wind speed characteristic value and the flood characteristic value is standardized to obtain the first initial risk value and the second initial risk value. The temperature and humidity index is calculated based on the temperature and humidity values ​​in the synchronous meteorological dataset. The temperature and humidity index is then standardized by ratio calculation according to the temperature and humidity safety threshold in the numerical mapping rule to obtain a third initial risk value. The lightning proximity time is calculated based on the real-time lightning distance value and the storm movement speed. The lightning proximity time is then standardized by ratio calculation according to the time safety threshold in the numerical mapping rule to obtain the lightning time risk value. The lightning time risk value and the lightning prediction feature value are then weighted and summed according to the preset lightning risk weight coefficient to obtain the fourth initial risk value. Based on the current operating location, the evaluation weight coefficients corresponding to each risk value are matched from the preset risk weight coefficients, and the first initial risk value, the second initial risk value, the third initial risk value, and the fourth initial risk value are weighted and summed according to the evaluation weight coefficients to obtain the basic meteorological risk value.

[0009] In an optional implementation, the step of performing scenario correction processing on the basic meteorological risk value based on preset power distribution operation data and the current operation location to obtain a comprehensive risk value includes: Based on the voltage level and operation type in the preset power distribution operation data, a first correction coefficient is obtained from the preset scenario correction coefficient library, and based on the line parameters in the power distribution operation data, a second correction coefficient is obtained from the scenario correction coefficient library. Based on the current work location and current time, the terrain type and season type are determined from a preset geographic database, and based on the terrain type and season type, a third correction coefficient and a fourth correction coefficient are obtained from the scene correction coefficient library; Based on the first correction coefficient, the second correction coefficient, the third correction coefficient, and the fourth correction coefficient, the basic meteorological risk value is subjected to scenario correction processing to obtain a comprehensive risk value.

[0010] On one hand, embodiments of this application provide a dynamic early warning device for meteorological risks in power distribution operations, the device comprising: The time alignment module is used to query meteorological forecast data based on the received current operation location, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. The feature fusion module is used to extract features and perform weighted fusion on the synchronous meteorological dataset according to preset fusion weight coefficients to obtain meteorological feature values ​​of each data type. The risk assessment module is used to convert the meteorological characteristic values ​​into corresponding initial risk values ​​according to preset numerical mapping rules, and to perform weighted summation of the initial risk values ​​according to preset risk weight coefficients to obtain basic meteorological risk values. The scenario correction module is used to perform scenario correction processing on the basic meteorological risk value based on preset power distribution operation data and the current operation location to obtain a comprehensive risk value; The risk level instruction module is used to match the comprehensive risk value with a preset risk and instruction relationship to obtain the risk level and warning instruction.

[0011] The embodiments of this application employing the above-described technical solution may have the following advantages: 1. By aligning meteorological forecast data with real-time on-site monitoring data and further integrating information such as wind speed, rainfall, temperature and humidity, icing, and lightning, the system can more timely reflect the actual weather changes at the work site and improve the timeliness of early warnings.

[0012] 2. Map various meteorological characteristic values ​​to initial risk values, then form basic meteorological risk values ​​through weighted averages, and finally output different levels of early warning instructions based on risk threshold ranges. This makes the early warning results no longer a rough judgment of "whether there is risk" but have clear levels and corresponding actions, thus improving the accuracy and enforceability of early warnings.

[0013] 3. By incorporating scenario factors such as voltage level, work type, line parameters, terrain, and season, the basic meteorological risk value is corrected. Therefore, the same meteorological conditions can lead to different risk assessments in different work scenarios, which is more in line with the actual differences in hazards in power distribution work, reduces misjudgments and omissions, and improves the on-site work safety assurance capabilities. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a flowchart of a dynamic early warning method for meteorological risks in power distribution operations provided in an embodiment of this application; Figure 2 This is a functional block diagram of a dynamic early warning device for meteorological risks in power distribution operations provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] It should be noted that the message processing solution provided in this application requires special explanation of the following two points: 1. This application involves relevant data during message processing (such as current job location). When the above embodiments of this application are applied to specific products or technologies, permission or consent from the target audience is required, and the collection, use, and processing of relevant data must comply with the relevant laws, regulations, and standards of the region, conforming to the principles of legality, legitimacy, and necessity, and not involving the acquisition of data types prohibited or restricted by laws and regulations. In some optional embodiments, the relevant data involved in the embodiments of this application is obtained after separate authorization from the target audience. In addition, when obtaining separate authorization from the target audience, the purpose of the relevant data is explained to the target audience.

[0018] 2. It is understood that in this application, the term "at least one" refers to one or more, and "multiple" means two or more; for example, "at least one notification method" means one, two, or more notification methods. The terms "first," "second," etc., are used to distinguish identical or similar items with essentially the same function. It should be understood that there is no logical or temporal dependency between "first," "second," and "nth," nor is there any limitation on the quantity or execution order.

[0019] like Figure 1 The diagram shown is a flowchart of the dynamic early warning method for meteorological risks in power distribution operations provided in this embodiment of the application. The dynamic early warning method for meteorological risks in power distribution operations provided in this embodiment of the application includes the following steps.

[0020] Step S1: Query meteorological forecast data based on the received current work location, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset.

[0021] After the power distribution work team arrives at the designated work site, they first need to obtain their current work location using a sensor terminal worn by each worker. This sensor terminal, fixed to a safety helmet or work uniform, integrates a Global Navigation Satellite System (GNSS) positioning module. By receiving BeiDou or GPS satellite signals, it calculates the worker's longitude, latitude, and altitude in real time. The positioning module updates once per second, achieving a positioning accuracy better than two meters in open areas. An edge computing terminal receives the positioning data from this sensor terminal once per second and uses this data as the reference point for all subsequent spatial queries and calculations—the current work location. This current work location contains the precise geographic coordinates of the work point, which is crucial for determining the scope of meteorological forecast data queries, the spatial attribution of real-time meteorological data, and the matching basis for terrain features in subsequent scene correction processing. For example, in power distribution line maintenance work in mountainous areas, different towers of the same power distribution line may be located at different altitudes and slopes. By obtaining the precise coordinates of each work team's tower in real time, it is possible to avoid misapplying general regional meteorological forecast data applicable to the entire township to work points with significant local micro-meteorological differences.

[0022] After obtaining the current work location, the edge computing terminal generates a square meteorological observation area centered on that location, with a preset center distance. The value of this center distance is determined based on the actual coverage area of ​​the power distribution operation; for example, if set to 5km, the side length of the meteorological observation area is 10km. This meteorological observation area limits the query range of meteorological forecast data, avoiding the acquisition of large-scale redundant data unrelated to the work site, while ensuring that the queried meteorological forecast data covers the movement range of meteorological systems around the work site that may affect operational safety. The edge computing terminal initiates a query request through a preset meteorological data service interface based on the latitude and longitude coordinates of the four corner points of the meteorological observation area. This interface connects to the data server of the China Meteorological Administration or an authorized commercial meteorological service provider. The requested data types include predicted instantaneous wind speed, predicted rainfall intensity, and predicted lightning probability values ​​for every ten minutes within the next two hours. The forecast data returned by the meteorological data service interface is generated using optical flow and a convolutional neural network extrapolation model, based on geostationary meteorological satellite cloud images (such as infrared, visible light, and water vapor cloud images from the Fengyun-4 satellite) and weather radar reflectivity mosaics. The edge computing terminal stores the returned weather forecast data in timestamp order.

[0023] Meanwhile, edge computing terminals collect meteorological data in real time through various sensors deployed at the work site. In this embodiment, a portable automatic weather station is set up in an open, unobstructed location around the work site. This weather station integrates wind speed and direction sensors, a tipping bucket rain gauge, a temperature and humidity sensor, and a visibility sensor. The wind speed and direction sensor collects instantaneous wind speed values ​​with a period of 100 milliseconds, the tipping bucket rain gauge collects real-time rainfall values ​​with a period of 1 minute, the temperature and humidity sensor collects ambient temperature and relative humidity values ​​with a period of 1 second, and the visibility sensor collects visibility values ​​with a period of 1 second. These data are collectively referred to as the first meteorological data, which directly reflects the real meteorological conditions at the work site at the current moment, compensating for the lack of spatial resolution in meteorological forecast data. For example, when carrying out power outage maintenance work in a river valley, the regional weather forecast may show a wind speed of 3 m / s, but the instantaneous wind speed measured by the portable weather station on site may reach 8 m / s due to the funneling effect. Relying solely on forecast data would severely underestimate the risk.

