A smart early warning method and system for signal surge arresters

Abstract

This application relates to the field of equipment operation status analysis technology, specifically disclosing an intelligent early warning method and system for signal surge arresters. This method collects temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area to generate a climate feature vector, which is then mapped to a geographic information system using location latitude and longitude information to obtain the lightning strike probability distribution. Furthermore, it combines the location latitude and longitude information, installation height, voltage fluctuation amplitude, and insulation resistance readings of the signal surge arrester to generate a risk score, further forming an early warning signal sequence, protection execution path, and final protection response chain. This application improves the accuracy of risk identification for signal surge arresters in different deployment locations.

Description

Technical Field

[0001] This application relates to the field of equipment operation status analysis technology, and more specifically, to an intelligent early warning method and system for signal surge arresters. Background Technology

[0002] In communication systems, power systems, and other applications involving signal transmission protection, existing early warning technologies for signal surge arresters primarily rely on equipment-side status information such as voltage anomalies, insulation changes, or localized operational anomalies for alarm judgment. While these technologies can reflect the abnormal states that have occurred or are about to occur for signal surge arresters to some extent, they lack effective characterization of the external environmental changes that precede the formation of lightning strike risks. It is difficult to transform environmental information closely related to lightning activity, such as temperature, humidity, wind speed, and atmospheric electric field strength, into risk assessment data specific to the deployment location. Furthermore, existing technologies typically lack a processing mechanism to combine environmental information with location latitude and longitude information and map it into a geographic information system for spatial representation. This makes it difficult to generate distribution results that reflect the differences in lightning strike risks across different regions. It is also difficult to further combine the signal surge arrester's location latitude and longitude information, installation height, voltage fluctuation amplitude, insulation resistance reading, operating time, historical fault records, and grounding status—deployment attributes and operating states—to generate targeted risk scores, high-risk period results, and protective action arrangements for specific equipment. Consequently, early warning results generally suffer from delayed timing, inaccurate targets, insufficient regional specificity, and a disconnect from subsequent protection implementation. Summary of the Invention

[0003] To address the shortcomings of existing early warning technologies for signal surge arresters, which struggle to translate changes in the external environment into risk assessments specific to the deployment location and to further integrate equipment deployment attributes and operational status to generate early warning results that are consistent with subsequent protection implementation, this application provides the following technical solution: In the first aspect, this application discloses an intelligent early warning method for signal surge arresters, comprising: Temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area are collected, fused, and used to generate a climate feature vector. This vector is then mapped to a geographic information system based on the latitude and longitude information of the target area to obtain the lightning strike probability distribution. Based on the location latitude and longitude information, installation height, voltage fluctuation amplitude and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, the environmental exposure level and lightning strike exposure degree are determined, and a risk score is generated. When the risk score is higher than the preset risk score threshold, a warning signal sequence for high-risk periods of lightning strikes is generated by combining the operating time of the signal arrester, historical fault records and grounding resistance status, as well as the climate feature vector of the target area. Based on the early warning signal sequence, the peak risk period is extracted, and combined with the preset protection execution constraints, a dynamic protection instruction sequence and a preliminary protection requirement ranking are generated. Based on the preliminary protection requirements and preset protection execution constraints, the dynamic protection instruction sequence is optimized to obtain the protection execution path; The feedback signal data of the corresponding signal surge arrester is collected according to the protection execution path. When there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, the command parameters of the dynamic protection command sequence are adjusted and the protection execution path is updated to determine the final protection response chain.

[0004] Secondly, this application discloses an intelligent early warning system for signal surge arresters, comprising: The climate modeling module is used to collect temperature, humidity, wind speed and atmospheric electric field intensity information of the target area, perform fusion processing to generate climate feature vectors, and combine the location latitude and longitude information of the target area to map to the geographic information system to obtain the lightning strike probability distribution results. The risk assessment module is used to determine the environmental exposure level and lightning exposure degree based on the location latitude and longitude information, installation height, voltage fluctuation amplitude and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, and generate a risk score. The early warning generation module is used to generate an early warning signal sequence for high-risk periods of lightning strikes by combining the running time of the signal arrester, historical fault records and grounding resistance status, and the climate feature vector of the target area when the risk score is higher than a preset risk score threshold. The instruction orchestration module is used to extract the peak risk period based on the warning signal sequence, and generate a dynamic protection instruction sequence and a preliminary protection requirement sorting based on preset protection execution constraints. The path optimization module is used to optimize the dynamic protection instruction sequence according to the initial protection requirements and preset protection execution constraints to obtain the protection execution path; The response correction module is used to collect feedback signal data of the corresponding signal surge arrester according to the protection execution path, and when there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, adjust the instruction parameters of the dynamic protection instruction sequence and update the protection execution path to determine the final protection response chain.

[0005] Compared with related technologies, this application has the following advantages: This application generates climate feature vectors by collecting temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area. These vectors are then combined with the latitude and longitude information of the target area and mapped onto a geographic information system to form a lightning strike probability distribution. This transforms discrete environmental information into a basic result of lightning strike risk that can be expressed by spatial location, thereby improving the spatial specificity and regional differentiation ability of lightning strike risk characterization.

[0006] This application spatially matches the location latitude and longitude information and installation height of the signal surge arrester with the lightning strike probability distribution results in the geographic information system, and generates a risk score by combining voltage fluctuation amplitude and insulation resistance readings. This links the environmental lightning strike risk with the equipment operating status, thereby improving the accuracy of signal surge arrester risk identification in different deployment locations.

[0007] This application generates a sequence of warning signals for high-risk periods of lightning strikes by combining runtime, historical fault records, climate feature vectors, and grounding resistance status with risk scoring. This extends equipment risk assessment to the identification of high-risk periods in the time dimension, improving the adaptability of the warning results to the actual lightning strike risk process.

[0008] This application extracts peak risk periods based on early warning signal sequences and generates dynamic protection instruction sequences, preliminary protection requirement rankings, and protection execution paths by combining preset protection execution constraints. This enables early warning results to continue to be transmitted to subsequent protection execution processes, improving the connection and executability between early warning processing and protection action arrangements.

[0009] This application collects feedback signal data according to the protection execution path, and adjusts the parameters of the dynamic protection command sequence and updates the protection execution path when there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude. This forms a dynamic correction process oriented towards the execution result, which improves the adaptability, stability and overall operational reliability of the protection response process. Attached Figure Description

[0010] Figure 1 A flowchart illustrating an intelligent early warning method for a signal surge arrester provided in this application; Figure 2 This is a structural schematic diagram of an intelligent early warning system for a signal surge arrester provided in this application. Detailed Implementation

[0011] 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.

[0012] Example 1 Please see Figure 1 As shown, this embodiment provides an intelligent early warning method for signal surge arresters, including the following steps: Step S10: Collect temperature, humidity, wind speed and atmospheric electric field intensity information of the target area, perform fusion processing, and generate climate feature vector.

[0013] In some implementations, the detailed steps for generating climate feature vectors include: Step 101: Determine the collection range and sampling time of the target area; specifically, read the area identifier corresponding to the area to be warned, extract the boundary range of the target area based on the area identifier, and determine the boundary range as the unified collection range of temperature information, humidity information, wind speed information, and atmospheric electric field intensity information; at the same time, determine the current sampling time according to the warning update time requirements; when it is a high-incidence weather for thunderstorms, shorten the time interval between two adjacent sampling times, and when it is not a high-incidence weather for thunderstorms, extend the time interval between two adjacent sampling times.

[0014] Step 102: Collect temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area at the current sampling time. Specifically, within a unified collection range, read the temperature, humidity, wind speed, and atmospheric electric field intensity measurements corresponding to the current sampling time. When there are multiple collection points at the same sampling time, summarize the similar measurements corresponding to the multiple collection points according to their location to obtain the original temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area.

[0015] Step 103: Perform time consistency correction on the original temperature information, original humidity information, original wind speed information, and original atmospheric electric field strength information; specifically, when the acquisition time of a certain type of information is earlier than or later than the current sampling time, select the sampling value with the smallest time difference from the current sampling time as the valid sampling value corresponding to that type of information at the current sampling time; thereby generating time-aligned temperature information, time-aligned humidity information, time-aligned wind speed information, and time-aligned atmospheric electric field strength information.

