Low-voltage area group hidden danger identification and quantification method based on middle platform multi-element data

Abstract

The application discloses a low-voltage transformer area group hidden danger identification and quantification method based on a middle station multi-element data, comprising the following steps: S1, data acquisition and cleaning; S2, power failure event merging; S3, multi-household power failure event splitting; S4, hidden danger identification and classification; S5, hidden danger related data retention and suspicious user confirmation; S6, node type opening and closing hidden danger risk assessment; S7, risk value quantification of classified multi-household hidden dangers; and S8, intensified monitoring and active response. The low-voltage transformer area group hidden danger identification and quantification method based on the middle station multi-element data accurately identifies low-voltage transformer area group hidden dangers and quantifies risks through data cleaning, event merging, splitting, hidden danger classification, risk quantification and intensified monitoring and the like, realizes active repair, and improves low-voltage power supply reliability and user satisfaction.

Description

Technical Field

[0001] This invention belongs to the field of low-voltage power supply safety monitoring technology, specifically involving a method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-dimensional data from a central data center. Background Technology

[0002] The complex operating environment of low-voltage systems, frequent occurrences of group-related hazards, and large-scale power outages are the main reasons affecting the reliability of low-voltage power supply. Group-related hazards can be mainly divided into two categories: one is faults caused by switch interruptions. These are caused by overload (mostly occurring in summer) and leakage, leading to switch tripping. Even after power station personnel quickly restore power, the difficulty in identifying the cause of leakage and resolving overload issues means the hazard often persists, easily causing repeated power outages for multiple users. The other category is cold-shrinkage type interruption faults, such as oxidation forming high resistance at the joints of RTO fuses and line clamps in junction boxes. This leads to thermal expansion when current flows and contraction when current is lost, manifesting as a momentary power outage after a high-load period. These outages are short-lived and may be imperceptible to users, but over time, they can cause the complete burnout of RTO fuses, line clamps, and other equipment, resulting in widespread power outages.

[0003] The data on these two types of potential hazards are hidden in the power outage records, work order records, and various measurement data of low-voltage distribution areas for group users. There is an urgent need for algorithmic identification, quantification of hazard risks, and monitoring to ensure timely handling and prevent permanent power outages for group users. While existing technologies have patents related to the identification of oxidation-related faults and frequent power outage hazards, there is still room for improvement in areas such as multi-household power outage event segmentation, hazard feature extraction, risk quantification, and proactive repair monitoring. Therefore, an optimized method for identifying and quantifying group-related hazards is urgently needed to improve the reliability of low-voltage power supply and user satisfaction. Summary of the Invention

[0004] The main objective of this invention is to provide a method for identifying and quantifying group hazards in low-voltage distribution areas based on multi-dimensional data from a central platform. By utilizing multi-system multi-dimensional data and through steps such as data cleaning, event merging, splitting, hazard classification, risk quantification, and enhanced monitoring, this method accurately identifies group hazards in low-voltage distribution areas and quantifies the risks, enabling proactive emergency repairs and improving the reliability of low-voltage power supply and user satisfaction.

[0005] To achieve the above objectives, this invention discloses a method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data center, comprising the following steps: Step S1: Collect relevant data and aggregate historical low-voltage user power outage records by transformer area, and remove relevant power outage records to obtain cleaned power outage records; Step S2: Sort the cleaned power outage records in ascending order of time sequence, and merge the power outage events according to the start and end time deviation and the threshold. Step S3: Deduplicatize and split the multi-household power outage events. After splitting, each user appears only once in each multi-household power outage event. Step S4: For multiple power outages that occur only once for a single user, identify potential risks of repeated power outages for user groups, and classify potential events according to the scope of the power outage; Step S5: Record key measurement point data according to the type of switch hazard to identify users suspected of leakage; Step S6: Calculate the risk value of node-type breakage hazards; Step S7: Quantify the risk values ​​of the classified multi-household hazards; Step S8: Enhance monitoring and proactive response.

[0006] As a further preferred technical solution to the above technical solution, the relevant data in step S1 includes the power data platform, safety monitoring domain, real-time measurement center, meteorological center, PMS ledger, and low-voltage distribution area data in the power supply service command system. The low-voltage distribution area data includes power outage records, measurement data, branch main protection action data, and user repair work order data.

[0007] As a further preferred technical solution to the above technical solution, the relevant power outage records in step S1 include power outage records corresponding to renovation work in the power security domain, power outage records corresponding to meter replacement or business expansion work in the marketing system, power outage records corresponding to power outage events in the main meter terminal, and power outage records corresponding to planned power outage work in the marketing system or power supply service command system.

