Ice event determination method, apparatus, device, and medium

Abstract

The method provided by the embodiment of the present disclosure comprises the following steps: constructing a three-level progressive identification framework of "multi-source independent screening-time and space correlation matching-fusion decision", overcoming the limitations of a single meteorological data source in coverage, observation accuracy or record standardization, converting historical meteorological data into candidate icing events and confidence levels with time and space attributes, correlating cross-sources based on strict time and space consistency constraints, fusing multi-source evidence, and generating a comprehensive judgment with high credibility. Finally, a set of scientific, unified, traceable and high-confidence historical icing event data set is output, which not only improves the accuracy and robustness of icing event identification, but also provides a solid and reliable data foundation for in-depth study of icing climate evolution, evaluation of future ice disaster risk of power grid and optimization of disaster prevention and mitigation strategies.

Description

Technical Field

[0001] This disclosure relates to the field of power grid disaster prevention and mitigation, and in particular to a method, apparatus, equipment and medium for determining icing events. Background Technology

[0002] Power grid icing is a critical meteorological disaster threatening the safe operation of the power grid. When freezing rain freezes on the surfaces of transmission lines and towers, it rapidly forms icing, triggering a series of cascading ice-related faults, and in extreme cases, even causing widespread power grid paralysis and power outages. Studying the changing characteristics of icing events not only helps to deepen our understanding of the formation mechanism of freezing rain but also provides a scientific basis for predicting the future development trend of power grid icing disasters. However, current methods for identifying icing events still have limitations, resulting in insufficient accuracy in identifying icing events and failing to meet the actual needs of safe power grid operation. Summary of the Invention

[0003] To address, or at least partially address, the aforementioned technical problems, this disclosure provides a method, apparatus, device, and medium for determining icing events. The technical solutions provided by this disclosure are as follows: This disclosure provides a method for determining icing events, the method comprising: Acquire multiple historical meteorological data for the target area within a historical time period, wherein the multiple historical meteorological data are obtained based on different acquisition methods; For each type of historical meteorological data, based on the historical meteorological data, candidate icing events with a time range and / or spatial range are determined, and the confidence level of the candidate icing events is determined. The candidate icing events are used to indicate that icing has occurred on power facilities within the time range and / or spatial range, and the confidence level is used to indicate the degree of credibility of the candidate icing events. Spatiotemporal matching is performed on candidate icing events derived from different historical meteorological data to obtain icing events to be verified. Each icing event to be verified corresponds to a candidate icing event in at least two types of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are associated in the time range and / or the spatial range. The confidence levels of the candidate icing events corresponding to the icing event to be verified are fused to obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

[0004] This disclosure provides an icing event determination apparatus, the apparatus comprising: The acquisition module is used to acquire various historical meteorological data of the target area within a historical time period, wherein the various historical meteorological data are obtained based on different acquisition methods; A determination module is configured to, for each type of historical meteorological data, determine, based on the historical meteorological data, candidate icing events having a time range and / or spatial range and the confidence level of the candidate icing events, wherein the candidate icing events are used to indicate that icing has occurred on power facilities within the time range and / or spatial range, and the confidence level is used to indicate the degree of credibility of the candidate icing events; The matching module is used to perform spatiotemporal matching on candidate icing events from different historical meteorological data to obtain icing events to be verified. The icing events to be verified correspond to candidate icing events in at least two types of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are associated in the time range and / or the spatial range. The determining module is further configured to fuse the confidence levels of the candidate icing events corresponding to the icing event to be verified to obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

[0005] This disclosure provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the above-described method for determining icing events.

[0006] This disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for determining icing events.

[0007] The solution provided in this disclosure overcomes the limitations of single meteorological data sources in terms of coverage, observation accuracy, or record standardization by constructing a three-level progressive identification framework of "multi-source independent screening—spatiotemporal correlation matching—fusion decision-making." It uniformly transforms historical meteorological data into candidate icing events and their confidence levels with spatiotemporal attributes, performs cross-source correlation based on strict spatiotemporal consistency constraints, and then fuses multi-source evidence to generate a highly credible comprehensive judgment. The final output is a scientific, unified, traceable, and highly confident historical icing event dataset, which not only improves the accuracy and robustness of icing event identification but also provides a solid and reliable data foundation for in-depth research on the evolution of icing climate, assessment of future ice disaster risks in power grids, and optimization of disaster prevention and mitigation strategies. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating a method for determining icing events provided in an embodiment of this disclosure; Figure 2 This is a schematic diagram of the structure of an icing event determination device provided in an embodiment of this disclosure; Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this disclosure. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this disclosure clearer, the embodiments of this disclosure will be described in further detail below with reference to the accompanying drawings.

[0011] The terms “first,” “second,” etc., as used in this disclosure may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are used only to distinguish one concept from another. For example, without departing from the scope of this disclosure, a first presupposed scope may be referred to as a second presupposed scope, and similarly, a second presupposed scope may be referred to as a first presupposed scope.

[0012] As used in this disclosure, the terms "at least one," "multiple," "each," and "any" mean that there is one, two, or more than two; "multiple" means two or more than two; "each" refers to each of the corresponding multiples; and "any" means any one of the multiples. For example, multiple meteorological observation units include three meteorological observation units, where "each" means every single one of the three meteorological observation units, and "any" means any one of the three meteorological observation units, which could be the first, the second, or the third.

[0013] It should be noted that the information, data and signals involved in this disclosure are all authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0014] The icing event determination method provided in this disclosure can be executed by a computer device. Optionally, the computer device is a terminal or a server. The terminal can be a mobile phone, tablet computer, computer, or other types of terminal, and the server can be a single server, a server cluster consisting of several servers, or a cloud computing service center.

[0015] Figure 1 This is a flowchart illustrating a method for determining icing events provided in an embodiment of this disclosure, as shown below. Figure 1 As shown, the method includes: 101. Obtain multiple historical meteorological data for the target area within a historical time period. These multiple historical meteorological data are obtained based on different acquisition methods.

[0016] In this embodiment, to deeply study the evolutionary characteristics of icing events, which not only helps to reveal the formation mechanism of icing-related weather phenomena such as freezing rain, but also provides scientific support for predicting the future development trend of power grid icing disasters, it is necessary to systematically identify icing events occurring in specific geographical areas within a historical time period. An icing event, also known as a historical icing process, refers to a natural process within a target geographical area during a certain historical time period, triggered by specific meteorological conditions such as freezing rain and rime ice, resulting in the formation of ice layers on the surface of power facilities such as transmission lines and towers. A historical icing process not only includes the initial occurrence of icing but also covers its continuous, developing, and eventual dissipation time series, and has a clearly defined spatial impact range. To accurately identify icing events in the target area within a historical time period, this scheme collects multi-source historical meteorological data through various acquisition methods. Historical meteorological data from different sources have characteristics in observation methods, spatiotemporal resolution, and information dimensions, complementing each other to improve the comprehensiveness and reliability of icing event identification, ensuring that the constructed historical icing event record has high accuracy and credibility.

[0017] The target area and historical time period refer to the specific geographical range (usually defined by latitude and longitude boundaries) and time span (e.g., from one year to another) set by the user for researching and constructing historical icing records, respectively. Together, they constitute the spatiotemporal boundaries for subsequent data collection and analysis. Historical meteorological data are various types of meteorological information reflecting atmospheric conditions and their evolution within the target area and historical time period. They can characterize the occurrence conditions, development processes, and impact characteristics of historical weather events, and are an important foundation for conducting climate analysis, retrospective analysis of icing disasters, and power grid risk assessment.