[0024] The edge computing terminal also acquires third-party meteorological data through an icing monitoring terminal. This icing monitoring terminal is installed on adjacent towers along the line where the work site is located. It uses a combination of weighing and image recognition methods to measure the ice thickness on the conductors, achieving a measurement accuracy of 0.1 mm. The edge computing terminal reads this ice thickness value every minute, and this data is used to assess the icing risk. Icing is a slow process, and real-time data acquisition reflects the current load status of the line, which is particularly important for power distribution operations in high-altitude areas during winter.

[0025] The edge computing terminal further generates a lightning warning zone based on the preset warning distance and the existing meteorological observation area. Due to the natural characteristics of lightning, a larger monitoring range is required: thunderstorm clouds typically take 10 to 30 minutes to reach the work site, while lightning movement speeds can reach 10 to 20 m / s (0.6 to 1.2 km / min). If only a 10 km side meteorological observation area is monitored, the evacuation time for workers after lightning enters the area is less than 10 minutes, which is far from sufficient for high-altitude live-line work. Therefore, the warning distance is set to 15 km, resulting in a 30 km side square area centered on the current work location. The edge computing terminal uses a lightning positioning system to collect real-time data on the lightning location and storm movement speed within this warning zone. The lightning positioning system is based on the multi-station time difference method, calculating the precise coordinates of the lightning occurrence by the time difference between the lightning electromagnetic waves received by multiple monitoring stations, with a positioning error controlled within 500 m. Simultaneously, the storm movement speed is calculated by dividing the position difference between two consecutive lightning positioning measurements by the time difference. The edge computing terminal calculates the real-time lightning distance value based on the coordinates of the current work location, the coordinates of the lightning occurrence location, and the storm's movement speed. The specific calculation method is as follows: first, the straight-line distance between the lightning occurrence point and the current work location is calculated, using the Haversine formula to account for the Earth's curvature; then, the remaining time for the thunderstorm cloud to reach the work location is estimated based on the storm's movement speed. This real-time lightning distance value is a core component of the second meteorological monitoring data; a smaller value indicates a higher risk.

[0026] After collecting meteorological forecast data, first meteorological observation data, third meteorological observation data, and second meteorological observation data, the edge computing terminal needs to perform time alignment processing on these four types of data. Since the timestamps of the meteorological forecast data are future points in time (e.g., 10 minutes or 20 minutes after the current moment), while the timestamps of the three types of real-time observation data are the current moment or the most recent collection moment, the goal of time alignment is to extract the future time node closest to the current moment from the forecast data and merge it with the real-time data of the current moment into a synchronized meteorological dataset. The edge computing terminal first reads the unified timestamp of the current system, then finds the time node in the meteorological forecast data that is greater than the current time and has the smallest time difference, and extracts the predicted instantaneous wind speed value, predicted rainfall intensity value, and predicted lightning probability value of that node. Next, the instantaneous wind speed value, rainfall value, ambient temperature value, relative humidity value, visibility value, icing thickness value, real-time lightning distance value, and storm movement speed value collected at the current moment are associated with the extracted forecast data according to the same timestamp (i.e., the current moment). All data is stored in a structured data table, with each row corresponding to a timestamp and each column corresponding to a meteorological parameter type. This synchronized meteorological dataset ensures the temporal comparability of the real-time and forecast data used for subsequent feature extraction and risk calculation; that is, it uses current measured values ​​and short-term future forecasts to jointly describe the dynamic trend of meteorological conditions at the work site. For example, if the current measured wind speed is 4 m / s, and the forecast shows that the wind speed will rise to 9 m / s in 10 minutes, the time-aligned dataset will contain both values, enabling subsequent weighted fusion to be performed correctly. After time alignment, the edge computing terminal temporarily stores the synchronized meteorological dataset in its local memory.

[0027] Step S2: Perform feature extraction and weighted fusion on the synchronous meteorological dataset according to the preset fusion weight coefficient to obtain meteorological feature values ​​of each data type.

[0028] To convert data from different sources and time scales into unified meteorological feature values, the edge computing terminal first needs to perform feature extraction on the synchronous meteorological dataset according to preset feature types. These feature types are determined based on meteorological factors that pose a direct threat to the safety of workers, as explicitly listed in the power distribution operation safety regulations, including wind speed type, rainfall type, and lightning type. For wind speed, the edge computing terminal extracts two specific values ​​from the synchronous meteorological dataset: one is the instantaneous wind speed value collected in real time by a portable automatic weather station, and the other is the predicted instantaneous wind speed value for the nearest future time point obtained from meteorological forecast data. These two values ​​together constitute the wind speed feature set. For rainfall, the edge computing terminal extracts two specific values: one is the rainfall amount value collected in real time by a portable automatic weather station, and the other is the predicted rainfall intensity value for the nearest future time point obtained from meteorological forecast data. These two values ​​together constitute the rainfall feature set. For lightning types, the edge computing terminal extracts three specific values: one is the real-time lightning distance value provided by the lightning location system, another is the storm movement speed value provided by the lightning location system, and the third is the predicted lightning probability value for the nearest future time point obtained from meteorological forecast data. These three values ​​together constitute the lightning feature set. The original multi-source heterogeneous data is classified according to risk type, ensuring that subsequent calculations for each risk type rely only on data directly related to that risk, avoiding data confusion between different risk types. For example, when conducting live-line work in a plain area on a summer afternoon, wind speeds may be low, but the risk of lightning is high. By extracting wind speed data and lightning data into different feature sets, various risks can be independently assessed without mutual interference.

[0029] After obtaining the wind speed feature set, the edge computing terminal needs to fuse the real-time instantaneous wind speed values ​​and predicted instantaneous wind speed values ​​in the feature set into a single wind speed feature value. Since real-time data reflects the actual situation at the current moment while predicted data reflects the trend of change in the near future, the contribution of both to the final risk value should be dynamically adjusted according to the actual scenario. The edge computing terminal searches for the first fusion weight coefficient corresponding to the wind speed type from a preset fusion weight coefficient library. This fusion weight coefficient is a decimal between 0 and 1, where the first fusion weight coefficient itself represents the weight of real-time data in the fusion result, while the weight of the predicted data is equal to 1 minus the first fusion weight coefficient. During initial system operation, the first fusion weight coefficient is set to 0.6, meaning real-time data accounts for 60% and predicted data accounts for 40%. This ratio is determined based on extensive field experiments, ensuring both accurate response to current risks and providing a certain lead time for early warning. The edge computing terminal fuses two values ​​from the wind speed feature set according to a weighted summation formula: the real-time instantaneous wind speed value is multiplied by a first fusion weight coefficient, and the predicted instantaneous wind speed value is multiplied by (1 minus the first fusion weight coefficient). The two products are then added together, and the result is the wind speed feature value. The unit of this wind speed feature value is the same as the original data, in m / s. For example, in a power distribution maintenance operation at a windy location in a mountainous area, the real-time instantaneous wind speed is 5 m / s, while the prediction shows that the wind speed will rise to 12 m / s in 10 minutes. After fusion with weights of 0.6 and 0.4, the wind speed feature value is 5 × 0.6 + 12 × 0.4 = 3 + 4.8 = 7.8 m / s. This fused value of 7.8 m / s reflects both the current strong wind conditions and the impending sudden increase in wind speed, allowing for proactive risk calculations.

[0030] After obtaining the rainfall feature set, the edge computing terminal uses a method similar to that used for wind speed to fuse it into a single flood feature value. The edge computing terminal searches for a second fusion weight coefficient corresponding to the rainfall type from a pre-set fusion weight coefficient library. This second fusion weight coefficient also falls between 0 and 1, and is set to 0.6 during initial system operation, meaning the real-time rainfall value accounts for 60% and the predicted rainfall intensity value accounts for 40%. The edge computing terminal multiplies the real-time rainfall value by the second fusion weight coefficient and multiplies the predicted rainfall intensity value by (1 minus the second fusion weight coefficient), then adds the two products to obtain the flood feature value, in mm / min. Since the formation of rainstorms often involves a development process, the real-time rainfall may still be within a safe range, but radar echo maps show that a strong rain cloud is approaching. In this situation, relying solely on real-time data may lead to missed reports, while relying solely on predicted data may result in false alarms due to prediction errors. Weighted fusion strikes a balance between these two approaches. For example, during power distribution emergency repair operations in a river valley, the real-time rainfall was 0.2 mm / min, which is within the safe range. However, the forecast showed that the rainfall intensity would reach 1.5 mm / min in 20 minutes, far exceeding the safe threshold. The combined flood characteristic value was 0.2×0.6+1.5×0.4=0.12+0.6=0.72 mm / min, which is close to the safe threshold and can trigger an alert level warning, reminding the workers to prepare for evacuation in advance.