[0016] Step 104: Perform dimensional correction on the time-aligned temperature information, time-aligned humidity information, time-aligned wind speed information, and time-aligned atmospheric electric field intensity information. Specifically, read the historical maximum and minimum values ​​of the corresponding information of the same type in the target area within a preset historical period. Subtract the historical minimum temperature value from the time-aligned temperature information, and then divide by the difference between the historical maximum and historical minimum temperature values ​​to generate the corrected temperature information. Generate the corrected humidity information, corrected wind speed information, and corrected atmospheric electric field intensity information in the same way. When the historical maximum and historical minimum values ​​of a certain type of information are the same, process the information of that type directly according to the preset fixed correction value to avoid invalid division.

[0017] Step 105: Generate temperature weight, humidity weight, wind speed weight, and atmospheric electric field strength weight based on historical lightning strike event records. Specifically, read historical lightning strike event records of the target area within a preset historical period, extract the temperature, humidity, wind speed, and atmospheric electric field strength information corresponding to each historical lightning strike event, and calculate the correlation between the four types of information and the number of lightning strike events. Then, divide the four correlation degrees by the sum of the four correlation degrees to generate temperature weight, humidity weight, wind speed weight, and atmospheric electric field strength weight, ensuring that the sum of the four weights is 1.

[0018] Step 106: Generate a climate feature vector based on the normalized temperature information, normalized humidity information, normalized wind speed information, normalized atmospheric electric field intensity information, and corresponding weights. Specifically, multiply the normalized temperature information by the temperature weight, multiply the normalized humidity information by the humidity weight, multiply the normalized wind speed information by the wind speed weight, and multiply the normalized atmospheric electric field intensity information by the atmospheric electric field intensity weight. Then, sum the four product results to generate the climate feature vector corresponding to the target area at the current sampling time. The climate feature vector serves as the input for location mapping and lightning strike probability calculation.

[0019] As demonstrated by the above steps, this process does not directly use a single environmental quantity as the basis for lightning strike risk assessment. Instead, it first performs unified collection, time consistency correction, dimension reduction, and weight determination based on historical lightning event records on temperature, humidity, wind speed, and atmospheric electric field intensity information, before generating a climate feature vector. This processing method compresses multiple types of environmental information with different dimensions, potentially inconsistent sampling times, and varying degrees of contribution to lightning strike risk into a uniformly accessible climate characterization result. This provides a consistent input basis for subsequent spatial location unit mapping and lightning strike probability calculation. Consequently, it reduces judgment bias caused by fluctuations in single environmental quantities or inconsistencies in the temporal sequence of multi-source environmental information, improving the consistency and usability of environmental risk characterization.

[0020] In some embodiments, the generation process of the climate feature vector provided in this application is described in detail below: Assuming that at a given sampling moment, the normalized temperature is 0.68, the normalized humidity is 0.83, the normalized wind speed is 0.41, and the normalized atmospheric electric field strength is 0.92; and the corresponding weights are 0.22 for temperature, 0.26 for humidity, 0.14 for wind speed, and 0.38 for atmospheric electric field strength; then multiplying the normalized temperature by its weight yields 0.1496, multiplying the normalized humidity by its weight yields 0.2158, multiplying the normalized wind speed by its weight yields 0.0574, and multiplying the normalized atmospheric electric field strength by its weight yields 0.3496. Summing these products yields a climate feature vector of 0.7724 corresponding to the current sampling moment. This generated climate feature vector serves as the input for subsequent spatial location unit mapping and lightning strike probability calculation.

[0021] Step S20 involves mapping the location latitude and longitude information of the target area into a geographic information system to obtain the lightning strike probability distribution result; in some embodiments, the implementation steps for obtaining the lightning strike probability distribution result include: Step 201: Read the climate feature vector and determine the climate feature vector as the climate quantity to be mapped for the current target area.

[0022] Step 202: Extract the location latitude and longitude information of the target area and divide it into spatial location units; specifically, extract the location latitude and longitude information corresponding to each location within the target area according to the boundary range of the target area, and divide the target area into multiple spatial location units according to the preset spatial division precision; the preset spatial division precision is determined based on the area of ​​the target area, the distribution density of signal arresters, and the spatial dispersion of historical lightning strike events; the spatial location unit is the smallest analysis area with a fixed longitude range and a fixed latitude range in the geographic information system.

[0023] Step 203: Map the climate feature vector to the spatial location unit; specifically, determine the center latitude and longitude of each spatial location unit as the corresponding location, load the climate feature vector into each spatial location unit, and generate the current climate feature value corresponding to each spatial location unit.

[0024] Step 204: Extract historical climate feature values ​​corresponding to historical lightning strike events; specifically, read historical lightning strike event records of the target area within a preset historical period, extract the occurrence location, occurrence time, and temperature, humidity, wind speed and atmospheric electric field intensity information corresponding to each historical lightning strike event, and generate corresponding historical climate feature values ​​according to the same rounding method and the same weighting as in the aforementioned steps.

[0025] Step 205: Screen similar lightning strike events based on the difference between the current climate characteristic value and the historical climate characteristic value; specifically, calculate the difference between the current climate characteristic value and the historical climate characteristic value corresponding to each spatial location unit; when the difference is less than the preset similarity condition, the corresponding historical lightning strike event is identified as a similar lightning strike event; the preset similarity condition is determined based on the upper limit of the statistical distribution of the difference between the climate characteristic values ​​corresponding to the historical lightning strike events.

[0026] Step 206: Generate a lightning strike probability value based on the number of similar lightning strike events; specifically, for each spatial location unit, count the number of similar lightning strike events selected, and divide the number of similar lightning strike events by the total number of observations corresponding to that spatial location unit within the same historical period to generate the lightning strike probability value corresponding to the current spatial location unit.

[0027] Step 207: Generate lightning strike probability distribution results based on lightning strike probability values. Specifically, compare the lightning strike probability value corresponding to each spatial location unit with a preset lightning strike probability grading threshold to generate the corresponding probability grading result. Then, load the lightning strike probability value and probability grading result into the geographic information system according to the spatial location relationship to generate the lightning strike probability distribution result corresponding to the target area. The preset lightning strike probability grading threshold is determined based on the historical lightning strike probability statistical distribution of the target area. The lightning strike probability distribution result is used as the basic result of spatial risk in subsequent steps.

[0028] The above steps combine the generated climate feature vector with the latitude and longitude information of the target area. Through spatial unit division, historical lightning strike event comparison, and lightning strike probability value grading, a lightning strike probability distribution result is generated. This processing method further transforms the climate characterization results obtained in the previous step into regionalized lightning strike risk results corresponding to specific spatial locations. This allows subsequent risk assessments to move beyond a rough judgment at the overall level of the target area and instead establish differentiated risk bases for different spatial units. Consequently, it improves the regional specificity and regional differentiation capability of lightning strike risk identification and provides a direct basis for determining environmental exposure levels based on the deployment location of signal surge arresters.

[0029] Furthermore, the generation process of the lightning strike probability value and lightning strike probability distribution results provided in the embodiments of this application will be described in detail below: Assuming that a certain spatial location unit has a total of 300 observations within a preset historical period, and the number of similar lightning strike events selected in step 205 is 54, then dividing 54 by 300 yields a lightning strike probability value of 0.18 for that spatial location unit. Furthermore, if 0.10 to 0.20 corresponds to a medium probability level in the preset lightning strike probability grading threshold, then the probability grading result for that spatial location unit is a medium probability level. The lightning strike probability value and probability grading result for that spatial location unit are then loaded into the geographic information system according to their spatial location relationship to obtain the lightning strike probability distribution result for that spatial location unit, which serves as the spatial risk basis for subsequent determination of environmental exposure levels.

[0030] Step S30: Based on the location latitude and longitude information, installation height, voltage fluctuation amplitude, and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, determine the environmental exposure level and lightning strike exposure degree, and generate a risk score.

[0031] In some implementations, the steps for generating a risk score include: Step 301: Read the lightning strike probability distribution results; determine the lightning strike probability distribution results as the basic results of spatial risk, and extract the lightning strike probability value and probability classification results corresponding to each spatial location unit.