[0018] In this embodiment of the disclosure, the area where the power grid icing process needs to be constructed (i.e., the target area) is determined, and the latitude and longitude range of the specific area is defined; the time period where the power grid icing process needs to be constructed (i.e., the historical time period) is determined; and multi-source historical meteorological data of the target area and the historical time period are collected so that historical icing events can be determined for the target area and the historical time period in the future.

[0019] In some embodiments, various historical meteorological data include meteorological station observation data, gridded meteorological data, and historical disaster records. Meteorological station observation data refers to daily meteorological information actually observed and recorded by ground meteorological stations located within the target area over a historical period, including daily weather phenomena (such as whether frost was recorded) and daily precipitation. Meteorological station observation data has high observational reliability, but its spatial coverage is limited by the distribution of stations. Gridded meteorological data (also known as reanalysis gridded data) is a spatiotemporally continuous and structured global or regional climate dataset reconstructed using data assimilation techniques through numerical weather prediction models, combined with multi-source data such as historical satellite remote sensing, radiosonde, and ground observations. For example, ERA5 reanalysis data can provide hourly resolution gridded meteorological elements, including precipitation, temperature, and weather phenomena closely related to icing (such as whether freezing rain occurred), exhibiting good spatiotemporal consistency and coverage. Historical disaster records are unstructured, descriptive disaster information, typically derived from disaster reports issued by government or emergency management departments, power grid company logs of tripping or equipment failures caused by icing, academic literature, local chronicles, and authoritative news reports. Although historical disaster records are not standardized in format and have limited spatiotemporal precision, they directly reflect the actual impact of icing events on the power system and have important corroborating value.

[0020] In this embodiment, meteorological station observation data, grid meteorological data, and historical disaster records are obtained through different methods such as field observation, numerical reanalysis, and manual recording. They complement each other and together constitute a multi-source heterogeneous historical meteorological information foundation, providing support for the accurate identification of subsequent icing events.

[0021] 102. For each type of historical meteorological data, based on the historical meteorological data, determine candidate icing events with a time range and / or spatial range and the confidence level of the candidate icing events. The candidate icing events are used to represent the occurrence of icing on power facilities within the time range and / or spatial range, and the confidence level is used to represent the degree of credibility of the candidate icing events.

[0022] In this embodiment of the disclosure, a candidate icing event refers to a possible icing phenomenon initially identified based on historical meteorological data from a single source (such as meteorological station observations, grid reanalysis data, or historical disaster records). Candidate icing events are defined by a specific time range (e.g., from one day to another) and / or spatial range (e.g., a specific latitude and longitude region or a specific power line segment), indicating that power facilities (e.g., transmission lines, towers, etc.) may experience icing within that spatiotemporal range. Because different data sources vary in accuracy, coverage, and reliability, icing events identified based on a single data source may have uncertainties; therefore, confidence level is introduced as a quantitative indicator.

[0023] In this embodiment of the disclosure, for each type of historical meteorological data, based on preset identification rules, candidate icing events with a clear time range and / or spatial range are extracted from the historical meteorological data, and a corresponding confidence level is assigned to each candidate event to quantify its credibility. Taking various types of historical meteorological data, including meteorological station observation data, grid meteorological data, and historical disaster records, as examples, different types of candidate icing events are identified from the three types of historical meteorological data: 1) The first candidate icing event is identified from meteorological station observation data, which is defined as the simultaneous or continuous recording of rime ice phenomena by multiple meteorological stations in the target area, and meets the preset minimum spatial coverage and duration thresholds; 2) The second candidate icing event is identified from reanalysis grid data, which is defined as the detection of a continuous freezing rain process at the grid scale, and meets the specified spatiotemporal persistence and affected area requirements; 3) The third candidate icing event is identified from historical disaster record data, which is defined as the description of icing disasters from channels such as power grid fault reports, government disaster reports, or authoritative documents, and the impact range or severity of the event reaches the preset scale standard.

[0024] For each type of candidate icing event, an initial confidence level is calculated based on its performance characteristics in the corresponding data source. For example, frost events with more coverage sites and longer durations have higher confidence levels; reanalysis events with stronger freezing rain signals and better grid consistency have higher confidence levels; and disaster records with authoritative sources, specific descriptions, and clear impacts are assigned even higher confidence levels. Different data sources employ confidence assessment rules adapted to their characteristics to ensure that the scoring results objectively reflect the reliability of the candidate event from a single data perspective.

[0025] 103. Spatiotemporal matching is performed on candidate icing events from different historical meteorological data to obtain icing events to be verified. Each icing event to be verified corresponds to a candidate icing event in at least two types of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are related in terms of time range and / or spatial range.

[0026] In this embodiment, spatiotemporal matching refers to the operation of comparing, associating, and aggregating candidate icing events identified from different historical meteorological data sources (such as meteorological station observations, reanalysis grid data, and historical disaster records) in the temporal and / or spatial dimensions. Its core purpose is to determine whether icing signals detected from multiple independent data sources point to the same actual physical icing process. The icing event to be verified is a potential icing process aggregated from candidate icing events from at least two different data sources through spatiotemporal matching. It represents a more credible and complete icing event hypothesis supported by multi-source evidence, but still needs to be confirmed through subsequent confidence fusion and threshold determination to determine whether it constitutes the final historical icing event. This spatiotemporal matching mechanism effectively integrates icing clues from heterogeneous data, avoiding missed or incorrect judgments due to the limitations of a single data source, thereby improving the overall accuracy and robustness of icing event identification.

[0027] In some embodiments, a joint spatiotemporal overlap analysis is performed on a first candidate icing event (derived from meteorological station observation data), a second candidate icing event (derived from reanalysis grid data), and a third candidate icing event (derived from historical disaster records). If two or more candidate icing events are temporally close (e.g., start and end times partially overlap or the interval does not exceed a preset threshold) and spatially intersect or are adjacent (e.g., located in the same power grid area or geographical buffer zone), they are considered to be spatiotemporally "correlated" and can be merged into the same icing event to be verified.

[0028] 104. The confidence levels of the candidate icing events corresponding to the icing event to be verified are fused to obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

[0029] In this embodiment of the disclosure, the overall confidence score refers to the unified confidence score obtained by fusing the individual confidence scores of multiple candidate icing events (such as the first, second, and third candidate icing events) associated with the icing event to be verified. This score is used to comprehensively assess whether the event to be verified actually occurred. The preset confidence score (or judgment threshold) is a pre-set numerical threshold used to determine whether the overall confidence score is sufficient to support the confirmation of the icing event to be verified as a real historical icing event. For example, the preset confidence score is 0.85. Historical icing events refer to icing processes that, after multi-source data identification, spatiotemporal matching, and confidence score fusion, are ultimately confirmed as credible and actually occurred within the target area and historical time period.

[0030] In some embodiments, a probabilistic complementary model can be used to calculate the overall confidence level of each icing event to be verified. The core idea of ​​the probabilistic complementary model is that if an icing process is reflected in multiple independent data sources (such as weather station observations, reanalysis grids, and disaster records), the probability of these independent pieces of evidence simultaneously reporting false alarms is extremely low, thus its overall confidence level is higher than that of any single source. Specifically, the probabilistic complementary model uses mathematical formulas (such as a fusion method based on the probability of independent event complements) to fuse the first, second, and third confidence levels to generate a comprehensive confidence level that reflects the overall confidence level. Subsequently, this comprehensive confidence level is compared with a preset confidence level. If the comprehensive confidence level is greater than or equal to the preset confidence level, the corresponding icing event to be verified is formally confirmed as a reconstructed historical icing event. Historical icing events, after multi-source independent identification, spatiotemporal consistency verification, and confidence quantification assessment, have high reliability and representativeness, and can serve as a basic dataset for studying the patterns of power grid icing disasters, risk modeling, and building future early warning capabilities.