[0031] After obtaining the lightning feature set, the edge computing terminal needs to calculate the lightning prediction feature value based on three values ​​within it. Unlike wind speed and rainfall, lightning risk is sudden and highly lethal. Therefore, it's not possible to simply weight and sum the real-time lightning distance value with the predicted lightning probability value. This is because the real-time lightning distance value has been accurately measured by the lightning location system, while the predicted lightning probability value is a statistical estimate based on a satellite cloud image extrapolation model; their physical meanings differ. The edge computing terminal first uses the real-time lightning distance value and storm movement speed value from the lightning feature set to calculate the estimated time of arrival of the thunderstorm cloud at the work site. Specifically, the calculation method is to divide the real-time lightning distance value by the storm movement speed value to obtain the lightning proximity time in minutes. Then, the edge computing terminal directly uses the predicted lightning probability value as the basis for the lightning prediction feature value. The generation of lightning prediction feature values ​​follows this logic: when the lightning approach time is less than a preset safe time threshold (e.g., 15 minutes), a higher lightning prediction feature value should be output even if the predicted lightning probability is low, because the lightning is approaching; conversely, when the lightning approach time is large but the predicted probability is high, a higher lightning prediction feature value should also be output, because a distant thunderstorm is developing and may move towards the work site. The edge computing terminal calculates the lightning prediction feature value through a nonlinear mapping function. The input to this function is the lightning approach time and the predicted lightning probability, and the output is a dimensionless value between 0 and 100. The specific implementation method is as follows: First, the lightning proximity time is normalized, with a safe time threshold of 15 minutes. When the lightning proximity time is less than or equal to 0, the normalized value is 1; when it is greater than or equal to 15 minutes, the normalized value is 0. Intermediate values ​​are linearly interpolated. Then, the normalized lightning proximity time is inverted (i.e., 1 is subtracted from the value), and then a weighted sum is performed with the predicted lightning occurrence probability value. The weight of the real-time proximity time is set to 0.7, and the weight of the predicted probability is set to 0.3. The result of this weighted sum is then multiplied by 100 to obtain the lightning prediction feature value. For example, if the lightning location system shows that lightning occurred 9km from the work site and the storm's movement speed is 0.9km / min, then the lightning proximity time is 10 minutes. After normalization, this is (15-10) / 15 = 0.333, which, when inverted, becomes 0.667. The predicted lightning probability is 0.8. Therefore, the lightning prediction characteristic value is (0.667×0.7+0.8×0.3)×100 = (0.4669+0.24)×100 = 70.69. This characteristic value is directly used for subsequent initial risk value calculations; a larger value indicates a higher lightning risk. The significance of generating the lightning prediction characteristic value lies in unifying the predicted probability extrapolated from real-time lightning location data and satellite cloud images into a single quantitative index. This allows the system to identify risks before thunderstorm clouds enter the safe distance threshold, providing valuable evacuation time for personnel working with live electrical equipment at high altitudes.For example, during 10kV live-line work in the summer thunderstorm season, the real-time lightning distance is 15km, which has not yet reached the 10km safety warning threshold. However, the predicted lightning probability is as high as 0.9, indicating that the distant thunderstorm is rapidly developing and moving towards the work site. The lightning prediction characteristic value obtained through the above calculation is high, thus triggering an early warning and avoiding the situation where it is too late to evacuate due to waiting for the lightning to enter the 10km range before issuing an alarm.

[0032] After generating the three feature values ​​mentioned above, the edge computing terminal outputs the wind speed feature value, flood feature value, and lightning prediction feature value as the output of this step. It also retains the un-fused ambient temperature value, relative humidity value, icing thickness value, and visibility value from the synchronous meteorological dataset. At this point, the edge computing terminal has completed the conversion from the synchronous meteorological dataset to meteorological feature values ​​of various data types.

[0033] Step S3: Convert the meteorological feature values ​​into corresponding initial risk values ​​according to the preset numerical mapping rules, and perform weighted summation on the initial risk values ​​according to the preset risk weight coefficients to obtain the basic meteorological risk values.

[0034] After generating meteorological feature values ​​for each data type in step S2, the edge computing terminal obtains a set of quantitative indicators for risk assessment, including wind speed feature values, flood feature values, lightning prediction feature values, and un-fused ambient temperature, relative humidity, real-time lightning distance, storm movement speed, icing thickness, and visibility values ​​from the synchronous meteorological dataset. These indicators have different dimensions and physical meanings and cannot be directly added or compared. Therefore, they need to be uniformly converted into initial risk values ​​with consistent dimensions and the same value range according to a preset numerical mapping rule. The preset numerical mapping rule in this application adopts a ratio standardization method: the measured or fused value of each meteorological parameter is compared with the corresponding safety threshold, and the ratio reflects the degree of deviation of the current meteorological conditions from the safety boundary. The higher the ratio, the greater the risk.

[0035] The edge computing terminal first processes the wind speed and flood characteristic values. It reads the wind speed safety threshold and rainfall safety threshold from preset numerical mapping rules; these two thresholds are pre-stored according to the regulations in the "Electric Power Safety Work Regulations" for the current work type and voltage level. For the wind speed characteristic value, the edge computing terminal divides it by the wind speed safety threshold and then multiplies the quotient by 100 to obtain the first initial risk value. For the flood characteristic value, the edge computing terminal divides it by the rainfall safety threshold and then multiplies the quotient by 100 to obtain the second initial risk value. Ratio standardization calculations unify meteorological parameters of different dimensions into a numerical range of 0 to 100, making the subsequent weighted sum mathematically comparable. For example, in a 10kV live-line working scenario, the wind speed safety threshold is 5 m / s. If the wind speed characteristic value output in step S2 is 7.8 m / s, then the first initial risk value is 7.8 / 5 × 100 = 156. However, since the upper limit of the risk value is truncated to 100, the actual value is 100, indicating that the wind speed risk has reached the highest level. If the wind speed characteristic value is 3 m / s, then the first initial risk value is 60, which is in the attention level range. The above mapping method reflects the degree of deviation of the measured value from the safety standard; the larger the value, the closer it is to or beyond the safety boundary.

[0036] The edge computing terminal then handles the high temperature and humidity risks. Ambient temperature and relative humidity values ​​are extracted from the synchronous meteorological dataset. Both values ​​are obtained from real-time data collection by portable automatic weather stations, without corresponding forecast data for fusion. Ambient temperature and relative humidity together determine the degree of heat stress for workers, but there is a non-linear coupling relationship between them: at the same temperature, higher humidity leads to slower sweat evaporation and more severe heat stress; at the same humidity, higher temperature also leads to more severe heat stress. Therefore, temperature and humidity cannot be independently mapped to risk values ​​and then simply added together; instead, a temperature-humidity index (THI) needs to be calculated first. The THI calculation formula adopts the human heat stress correction form recommended by the World Meteorological Organization: the THI equals the ambient temperature value minus 0.55 multiplied by (1 minus the relative humidity value) and then multiplied by (ambient temperature value minus 14.5). In this formula, 14.5 is the baseline temperature constant. When the ambient temperature is below 14.5 degrees Celsius, high humidity can actually exacerbate the feeling of cold. However, the high temperature and humidity risks in power distribution work safety regulations mainly focus on high-temperature scenarios in summer, so this formula is widely used in engineering practice. The edge computing terminal divides the calculated temperature and humidity index by the temperature and humidity index safety threshold read from the preset numerical mapping rules, and then multiplies the quotient by 100 to obtain the third initial risk value. The temperature and humidity index safety threshold is determined according to the type of work. For example, for high-altitude live-line work, this threshold is set to 28. When the temperature and humidity index exceeds 28, it indicates a risk of heatstroke. The ratio standardization calculation maps this dimensionless composite index of temperature and humidity to the range of 0 to 100, so that the third initial risk value has the same range as other risk values.