[0032] Step 302: Obtain the location latitude and longitude information and installation height of each signal surge arrester; specifically, read the location latitude and longitude information and installation height of each signal surge arrester in the target area, and establish a one-to-one correspondence between the location latitude and longitude information and installation height and the corresponding signal surge arrester.

[0033] Step 303: Perform spatial matching between each signal surge arrester and the lightning strike probability distribution results; specifically, according to the latitude and longitude information of each signal surge arrester, determine the corresponding spatial location unit in the lightning strike probability distribution results, read the lightning strike probability value and probability classification result corresponding to the spatial location unit, and generate the spatial lightning strike probability value and spatial probability classification result corresponding to each signal surge arrester.

[0034] Step 304: Generate the environmental exposure level based on the spatial lightning strike probability value and the installation height. Specifically, first, determine the spatial exposure score based on the spatial probability classification results; high probability classification corresponds to a high spatial exposure score, medium probability classification corresponds to a medium spatial exposure score, and low probability classification corresponds to a low spatial exposure score. Then, determine the height exposure score based on the height range in which the installation height is located; the higher the installation height, the greater the height exposure score. Then, add the spatial exposure score and the height exposure score to obtain the total environmental exposure score. Finally, compare the total environmental exposure score with the preset exposure level range to generate the environmental exposure level.

[0035] Step 305: Obtain the voltage fluctuation amplitude and insulation resistance reading of each signal surge arrester; specifically, collect the real-time voltage data of the line where each signal surge arrester is located within a preset statistical period, and generate the voltage fluctuation amplitude of the corresponding signal surge arrester based on the difference between the maximum and minimum voltage values ​​within the statistical period; at the same time, read the insulation resistance detection value of the corresponding signal surge arrester and generate the corresponding insulation resistance reading.

[0036] Step 306: Generate the lightning exposure level based on the environmental exposure level, voltage fluctuation amplitude, and insulation resistance reading. Specifically, first, convert the environmental exposure level into an environmental exposure value; then, divide the voltage fluctuation amplitude by the corresponding line's allowable voltage fluctuation upper limit to generate a voltage disturbance value; then, divide the insulation resistance safety lower limit by the insulation resistance reading to generate an insulation degradation value. Wherein, when the insulation resistance reading is higher than the insulation resistance safety lower limit, the insulation degradation value is no higher than 1; when the insulation resistance reading is lower than the insulation resistance safety lower limit, the insulation degradation value is higher than 1. Then, perform a weighted summation of the environmental exposure value, voltage disturbance value, and insulation degradation value according to a preset combination weight to generate the lightning exposure level.

[0037] Step 307: Generate a risk score based on the environmental exposure level and the degree of lightning exposure; specifically, use the environmental exposure value corresponding to the environmental exposure level as the corrected score and the degree of lightning exposure as the main score. First, the lightning exposure level is multiplied by the main scoring coefficient, then the environmental exposure value is multiplied by the corrected scoring coefficient, and finally the two are summed to generate a risk score. Here, the two refer to: the result of multiplying the lightning exposure level by the main scoring coefficient, plus the result of multiplying the environmental exposure value by the corrected scoring coefficient; the main scoring coefficient and the corrected scoring coefficient are determined based on the degree of fit between the exposure level and the actual damage result in historical lightning damage events; the risk score serves as the input for screening target signal surge arresters in subsequent steps.

[0038] Through the above steps, based on the generated lightning strike probability distribution, the location latitude and longitude information, installation height, voltage fluctuation amplitude, and insulation resistance reading of the signal surge arrester are further incorporated. First, the environmental exposure level is determined, then the lightning strike exposure degree is generated, and finally, a risk score is formed. This processing method can correlate the regional environmental lightning strike risk results with the operational status results of specific equipment, allowing the risk assessment object to converge from spatial location units to specific signal surge arresters. That is, signal surge arresters with different installation heights, voltage disturbance levels, and insulation states within the same area can obtain differentiated risk score results. This improves the targeting and accuracy of risk identification for specific equipment and provides stable input for subsequent screening of target signal surge arresters and generating high-risk period warning signal sequences. In some embodiments, the risk score generation process provided by the embodiments of this application is described in detail below: Assuming the environmental exposure level of a certain signal surge arrester is converted to an environmental exposure value of 0.64, the corresponding voltage fluctuation amplitude within a preset statistical period is 8.4 volts, and the upper limit of the allowable voltage fluctuation of the corresponding line is 12 volts, then the voltage disturbance value is 8.4 divided by 12, resulting in 0.70; the lower limit of insulation resistance safety is 5 megohms, and the current insulation resistance reading is 4 megohms, then the insulation degradation value is 5 divided by 4, resulting in 1.25; furthermore, if the combined weights of the environmental exposure value, voltage disturbance value, and insulation degradation value are 0.35, 0.30, and 0.35 respectively, then the lightning exposure degree is 0.64 multiplied by 0.35, 0.70 multiplied by 0.30, and 1.25 multiplied by 0.35, summed to obtain 0.8715. If the main scoring coefficient is 0.80 and the corrected scoring coefficient is 0.20, then the risk score is 0.8715 multiplied by 0.80 and 0.64 multiplied by 0.20, and then summed to get 0.8252; the generated risk score is used for subsequent target signal surge arrester screening.

[0039] Step S40: When the risk score is higher than the preset risk score threshold, a warning signal sequence for high-risk lightning strike periods is generated by combining the operating time of the signal surge arrester, historical fault records and grounding resistance status, as well as the climate feature vector of the target area. Step 401: Read the risk score and iterate through the risk scores corresponding to each signal surge arrester.

[0040] Step 402: Screen target signal surge arresters with risk scores higher than the preset risk score threshold; specifically, compare the risk score corresponding to each signal surge arrester with the preset risk score threshold, and screen target signal surge arresters with risk scores higher than the preset risk score threshold; the preset risk score threshold is determined based on the risk score distribution range before the occurrence of historical lightning damage events.

[0041] Step 403: Obtain the running time and historical fault records of the target signal surge arrester; specifically, read the cumulative running time of the target signal surge arrester since it was put into operation, and read the corresponding historical fault records of the target signal surge arrester; the historical fault records include the fault occurrence time, fault type and fault number.

[0042] Step 404: Generate operational risk results based on runtime and historical fault records. Specifically, first, divide the runtime by the average runtime of the same type of signal surge arrester to generate a runtime deviation value; divide the number of faults in the historical fault records by the average number of faults of the same type of signal surge arrester in the same operating cycle to generate a fault frequency deviation value; divide the baseline fault interval of the same type of signal surge arrester by the average fault interval between two adjacent faults of the target signal surge arrester to generate a fault density value; then, calculate the correlation between the runtime deviation value, fault frequency deviation value, and fault density value and the historical lightning damage results, and determine the runtime combination weight, fault frequency combination weight, and fault density combination weight based on the proportion of each correlation degree; then, multiply the runtime deviation value, fault frequency deviation value, and fault density value by the corresponding combination weight, and perform summation processing on the product results to generate operational risk results.

[0043] Step 405: Read the climate feature vector and generate the grounding resistance status of the target signal surge arrester; specifically, read the climate feature vector and simultaneously read the current grounding resistance detection value of the target signal surge arrester; compare the current grounding resistance detection value with the preset safe grounding resistance range; if the current grounding resistance detection value falls within the preset safe grounding resistance range, a normal grounding state is generated; if the current grounding resistance detection value exceeds the preset safe grounding resistance range, an abnormal grounding state is generated; the preset safe grounding resistance range is determined based on the grounding safety requirements of the line where the corresponding signal surge arrester is located.

[0044] Step 406: Generate time-period risk values ​​for each time window based on operational risk results, climate feature vectors, and grounding resistance status. Specifically, the subsequent warning period is first divided into multiple consecutive time windows. For each time window, the climate feature vector corresponding to the current time window is read. When the current time window exceeds the sampling time of the climate feature vector, the acquisition and fusion processing is re-executed at the update time corresponding to the time window to generate the updated climate feature vector corresponding to the time window. Then, the operational risk results are converted into equipment-side risk values, the climate feature vector corresponding to the current time window is converted into weather-side risk values, and the grounding resistance status is converted into grounding correction values. Finally, the equipment-side risk values, weather-side risk values, and grounding correction values ​​are weighted and summed according to preset time-period combination weights to generate the time-period risk value corresponding to the time window.