[0031] The solution provided in this disclosure overcomes the limitations of single meteorological data sources in terms of coverage, observation accuracy, or record standardization by constructing a three-level progressive identification framework of "multi-source independent screening—spatiotemporal correlation matching—fusion decision-making." It uniformly transforms historical meteorological data into candidate icing events and their confidence levels with spatiotemporal attributes, performs cross-source correlation based on strict spatiotemporal consistency constraints, and then fuses multi-source evidence to generate a highly credible comprehensive judgment. The final output is a scientific, unified, traceable, and highly confident historical icing event dataset, which not only improves the accuracy and robustness of icing event identification but also provides a solid and reliable data foundation for in-depth research on the evolution of icing climate, assessment of future ice disaster risks in power grids, and optimization of disaster prevention and mitigation strategies.

[0032] In some embodiments, historical meteorological data indicates the meteorological conditions observed by multiple meteorological observation units within a target area during a historical time period; then step 102 includes the following steps 1021-1024.

[0033] 1021. Based on historical meteorological data, identify the target meteorological observation units where the target weather phenomenon was observed within the historical time period.

[0034] In this embodiment, based on multiple meteorological observation units in historical meteorological data, target meteorological observation units that have observed the target weather phenomenon (such as freezing rain or rime) within a historical time period are first selected, and these observation records are aggregated into temporally continuous and spatially adjacent meteorological events. The spatial coverage of each meteorological event is defined as the ratio of its observation frequency or effective observation value to the total number of meteorological observation units in the region, used to quantify the relative impact breadth of the target weather phenomenon. Only when the spatial coverage of a meteorological event is not less than a first preset range and the duration does not exceed a preset maximum duration will it be identified as a candidate icing event. Finally, the confidence level of the candidate icing event is quantified by comparing the actual spatial coverage with a preset daily spatial coverage ratio, ensuring that the confidence level can reflect the credibility of the candidate icing event.

[0035] In this embodiment of the disclosure, historical meteorological data refers to the set of meteorological information collected or recorded by multiple meteorological observation units within a target area during a specified historical period. The target weather phenomenon specifically refers to meteorological conditions directly related to icing of power facilities, such as freezing rain, rime ice, wet snow, or precipitation accompanied by temperatures below 0°C. The target meteorological observation unit refers to the meteorological observation unit that actually observed or recorded the aforementioned target weather phenomenon during that historical period.

[0036] In some embodiments, historical meteorological data are meteorological station observation data or grid meteorological data, and the meteorological observation unit is a meteorological station or a grid area.

[0037] In this embodiment, a meteorological observation unit refers to a basic spatial unit used to collect or characterize meteorological information: when meteorological station observation data is used, the meteorological observation unit is a specific ground meteorological station; when gridded meteorological data (such as reanalysis grid data) is used, the meteorological observation unit corresponds to a regularly divided grid area (such as latitude and longitude grid points in ERA5 data). Different types of observation units reflect the spatial structure and observation methods of different data sources, providing a foundation for subsequent unified processing and icing event identification.

[0038] 1022. Based on the time when the target weather phenomenon is observed by the target meteorological observation unit, determine at least one meteorological event with spatial coverage and duration. The meteorological event indicates the process of the target weather phenomenon occurring continuously in time. The spatial coverage is the ratio of the sum of the observation frequencies of the target weather phenomenon observed by each target meteorological observation unit corresponding to the meteorological event to the total number of meteorological observation units in the target area.

[0039] In this embodiment, a meteorological event refers to a spatiotemporal process consisting of the continuous occurrence of a target weather phenomenon (such as freezing rain or rime), identified through observation records from target meteorological observation units. Duration refers to the length of time the meteorological event spans from start to finish (e.g., 2 consecutive days or 6 hours), characterizing the event's temporal persistence. Spatial coverage is not simply a statistical count of the number of observation units involved, but rather quantified as the ratio of the sum of the observation frequencies or effective observation values ​​of all target meteorological observation units for the target weather phenomenon during the event's duration to the total number of meteorological observation units within the target area. This reflects the relative breadth and density of the weather phenomenon's influence within the area. This approach balances observation density and regional scale, ensuring comparability of meteorological events from different regions or data sources, and providing a unified basis for subsequent screening of candidate icing events and confidence assessment.

[0040] 1023. Meteorological events with a spatial coverage area not less than the first preset range and a duration not greater than the preset duration are identified as candidate icing events. The first preset range is determined based on the duration and the preset daily spatial coverage ratio.

[0041] In this embodiment, a candidate icing event refers to a meteorological event that meets specific spatiotemporal conditions and is preliminarily considered to potentially correspond to a real icing process. To screen for meaningful candidate icing events, two key thresholds are set: a first preset range and a preset duration. The preset duration is the maximum allowed duration of icing-related meteorological phenomena (e.g., no more than 7 days), used to exclude abnormal or atypical long-term signals. The first preset range is the minimum coverage threshold for determining whether the spatial impact is sufficient. The first preset range is not a fixed value but is dynamically calculated based on the event's duration and a preset daily spatial coverage ratio. Only when the spatial coverage of a meteorological event is not less than the first preset range and the duration does not exceed the preset duration is it confirmed as a candidate icing event, thus balancing the spatial representativeness and temporal rationality of the event.

[0042] In some embodiments, the definitions of the first preset range and the preset duration are adjusted according to the different types of historical meteorological data: When the historical meteorological data is meteorological station observation data, the first preset range refers to the minimum proportion of the number of meteorological stations with effective rime ice out of the total number of meteorological stations in the target area. Here, a "effective rime ice" meteorological station refers to a station that recorded rime ice on a given day and whose daily precipitation was not less than the first preset value; the preset duration is the maximum number of consecutive days that rime ice is allowed to occur. When the historical meteorological data is gridded meteorological data (such as reanalysis grid data), the first preset range refers to the minimum proportion of the number of grids with effective freezing rain out of the total number of grids in the target area. Here, a "effective freezing rain" grid refers to a grid whose daily cumulative freezing rain calculated based on reanalysis grid data is not less than the second preset value; the preset duration is the maximum number of consecutive days that freezing rain is allowed to occur.

[0043] In this embodiment of the disclosure, due to the differences in scale and accuracy between site observation and grid simulation, the first preset value is greater than the second preset value to ensure the physical consistency and comparability of the identification standards under different data sources.

[0044] 1024. Determine the confidence level of candidate icing events based on spatial coverage range and daily spatial coverage ratio.

[0045] In this embodiment, the confidence level of a candidate icing event is used to quantify how credibly it reflects an icing process. The confidence level is calculated based on two indicators: spatial coverage and daily spatial coverage ratio. Spatial coverage represents the cumulative impact breadth of effective meteorological observation units (such as stations or grids) that observe the target weather phenomenon (such as rime or freezing rain) throughout the entire event duration, typically expressed as a proportion of the total number of observation units within the target area. Daily spatial coverage ratio is a preset benchmark threshold, representing the minimum regional impact ratio that a typical icing event should achieve within a single day. By comparing the actual spatial coverage with the daily spatial coverage ratio calculated based on the duration, the confidence level of the candidate icing event is evaluated, ensuring that the confidence level not only reflects the intensity of the candidate icing event but also its spatiotemporal rationality, providing a reliable weighting basis for subsequent multi-source fusion.