[0037] The edge computing terminal then processes lightning risk, the only risk type among the six categories that simultaneously uses real-time and forecast data for fusion calculations. Real-time lightning distance and storm movement speed values ​​are extracted from the synchronous meteorological dataset; both values ​​originate from real-time monitoring by the lightning location system. The edge computing terminal first calculates the lightning proximity time based on these values: dividing the real-time lightning distance by the storm movement speed yields the lightning proximity time in minutes, representing the estimated time required for the thunderstorm cloud to reach the work site at its current speed. Then, a time safety threshold is read from a preset numerical mapping rule. This threshold, set at 15 minutes according to power distribution work safety regulations, indicates that a warning should be issued when the lightning proximity time is less than 15 minutes. The edge computing terminal divides the time safety threshold by the lightning proximity time and multiplies the quotient by 100 to obtain the lightning time risk value. This calculation method means that the shorter the lightning proximity time, the greater the lightning time risk value. When the lightning proximity time equals the time safety threshold, the risk value is 100; when the lightning proximity time is less than the safety threshold, the risk value exceeds 100 and is truncated to 100. The lightning time risk value reflects the immediate threat revealed by real-time lightning location data. The edge computing terminal simultaneously obtains the lightning prediction feature value from the output of step S2. This feature value already incorporates information about the predicted probability of lightning occurrence. To obtain the final fourth initial risk value, the edge computing terminal performs a weighted sum of the lightning time risk value and the lightning prediction feature value according to a preset lightning risk weighting coefficient. The lightning risk weighting coefficient is also between 0 and 1, set to 0.7 during initial system operation, meaning the lightning time risk value accounts for 70% and the lightning prediction feature value accounts for 30%. The weighted summation is calculated as follows: multiply the lightning time risk value by the lightning risk weighting coefficient, multiply the lightning prediction feature value by (1 minus the lightning risk weighting coefficient), and then add the two products to obtain the fourth initial risk value. When thunderstorms are approaching the work site, the accurate information provided by real-time lightning location data should be dominant; when thunderstorms are developing in the distance but have not yet entered the warning range, predictive data can trigger an increase in risk in advance, giving personnel working on live energized heights time to evacuate.

[0038] In this embodiment, after calculating the first, second, third, and fourth initial risk values, the edge computing terminal also needs to process icing risk and low visibility risk. Icing risk is calculated using the icing thickness value from the synchronous meteorological dataset. The icing thickness value is divided by a preset icing thickness safety threshold and then multiplied by 100 to obtain the fifth initial risk value. Low visibility risk is calculated using the visibility value from the synchronous meteorological dataset. The preset visibility safety threshold is divided by the visibility value and then multiplied by 100 to obtain the sixth initial risk value. The calculation of these two types of risks also follows the ratio-standardized numerical mapping rule.

[0039] In one optional implementation, the edge computing terminal matches the assessment weight coefficients corresponding to each risk value from a preset risk weight coefficient library based on the current operating location. Since the climate characteristics and meteorological risk probabilities vary significantly across different geographical locations—for example, power distribution operations in coastal areas are more susceptible to typhoons, so the weight of strong wind risk should be higher than in inland areas; while power distribution operations in mountainous and valley areas are prone to thunderstorms and flash floods in summer, so the weights of lightning risk and rainstorm flood risk are correspondingly increased—the edge computing terminal queries the preset risk weight coefficient library using the latitude and longitude coordinates of the current operating location as an index. This risk weight coefficient library is pre-constructed based on different climate zones, terrain features, and historical meteorological accident data across the country, dividing geographical areas into multiple risk zoning units. Each unit stores corresponding assessment weight coefficients for different meteorological types. The edge computing terminal matches six assessment weight coefficients from the risk weight coefficient library, corresponding to six risk categories: strong wind, rainstorm flood, high temperature and humidity, lightning, icing, and low visibility. The sum of the six assessment weight coefficients equals 1. In another optional implementation, the edge computing terminal extracts the voltage level and operation type from preset power distribution operation data. These two parameters determine the relative importance of different meteorological risks in the overall risk. For example, in 10kV live-line work, the weight of strong wind and lightning risks is much higher than that of high temperature and humidity risks; while in ground power outage maintenance work, the weight of rainstorm and flood risks and low visibility risks is relatively high. Based on the combination of voltage level and operation type, the edge computing terminal matches six evaluation weight coefficients from a preset risk weight coefficient library, corresponding to the six risk categories of strong wind, rainstorm and flood, high temperature and humidity, lightning, icing, and low visibility, respectively. The sum of the six evaluation weight coefficients equals 1. Then, the edge computing terminal multiplies the first, second, third, fourth, fifth, and sixth initial risk values ​​by their corresponding evaluation weight coefficients, and then adds the six products together to obtain the basic meteorological risk value. The basic meteorological risk value is a dimensionless value between 0 and 100. It integrates real-time monitoring information, short-term forecast information, and the importance of each risk in the current operational scenario for six types of meteorological risks. For example, when performing 10kV live-line work in a mountainous area in the summer afternoon, the assessment weight coefficients for strong wind risk and lightning risk are relatively high, while the weight coefficient for icing risk is 0. The basic meteorological risk value is mainly determined by the first and fourth initial risk values. If the wind speed characteristic value is 7 m / s and the lightning prediction characteristic value is 70, the basic meteorological risk value will be at a high level, triggering the subsequent warning process.

[0040] Step S4: Based on the preset power distribution operation data and the current operation location, perform scenario correction processing on the basic meteorological risk value to obtain a comprehensive risk value.

[0041] After the edge computing terminal calculates the basic meteorological risk value, the obtained value only reflects the deviation of real-time meteorological parameters and short-term forecast data from the general safety threshold at the work site. It does not yet consider the amplification or reduction effects of different work scenarios on meteorological risk. Since power distribution operations cover various terrains including urban and rural areas, mountainous regions, and valleys, and the types of operations include live-line work, power outage maintenance, high-altitude work, and ground work, and line parameters also vary in voltage level, tower height, and service life, the threat posed by the same meteorological conditions to the safety of workers varies significantly in different scenarios. Therefore, the edge computing terminal needs to perform scenario-correction processing on the basic meteorological risk value based on preset power distribution operation data and the current work location to obtain a comprehensive risk value that truly reflects the actual risk level of the current work scenario.

[0042] The edge computing terminal first extracts two parameters—voltage level and operation type—from pre-set power distribution operation data. This data originates from the work permit system and is pre-entered into the edge computing terminal's local memory before the operation begins. Voltage levels include 0.4kV, 10kV, 35kV, and 110kV, while operation types include specific categories such as live-line work, power outage maintenance, high-altitude work, ground work, and cable well work. Based on the combination of voltage level and operation type, the edge computing terminal retrieves the corresponding first correction coefficient from a pre-set scenario correction coefficient library. This library is pre-built through extensive field testing, historical accident data statistics, and analysis of power industry safety regulations. The first correction coefficient is specifically used to quantify the sensitivity of different operation types to weather risks. For example, for 10kV live-line work, since workers are in direct contact with live conductors, strong winds could cause excessive conductor swaying, leading to phase-to-phase short circuits or electric shocks, and lightning poses a direct threat. Therefore, the first correction coefficient is set to 1.30, indicating that the risk of live-line work under the same weather conditions is 30% higher than the baseline scenario. For routine ground-based power outage maintenance work, workers maintain a safe distance from live conductors, and the impact of strong winds and lightning is relatively small. The first correction factor is set to 1.00, which is the baseline value. For indoor ring main unit work, since the work site is located inside a building, meteorological factors such as wind speed, rainfall, and visibility have minimal impact. The first correction factor is set to 0.90, indicating a 10% reduction in risk. The edge computing terminal reads the voltage level and work type from the power distribution work data, performs precise matching in the scene correction factor library, and retrieves the corresponding first correction factor.

[0043] The edge computing terminal then extracts line parameters from the distribution operation data, including line voltage level, tower height, conductor span, design icing thickness, and the line's service life and health status. These parameters determine the line's tolerance to meteorological risks. Based on the combination of line parameters, the edge computing terminal retrieves a second correction coefficient from a scenario correction coefficient library. This second correction coefficient is specifically used to quantify the amplification effect of the line's own condition on meteorological risks. For example, for a newly built 10kV standard overhead line, the tower structure is robust, and the conductor sag meets design standards, exhibiting good tolerance under strong winds and icing conditions; the second correction coefficient is set to 1.00. For a 35kV long-span overhead line, due to the larger distance between adjacent towers, the conductor is more prone to wind-induced swaying, potentially leading to insufficient phase-to-phase distance and discharge accidents; therefore, the second correction coefficient is set to 1.26. For older 10kV lines with over 15 years of service life, tower corrosion, aging hardware, and decreased conductor strength significantly increase the probability of line breakage and tower collapse under icing or strong wind conditions compared to newly built lines; the second correction coefficient is set to 1.21. For lines with existing defects, such as broken insulators or tilted towers, the second correction factor is set to 1.38, indicating that the risk is amplified by 38%. For 10kV cable lines, since the cables are buried underground, meteorological factors such as wind speed and icing do not directly affect the cables themselves, so the second correction factor is set to 0.90. After the edge computing terminal reads the line parameters from the power distribution operation data, it matches the corresponding second correction factor in the scene correction factor library.