[0045] In some implementations, to illustrate the method of using operational risk results, climate feature vectors, and grounding resistance status in the calculation of time-period risk values, the conversion rules for converting operational risk results to equipment-side risk values, climate feature vectors to weather-side risk values, and grounding resistance status to grounding correction values ​​include: first, reading the maximum and minimum operational risk results corresponding to the same type of signal surge arrester within a preset historical period, subtracting the minimum operational risk result from the current operational risk result, and then dividing by the difference between the maximum and minimum operational risk results to generate the equipment-side risk value; then, reading the maximum and minimum climate feature vectors corresponding to the target area within a preset historical period, subtracting the minimum climate feature vector from the climate feature vector corresponding to the current time window, and then dividing by the difference between the maximum and minimum climate feature vectors to generate the weather-side risk value; for grounding resistance status, first... The system reads the current grounding resistance detection value and the upper limit of the preset safe grounding resistance range. If the current grounding resistance detection value falls within the preset safe grounding resistance range, the grounding correction value is determined as the baseline correction value. If the current grounding resistance detection value exceeds the preset safe grounding resistance range, the deviation between the current grounding resistance detection value and the upper limit of the preset safe grounding resistance range is divided by the upper limit of the preset safe grounding resistance range to generate an abnormal deviation value. This abnormal deviation value is then combined with the baseline correction value to generate the grounding correction value. The larger the abnormal deviation value, the larger the grounding correction value. When the maximum and minimum values ​​of the operational risk results are the same, or the maximum and minimum values ​​of the climate feature vector are the same, the corresponding converted values ​​are directly determined as the preset baseline value. Subsequently, the equipment-side risk value, weather-side risk value, and grounding correction value are used as inputs within the same time window to generate the time period risk value corresponding to that time window.

[0046] Step 407: Generate a warning signal sequence based on the time period risk value corresponding to each time window. Specifically, compare the time period risk value corresponding to each time window with a preset time period risk threshold. When the time period risk value is higher than the preset time period risk threshold, encode the time window as a high-risk signal value. When the time period risk value is not higher than the preset time period risk threshold, encode the time window as a non-high-risk signal value. Then, arrange the encoded values ​​corresponding to each time window in chronological order to generate a warning signal sequence. The warning signal sequence serves as the input for extracting peak risk periods in subsequent steps.

[0047] The rules for setting the preset time period risk threshold include: reading the time period risk value of the corresponding time window before the occurrence of historical lightning damage events within the preset historical period, and the time period risk value corresponding to the control time window where no lightning damage events occurred; sorting the two types of time period risk values ​​according to their numerical values, and calculating the overlapping interval of the two types of time period risk values; determining the boundary value that can distinguish the time window of historical lightning damage events from the control time window as the time period risk threshold; when new historical lightning damage event records are added subsequently, recalculating the distribution of the two types of time period risk values, and updating the preset time period risk threshold.

[0048] As can be seen from the above steps, this step does not directly output alarm results based solely on the single-point score at the current moment. Instead, it continues to combine runtime, historical fault records, grounding resistance status, and climate feature vectors updated according to time windows to generate time-period risk values ​​for each time window, and further forms a warning signal sequence. Through this processing method, the static equipment risk judgment results obtained in the previous step can be extended to the risk evolution judgment in the continuous time dimension, so that the warning results not only reflect "which equipment has a higher risk", but also "in which time period the risk increases or persists". As a result, the time period identification capability of lightning strike risk identification can be improved, so that the subsequent peak risk period extraction and protection action arrangement have a clear time basis, reducing the problem of delayed warning time or unclear warning period.

[0049] Furthermore, the generation process of the warning signal sequence provided in the embodiments of this application will be described in detail below: Assuming the subsequent warning period is divided into four consecutive time windows, with corresponding risk values ​​of 0.61, 0.78, 0.86, and 0.69, and a preset risk threshold of 0.75, the first time window is encoded as a non-high-risk signal value, the second time window as a high-risk signal value, the third time window as a high-risk signal value, and the fourth time window as a non-high-risk signal value. Arranged in chronological order, the corresponding warning signal sequence is generated as 0, 1, 1, 0. The generated warning signal sequence serves as the input for extracting the peak risk period.

[0050] Step S50: Extract the peak risk period based on the early warning signal sequence, and generate a dynamic protection instruction sequence and preliminary protection requirement sorting by combining the preset protection execution constraints.

[0051] In some implementations, the preset protection execution constraints include instruction priority classification criteria and isolation action timing constraints.

[0052] The method for generating instruction priority classification is as follows: Read the execution success rate, average response time, and risk reduction magnitude of each protective action from historical execution records; read the execution success rate distribution interval, average response time distribution interval, and risk reduction magnitude distribution interval corresponding to each protective action within a preset historical period; determine the relative position of the current execution success rate within the execution success rate distribution interval as the execution success rate rounding value, and determine the relative position of the current risk reduction magnitude within the risk reduction magnitude distribution interval as the risk reduction magnitude rounding value; for the average response time, first determine the position of the current average response time within the average response time distribution interval. If the current average response time is within a shorter response time interval, a larger response time score is generated; if the current average response time is within a longer response time interval... If the response time is short, a smaller response time score is generated, and the position of the average response time in the average response time distribution interval is reversed with the preset time score interval to generate a response time score. Then, the correlation between the execution success rate, response time score, and risk reduction magnitude, respectively, and the effectiveness of the protective action is calculated. Based on the proportion of each correlation degree to the total correlation degree, success rate weight, time weight, and risk reduction magnitude weight are generated. Then, the execution success rate is multiplied by the success rate weight, the response time score is multiplied by the time weight, and the risk reduction magnitude is multiplied by the risk reduction magnitude weight. The product results are summed to generate the action priority score for each protective action. Finally, the actions are sorted from high to low according to the action priority score to generate the instruction priority classification basis.

[0053] The method for generating a response time score includes: dividing the average response time distribution interval into multiple response time sub-intervals, and dividing the preset time score interval into multiple score sub-intervals with the same number of response time sub-intervals; when the current average response time falls into a shorter response time sub-interval, reading the higher score sub-interval corresponding to the opposite side of the response time sub-interval, and generating a response time score; when the current average response time falls into a longer response time sub-interval, reading the lower score sub-interval corresponding to the opposite side of the response time sub-interval, and generating a response time score; wherein, the shorter the average response time, the larger the response time score.

[0054] The method for generating isolation action timing constraints is as follows: read the completion time of the corresponding line interruption action, isolation action and recovery action, and read the conflict relationship between each action; when the latter action can only be executed after the former action is completed, the minimum interval between the completion time of the former action and the trigger time of the latter action is determined as the sum of the completion time of the two actions, and the isolation action timing constraints are generated.

[0055] In some implementations, the steps of generating a dynamic protection instruction sequence and a preliminary protection requirement ordering include: Step 501: Read the warning signal sequence and extract the risk signal value and time position corresponding to each time window in the warning signal sequence.

[0056] Step 502: Extract peak risk periods from the warning signal sequence. Specifically, the warning signal sequence is segmented according to a preset time window to extract the maximum risk signal value and the duration of continuous high values ​​corresponding to each time window. When the maximum risk signal value is higher than the preset peak condition and the duration of continuous high values ​​is not lower than the preset duration condition, the corresponding time window is determined as the peak risk period. The preset peak condition and preset duration condition are determined based on the distribution of warning signal values ​​during historical high-risk periods of lightning strikes.

[0057] Step 503: Generate protection requirement categories based on peak risk periods. Specifically, for each peak risk period, read the corresponding maximum risk signal value, duration, and line operation status. When the maximum risk signal value is higher than the first requirement condition, generate an isolation protection requirement category. When the maximum risk signal value is higher than the second requirement condition, generate a protection enhancement requirement category. When the line has a backup power supply source and the duration is higher than the preset power supply switching condition, generate a backup power supply requirement category. Then, summarize all requirement categories corresponding to the same peak risk period to generate the corresponding protection requirement category result.