[0008] As a further preferred technical solution to the above technical solution, in step S2, the power outage start time deviation Spc and the power outage end time deviation Tpc between each power outage record and the previous record are calculated. A threshold Tyz is set. If Spc≤Tyz and Tpc≤Tyz, it is marked as the same power outage event; otherwise, it is marked as the start of a new event. Multiple power outage event sets are generated within the transformer area, resulting in power outage event SJ, where: Outage start time deviation Spc = current outage start time minus the previous outage start time; Outage end time deviation Tpc = current outage end time minus the previous outage end time.

[0009] As a further preferred technical solution to the above technical solution, in step S3, for a power outage event SJ containing multiple household power outage records, if there are identical users, it is defined as an event SJM to be split and split. After splitting, each user appears only once in each multi-household power outage event SJ. The specific implementation of the splitting process is as follows: Step S3.1: Traverse the power outage records, count the frequency of power outages for each user, and take the highest frequency as the number K of the multi-user power outage events SJ after splitting; Step S3.2: Sort the power outage records of N users with K power outage counts in ascending order by the start time of the power outage, and assign a sequence number to each record. Set user 1's power outage record as And the power outage record of user N is Each power outage record includes the start time and duration of the outage. The set of power outage records for different users corresponding to the same serial number X }, take the average of the power outage duration and power outage start time within the preset quantile range, and use them as the power outage duration and power outage start time of the power outage event SJ corresponding to the sequence number seq; Step S3.3: Merge the power outage records where the number of power outages for a user is not equal to K. Compare the power outage duration and start time of each record with the power outage event SJ of each sequence number seq. If the power outage duration deviates from the power outage event SJ of a certain sequence number seq by no more than a preset value, then include it in the power outage event SJ of that sequence number seq. If there are multiple records that meet the conditions for the same user, take the record with the start time closest to that power outage event and include it in the power outage event SJ of that sequence number seq. Step S3.4: Add the final sequence numbers SJ to the SJ set.

[0010] As a further preferred technical solution to the above technical solution, in step S4, for a multi-household power outage event SJ that occurs only once for a single user, if there are repeated power outages for the same user group, it is defined as a hidden danger and the corresponding power outage event set YHSJ. Power outage range identification and classification are performed on YHSJ, including multiple access points, single access point, multiple meter boxes, and single meter box. For multiple access points, the cause of the power outage is identified: For the transformer substation with the main protection device installed, determine whether the main protection device in the same substation has been activated within the range of the power outage start time ± threshold. If it has, obtain the reason for the activation and mark the power outage reason for this transformer substation, and mark it as a switch-type hidden danger KGYH. If there is no activation of the main protection device, mark it as a node-type disconnection hidden danger JDYH. For transformer substations without a main protection device, the following criteria are used to classify the substation based on the description in the repair request, the duration of the power outage, the percentage of affected users (ZB), and the loss of load current after the power outage: If the repair report text describes the switch or ZB as being greater than the threshold for the number of users in the transformer area, it is judged as a switch-related fault. Among them: if the power outage duration is the first time, it is a leakage current switch hazard; if the power outage duration is the second time, it is an overload switch hazard; if the repair report text describes the non-switch or ZB as being less than the threshold for the number of users in the transformer area, it is identified as a node-related disconnection hazard JDYH.

[0011] As a further preferred technical solution of the above technical solution, in step S5, for overload-type switch hidden dangers, the station area load current TQmaxFH at the measurement point immediately before the overload power outage, and the current ZBmaxFH at the measurement point immediately before the overload tripping of the main protection station area device are retained; for leakage-type switch hidden dangers, the zero-fire difference LHC of all users under the station area at the measurement point immediately before the leakage power outage is obtained and suspicious users are determined; If it is a main protection station area, it is also necessary to obtain the leakage current operating value DZZ at the moment of main protection tripping, wherein: In the main protection station area, if there is a user satisfying DZZ / 2<LHC<DZZ, the user is determined as a suspicious user and the risk value is increased, otherwise it is determined as three-phase leakage or line leakage; In a non-main protection station area, the three users with the largest LHC at the measurement point immediately before the power outage are returned, and if there is a user with LHC greater than a preset value, the risk value is increased.