[0044] After acquiring the first and second correction coefficients, the edge computing terminal determines the terrain type and season type based on the current work location and time. The current work location is obtained through the GPS positioning module of the wearable sensing terminal, which includes the longitude, latitude, and altitude information of the work point. The edge computing terminal uses the latitude and longitude coordinates of the current work location as an index to initiate a query request to a pre-set geographic database. The geographic database stores terrain and landform classification data nationwide, including types such as mountains, plains, valleys, high altitudes, urban built-up areas, and hilly slopes. The geographic database returns the terrain type matching the current work location. For example, if the work point is located in a narrow passage between two ridges, the geographic database returns the "valley pass" type; if the work point is located in a low-lying area along a river, it returns the "river valley low-lying area" type; if the work point is located in an area with an altitude exceeding 1000 meters, it returns the "high-altitude mountainous area" type. Based on the returned terrain type, the edge computing terminal retrieves the third correction coefficient from the scene correction coefficient library. The third correction coefficient is specifically used to quantify the amplification or reduction effect of different terrains on micro-meteorological effects. For example, due to the funneling effect, the actual wind speed in valley pass areas can be 40% to 50% higher than the regional weather forecast, so the third correction factor is set to 1.54; low-lying river valleys are prone to flooding and flash floods during heavy rains, hindering the evacuation of workers, so the third correction factor is set to 1.43; high-altitude mountainous areas have thin air, low temperatures, and frequent lightning activity, so the third correction factor is set to 1.37; open plains have flat terrain and stable airflow, so the third correction factor is set to 1.00; densely built-up urban areas have a blocking and weakening effect on wind speed, so the third correction factor is set to 0.90. The edge computing terminal also determines the season type based on the current time. The current time is read from the edge computing terminal's system clock, including year, month, day, hour, minute, and second. The edge computing terminal determines the season based on the month: March to May is spring, June to August is summer, September to November is autumn, and December to February of the following year is winter. Simultaneously, the system determines whether a period is considered special based on the specific time period: the afternoon period is defined as 12:00 to 16:00, the night period as 18:00 to 6:00 the next day, and holidays are determined according to the national holiday schedule. The combination of seasonal type and time period type determines the probability and severity of meteorological risks. The edge computing terminal obtains a fourth correction coefficient from the scene correction coefficient library based on the seasonal type and time period type. For example, during the summer thunderstorm season in the afternoon, severe convective weather occurs frequently, and the risk of lightning and strong winds is significantly higher than at other times, so the fourth correction coefficient is set to 1.30; during the winter icing period at night, low temperatures combined with the risk of icing reduce the operational flexibility of workers, so the fourth correction coefficient is set to 1.30; during the spring and autumn weekdays, meteorological conditions are relatively stable, so the fourth correction coefficient is set to 1.00.

[0045] After obtaining the first, second, third, and fourth correction coefficients, the edge computing terminal multiplies these four types of correction coefficients to obtain the comprehensive scenario correction coefficient. Each scenario factor has a multiplicative impact on the basic risk; when multiple factors exist simultaneously, their combined impact is the product of the correction coefficients of each factor. The formula for calculating the comprehensive scenario correction coefficient is: multiply the first, second, third, and fourth correction coefficients consecutively. After the multiplication operation is completed, the edge computing terminal performs a boundary check on the comprehensive scenario correction coefficient. According to predefined constraint rules, the value range of the comprehensive scenario correction coefficient is limited to between 0.3 and 1.7. If the multiplication result is less than 0.3, it is forcibly set to 0.3; if the multiplication result is greater than 1.7, it is forcibly set to 1.7. The above boundary check operation avoids the risk amplification or reduction effect beyond the physical reality caused by multiplying multiple extreme correction coefficients. For example, when four factors exist simultaneously: live-line work (first correction factor 1.30), valley wind gap (third correction factor 1.54), old lines (second correction factor 1.21), and summer afternoon (fourth correction factor 1.30), the product of these factors is 1.30×1.54×1.21×1.30≈3.15, which far exceeds the physically reasonable upper limit of 1.7. In this case, the value is forcibly truncated to 1.7 to avoid the risk value being excessively amplified, leading to frequent false alarms.

[0046] Finally, the edge computing terminal multiplies the basic meteorological risk value output in step S3 by the comprehensive scenario correction coefficient to obtain the comprehensive risk value. The formula for calculating the comprehensive risk value is: multiply the basic meteorological risk value by the comprehensive scenario correction coefficient. If the product exceeds 100, it is forcibly truncated to 100; if the product is less than 0, it is forcibly truncated to 0. The comprehensive risk value is a dimensionless value between 0 and 100, which simultaneously integrates real-time meteorological monitoring data, short-term forecast data, weighted summation of multiple types of risks, and correction effects from four types of scenario factors. For example, when performing 10kV live-line work in a windy valley on a summer afternoon, assuming the basic meteorological risk value is 65, and the comprehensive scenario correction coefficient, after multiplication and boundary checks, is 1.7, then the comprehensive risk value is 65 × 1.7 = 110.5, truncated to 100, indicating that the risk has reached the highest level and a red alert should be triggered immediately. When conducting power outage maintenance work on a plain under the same meteorological conditions, the first correction factor is 1.00, the third correction factor is 1.00, and the comprehensive scenario correction factor may only be 1.21. The comprehensive risk value is 65 × 1.21 = 78.65, corresponding to an orange alert, which differs by one alert level. Because the same micro-meteorological conditions produce different comprehensive risk values ​​under different operational scenarios, differentiated alert responses are triggered. This ensures operational safety in high-risk scenarios while avoiding over-control in low-risk scenarios. The edge computing terminal temporarily stores the calculated comprehensive risk value in its local memory.

[0047] Step S5: Match the comprehensive risk value with the preset risk and instruction relationship to obtain the risk level and early warning instruction.

[0048] After the edge computing terminal completes scene correction processing and obtains a comprehensive risk value, this comprehensive risk value is a dimensionless numerical value between 0 and 100. It integrates real-time meteorological monitoring data, short-term forecast data, weighted summation of multiple types of risks, and corrections for four types of scene factors. However, this value itself cannot directly guide the actions of on-site personnel and needs to be converted into a risk level with clear operational implications and specific warning instructions. To this end, the edge computing terminal matches the comprehensive risk value with a preset risk-instruction relationship. The risk-instruction relationship is stored in the local hard memory of the edge computing terminal. This relationship includes a set of risk threshold intervals and the corresponding risk level and warning instruction for each interval. The set of risk threshold intervals consists of three consecutive numerical intervals: the first risk interval, the second risk interval, and the third risk interval. According to the safety regulations of the power industry and the actual requirements of power distribution operations, the lower limit of the first risk interval is set to 30 and the upper limit to 60, the lower limit of the second risk interval is set to 60 and the upper limit to 90, and the lower limit of the third risk interval is set to 90 and the upper limit to 100. A comprehensive risk value below the lower limit of the first risk range (i.e., less than 30) is defined as a safe state and does not trigger any warning. The division of this threshold range is based on the classification standards for meteorological warning signals in the "Electric Power Safety Work Regulations" and the statistical analysis results of accident data from power distribution work sites over many years.

[0049] The edge computing terminal first reads the comprehensive risk value and compares it with the boundary values ​​in the risk threshold interval set. The comparison process is performed sequentially from low risk to high risk. The edge computing terminal determines whether the comprehensive risk value is less than the lower limit of 30 in the first risk interval. If the result is true, it means that the current weather conditions do not pose a threat to operational safety, the operational window is valid, and the edge computing terminal does not generate any warning commands, trigger audible and visual alarms, or send any prompts to any terminals. The work team continues the operation according to the normal procedure. For example, when performing ground power outage maintenance work in an open plain, the comprehensive risk value is 25, which is lower than 30, indicating a safe state. Workers do not need to pay attention to weather changes and can focus on the maintenance task.