[0058] Step 504: Generate a set of candidate protection actions based on the protection requirement categories; specifically, for each protection requirement category, invoke the pre-established requirement action mapping relationship; wherein, the isolation protection requirement category corresponds to the isolation preparation action, the protection enhancement requirement category corresponds to the protection status enhancement action, and the backup power supply requirement category corresponds to the backup power supply preparation action; then, summarize all the protection actions mapped under the same peak risk period to generate a set of candidate protection actions.

[0059] Step 505: Generate action sorting results based on preset protection execution constraints. Specifically, first, read the action priority score corresponding to each protection action in the candidate protection action set and sort them from high to low according to the action priority score; then, correct the execution order relationship and minimum execution interval between adjacent protection actions based on the isolation action timing constraints to generate action sorting results.

[0060] Step 506: Generate a dynamic protection instruction sequence based on the action sorting result. Specifically, first read the order of each protection action in the action sorting result, and read the start time, end time, duration, and maximum risk signal value of the current peak risk period. Determine the trigger time of the first protection action in the sorting order as the start time of the peak risk period. For the second and subsequent protection actions in the sorting order, read the trigger time and duration of the previous protection action, and read the minimum execution interval between two adjacent protection actions determined in step 505. Add the trigger time of the previous protection action, the duration of the previous protection action, and the minimum execution interval to generate the trigger time of the current protection action.

[0061] The system reads the basic duration and basic execution intensity corresponding to each protective action category. Different protective action categories correspond to different ranges of basic duration and execution intensity. The maximum risk signal value is then compared with a preset risk grading range. If the maximum risk signal value falls into a higher risk grading range, a preset duration correction is added to the basic duration of the corresponding protective action category, and a preset intensity correction is added to the basic execution intensity. If the maximum risk signal value falls into a lower risk grading range, the basic duration and basic execution intensity remain unchanged. Next, it determines whether the sum of the trigger time and duration of the current protective action exceeds the end time of the peak risk period. If it does, the duration of the current protective action is corrected to the difference between the end time of the peak risk period and the trigger time of the current protective action. Finally, the trigger time, duration, and execution intensity of each protective action are matched one-to-one with the corresponding protective action in order, generating a dynamic protective instruction sequence.

[0062] In some embodiments, the generation process of the dynamic protection command sequence provided in this application is described in detail below: Assuming a peak risk period starts at 14:00 and ends at 14:12, lasting 12 minutes, with a maximum risk signal value of 0.93; after sorting in step 505, the protective actions are arranged in the following order: isolation preparation action, protection status enhancement action, and backup power supply preparation action. If the basic duration of the isolation preparation action is 3 minutes, and the minimum execution interval between the isolation preparation action and the protection status enhancement action is 1 minute, then the trigger time for the protection status enhancement action is 14:04; if the duration of the protection status enhancement action is 2 minutes, and the minimum execution interval between it and the backup power supply preparation action is 1 minute, then the trigger time for the backup power supply preparation action is 14:07. Furthermore, when the maximum risk signal value corresponds to a higher risk level range, a corresponding preset intensity correction is added to the basic execution intensity of the protection status enhancement action, and a corresponding preset duration correction is added to the basic duration of the backup power supply preparation action, ultimately generating a dynamic protection instruction sequence corresponding to the peak risk period. The generated dynamic protection instruction sequence is used for subsequent preliminary protection requirement sorting and optimization processing.

[0063] Step 507: Generate a preliminary protection requirement ranking based on the risk index and highest action priority score of the peak risk period. Specifically, read the duration and maximum risk signal value corresponding to each peak risk period; compare the maximum risk signal value with the preset risk classification interval, and read the risk amplitude coefficient corresponding to the risk classification interval; wherein, the preset risk classification interval is divided according to the statistical distribution of the maximum risk signal value in the historical warning signal sequence, and the risk amplitude coefficient is determined according to the statistical results of the historical lightning damage degree corresponding to each risk classification interval. The higher the risk classification, the larger the risk amplitude coefficient; then multiply the duration by the risk amplitude coefficient, and multiply the product by the maximum risk signal value to generate the peak risk period. The risk index corresponding to each segment is determined. Then, the correlation between the risk index and the highest action priority score in historical protection execution records and the risk reduction results after the protection action is executed is read. Based on the proportion of the two correlations to the total correlation, risk index ranking weights and action priority score ranking weights are generated. The risk index is then multiplied by the risk index ranking weights, and the highest action priority score is multiplied by the action priority score ranking weights. The product results are summed to generate protection demand priority values ​​corresponding to each peak risk period. Finally, the protection demand priority values ​​are arranged from high to low to generate a preliminary protection demand ranking. The dynamic protection instruction sequence and the preliminary protection demand ranking serve as inputs for optimization processing in subsequent steps.

[0064] As can be seen from the above steps, this step, based on the early warning signal sequence generated in step S40, further extracts the peak risk period and generates a dynamic protection instruction sequence and preliminary protection requirement ranking by combining preset protection execution constraints. This processing method can further transform the high-risk identification results in the time dimension into executable protection action arrangements, so that the early warning output no longer remains at the level of risk indication, but forms action candidates, action order, and requirement priority relationships for subsequent execution. That is, different peak risk periods can form different protection requirement priority values ​​based on their duration, maximum risk signal value, and action priority score; thereby, the connection between early warning results and protection processing can be strengthened, improving the feasibility and targeting of subsequent protection execution.

[0065] Furthermore, the generation process of the preliminary protection requirement ranking provided by the embodiments of this application will be described in detail below: Assuming there are peak risk periods A and B, where peak risk period A lasts for 10 minutes, has a maximum risk signal value of 0.92, and a risk amplitude coefficient of 1.20, then the risk index for peak risk period A is 10 multiplied by 1.20 and then by 0.92, resulting in 11.04. Peak risk period B lasts for 8 minutes, has a maximum risk signal value of 0.88, and a risk amplitude coefficient of 1.10, then the risk index for peak risk period B is 8 multiplied by 1.10 and then by 0.88, resulting in 7.744. If the highest action priority score in the dynamic protection command sequence corresponding to peak risk period A is 0.86, and the highest action priority score in the dynamic protection command sequence corresponding to peak risk period B is 0.91, and the risk index ranking weight is 0.70, and the action priority score ranking weight is 0.30, then the protection demand priority value corresponding to peak risk period A is higher than the protection demand priority value corresponding to peak risk period B. This generates a preliminary protection demand ranking where peak risk period A takes precedence over peak risk period B. The generated preliminary protection requirement ranking is used for subsequent dynamic protection instruction sequence optimization.

[0066] Step S60: Based on the initial protection requirements and preset protection execution constraints, optimize the dynamic protection instruction sequence to obtain the protection execution path.

[0067] In some implementations, the preset protection execution constraints also include backup power switching criteria, overload protection criteria, switching delay criteria, and instruction compatibility verification criteria.

[0068] The method for generating the basis for backup power supply switching is as follows: read the current voltage value of the backup power supply, the remaining capacity of the backup power supply, and the current power supply status of the main line; when the current voltage value of the backup power supply is higher than the preset minimum power supply voltage, the remaining capacity of the backup power supply is higher than the preset minimum capacity, and the power supply status of the main line is abnormal, a switching permission result is generated; otherwise, a switching prohibition result is generated.

[0069] The method for generating overload protection criteria is as follows: read the rated current and current real-time current of the corresponding line; pre-establish a current change mapping value for each dynamic protection command; add the current real-time current to the current change mapping value of the corresponding dynamic protection command to generate the expected line current value; when the expected line current value is higher than the rated current, an overload result is generated; otherwise, a non-overload result is generated.

[0070] The method for generating the switching delay basis is as follows: read the completion time of the corresponding actions of two adjacent dynamic protection commands, and determine the sum of the completion times of the two actions as the minimum waiting time between the two adjacent dynamic protection commands.

[0071] The method for generating the basis for instruction compatibility verification is as follows: read the execution object, execution action, and execution time of two adjacent dynamic protection instructions; when the execution objects are the same, the execution actions are opposite, and the execution times overlap, an incompatible result is generated; otherwise, a compatible result is generated.

[0072] In some implementations, the steps for obtaining the protection execution path include: Step 601: Read the dynamic protection instruction sequence and the preliminary protection requirement ranking; determine the dynamic protection instruction sequence with the highest preliminary protection requirement ranking as the priority optimization target.