[0046] In some embodiments, step 1024 includes: querying a preset correspondence based on the ratio of spatial coverage range to daily spatial coverage ratio to obtain the confidence level corresponding to the numerical interval to which the ratio belongs; wherein, the preset correspondence includes multiple non-overlapping numerical intervals and the confidence level corresponding to each numerical interval.

[0047] In this embodiment, the ratio of the spatial coverage area of ​​a candidate icing event to the daily spatial coverage ratio (i.e., the multiple of the actual coverage intensity relative to the baseline expectation) is calculated, and then a query is performed in a preset correspondence based on this ratio. The preset correspondence is a set of predefined, non-overlapping numerical intervals and their respective mapped confidence values. By assigning the calculated ratio to the corresponding interval, the confidence level of the candidate icing event can be directly obtained. This method transforms continuous coverage intensity into discrete but reasonable confidence levels, simplifying the evaluation logic while ensuring the objectivity and operability of confidence assignment.

[0048] In this embodiment, by refining historical meteorological data to the meteorological observation unit level and constructing meteorological events based on the spatiotemporal continuity of the target weather phenomenon, refined identification of icing processes is achieved. By introducing dual constraints of a first preset range and a preset duration, sporadic, localized, or abnormally persistent noise signals are eliminated, improving the physical plausibility of candidate icing events. Simultaneously, the confidence level is quantified based on the ratio of spatial coverage to daily spatial coverage ratio, making the evaluation results more objective and comparable, and improving the accuracy of historical icing event identification.

[0049] For example, when historical meteorological data is meteorological station observation data, candidate icing events are identified according to the above steps 1021-1024: First, the total number of meteorological stations A_station in the target area is determined according to the defined latitude and longitude range, and meteorological stations of the same level are selected as the analysis objects. In order to exclude local and small-scale atypical icing processes, it is necessary to set a lower limit for the influence range of typical icing processes, that is, define the minimum effective influence ratio I_station. Only when the coverage of the rime ice phenomenon reaches or exceeds this ratio I_station is it considered to have regional representativeness. Specifically, meteorological stations that meet the following conditions are defined as effective rime ice meteorological stations: (1) rime ice phenomenon is recorded on the day; (2) the daily precipitation is not less than 2 mm. For a typical icing process, the number of all effective rime ice meteorological stations A_station_ys during its duration is calculated, and its proportion to the total number of meteorological stations in the area is defined as the event spatial influence ratio I_event_ys, that is: I_event_ys=(A_station_ys / A_station)×100%.

[0050] Further, an upper limit L_station is set for the duration of a typical icing process. Generally, the duration of rime ice is relatively short, so the value is usually small to exclude long-term sporadic rime ice phenomena that last for several consecutive days. If any of the following conditions are met within a continuous period, it can be determined as a possible icing process based on rime ice monitoring: 1) The percentage of meteorological stations with rime ice on a single day exceeds I_station; 2) Rime ice occurs for two consecutive days, and the cumulative percentage of stations with rime ice exceeds 1.5 times I_station (where a station with rime ice for two consecutive days is counted as 2 stations); 3) Rime ice occurs for three consecutive days, and the cumulative percentage of stations with rime ice exceeds 2 times I_station (where the same meteorological station with rime ice for two out of three days is counted as 2 stations); 4) And so on, when rime ice occurs for L consecutive days (L is not greater than L_station), the cumulative percentage of stations with rime ice I_event_ys must satisfy I_event_ys greater than 0.5×L×I_station+0.5×I_station. Ultimately, all continuous processes that satisfy I_event_ys≥0.5×L×I_station+0.5×I_station are identified as icing processes based on rime monitoring in the corresponding time period in that region, and are thus identified as candidate icing events.

[0051] To further quantify the credibility of candidate icing events, for each icing process identified in a round (i.e., each candidate icing event), the maximum value of the percentage of stations where rime ice occurred during the entire duration, I_event_ysmax, is calculated. Based on the relationship between I_event_ysmax and I_station, the confidence score CL_ys is determined: if I_event_ysmax ≥ 3 × I_station, then CL_ys equals 0.90; if 3 × I_station > I_event_ysmax ≥ 2 × I_station, then CL_ys equals 0.75; if 2 × I_station > I_event_ysmax ≥ I_station, then CL_ys equals 0.60; if I_event_ysmax < I_station, then CL_ys equals 0.00.

[0052] The confidence value CL_ys serves as the initial confidence level for the candidate icing event, used for subsequent multi-source fusion and verification. By applying these rules, regionally representative and physically plausible icing events can be scientifically and consistently identified from meteorological station observation data, and assigned reasonable confidence scores, providing a reliable basis for constructing a high-quality historical icing event dataset.

[0053] For example, when historical meteorological data is reanalysis grid data, candidate icing events are identified according to steps 1021-1024 above: First, determine the total number of grids A_grid within the target area based on the defined latitude and longitude range. To exclude localized, small-scale atypical icing processes, a lower limit for the impact range of a typical icing process needs to be set, i.e., a minimum effective impact ratio I_grid is defined. Only when the coverage area of ​​freezing rain reaches or exceeds this ratio I_grid is it considered regionally representative. Specifically, the specific hourly time of freezing rain occurrence and its corresponding precipitation for each grid between 20:00 yesterday and 20:00 today are statistically analyzed. The precipitation of all hours with freezing rain is accumulated to obtain the daily freezing rain amount for that grid on that day. Grids with a daily freezing rain amount of not less than 1 mm are defined as effective freezing rain grids. For a typical icing process, calculate the number of all effective freezing rain grids A_grid_ys during its duration, and define the proportion of the number of grids A_grid_ys to the total number of grids in the region as the event space influence proportion I_event_grid, i.e.: I_event_grid=(A_grid_ys / A_grid)×100%.

[0054] Further, an upper limit L_grid is set for the duration of a typical icing process. Generally, freezing rain events are relatively short-lived, so a smaller value is typically used to exclude prolonged freezing rain events lasting several days. A possible icing process based on reanalysis freezing rain data is determined if any of the following conditions are met within a continuous period: 1) The percentage of valid freezing rain grids on a single day exceeds I_grid; 2) Valid freezing rain grids appear for two consecutive days, and the cumulative percentage of valid freezing rain grids exceeds 1.5 times I_grid (where a grid that appears as valid freezing rain for two consecutive days is counted as 2 valid freezing rain grids); 3) Valid freezing rain grids appear for three consecutive days, and the cumulative percentage of valid freezing rain grids exceeds 1.5 times I_grid. The proportion of rain-affected grids (I_event_grid) exceeds twice the value of I_grid (where a grid that is valid for three consecutive days is counted as 3 valid rain-affected grids; if a grid is valid for two out of three days, it is counted as 2 valid rain-affected grids); 4) Similarly, when valid rain-affected grids appear for L consecutive days (L ≤ L_grid), the cumulative proportion of valid rain-affected grids (I_event_grid) must satisfy I_event_grid ≥ 0.5 × L × I_grid + 0.5 × I_grid. Finally, all continuous processes that satisfy I_event_grid ≥ 0.5 × L × I_grid + 0.5 × I_grid are identified as icing processes based on reanalysis of freezing rain data in the corresponding time period in that region, i.e., they are all identified as candidate icing events.