[0050] If the overall risk value is greater than or equal to 30, the edge computing terminal further determines whether the value falls within the first risk range, i.e., whether it is greater than or equal to 30 and less than or equal to 60. If the determination result is true, it indicates that there is a potential meteorological hazard, and the risk has not yet reached the level of prohibiting operations, but the operators need to remain vigilant and pay attention to weather changes. The edge computing terminal converts the overall risk value into a first-level warning instruction and a risk level of attention based on the relationship between risk and instruction. The first-level warning instruction includes the following specific control content: controlling the preset target equipment to emit a preset prompt sound and a preset yellow light. The target equipment includes the sound and light alarm integrated on the portable weather station deployed at the work site, and the vibration motor and indicator light on the wearable terminal of each operator. The preset prompt sound is set to a short "beep" sound, once every 5 seconds, lasting 0.5 seconds each time, with the volume controlled at 70 decibels, in order to remind the operators to pay attention without causing panic. The preset yellow light is set to a yellow LED light that flashes 30 times per minute, installed on the top of the portable weather station for easy observation by all personnel on site. Simultaneously, the Level 1 warning also sends text prompts to the on-site supervisor's handheld terminal, including the current overall risk value, the main sources of risk (e.g., "Wind speed will rise to 8 m / s in 10 minutes"), and suggested countermeasures (e.g., "Suspend high-altitude operations and strengthen monitoring"). This information is transmitted via 4G / 5G network and displayed on the supervisor's terminal screen. For example, when conducting 10kV power outage maintenance work in a river valley, if the overall risk value is 55, falling within the first risk zone, the Level 1 warning triggers a flashing yellow light and an audible alert. The supervisor receives the message, "Current wind speed is 6 m / s, predicted to reach 9 m / s in 15 minutes; it is recommended to suspend pole work." This warning allows workers to proactively lower their working height or suspend high-risk operations before the risk worsens, avoiding the passive situation of being forced to evacuate during a subsequent orange warning.

[0051] If the overall risk value is greater than 60, the edge computing terminal further determines whether the value belongs to the second risk range, i.e., whether it is greater than 60 and less than or equal to 90. If the determination result is true, it indicates that the meteorological parameters have approached or reached the safety threshold, indicating a medium-to-high risk, and immediate control measures are required. The edge computing terminal converts the overall risk value into a level-two warning instruction and a warning risk level based on the relationship between risk and instruction. The level-two warning instruction includes the following specific control content: controlling the target equipment to emit a preset alarm sound and a preset orange light, and sending a preset warning message to a preset on-site monitoring terminal. The preset alarm sound is set to a rapid "beep beep" sound, sounding twice every 1 second, each lasting 0.2 seconds, with the volume increased to 85 decibels to forcibly attract the attention of the operators. The preset orange light is set to an orange LED light that flashes 60 times per minute, which is more conspicuous than yellow light. Simultaneously, the Level II early warning command sends detailed warning information to the on-site supervisor's handheld terminal, including the current comprehensive risk value, the types of meteorological parameters exceeding the standard, the predicted deterioration trend, and specific control requirements (such as "suspend all high-altitude operations, evacuate personnel to the ground, and prepare for emergency evacuation"). In addition, the Level II early warning command also sends medium-to-high risk warning records to the backend safety management platform via the 4G / 5G network. These records include the location of the work site, work team information, comprehensive risk value, measured values ​​of various meteorological parameters, and predicted values. The work site is highlighted in orange on the backend platform interface. For example, when performing 10kV live-line work in a windy valley, with a comprehensive risk value of 78, falling within the second risk range, the Level II early warning command triggers flashing orange lights and a rapid alarm sound. The supervisor's terminal displays "Wind speed has reached 7.5m / s, predicted to reach 11m / s in 10 minutes. Immediately stop live-line work and personnel should descend the pole." Upon hearing the alarm, the workers immediately stop operation, gather their tools, and evacuate the tower. The entire process is completed within 2 minutes, avoiding the risk of electric shock due to a subsequent surge in wind speed.

[0052] If the overall risk value is greater than 90, the edge computing terminal determines that the value belongs to the third risk range, i.e., greater than 90 and less than or equal to 100. If the determination result is true, it means that the meteorological parameters have exceeded the safety threshold, posing an immediate risk to personnel and equipment safety, and all operations must be stopped immediately and a forced evacuation must be carried out. The edge computing terminal converts the overall risk value into a three-level warning instruction and a danger level based on the relationship between risk and instruction. The three-level warning instruction includes the following specific control content: controlling the target device to emit a preset shriek sound and a preset red light, and sending a preset emergency alarm message to the preset background management platform. The preset shriek sound is set to a continuously changing sharp sound (frequency periodically sweeping between 2kHz and 4kHz), with the volume increased to 100 decibels to produce the strongest warning effect. The preset red light is set to a continuously lit (non-flickering) high-brightness red LED light, or a red strobe light that flashes extremely rapidly (more than 120 times per minute), installed on the top of the portable weather station to ensure that it can still be observed from a distance in severe weather. Simultaneously, the Level 3 warning command sends a continuous high-intensity vibration alarm to the wearable terminals of all on-site workers, displaying the words "Evacuate Immediately" in red and a countdown timer on the terminal screen. An emergency evacuation message is sent to the handheld terminal of the on-site supervisor, including "Immediately stop all work, all personnel evacuate to a safe area," along with a pre-planned evacuation route (based on a GIS system). An emergency warning record is sent to the backend safety management platform, which automatically triggers a platform-level fault alarm pop-up and simultaneously links to the emergency command system, sending SMS and voice call notifications to the dispatch center and maintenance department. For example, during live-line work in a high-altitude mountainous area on a summer afternoon, if the lightning location system shows that the lightning distance has entered the 8km range and the predicted lightning probability is 0.9, with a comprehensive risk value of 96, belonging to the third risk zone, the Level 3 warning command triggers a red flashing light and a whistling sound, all wearable terminals vibrate at high intensity and display "Evacuate Immediately," and the on-site supervisor confirms that all personnel quickly descend the pole and evacuate to a safe area 50 meters away from the tower. Approximately 5 minutes later, the thunderstorm cloud reached the vicinity of the work site. Since all personnel had already evacuated, a lightning strike injury or death was avoided. Simultaneously, the back-end control platform received an emergency alarm, the dispatch center retrieved on-site video to confirm the situation, and notified neighboring work teams to cease operations.

[0053] After completing the matching process described above, the edge computing terminal sends the generated risk level code and warning command to each target device for execution via wired or wireless means. Simultaneously, the edge computing terminal stores information such as the timestamp of this matching, the comprehensive risk value, the trigger interval, and the issued warning command level in a local log file for subsequent security audits and incident tracing. Thus, step S5 completes the full conversion from comprehensive risk values ​​to executable on-site warning actions, enabling meteorological risk warning information to be accurately conveyed to every on-site worker and back-end administrator in multiple forms such as light, sound, text, and vibration. This ensures differentiated response measures corresponding to different risk levels, achieving the safety goal of tiered management.

[0054] The following describes the relevant devices of the dynamic early warning scheme for meteorological risks in power distribution operations provided in the embodiments of this application.

[0055] It should be noted that, in the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program with a predetermined function, which works together with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0056] Please see Figure 2 This is a functional block diagram of a dynamic early warning device for meteorological risks in power distribution operations provided in this application embodiment. This dynamic early warning device 2 for meteorological risks in power distribution operations can be used to perform the functions described in this application. Figure 1 The corresponding steps in the dynamic early warning method for meteorological risks in power distribution operations provided in the embodiment. Specifically, the dynamic early warning device 2 for meteorological risks in power distribution operations may include: The time alignment module 21 is used to query meteorological forecast data based on the received current operation location, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. The feature fusion module 22 is used to extract features and perform weighted fusion on the synchronous meteorological dataset according to the preset fusion weight coefficients to obtain meteorological feature values ​​of each data type. The risk assessment module 23 is used to convert the meteorological characteristic values ​​into corresponding initial risk values ​​according to preset numerical mapping rules, and to perform weighted summation of the initial risk values ​​according to preset risk weight coefficients to obtain basic meteorological risk values. The scene correction module 24 is used to perform scene correction processing on the basic meteorological risk value based on the preset power distribution operation data and the current operation location to obtain a comprehensive risk value; The level instruction module 25 is used to match the comprehensive risk value with the preset risk and instruction relationship to obtain the risk level and warning instruction.

[0057] In an optional implementation, the time alignment module 21 is used for: A meteorological observation area is generated based on the current working location and a preset center distance, and meteorological forecast data is queried through a preset meteorological data service interface based on the meteorological observation area. Real-time data collection of instantaneous wind speed, rainfall, ambient temperature, relative humidity, and visibility at the current work location to obtain the first meteorological detection data; A lightning warning area is generated based on the preset warning distance and the meteorological observation area. The location of lightning occurrence and the speed of storm movement within the lightning warning area are collected in real time. The real-time lightning distance value is calculated based on the current working position, the location of lightning occurrence, and the speed of storm movement to obtain the second meteorological detection data. Based on the collection timestamps in each data set, the meteorological forecast data, the first meteorological detection data, and the second meteorological detection data are time-aligned to obtain a synchronized meteorological dataset.