[0073] Step 602: Read the backup power switching criteria, overload protection criteria, switching delay criteria, and instruction compatibility verification criteria; match the above criteria with the priority optimization objects one by one.

[0074] Step 603: Filter executable dynamic protection commands related to backup power supply based on backup power supply switching criteria; specifically, judge each dynamic protection command related to backup power supply in the dynamic protection command sequence; when the backup power supply switching criteria is a switch-allowed result, retain the corresponding dynamic protection command; when the backup power supply switching criteria is a switch-prohibited result, delete or postpone the corresponding dynamic protection command.

[0075] Step 604: Modify the dynamic protection instruction sequence according to the overload protection criteria. Specifically, sequentially read the execution object, action category, and action execution intensity corresponding to each retained dynamic protection instruction, and read the current real-time current value of the corresponding line and the current change value corresponding to the action category under the current action execution intensity. The current change value is determined based on the line current change corresponding to the same action category and the same action execution intensity in the historical execution record. Add the current real-time current value and the current change value to generate the expected line current value after the execution of the dynamic protection instruction. Then compare the expected line current value with the rated current of the corresponding line. When the expected line current value is higher than the rated current, the dynamic protection instruction is determined as an overload dynamic protection instruction. For overload dynamic protection instructions, first determine whether it is related to the previous or subsequent dynamic protection instruction. If the overload dynamic protection command and the adjacent dynamic protection command are not applied to the same line, the overload dynamic protection command and the adjacent dynamic protection command are split into two segments for execution. The minimum waiting time corresponding to the switching delay in step 602 is read and used as the insertion waiting time between the two segments. If the overload dynamic protection command and the adjacent dynamic protection command are not applied to the same line, the original execution order remains unchanged. If the expected line current value after the execution of the dynamic protection command is still higher than the rated current after splitting, the action execution intensity of the dynamic protection command is reduced, and the current change value corresponding to the action execution intensity is read again. The expected line current value is recalculated until the expected line current value is not higher than the rated current. After all overload dynamic protection commands have been split, ordered, or the action execution intensity has been corrected, a corrected dynamic protection command sequence is generated.

[0076] Step 605: Adjust the execution interval between adjacent dynamic protection instructions according to the switching delay basis. Specifically, after generating the corrected dynamic protection instruction sequence in step 604, sequentially read the action category, execution object, execution start time, and execution end time corresponding to two adjacent dynamic protection instructions in the corrected dynamic protection instruction sequence, and read the switching delay basis in step 602. When the time interval between the execution start time of the subsequent dynamic protection instruction and the execution end time of the preceding dynamic protection instruction is less than the minimum waiting time corresponding to the switching delay basis, adjust the execution start time of the subsequent dynamic protection instruction to the sum of the execution end time of the preceding dynamic protection instruction and the minimum waiting time, and correspondingly postpone the execution start time and execution end time of subsequent dynamic protection instructions. When the time interval between the execution start time of the subsequent dynamic protection instruction and the execution end time of the preceding dynamic protection instruction is not less than the minimum waiting time, keep the current execution interval unchanged. After all adjacent dynamic protection instructions have completed the execution interval adjustment, generate a dynamic protection instruction sequence with the execution interval written in.

[0077] Step 606: Verify the dynamic protection instruction sequence written into the execution interval according to the instruction compatibility verification criteria. Specifically, read the dynamic protection instruction sequence written into the execution interval generated in step 605, extract the execution object, action category, execution start time, and execution end time corresponding to two adjacent dynamic protection instructions in sequence, and read the instruction compatibility verification criteria in step 602. When two adjacent dynamic protection instructions have the same execution object, opposite action categories, and overlapping execution periods, the two adjacent dynamic protection instructions are determined to be incompatible dynamic protection instructions. For incompatible dynamic protection instructions, first compare the action priority scores corresponding to the two dynamic protection instructions, retain the dynamic protection instruction with the higher action priority score, and delete the dynamic protection instruction with the lower action priority score. In addition to processing, replacing, or delaying processing; among them, when there is a substitute action of the same type for a dynamic protection instruction with a lower action priority score, the dynamic protection instruction is replaced; when there is no substitute action of the same type but the overlapping execution time can be eliminated by adjusting the execution start time, the dynamic protection instruction is delayed; when there is no substitute action of the same type and the overlapping execution time cannot be eliminated by adjusting the execution start time, the dynamic protection instruction is deleted; when two adjacent dynamic protection instructions have different execution objects, or their action categories are not opposite, or their execution time periods do not overlap, the current order and execution time of the two dynamic protection instructions remain unchanged; after all adjacent dynamic protection instructions have completed compatibility verification, an optimized dynamic protection instruction sequence is generated.

[0078] Step 607: Generate a protection execution path based on the optimized dynamic protection instruction sequence; specifically, read the optimized dynamic protection instruction sequence generated in step 606, extract the execution object, execution start time and execution end time in sequence according to the execution order of each dynamic protection instruction, and expand each dynamic protection instruction in chronological order to generate a protection execution path.

[0079] As can be seen from the processing in step S60 above, this step does not directly execute the dynamic protection command sequence generated in step S50 sequentially. Instead, it further combines the backup power switching basis, overload protection basis, switching delay basis, and command compatibility verification basis to perform filtering, correction, interval adjustment, and compatibility verification on the dynamic protection command sequence, ultimately generating a protection execution path. Through this processing method, the protection action arrangement results formed in the previous step can further meet the actual execution conditions, reducing execution failures caused by backup power not being switchable, line overload, action timing conflicts, or command incompatibility. As a result, the executability, sequential rationality, and action coordination of the protection path can be improved, and a clear path basis can be provided for subsequent dynamic correction based on execution feedback.

[0080] Step S70: Collect feedback signal data of the corresponding signal surge arrester according to the protection execution path, and when there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, adjust the instruction parameters of the dynamic protection instruction sequence and update the protection execution path to determine the final protection response chain.

[0081] In some implementations, the steps for determining the final protection response chain include: Step 701: Read the protection execution path generated in step 607, extract the execution object, action category, execution start time, execution end time and action execution intensity corresponding to each dynamic protection instruction in the protection execution path in sequence, and determine the execution order of each dynamic protection instruction as the feedback collection order.

[0082] Step 702: Collect feedback signal data according to the protection execution path; specifically, according to the feedback collection sequence determined in step 701, during and after the execution of each dynamic protection command, collect feedback signal data of the corresponding signal surge arrester; the feedback signal data includes one or more of the following: action execution status information, execution completion time information, line voltage value after action, line current value after action, and surge arrester status value after action; wherein, the action execution status information is used to characterize whether the current dynamic protection command has been completed, the execution completion time information is used to characterize the actual completion time of the current dynamic protection command, the line voltage value after action is used to characterize the voltage change result of the corresponding line after the execution of the current dynamic protection command, the line current value after action is used to characterize the current change result of the corresponding line after the execution of the current dynamic protection command, and the surge arrester status value after action is used to characterize the operating status of the corresponding signal surge arrester after the execution of the current dynamic protection command.

[0083] Step 703: Generate the real-time voltage fluctuation amplitude of the corresponding signal surge arrester according to the feedback acquisition sequence; specifically, collect the real-time voltage data of the line where the corresponding signal surge arrester is located according to the same statistical period as in step 702; for each statistical period, extract the maximum voltage value and the minimum voltage value within the statistical period, and determine the difference between the maximum voltage value and the minimum voltage value as the real-time voltage fluctuation amplitude corresponding to the statistical period; then arrange the real-time voltage fluctuation amplitudes corresponding to each statistical period in chronological order to generate a real-time voltage fluctuation amplitude sequence.