[0055] To further quantify the credibility of candidate icing events, for each icing process identified in each round (i.e., each candidate icing event), the maximum value of the percentage of effective freezing rain grids that occurred cumulatively over the entire duration, I_event_gridmax, is calculated. Based on the numerical relationship between I_event_gridmax and I_grid, the confidence score CL_grid is determined: if I_event_gridmax ≥ 3 × I_grid, then CL_grid is 0.85; if 3 × I_grid > I_event_gridmax ≥ 2 × I_grid, then CL_grid is 0.70; if 2 × I_grid > I_event_gridmax ≥ I_grid, then CL_grid is 0.55; if I_event_gridmax < I_grid, then CL_grid is 0.00.

[0056] The confidence value CL_grid serves as the initial confidence score for the candidate icing event, used for subsequent multi-source fusion and verification. By applying these rules, regionally representative and physically plausible icing events can be scientifically and stably identified from the reanalysis grid data, and assigned reasonable confidence scores, providing a reliable basis for constructing a high-quality historical icing event dataset.

[0057] In some embodiments, historical meteorological data are historical disaster records, which are used to record the location and impact range of abnormal power system events caused by icing; then step 102 includes the following steps 1025-1027: 1025. Based on historical disaster record data, determine the candidate areas for ice accumulation disasters and the corresponding occurrence times. The candidate areas are located in the target area, and the occurrence times belong to the historical time period.

[0058] In this embodiment, when historical meteorological data uses historical disaster records, these records are derived from power grid company fault reports, government disaster reports, academic literature, local chronicles, or authoritative news reports. These records document the specific location and impact range of abnormal power system events (such as line tripping or tower collapse) caused by icing. Icing disaster events located within the target area and belonging to a specific historical time period are extracted from the historical disaster records to determine their candidate areas and occurrence times. A second preset range is set as an area threshold; candidate icing events are generated only when the candidate area is not smaller than this threshold, thus excluding isolated events with too small an impact. Confidence levels are assigned based on the type of record source (such as official reports, power grid logs, academic literature, or media reports); the more authoritative the source and the more detailed the information, the higher the confidence level. This transforms non-standardized text information into structured candidate events with spatiotemporal attributes and confidence scores, providing effective input for multi-source fusion.

[0059] Historical disaster record data refers to textual or structured information on power system faults or natural disasters related to icing, recorded by power departments, meteorological departments, emergency management agencies, academic research institutions, or authoritative media reports. This information typically includes the time, location, phenomenon description, and impact of the event. Candidate areas for icing disasters are geographical areas (e.g., counties, townships, line sections) extracted from historical disaster records that clearly document icing or related power faults. These candidate areas must be located within the user-defined target area. Occurrence time refers to the specific date or time period (e.g., a specific year, month, and day) indicated in the record of the icing disaster event. This time must fall within the user-defined historical time period to ensure the event's temporal relevance. The target area refers to the geographical area of ​​interest to the user, such as a province, power grid zone, or specific transmission corridor, used to define the spatial boundaries of the analysis. The historical time period refers to a specific past time interval (e.g., 2000-2023) set by the user, used to filter historical events that meet the criteria.

[0060] In this embodiment of the disclosure, historical disaster records are analyzed to extract event information related to icing. Combined with spatial and temporal constraints, the possible locations and times of icing events are identified. Through standardization of geographical locations and time alignment in the records, events located within the target area and occurring within a historical time period are selected, forming a preliminary spatiotemporal candidate set.

[0061] 1026. If the range of the candidate region is not less than the second preset range, generate candidate icing events based on the range of the candidate region and the occurrence time.

[0062] The candidate region's scope refers to the size of the geographical area affected by the icing disaster, which can be expressed as an area (e.g., square kilometers), the number of administrative regions covered, or the length of affected transmission lines. The second preset scope is a pre-defined minimum impact threshold used to determine whether a candidate region is regionally representative. For example, it can be set to "affecting no fewer than two county-level administrative regions" or "having an area of ​​no less than 1000 square kilometers." In some embodiments, the second preset scope is the minimum number of affected administrative regions, used to measure the spatial breadth of the candidate region's impact. For example, this threshold can be set to two county-level administrative regions; that is, only when the impact range of an icing event described in historical disaster records covers no fewer than two county-level administrative units is it considered to have regional characteristics, thus generating a candidate icing event. A candidate icing event refers to a structured event with clear spatiotemporal attributes, generated after meeting spatial and temporal conditions and passing scope screening, representing a possible regional icing process.

[0063] In this embodiment, a "second preset range" is introduced as a screening criterion to exclude isolated fault events with too small an impact range, limited to local points, and to avoid misjudging them as regional icing processes. Only when the range of the candidate area reaches or exceeds this threshold is a formal "candidate icing event" generated based on its spatial range and occurrence time.

[0064] 1027. Determine the confidence level of candidate icing events based on the source type of historical disaster records corresponding to the candidate icing events.

[0065] The source type of historical disaster records refers to the category of the original records on which the candidate icing event was based. For example, source types include: power grid company operation and maintenance logs, government disaster reports, meteorological station reports, academic research results, local chronicles, and authoritative media reports. The mapping relationship between source type and confidence level involves pre-setting confidence levels for different source types. For example, the confidence level for official power grid fault reports is 0.9, for government reports it is 0.8, for academic papers it is 0.7, for local chronicles it is 0.6, for authoritative media reports it is 0.5, and for non-authoritative online information it is 0.3 or below.

[0066] In this embodiment, the confidence level of candidate icing events is quantitatively assessed based on the historical records they rely on. Since the reliability of information from different sources varies, a mapping rule between source type and confidence level is established to transform qualitative information into quantitative indicators that can participate in subsequent fusion calculations. For example, fault logs from the power grid company's internal system have high accuracy and traceability, thus receiving a high confidence level; while unverified social media information has a lower confidence level. This mechanism effectively improves the scientific rigor and objectivity of multi-source data fusion, providing a reliable basis for subsequent event verification and risk assessment.

[0067] In some embodiments, step 103 includes: determining the icing event to be verified based on candidate icing events derived from different historical meteorological data that overlap or are adjacent in time and / or spatial ranges.

[0068] The phrase "originating from different historical meteorological data" indicates that these candidate icing events originate from at least two heterogeneous data sources. For example, one source might be from meteorological station frost records, and the other from reanalysis of freezing rain data, or from disaster reports. Overlapping or adjacent temporal and / or spatial ranges refer to two or more candidate icing events partially or completely overlapping in time (e.g., both occurring between January 10th and 12th), and / or having an intersection or proximity in geographic space (e.g., affected areas are adjacent or the distance is less than a preset buffer distance). An icing event to be verified refers to a fused event formed by aggregating candidate icing events from multiple sources through spatiotemporal matching. It represents a potentially real icing process and requires further verification through confidence fusion.

[0069] In this embodiment, spatiotemporal correlation analysis is performed on candidate icing events from different historical meteorological data sources. If two or more candidate icing events are similar in time (e.g., start and end dates overlap or the interval does not exceed the tolerance threshold) and overlap or are adjacent in space (e.g., affected areas intersect or boundary distance is within a reasonable range), they are considered to likely describe the same real icing process. These spatiotemporally correlated candidate events are then clustered to form a unified icing event to be verified. By integrating icing clues from heterogeneous data sources such as observations, simulations, and disaster records through a matching mechanism based on spatiotemporal overlap or adjacency, the accuracy of event identification is improved.

[0070] In this embodiment, by introducing historical disaster records as an independent data source and combining spatial range filtering and source credibility assessment mechanisms, evidence of actual icing disasters is mined, making up for the problems that pure meteorological data may have, such as "meteorological conditions exist but there is no actual icing" or "observation blind spots are missed". At the same time, the confidence assignment method based on source type takes into account both data authenticity and usability, enabling unstructured text information to be quantitatively participated in subsequent fusion decision-making, thereby improving the completeness, objectivity and practicality of historical icing event identification.