[0058] In an optional implementation, the feature fusion module 22 is used to: Based on the preset feature types, feature extraction is performed on the synchronous meteorological dataset to obtain wind speed feature set, rainfall feature set and lightning feature set; Based on the wind speed feature set, a first fusion weight coefficient is determined from the preset fusion weight coefficients, and the wind speed feature set is weighted and summed according to the first fusion weight coefficient to generate wind speed feature values. Based on the rainfall feature set, a second fusion weight coefficient is determined from the fusion weight coefficients, and the rainfall feature set is weighted and summed according to the second fusion weight coefficient to generate flood feature values; The probability of lightning occurrence within the lightning warning area is calculated based on the lightning feature set, and lightning prediction feature values ​​are generated.

[0059] In an optional implementation, the risk assessment module 23 is used for: Based on the wind speed safety threshold and rainfall safety threshold in the preset numerical mapping rules, the ratio of the wind speed characteristic value and the flood characteristic value is standardized to obtain the first initial risk value and the second initial risk value. The temperature and humidity index is calculated based on the temperature and humidity values ​​in the synchronous meteorological dataset. The temperature and humidity index is then standardized by ratio calculation according to the temperature and humidity safety threshold in the numerical mapping rule to obtain a third initial risk value. The lightning proximity time is calculated based on the real-time lightning distance value and the storm movement speed. The lightning proximity time is then standardized by ratio calculation according to the time safety threshold in the numerical mapping rule to obtain the lightning time risk value. The lightning time risk value and the lightning prediction feature value are then weighted and summed according to the preset lightning risk weight coefficient to obtain the fourth initial risk value. Based on the current operating location, the evaluation weight coefficients corresponding to each risk value are matched from the preset risk weight coefficients, and the first initial risk value, the second initial risk value, the third initial risk value, and the fourth initial risk value are weighted and summed according to the evaluation weight coefficients to obtain the basic meteorological risk value.

[0060] In an optional implementation, the scene correction module 24 is used to: Based on the voltage level and operation type in the preset power distribution operation data, a first correction coefficient is obtained from the preset scenario correction coefficient library, and based on the line parameters in the power distribution operation data, a second correction coefficient is obtained from the scenario correction coefficient library. Based on the current work location and current time, the terrain type and season type are determined from a preset geographic database, and based on the terrain type and season type, a third correction coefficient and a fourth correction coefficient are obtained from the scene correction coefficient library; Based on the first correction coefficient, the second correction coefficient, the third correction coefficient, and the fourth correction coefficient, the basic meteorological risk value is subjected to scenario correction processing to obtain a comprehensive risk value.

[0061] In an optional implementation, the level instruction module 25 is used for: The comprehensive risk value is compared with the set of risk threshold intervals in the preset risk and instruction relationship; When the comprehensive risk value falls within the first risk range, the comprehensive risk value is converted into a first-level warning instruction and a risk level based on the relationship between risk and instruction. The first-level warning instruction is then used to control the preset target device to emit a preset prompt sound and a preset yellow light. When the comprehensive risk value falls within the second risk range, the comprehensive risk value is converted into a secondary early warning command and a warning risk level according to the relationship between risk and command. The target device is then controlled to emit a preset alarm sound and a preset orange light according to the secondary early warning command, and a preset warning message is sent to a preset on-site monitoring terminal. When the comprehensive risk value falls within the third risk range, the comprehensive risk value is converted into a level three warning instruction and a danger level according to the relationship between risk and instruction. Based on the level three warning instruction, the target device is controlled to emit a preset whistling sound and a preset red light, and a preset emergency alarm message is sent to a preset background management platform.

[0062] It should be understood that the various variations and specific embodiments of the methods provided in the above embodiments are also applicable to the dynamic early warning device for meteorological risks of power distribution operations in this embodiment. Through the foregoing detailed description of the dynamic early warning method for meteorological risks of power distribution operations, those skilled in the art can clearly understand the implementation method of the dynamic early warning device for meteorological risks of power distribution operations in this embodiment. For the sake of brevity, it will not be described in detail here.

[0063] Please see Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. The computer device 3 is used to execute the steps performed by the computer device in the aforementioned method embodiments. The computer device 3 may include one or more devices (e.g., a server, node, terminal device, etc.) or internal components (e.g., a chip, software module, or hardware module). The computer device may include at least one processor 31 and a communication interface 32. Further optionally, the computer device may also include at least one memory 33 and a bus 34. Additionally, the processor 31, communication interface 32, and memory 33 are connected via the bus 34. Wherein: (1) The processor 31 is a module that performs arithmetic and / or logical operations. Specifically, it may be one or a combination of processing modules such as a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor unit (MPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a complex programmable logic device (CPLD), a coprocessor (to assist the central processing unit in completing corresponding processing and applications), and a micro controller unit (MCU).

[0064] (2) The communication interface 32 can be used to provide information input or output to at least one processor 31. And / or, the communication interface 32 can be used to receive data sent from outside and / or send data to outside, and can be a wired link interface including such as an Ethernet cable, or a wireless link interface (Wi-Fi, Bluetooth, general wireless transmission, vehicle short-range communication technology and other short-range wireless communication technologies, etc.). The communication interface 32 can serve as a network interface.

[0065] (3) The memory 33 is used to provide storage space, in which data such as the operating system and computer programs (including program instructions) can be stored. The memory 33 can be one or a combination of random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM), etc.

[0066] In specific implementation, processor 31 executes the following steps by running the computer program stored in memory 33: Based on the received current work location, query meteorological forecast data, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. Based on the preset fusion weight coefficients, feature extraction and weighted fusion are performed on the synchronous meteorological dataset to obtain meteorological feature values ​​for each data type. The meteorological characteristic values ​​are converted into corresponding initial risk values ​​according to the preset numerical mapping rules, and the initial risk values ​​are weighted and summed according to the preset risk weight coefficients to obtain the basic meteorological risk values. Based on the preset power distribution operation data and the current operation location, the basic meteorological risk value is modified according to the scenario to obtain a comprehensive risk value; The comprehensive risk value is matched with the preset risk and instruction relationship to obtain the risk level and early warning instruction.

[0067] In one possible implementation, processor 31 is also used to perform the following operations: A meteorological observation area is generated based on the current working location and a preset center distance, and meteorological forecast data is queried through a preset meteorological data service interface based on the meteorological observation area. Real-time data collection of instantaneous wind speed, rainfall, ambient temperature, relative humidity, and visibility at the current work location to obtain the first meteorological detection data; A lightning warning area is generated based on the preset warning distance and the meteorological observation area. The location of lightning occurrence and the speed of storm movement within the lightning warning area are collected in real time. The real-time lightning distance value is calculated based on the current working position, the location of lightning occurrence, and the speed of storm movement to obtain the second meteorological detection data. Based on the collection timestamps in each data set, the meteorological forecast data, the first meteorological detection data, and the second meteorological detection data are time-aligned to obtain a synchronized meteorological dataset.

[0068] In one possible implementation, processor 31 is also used to perform the following operations: Based on the preset feature types, feature extraction is performed on the synchronous meteorological dataset to obtain wind speed feature set, rainfall feature set and lightning feature set; Based on the wind speed feature set, a first fusion weight coefficient is determined from the preset fusion weight coefficients, and the wind speed feature set is weighted and summed according to the first fusion weight coefficient to generate wind speed feature values. Based on the rainfall feature set, a second fusion weight coefficient is determined from the fusion weight coefficients, and the rainfall feature set is weighted and summed according to the second fusion weight coefficient to generate flood feature values; The probability of lightning occurrence within the lightning warning area is calculated based on the lightning feature set, and lightning prediction feature values ​​are generated.

[0069] In one possible implementation, processor 31 is also used to perform the following operations: Based on the wind speed safety threshold and rainfall safety threshold in the preset numerical mapping rules, the ratio of the wind speed characteristic value and the flood characteristic value is standardized to obtain the first initial risk value and the second initial risk value. The temperature and humidity index is calculated based on the temperature and humidity values ​​in the synchronous meteorological dataset. The temperature and humidity index is then standardized by ratio calculation according to the temperature and humidity safety threshold in the numerical mapping rule to obtain a third initial risk value. The lightning proximity time is calculated based on the real-time lightning distance value and the storm movement speed. The lightning proximity time is then standardized by ratio calculation according to the time safety threshold in the numerical mapping rule to obtain the lightning time risk value. The lightning time risk value and the lightning prediction feature value are then weighted and summed according to the preset lightning risk weight coefficient to obtain the fourth initial risk value. Based on the current operating location, the evaluation weight coefficients corresponding to each risk value are matched from the preset risk weight coefficients, and the first initial risk value, the second initial risk value, the third initial risk value, and the fourth initial risk value are weighted and summed according to the evaluation weight coefficients to obtain the basic meteorological risk value.