[0084] Step 704: Generate deviation judgment results based on feedback signal data and real-time voltage fluctuation amplitude sequence. Specifically, following the feedback acquisition order determined in step 701, read the post-action line voltage value, post-action line current value, post-action surge arrester status value, and corresponding real-time voltage fluctuation amplitude within the same statistical period for each dynamic protection command. First, compare the post-action line voltage value with the rated line voltage within the statistical period to generate a post-action voltage deviation value. Then, combine the post-action voltage deviation value with the real-time voltage fluctuation amplitude within the corresponding statistical period to generate a post-action voltage fluctuation result. Afterward, compare the post-action voltage fluctuation result with the preset fluctuation range, and compare the post-action line current value with the corresponding rated current range of the line. Simultaneously, compare the post-action surge arrester status value. The status value is compared with the preset normal status set. If the voltage fluctuation result after the action does not fall back to the preset fluctuation range, or the line current value after the action exceeds the rated current range, or the arrester status value after the action does not belong to the preset normal status set, the corresponding dynamic protection command is determined as a deviation dynamic protection command, and a corresponding deviation judgment result is generated. If the above three comparison results all meet the corresponding allowable range, the corresponding dynamic protection command is determined as a non-deviation dynamic protection command, and a corresponding non-deviation judgment result is generated. Among them, the preset fluctuation range is determined based on the allowable voltage fluctuation range of the line where the corresponding signal arrester is located, the rated current range is determined based on the rated current and allowable deviation range of the corresponding line, and the preset normal status set is determined based on the corresponding signal arrester operating status definition table.

[0085] Step 705: Generate the command parameters to be corrected based on the deviation judgment result; specifically, for each deviation dynamic protection command for which the deviation judgment result was generated in step 704, read the current command parameters; the current command parameters include the action trigger time, action duration, and action execution intensity; then calculate the voltage difference between the real-time voltage fluctuation amplitude and the preset upper or lower limit of the fluctuation range, the current difference between the line current value after the action and the upper limit of the rated current range, and the state abnormality level corresponding to the surge arrester state value after the action; when the real-time voltage fluctuation amplitude is higher than the preset upper limit of the fluctuation range, the voltage difference is determined as high deviation. Differential voltage quantity; when the real-time voltage fluctuation amplitude is lower than the preset fluctuation range lower limit, the voltage difference is determined as a low deviation voltage quantity; then the high deviation voltage quantity or low deviation voltage quantity, current difference, and abnormal status level are converted into trigger time correction quantity, duration correction quantity, and execution intensity correction quantity, respectively; among them, the larger the voltage difference, the larger the trigger time correction quantity and execution intensity correction quantity; the larger the current difference, the larger the duration correction quantity; the higher the abnormal status level, the larger the three types of correction quantities; finally, a one-to-one correspondence is established between the trigger time correction quantity, duration correction quantity, and execution intensity correction quantity and the corresponding deviation dynamic protection command to generate the command parameters to be corrected.

[0086] As another example, the process of generating the instruction parameters to be corrected provided in the embodiments of this application will be described in detail below: Assuming the current action trigger time of a certain deviation dynamic protection command is 16:20, the current action duration is 4 minutes, and the current action execution intensity is 0.72; the real-time voltage fluctuation amplitude within the statistical period corresponding to this dynamic protection command is 13 volts, and the preset fluctuation range upper limit is 10 volts, then the corresponding high deviation voltage is 3 volts; the line current value after the action is 81 amps, and the rated current range upper limit is 75 amps, then the corresponding current difference is 6 amps; the state abnormality level corresponding to the surge arrester status value after the action is level 2; based on the pre-established conversion relationship, the trigger time correction is 1 minute, the duration correction is 2 minutes, and the execution intensity correction is 0.06; thus, the command parameters to be corrected corresponding to this deviation dynamic protection command are obtained and called in step 706.

[0087] Step 706: Generate a corrected dynamic protection command sequence based on the command parameters to be corrected. Specifically, for each deviation dynamic protection command, read its current command parameters and the command parameters to be corrected generated in step 705. When the real-time voltage fluctuation amplitude corresponding to the deviation dynamic protection command is higher than the upper limit of the preset fluctuation range, subtract the trigger time correction amount from the current action trigger time, add the duration correction amount to the current action duration, and add the execution intensity correction amount to the current action execution intensity. When the real-time voltage fluctuation amplitude corresponding to the deviation dynamic protection command is lower than the lower limit of the preset fluctuation range, add the current action trigger time... The trigger time correction is calculated by subtracting the duration correction from the current action duration and subtracting the execution intensity correction from the current action execution intensity. When the line current value after the action corresponding to the deviation dynamic protection command exceeds the upper limit of the rated current range, after completing the aforementioned voltage direction correction, the action execution intensity is then corrected downward to no higher than the maximum execution intensity corresponding to the upper limit of the allowable current of the corresponding line. For the non-deviation dynamic protection command, the current command parameters remain unchanged. Then, the deviation dynamic protection command after parameter correction and the non-deviation dynamic protection command without correction are rearranged in the original execution order to generate the corrected dynamic protection command sequence.

[0088] Furthermore, the generation process of the modified dynamic protection command sequence provided in the embodiments of this application will be described in detail below: Taking the aforementioned deviation dynamic protection command as an example, when the real-time voltage fluctuation amplitude corresponding to the dynamic protection command is higher than the upper limit of the preset fluctuation range, the current action trigger time is adjusted from 16:20 to 16:19, the current action duration is adjusted from 4 minutes to 6 minutes, and the current action execution intensity is adjusted from 0.72 to 0.78. Furthermore, when the line current value after the action is simultaneously higher than the upper limit of the rated current range, after completing the aforementioned correction, the action execution intensity is further corrected downward to no higher than the maximum execution intensity corresponding to the upper limit of the allowable current of the corresponding line. Then, the deviation dynamic protection command after parameter correction and the uncorrected deviation-free dynamic protection command are rearranged in the original execution order to generate a corrected dynamic protection command sequence, which is used for subsequent protection execution path updates.

[0089] Step 707: Update the protection execution path according to the corrected dynamic protection instruction sequence. Specifically, read the corrected dynamic protection instruction sequence and extract the execution object, action category, action trigger time, action duration, and action execution intensity corresponding to each dynamic protection instruction in sequence. Determine the execution start time of the first dynamic protection instruction as its action trigger time, and determine the sum of the execution start time and the action duration as the execution end time of the first dynamic protection instruction. For the second and subsequent dynamic protection instructions, first determine their action trigger time as the current execution start time, and then determine the sum of the current execution start time and the action duration as the current execution end time. When the current execution start time is earlier than the execution end time of the previous dynamic protection instruction, postpone the current execution start time to the execution end time of the previous dynamic protection instruction, and recalculate the current execution end time accordingly. When the current execution start time is not earlier than the execution end time of the previous dynamic protection instruction, keep the current execution start time unchanged. After all dynamic protection instructions have completed the recalculation of execution start and end times, unfold each dynamic protection instruction in the execution order to generate the updated protection execution path.

[0090] Step 708: Generate the final protection response chain based on the updated protection execution path. Specifically, read the updated protection execution path, extract the execution object, action category, execution start time, execution end time, and action execution intensity corresponding to each dynamic protection instruction in sequence, and read the feedback signal data collected in step 702 that corresponds one-to-one with each dynamic protection instruction and the deviation judgment result generated in step 704. Combine the execution object, action category, execution start time, execution end time, action execution intensity, corresponding feedback signal data, and corresponding deviation judgment result of each dynamic protection instruction into a response chain node. Then connect all response chain nodes in the order of their execution start times to generate the final protection response chain. Each response chain node in the final protection response chain includes at least the execution object, action category, execution start time, execution end time, action execution intensity, line voltage value after action, line current value after action, surge arrester status value after action, and deviation judgment result.

[0091] As can be seen from the processing in step S70 above, this step, based on the protection execution path generated in step S60, further collects feedback signal data, generates real-time voltage fluctuation amplitude, determines execution deviation, and corrects the action triggering time, action duration, and action execution intensity of the dynamic protection command sequence, then updates the protection execution path, and finally forms the final protection response chain. Through this processing method, the protection execution results formed by the aforementioned early warning, sorting, and optimization can be associated with the actual execution feedback closed loop, so that the protection processing is no longer a one-time static output, but can be corrected again according to the line status and surge arrester status after execution. As a result, the adaptability of the protection response process to changes in the field status can be improved, the stability of protection execution and overall operational reliability can be enhanced, thereby achieving the protection response reliability improvement described in the specification summary.