[0071] In this embodiment, to exclude localized icing events with too small an impact range, a lower limit can be set for the impact range of a typical icing process. For example, it can be stipulated that at least N county-level administrative regions within the area must be affected to be considered a regionally representative icing process. This threshold is used to ensure that the identified events have a certain spatial breadth and degree of disaster impact, avoiding misjudging isolated point faults as regional icing events. Based on disaster damage records released by relevant departments and historical ice-related power outage records of the power grid company, the icing process occurring in the target area within a specified time period is determined. Disaster records typically include the time of occurrence, affected area, degree of icing, and a description of related impacts. Based on this information, its credibility is further assessed, and a corresponding confidence level is assigned. In this embodiment, disaster records from different sources correspond to different confidence levels: official disaster reports released by emergency management departments or disaster information disclosed at press conferences are given a confidence level of 1 due to their strong authority and completeness; historical ice-induced power outage records provided by power grid companies have a confidence level ranging from 0.5 to 0.9, based on the specific information content in the outage investigation report (such as faulty lines, ice thickness, duration, etc.); and non-official sources such as academic documents, books, local chronicles, and news reports have a confidence level ranging from 0.4 to 0.8, based on their authority, detailed content, and information credibility. Through this method, unstructured disaster records are transformed into structured data with spatiotemporal attributes and quantified confidence levels, providing a reliable basis for subsequent multi-source fusion and icing event confirmation. This method effectively utilizes actual disaster damage evidence, compensates for the shortcomings of meteorological observation data in terms of spatial coverage and event verification, and improves the accuracy and practicality of historical icing event identification.

[0072] In some embodiments, the process of determining the overall confidence level includes: determining the difference between the confidence level and a preset value for each candidate icing event corresponding to the icing event to be verified; determining the product of the obtained differences; and determining the difference between the preset value and the product as the overall confidence level of the icing event to be verified.

[0073] The preset value is a pre-defined constant, typically set to 1, used to construct the mathematical basis for confidence fusion, representing the "completely unreliable" complement benchmark. The difference refers to the difference between the preset value and the confidence level of each candidate icing event, i.e., 1. Confidence level represents the probability or uncertainty that the evidence "does not support" the occurrence of the event. The product is the result of multiplying the differences between all candidate icing events, reflecting the joint uncertainty of multiple sources of evidence simultaneously "failing" or "misjudging." The overall confidence level is a quantitative assessment of the overall credibility of the icing event to be verified, obtained by subtracting the above product from a preset value, reflecting the overall credibility level of multiple independent sources of evidence jointly supporting the occurrence of the event.

[0074] In this embodiment, the calculation of the overall confidence level adopts a fusion model based on the idea of ​​probabilistic complementarity. First, for each candidate icing event associated with the icing event to be verified, the difference between its confidence level and a preset value (usually 1) is calculated, i.e., 1 is subtracted from the confidence level. The result can be understood as the degree of uncertainty of the data source failing to confirm or deny the occurrence of the event. Next, these differences corresponding to all candidate icing events are multiplied to obtain a product, which represents the joint probability that multiple independent data sources are simultaneously wrong (i.e., none of them truly reflect the icing event). Finally, the preset value 1 is subtracted from the product, and the result is the overall confidence level of the icing event to be verified. In this way, the advantages of multi-source cross-validation can be fully utilized. Even if the confidence level of a single data source is not high, as long as multiple sources consistently point to the same event, the overall confidence level can still be improved. This not only enhances the sensitivity to the identification of real icing events, but also suppresses false positive results caused by single noise or false alarms.

[0075] For example, if a certain icing event to be verified is associated with three candidate icing events with confidence levels of 0.9, 0.8 and 0.7 respectively, then we calculate 1 minus 0.9, 1 minus 0.8 and 1 minus 0.7 respectively to get 0.1, 0.2 and 0.3; multiply these three values ​​together to get 0.006; then subtract 0.006 from 1 to get the final overall confidence level of 0.994.

[0076] In this embodiment, the overall confidence level of each icing event to be verified can be calculated through cross-validation of multi-source data, denoted as CL_event. This overall confidence level is jointly determined by the independent confidence levels from different data sources: CL_ys is the confidence level of candidate icing events identified based on meteorological station observation data, CL_grid is the confidence level of candidate icing events identified based on reanalysis grid data, and CL_history is the confidence level of candidate icing events identified based on historical disaster records. The overall confidence level CL_event is calculated as follows: CL_event = 1 - (1 - CL_ys) × (1 - CL_grid) × (1 - CL_history), where 1 is the target value mentioned above. This formula reflects the idea of ​​probabilistic complementarity, that is, the probability of multiple independent pieces of evidence being wrong at the same time is extremely low. Therefore, when multiple sources support the same event, its overall credibility is significantly improved.

[0077] In some embodiments, if the overall confidence score CL_event of an icing process to be verified is greater than or equal to 0.85, it is determined to be a typical icing process, and ultimately identified as all typical icing processes in the target area within that time period. Furthermore, in the spatiotemporal matching stage, candidate events that meet the following conditions are grouped into the same icing process to be verified: the first, second, and third icing candidate events are temporally continuous or adjacent, and their geographically overlapping coverage areas are uniformly classified into the same icing process to be verified. This rule enables the effective aggregation of multi-source information in both temporal and spatial dimensions, improving the completeness and consistency of event identification.

[0078] Based on the above-described embodiment, taking Region A as an example, where the target area for constructing the power grid icing process is identified, Region A's latitude and longitude range is set from 119° to 136° east longitude and from 38.5° to 54.5° north latitude, spanning from 1956 to 2024. Based on this spatiotemporal range, the following steps are performed: (a) Data collection and analysis: Collect ERA5 reanalysis grid data for the region from 1956 to 2024, including hourly weather phenomena and precipitation; obtain daily precipitation and weather phenomenon observation data from national meteorological stations; and integrate historical disaster records from emergency management departments, power grid companies, academic literature and local chronicles.

[0079] (II) Identification of icing processes based on rime ice phenomena at meteorological stations: Based on the defined latitude and longitude range, a total of 255 national meteorological stations were identified in the region. To exclude localized icing events with small impact areas, the lower limit of the impact of a typical icing process was set: (1) A meteorological station with rime ice phenomena and a daily precipitation of not less than 2 mm is defined as a valid rime ice meteorological station; (2) During a typical icing process, the minimum ratio of the number of all valid rime ice meteorological stations to the total number of meteorological stations in the region, I_station, is defined as 4%.

[0080] The upper limit L_station for the duration of a typical icing process is further set to 4 days, because rime ice usually lasts for a short period of time, to avoid misjudging sporadic rime ice that appears over several consecutive days as a complete icing process.

[0081] Based on meteorological station rime monitoring data, a possible icing process is determined if any of the following conditions are met: (1) The proportion of meteorological stations with rime on a single day to the total number of meteorological stations in the area exceeds I_station; (2) Rime occurs for two consecutive days, and the cumulative proportion of stations with rime exceeds 1.5 times I_station (where a station with rime for two consecutive days is counted as 2 stations); (3) Rime occurs for three consecutive days. The cumulative number of stations with rime ice exceeding twice the number of stations with rime ice (a station with rime ice for three consecutive days is counted as 3 stations, and if it occurs for two out of three days, it is counted as 2 stations); (4) and so on, if rime ice occurs for L consecutive days, the cumulative number of stations with rime ice exceeding the number of stations with rime ice must satisfy: I_event_ys≥0.5×L×I_station+0.5×I_station, where the number of stations is calculated cumulatively based on the number of consecutive days. Wherein, I_event_ys=(A_station_ys / A_station)×100%, A_station_ys is the cumulative number of effective rime ice meteorological stations during the process, and A_station is the total number of meteorological stations in the area.