[0070] In one possible implementation, processor 31 is also used to perform the following operations: Based on the voltage level and operation type in the preset power distribution operation data, a first correction coefficient is obtained from the preset scenario correction coefficient library, and based on the line parameters in the power distribution operation data, a second correction coefficient is obtained from the scenario correction coefficient library. Based on the current work location and current time, the terrain type and season type are determined from a preset geographic database, and based on the terrain type and season type, a third correction coefficient and a fourth correction coefficient are obtained from the scene correction coefficient library; Based on the first correction coefficient, the second correction coefficient, the third correction coefficient, and the fourth correction coefficient, the basic meteorological risk value is subjected to scenario correction processing to obtain a comprehensive risk value.

[0071] In one possible implementation, processor 31 is also used to perform the following operations: The time alignment module is used to query meteorological forecast data based on the received current operation location, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. The feature fusion module is used to extract features and perform weighted fusion on the synchronous meteorological dataset according to preset fusion weight coefficients to obtain meteorological feature values ​​of each data type. The risk assessment module is used to convert the meteorological characteristic values ​​into corresponding initial risk values ​​according to preset numerical mapping rules, and to perform weighted summation of the initial risk values ​​according to preset risk weight coefficients to obtain basic meteorological risk values. The scenario correction module is used to perform scenario correction processing on the basic meteorological risk value based on preset power distribution operation data and the current operation location to obtain a comprehensive risk value; The risk level instruction module is used to match the comprehensive risk value with a preset risk and instruction relationship to obtain the risk level and warning instruction.

[0072] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product, which includes one or more computer programs. When the computer program is loaded and executed on a computer device, it generates, in whole or in part, the processes or functions described in the embodiments of this application; the computer device can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer program can be stored in or transmitted through a computer-readable storage medium; the computer program can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible to the computer device or a data processing device such as a server or data center that integrates one or more available media; wherein, the available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0073] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A method for dynamic early warning of meteorological risks in power distribution operations, characterized in that, The method includes: Based on the received current work location, query meteorological forecast data, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. Based on the preset fusion weight coefficients, feature extraction and weighted fusion are performed on the synchronous meteorological dataset to obtain meteorological feature values ​​for each data type. The meteorological characteristic values ​​are converted into corresponding initial risk values ​​according to the preset numerical mapping rules, and the initial risk values ​​are weighted and summed according to the preset risk weight coefficients to obtain the basic meteorological risk values. Based on the preset power distribution operation data and the current operation location, the basic meteorological risk value is modified according to the scenario to obtain a comprehensive risk value; The comprehensive risk value is matched with the preset risk and instruction relationship to obtain the risk level and early warning instruction.

2. The dynamic early warning method for meteorological risks in power distribution operations according to claim 1, characterized in that, The step of querying meteorological forecast data based on the received current work location and performing time alignment processing on the meteorological forecast data and the real-time collected meteorological monitoring data to obtain a synchronized meteorological dataset includes: A meteorological observation area is generated based on the current working location and a preset center distance, and meteorological forecast data is queried through a preset meteorological data service interface based on the meteorological observation area. Real-time data collection of instantaneous wind speed, rainfall, ambient temperature, relative humidity, and visibility at the current work location to obtain the first meteorological detection data; A lightning warning area is generated based on the preset warning distance and the meteorological observation area. The location of lightning occurrence and the speed of storm movement within the lightning warning area are collected in real time. The real-time lightning distance value is calculated based on the current working position, the location of lightning occurrence, and the speed of storm movement to obtain the second meteorological detection data. Based on the collection timestamps in each data set, the meteorological forecast data, the first meteorological detection data, and the second meteorological detection data are time-aligned to obtain a synchronized meteorological dataset.

3. The dynamic early warning method for meteorological risks in power distribution operations according to claim 2, characterized in that, in, The meteorological feature values ​​include wind speed feature values, flood feature values, and lightning prediction feature values. The process of extracting features from the synchronous meteorological dataset and performing weighted fusion based on preset fusion weight coefficients to obtain meteorological feature values ​​for each data type includes: Based on the preset feature types, feature extraction is performed on the synchronous meteorological dataset to obtain wind speed feature set, rainfall feature set and lightning feature set; Based on the wind speed feature set, a first fusion weight coefficient is determined from the preset fusion weight coefficients, and the wind speed feature set is weighted and summed according to the first fusion weight coefficient to generate wind speed feature values. Based on the rainfall feature set, a second fusion weight coefficient is determined from the fusion weight coefficients, and the rainfall feature set is weighted and summed according to the second fusion weight coefficient to generate flood feature values; The probability of lightning occurrence within the lightning warning area is calculated based on the lightning feature set, and lightning prediction feature values ​​are generated.

4. The dynamic early warning method for meteorological risks in power distribution operations according to claim 3, characterized in that, The step of converting the meteorological characteristic values ​​into corresponding initial risk values ​​according to preset numerical mapping rules, and then weighting and summing the initial risk values ​​according to preset risk weight coefficients to obtain basic meteorological risk values ​​includes: Based on the wind speed safety threshold and rainfall safety threshold in the preset numerical mapping rules, the ratio of the wind speed characteristic value and the flood characteristic value is standardized to obtain the first initial risk value and the second initial risk value. The temperature and humidity index is calculated based on the temperature and humidity values ​​in the synchronous meteorological dataset. The temperature and humidity index is then standardized by ratio calculation according to the temperature and humidity safety threshold in the numerical mapping rule to obtain a third initial risk value. The lightning proximity time is calculated based on the real-time lightning distance value and the storm movement speed. The lightning proximity time is then standardized by ratio calculation according to the time safety threshold in the numerical mapping rule to obtain the lightning time risk value. The lightning time risk value and the lightning prediction feature value are then weighted and summed according to the preset lightning risk weight coefficient to obtain the fourth initial risk value. Based on the current operating location, the evaluation weight coefficients corresponding to each risk value are matched from the preset risk weight coefficients, and the first initial risk value, the second initial risk value, the third initial risk value, and the fourth initial risk value are weighted and summed according to the evaluation weight coefficients to obtain the basic meteorological risk value.

5. The dynamic early warning method for meteorological risks in power distribution operations according to claim 1, characterized in that, The step of performing scenario correction processing on the basic meteorological risk value based on preset power distribution operation data and the current operation location to obtain a comprehensive risk value includes: Based on the voltage level and operation type in the preset power distribution operation data, a first correction coefficient is obtained from the preset scenario correction coefficient library, and based on the line parameters in the power distribution operation data, a second correction coefficient is obtained from the scenario correction coefficient library. Based on the current work location and current time, the terrain type and season type are determined from a preset geographic database, and based on the terrain type and season type, a third correction coefficient and a fourth correction coefficient are obtained from the scene correction coefficient library; Based on the first correction coefficient, the second correction coefficient, the third correction coefficient, and the fourth correction coefficient, the basic meteorological risk value is subjected to scenario correction processing to obtain a comprehensive risk value.

6. A dynamic early warning device for meteorological risks in power distribution operations, applied to the dynamic early warning method for meteorological risks in power distribution operations as described in claim 1, characterized in that, The device includes: The time alignment module is used to query meteorological forecast data based on the received current operation location, and perform time alignment processing on the meteorological forecast data and the real-time collected meteorological detection data to obtain a synchronized meteorological dataset. The feature fusion module is used to extract features and perform weighted fusion on the synchronous meteorological dataset according to preset fusion weight coefficients to obtain meteorological feature values ​​of each data type. The risk assessment module is used to convert the meteorological characteristic values ​​into corresponding initial risk values ​​according to preset numerical mapping rules, and to perform weighted summation of the initial risk values ​​according to preset risk weight coefficients to obtain basic meteorological risk values. The scenario correction module is used to perform scenario correction processing on the basic meteorological risk value based on preset power distribution operation data and the current operation location to obtain a comprehensive risk value; The risk level instruction module is used to match the comprehensive risk value with a preset risk and instruction relationship to obtain the risk level and warning instruction.

7. A computer device, characterized in that, include: Memory and processor: A memory, wherein one or more computer programs are stored; A processor is used to load one or more computer programs to implement the dynamic early warning method for meteorological risks of power distribution operations as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the dynamic early warning method for meteorological risks of power distribution operations according to any one of claims 1 to 5.

Images