[0092] Example 2 See Figure 2 As shown, this embodiment provides an intelligent early warning system for signal surge arresters. Since this system uses an intelligent early warning method for signal surge arresters from Embodiment 1, it has the same effect, which will not be repeated here. The system includes: The climate modeling module is used to collect temperature, humidity, wind speed and atmospheric electric field intensity information of the target area, perform fusion processing to generate climate feature vectors, and combine the location latitude and longitude information of the target area to map to the geographic information system to obtain the lightning strike probability distribution results. The risk assessment module is used to determine the environmental exposure level and lightning exposure degree based on the location latitude and longitude information, installation height, voltage fluctuation amplitude and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, and generate a risk score. The early warning generation module is used to generate a sequence of early warning signals for high-risk periods of lightning strikes by combining runtime, historical fault records, climate feature vectors and grounding resistance status when the risk score is higher than the preset risk score threshold. The instruction orchestration module is used to extract peak risk periods based on the early warning signal sequence, and generate dynamic protection instruction sequences and preliminary protection requirement sorting by combining preset protection execution constraints. The path optimization module is used to optimize the dynamic protection instruction sequence based on the initial protection requirements and preset protection execution constraints to obtain the protection execution path; The response correction module is used to collect feedback signal data of the corresponding signal surge arrester according to the protection execution path, and when there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, adjust the command parameters of the dynamic protection command sequence and update the protection execution path to determine the final protection response chain.

[0093] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims.

Claims

1. A smart early warning method for signal surge arresters, characterized in that, include: Temperature, humidity, wind speed, and atmospheric electric field intensity information of the target area are collected, fused, and used to generate a climate feature vector. This vector is then mapped to a geographic information system based on the latitude and longitude information of the target area to obtain the lightning strike probability distribution. Based on the location latitude and longitude information, installation height, voltage fluctuation amplitude and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, the environmental exposure level and lightning strike exposure degree are determined, and a risk score is generated. When the risk score is higher than the preset risk score threshold, a warning signal sequence for high-risk periods of lightning strikes is generated by combining the operating time of the signal arrester, historical fault records and grounding resistance status, as well as the climate feature vector of the target area. Based on the early warning signal sequence, the peak risk period is extracted, and combined with the preset protection execution constraints, a dynamic protection instruction sequence and a preliminary protection requirement ranking are generated. Based on the preliminary protection requirements and preset protection execution constraints, the dynamic protection instruction sequence is optimized to obtain the protection execution path; The feedback signal data of the corresponding signal surge arrester is collected according to the protection execution path. When there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, the command parameters of the dynamic protection command sequence are adjusted and the protection execution path is updated to determine the final protection response chain.

2. The intelligent early warning method for a signal surge arrester according to claim 1, characterized in that, The methods for obtaining the lightning strike probability distribution include: Read the climate feature vector, extract the latitude and longitude information of the target area, and divide it into spatial location units; The climate feature vector is mapped to each of the spatial location units to generate the current climate feature value; Extract historical climate feature values ​​corresponding to historical lightning strike events of the signal arrester, and filter the number of similar lightning strike events based on the difference between the current climate feature value and the historical climate feature value; A lightning strike probability value is generated based on the number of similar lightning strike events, and a lightning strike probability distribution result is generated based on the lightning strike probability value.

3. The intelligent early warning method for a signal surge arrester according to claim 2, characterized in that, The method for generating the risk score includes: Based on the latitude and longitude information of each signal surge arrester, spatial matching is performed between each signal surge arrester and the spatial location unit, and an environmental exposure level is generated by combining the installation height of the signal surge arrester; Obtain the voltage fluctuation amplitude and insulation resistance reading of each of the signal surge arresters, and generate the lightning exposure level based on the environmental exposure level, the voltage fluctuation amplitude, and the insulation resistance reading; A risk score is generated based on the environmental exposure level and the degree of lightning exposure.

4. The intelligent early warning method for a signal surge arrester according to claim 1, characterized in that, The method for generating the warning signal sequence for the high-risk period of lightning strikes includes: Read the risk score and filter target signal surge arresters whose risk scores are higher than the preset risk score threshold; The runtime and historical fault records of the target signal surge arrester are obtained, and an operational risk result is generated; based on the operational risk result, the climate feature vector, and the grounding resistance status, a time period risk value corresponding to each time window is generated; Based on the time period risk value corresponding to each time window, a warning signal sequence for the high-risk period of lightning strikes is generated.

5. The intelligent early warning method for a signal surge arrester according to claim 4, characterized in that, The method for generating the dynamic protection command sequence includes: Peak risk periods are extracted based on the aforementioned warning signal sequence; Based on the peak risk period, a protection requirement category is generated, and based on the protection requirement category, a set of candidate protection actions is generated; The candidate protection action set is sorted according to the preset protection execution constraints to generate an action sorting result; The dynamic protection instruction sequence is generated based on the action sorting results and the start time, end time, duration, and maximum risk signal value of the peak risk period.

6. The intelligent early warning method for a signal surge arrester according to claim 5, characterized in that, The method for generating the preliminary protection requirements ranking includes: Read the duration and maximum risk signal value corresponding to each of the peak risk periods, and generate a risk index based on the duration, the maximum risk signal value and the corresponding risk amplitude coefficient; Read the highest action priority score in the dynamic protection instruction sequence corresponding to each peak risk period, and generate a protection requirement priority value based on the risk index and the highest action priority score; The protection requirements are sorted in descending order of priority to generate the preliminary protection requirement ranking.

7. The intelligent early warning method for a signal surge arrester according to claim 6, characterized in that, The methods for obtaining the protection execution path include: Read the dynamic protection instruction sequence and the initial protection requirements sorting; Based on the backup power switching criteria in the preset protection execution constraints, the executable backup power related dynamic protection commands are selected, and the dynamic protection command sequence is modified according to the overload protection criteria in the preset protection execution constraints. Based on the switching delay criteria in the preset protection execution constraints, the execution interval between adjacent dynamic protection instructions is adjusted, and the modified dynamic protection instruction sequence is verified based on the instruction compatibility verification criteria in the preset protection execution constraints. The protection execution path is generated based on the verified and corrected dynamic protection instruction sequence.

8. The intelligent early warning method for a signal surge arrester according to claim 7, characterized in that, The method for determining the final protection response chain includes: The feedback signal data is collected according to the protection execution path, and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester is generated. A deviation determination result is generated based on the feedback signal data and the real-time voltage fluctuation amplitude, and a correction instruction parameter is generated based on the deviation determination result; A modified dynamic protection instruction sequence is generated based on the instruction parameters to be modified, and the protection execution path is updated based on the modified dynamic protection instruction sequence. Based on the updated protection execution path, generate the final protection response chain.

9. The intelligent early warning method for a signal surge arrester according to claim 5, characterized in that, The preset protection execution constraints include isolation action timing constraints, and the method for generating action sorting results includes: Read the execution success rate, average response time and risk reduction of each protective action in the candidate protective action set, and generate the action priority score corresponding to each protective action; The protective actions are sorted according to their priority scores, and the execution order and minimum execution interval of adjacent protective actions are corrected according to the isolation action timing constraints to generate the action sorting result.

10. An intelligent early warning system for a signal surge arrester, characterized in that, include: The climate modeling module is used to collect temperature, humidity, wind speed and atmospheric electric field intensity information of the target area, perform fusion processing to generate climate feature vectors, and combine the location latitude and longitude information of the target area to map to the geographic information system to obtain the lightning strike probability distribution results. The risk assessment module is used to determine the environmental exposure level and lightning exposure degree based on the location latitude and longitude information, installation height, voltage fluctuation amplitude and insulation resistance reading of the signal surge arrester, combined with the lightning strike probability distribution results, and generate a risk score. The early warning generation module is used to generate an early warning signal sequence for high-risk periods of lightning strikes by combining the running time of the signal arrester, historical fault records and grounding resistance status, and the climate feature vector of the target area when the risk score is higher than a preset risk score threshold. The instruction orchestration module is used to extract the peak risk period based on the warning signal sequence, and generate a dynamic protection instruction sequence and a preliminary protection requirement sorting based on preset protection execution constraints. The path optimization module is used to optimize the dynamic protection instruction sequence according to the initial protection requirements and preset protection execution constraints to obtain the protection execution path; The response correction module is used to collect feedback signal data of the corresponding signal surge arrester according to the protection execution path, and when there is a deviation between the feedback signal data and the real-time voltage fluctuation amplitude of the corresponding signal surge arrester, adjust the instruction parameters of the dynamic protection instruction sequence and update the protection execution path to determine the final protection response chain.

Images