[0082] Finally, all processes that met the above conditions were identified as icing processes in the region from 1956 to 2024 based on rime ice monitoring, with a total of 44 rounds identified. For each round of icing process, the maximum value of the percentage of stations with accumulated rime ice during the process, I_event_ysmax, was calculated, and its confidence level CL_ys was determined accordingly: (1) For a one-day process, I_event_ysmax equals I_event_ys; (2) For a multi-day process, the percentage of meteorological stations with rime ice on each day and the cumulative percentage of the process were calculated, and the maximum value was taken as I_event_ysmax; (3) The confidence level CL_ys was determined based on the ratio of I_event_ysmax to I_station.

[0083] (III) Identifying Icing Processes Based on Reanalysis Grid Data: Based on the defined latitude and longitude range, the number of reanalysis data grids in the area is determined as A_grid. To exclude local events, a lower limit is set for the impact range of a typical icing process: (1) Statistically count the specific hours and precipitation of freezing rain in each grid from 20:00 yesterday to 20:00 today, and sum up the precipitation of all hours with freezing rain to obtain the daily freezing rain amount of the grid; (2) Define the grid with a daily freezing rain amount of not less than 1 mm as an effective freezing rain grid. Define the minimum ratio of the number of all effective freezing rain grids to the total number of grids in the area during a typical icing process as I_grid as 4%. Set the upper limit of the duration of a typical icing process as L_grid as 4 days to exclude sporadic freezing rain phenomena over multiple consecutive days.

[0084] Based on the reanalysis data, if any of the following conditions are met, it is determined to be a possible icing process: (1) The proportion of effective freezing rain grids on a single day exceeds I_grid; (2) Effective freezing rain grids appear for two consecutive days, and the cumulative proportion of effective freezing rain grids exceeds 1.5 times I_grid (a grid that has effective freezing rain grids for two consecutive days is counted as 2); (3) Effective freezing rain grids appear for three consecutive days, and the cumulative proportion of effective freezing rain grids exceeds 2 times I_grid (a grid that has effective freezing rain grids for three consecutive days is counted as 3, and two out of the three days have effective freezing rain grids is counted as 2); (4) And so on, if effective freezing rain grids appear for L consecutive days, the cumulative proportion of effective freezing rain grids I_event_grid must satisfy: I_event_grid≥0.5×L×I_grid+0.5×I_grid, where the number of grids is calculated cumulatively based on the number of consecutive days. Where I_event_grid=(A_grid_ys / A_grid)×100%, where A_grid_ys is the number of effective freezing rain grids accumulated during the process, and A_grid is the total number of grids in the area.

[0085] Finally, all processes that met the above conditions were identified as icing processes in the region from 1956 to 2024 based on reanalysis freezing rain data, totaling 52. For each process, the maximum value of the cumulative effective freezing rain grid ratio, I_event_gridmax, was calculated, and its confidence level, CL_grid, was determined accordingly: (1) For a one-day process, I_event_gridmax equals I_event_grid; (2) For a multi-day process, the daily effective freezing rain grid ratio and the cumulative process ratio were calculated respectively, and the maximum value was taken as I_event_gridmax; (3) The confidence level, CL_grid, was determined based on the ratio between I_event_gridmax and I_grid.

[0086] (iv) Identifying the icing process in this area based on disaster record data: A lower limit for the impact of a typical icing process is set: at least four county-level administrative regions within the area must be affected. Based on disaster reports issued by emergency management departments, historical ice-related power outage records from power grid companies, academic literature, local chronicles, and news reports, information such as the occurrence time, affected area, and degree of icing events is extracted to determine the icing processes in this area between 1956 and 2024. The confidence level is determined based on the information sources in the disaster records.

[0087] (V) Identification of Typical Icing Processes in this Area Based on Cross-Validation of Multi-Source Data: The comprehensive confidence level CL_event was calculated through cross-validation of multi-source data. When the comprehensive confidence level CL_event is greater than or equal to 0.85, it is determined to be a typical icing process. Ultimately, it was determined that 28 typical icing processes occurred in region A between 1956 and 2024. This result integrates multi-source evidence from meteorological station observations, reanalysis data, and disaster records, possessing high reliability and representativeness, and can be used for subsequent power grid icing risk assessment and climate pattern research.

[0088] Figure 2 This is a schematic diagram of the structure of an icing event determination device provided in an embodiment of this disclosure, as shown below. Figure 2 As shown, the device includes: The acquisition module 201 is used to acquire various historical meteorological data of the target area within a historical time period. The various historical meteorological data are obtained based on different acquisition methods. The determination module 202 is used to determine, for each type of historical meteorological data, candidate icing events with a time range and / or spatial range and the confidence level of the candidate icing events based on the historical meteorological data. The candidate icing events are used to indicate that icing has occurred on power facilities within the time range and / or spatial range, and the confidence level is used to indicate the degree of credibility of the candidate icing events. The matching module 203 is used to perform spatiotemporal matching on candidate icing events from different historical meteorological data to obtain icing events to be verified. The icing events to be verified correspond to candidate icing events in at least two kinds of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are related in terms of time range and / or spatial range. The determination module 202 is also used to fuse the confidence levels of candidate icing events corresponding to the icing event to be verified, and obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

[0089] In some embodiments, historical meteorological data indicates the meteorological conditions observed by multiple meteorological observation units within a target area during a historical time period; the determining module 202 is used to determine, based on historical meteorological data, target meteorological observation units that observed the target weather phenomenon during the historical time period; based on the time when the target meteorological observation units observed the target weather phenomenon, determine at least one meteorological event with spatial coverage and duration, the meteorological event indicating the process of the target weather phenomenon occurring continuously in time, the spatial coverage being the ratio of the sum of the observation frequencies of the target weather phenomenon observed by each target meteorological observation unit corresponding to the meteorological event to the total number of meteorological observation units in the target area; meteorological events with a spatial coverage not less than a first preset range and a duration not greater than a preset duration are determined as candidate icing events, the first preset range being determined based on the duration and a preset daily spatial coverage ratio; the confidence level of the candidate icing event is determined based on the spatial coverage and the daily spatial coverage ratio.

[0090] In some embodiments, the determining module 202 is used to query a preset correspondence based on the ratio of spatial coverage range to daily spatial coverage ratio, and obtain the confidence level corresponding to the numerical interval to which the ratio belongs; wherein, the preset correspondence includes multiple non-overlapping numerical intervals and the confidence level corresponding to each numerical interval.

[0091] In some embodiments, historical meteorological data are meteorological station observation data or grid meteorological data, and the meteorological observation unit is a meteorological station or a grid area.

[0092] In some embodiments, historical meteorological data are historical disaster records, which are used to record the location and impact range of abnormal power system events caused by icing; the determination module 202 is used to determine candidate areas and corresponding occurrence times of icing disasters based on historical disaster record data, wherein the candidate areas are located in the target area and the occurrence times belong to historical time periods; if the range of the candidate area is not less than a second preset range, a candidate icing event is generated based on the range of the candidate area and the occurrence time; and the confidence level of the candidate icing event is determined based on the source type of the historical disaster record corresponding to the candidate icing event.

[0093] In some embodiments, the matching module 203 is used to determine the icing event to be verified based on candidate icing events derived from different historical meteorological data that overlap or are adjacent in time and / or spatial range.

[0094] In some embodiments, the determining module 202 is used to determine the difference between the confidence level and the preset value for each candidate icing event corresponding to the icing event to be verified; determine the product of the obtained differences; and determine the difference between the preset value and the product as the comprehensive confidence level of the icing event to be verified.

[0095] It should be noted that the icing event determination device provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the icing event determination device and the icing event determination method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

[0096] This disclosure also provides a computer device including a processor and a memory, wherein the memory stores at least one computer program, which is loaded and executed by the processor to perform the operations performed in the icing event determination method of the above embodiments. Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of this disclosure. Figure 3 As shown, the computer device 300 includes one or more processors 301 and a memory 302. The processor 301 may be a central processing unit (CPU) or other processing unit with icing event determination capability and / or instruction execution capability, and may control other components in the computer device 300 to perform desired functions. The memory 302 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 301 may execute the program instructions to implement the icing event determination method of the embodiments of this disclosure described above and / or other desired functions. Various contents such as input signals, signal components, and noise components may also be stored in the computer-readable storage medium.

[0097] In one example, the computer device 300 may further include an input device 303 and an output device 304, these components being interconnected via a bus system and / or other forms of connection mechanisms (not shown). Furthermore, the input device 303 may include, for example, a keyboard, a mouse, etc. The output device 304 may output various information to the outside, including determined distance information, direction information, etc. The output device 304 may include, for example, a monitor, speakers, a printer, and a communication network and its connected remote output devices, etc.

[0098] Of course, for the sake of simplicity, Figure 3Only some of the components of the computer device 300 relevant to this disclosure are shown in this illustration, omitting components such as buses, input / output interfaces, etc. In addition, the computer device 300 may include any other suitable components depending on the specific application.

[0099] In addition to the methods and devices described above, embodiments of this disclosure may also be computer program products, including computer program instructions that, when executed by a processor, cause the processor to perform the icing event determination method provided in the embodiments of this disclosure.

[0100] Computer program products can be written in any combination of one or more programming languages ​​to perform the operations of embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0101] Furthermore, embodiments of this disclosure may also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the icing event determination method provided in embodiments of this disclosure.

[0102] Computer-readable storage media may take the form of any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0103] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0104] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of ice accretion event determination, the method comprising: The method includes: Acquire multiple historical meteorological data for the target area within a historical time period, wherein the multiple historical meteorological data are obtained based on different acquisition methods; For each type of historical meteorological data, based on the historical meteorological data, candidate icing events with a time range and / or spatial range are determined, and the confidence level of the candidate icing events is determined. The candidate icing events are used to indicate that icing has occurred on power facilities within the time range and / or spatial range, and the confidence level is used to indicate the degree of credibility of the candidate icing events. Spatiotemporal matching is performed on candidate icing events derived from different historical meteorological data to obtain icing events to be verified. Each icing event to be verified corresponds to a candidate icing event in at least two types of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are associated in the time range and / or the spatial range. The confidence levels of the candidate icing events corresponding to the icing event to be verified are fused to obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

2. The method of claim 1, wherein, The historical meteorological data indicates the meteorological conditions observed by multiple meteorological observation units within the target area during the historical time period; the determination of candidate icing events with temporal and / or spatial ranges and the confidence levels of the candidate icing events based on the historical meteorological data includes: Based on the historical meteorological data, target meteorological observation units in which the target weather phenomenon was observed during the historical time period were identified. Based on the time when the target weather phenomenon is observed by the target meteorological observation unit, at least one meteorological event with spatial coverage and duration is determined. The meteorological event indicates the process of the target weather phenomenon occurring continuously in time. The spatial coverage is the ratio of the sum of the observation frequencies of the target weather phenomenon observed by each target meteorological observation unit corresponding to the meteorological event to the total number of meteorological observation units in the target area. Meteorological events whose spatial coverage is not less than a first preset range and whose duration is not greater than a preset duration are identified as candidate icing events. The first preset range is determined based on the duration and a preset daily spatial coverage ratio. The confidence level of the candidate icing event is determined based on the spatial coverage range and the daily spatial coverage ratio.

3. The method according to claim 2, characterized in that, Determining the confidence level of the candidate icing event based on the spatial coverage range and the daily spatial coverage ratio includes: Based on the ratio of the spatial coverage area to the daily spatial coverage ratio, a preset correspondence is queried to obtain the confidence level corresponding to the numerical interval to which the ratio belongs; wherein, the preset correspondence includes multiple non-overlapping numerical intervals and the confidence level corresponding to each numerical interval.

4. The method according to claim 2, characterized in that, The historical meteorological data refers to meteorological station observation data or grid meteorological data, and the meteorological observation unit refers to a meteorological station or grid area.

5. The method according to claim 1, characterized in that, The historical meteorological data are historical disaster records, which are used to record the location and impact range of abnormal power system events caused by icing; the determination of candidate icing events with temporal and / or spatial ranges and the confidence level of the candidate icing events based on the historical meteorological data includes: Based on the historical disaster record data, candidate areas for ice accumulation disasters and corresponding occurrence times are determined, wherein the candidate areas are located in the target area and the occurrence times belong to the historical time period; If the range of the candidate region is not less than the second preset range, the candidate icing event is generated based on the range of the candidate region and the occurrence time. The confidence level of the candidate icing event is determined based on the source type of the historical disaster records corresponding to the candidate icing event.

6. The method according to claim 1, characterized in that, The process of performing spatiotemporal matching on candidate icing events derived from different historical meteorological data to obtain icing events to be verified includes: The icing event to be verified is determined based on candidate icing events derived from different historical meteorological data that overlap or are adjacent in the time range and / or spatial range.

7. The method according to claim 1, characterized in that, The process of fusing the confidence levels of the candidate icing events corresponding to the icing event to be verified to obtain the comprehensive confidence level of the icing event to be verified includes: For each candidate icing event corresponding to the icing event to be verified, the difference between the confidence level and the preset value is determined respectively; Determine the product of the differences obtained; The difference between the preset value and the product is determined as the overall confidence level of the icing event to be verified.

8. A device for determining icing events, characterized in that, The device includes: The acquisition module is used to acquire various historical meteorological data of the target area within a historical time period, wherein the various historical meteorological data are obtained based on different acquisition methods; A determination module is configured to, for each type of historical meteorological data, determine, based on the historical meteorological data, candidate icing events having a time range and / or spatial range and the confidence level of the candidate icing events, wherein the candidate icing events are used to indicate that icing has occurred on power facilities within the time range and / or spatial range, and the confidence level is used to indicate the degree of credibility of the candidate icing events; The matching module is used to perform spatiotemporal matching on candidate icing events from different historical meteorological data to obtain icing events to be verified. The icing events to be verified correspond to candidate icing events in at least two types of historical meteorological data, and the candidate icing events corresponding to the icing events to be verified are associated in the time range and / or the spatial range. The determining module is further configured to fuse the confidence levels of the candidate icing events corresponding to the icing event to be verified to obtain the comprehensive confidence level of the icing event to be verified; if the comprehensive confidence level is greater than the preset confidence level, the icing event to be verified is determined as a historical icing event that occurred in the target area within the historical time period.

9. A computer device, characterized in that, include: A memory and a processor, wherein the memory stores a computer program that, when executed by the processor, implements the icing event determination method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the icing event determination method as described in any one of claims 1-7.

Images