[0012] As a further preferred technical solution of the above technical solution, in step S6, for the node-type breaking hidden danger JDYH, the risk value of the node-type breaking hidden danger JDYH is determined by calculating relevant information, and the specific implementation is the following steps: Step S6.1: Take the power outage user set of the node-type breaking hidden danger JDYH, take the average measured voltage DY of the user set in the week before the power outage, and simultaneously calculate the total current DL of JDYH; Step S6.2: Calculate the voltage change DYBH under the condition of severe current fluctuation DLBH, select events with reverse voltage and current changes, calculate the hidden danger resistance DZ=DLBH / DYBH, and obtain the average resistance PJDZ of all events; Step S6.3: Identify whether the hidden danger resistance increases at the moment of high current. In the severe current fluctuation event, take the maximum measured current value to calculate the hidden danger resistance DDLDZ. If DDLDZ is greater than the product of the average resistance PJDZ and a threshold, the hidden danger risk is increased; Step S6.4: If there is a high load moment in the 2 measurement points before the power outage moment, and the proportion of power outage moments meeting the condition is greater than a threshold, the hidden danger risk value is increased; Step S6.5: If the temperature at the power outage moment is within the preset minimum temperature section of the temperature measurement point, and the proportion of power outage moments meeting the condition is greater than a threshold, the hidden danger risk value is increased.

[0013] As a further preferred technical solution of the above technical solution, step S6 is specifically implemented as the following steps: Step S8.1: Hidden dangers with a quantified risk value greater than a preset value, or hidden dangers corresponding to multiple power outage events whose last event corresponds to a 95598 work order and a security domain repair ticket, are included in enhanced monitoring, with a monitoring period of 1 month; Step S8.2: Monitor the operation of the main protection system and user power outage records through the real-time measurement center. If a power outage event occurs in a similar user range, push information to the person in charge of emergency repair at the corresponding power supply station via SMS and app. The information includes the type of hazard, shared weight, exclusive weight of the hazard type, switch leakage hazard, switch overload hazard, and the load reduction after overload in non-main protection areas. Step S8.3: For overload hazards, when the total protection current value of the corresponding overload switch in the main protection area exceeds ZBmaxFH after a preset number of times, or when the current in the non-main protection area exceeds TQmaxFH after a preset number of times, monitoring shall be stopped. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the present invention. Detailed Implementation

[0015] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.

[0016] In the preferred embodiments of the present invention, those skilled in the art should note that the middle platform and the like involved in the present invention can be regarded as prior art.

[0017] Preferred embodiment.

[0018] like Figure 1 As shown, this invention discloses a method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, comprising the following steps: Step S1 (Data Acquisition and Cleaning): Collect relevant data and categorize historical low-voltage user power outage records by transformer area. After removing relevant power outage records, cleaned power outage records are obtained. Specifically, the relevant data in step S1 includes the power data center, safety monitoring domain, real-time measurement center, meteorological center, PMS ledger, and low-voltage distribution area data in the power supply service command system. The low-voltage distribution area data includes power outage records (from user meters and distribution area master meters), measurement data (from user meters, branch main protection devices, and distribution area master meters), action data of branch main protection devices, and user repair work order data.

[0019] More specifically, the relevant power outage records in step S1 include power outage records (safety supervision work tickets and work records, stored in the safety domain) corresponding to renovation work (first type and second type of work tickets), power outage records corresponding to meter replacement or business expansion work in the marketing system, power outage records corresponding to power outage events at the main meter terminal, and power outage records corresponding to planned power outage work in the marketing system or power supply service command system.

[0020] Step S2 (Power Outage Event Merging): Sort the cleaned power outage records in ascending order of time sequence, and merge power outage events according to the start and end time deviation and threshold. In step S2, the power outage start time deviation Spc and power outage end time deviation Tpc between each power outage record and the previous record are calculated. A threshold Tyz is set. If Spc ≤ Tyz and Tpc ≤ Tyz, it is marked as the same power outage event; otherwise, it is marked as the start of a new event. Multiple power outage event sets are generated within the transformer area, resulting in power outage event SJ, where: Outage start time deviation Spc = current outage start time minus the previous outage start time; Outage end time deviation Tpc = current outage end time minus the previous outage end time.

[0021] Step S3 (Multi-household power outage event splitting): Deduplicatively split the multi-household power outage events, so that each user appears only once in each multi-household power outage event; In step S3, for a power outage event SJ containing multiple household power outage records, if there are identical users, it is defined as an event SJM to be split and split. After splitting, each user appears only once in each multi-household power outage event SJ. The specific implementation of the splitting process is as follows: Step S3.1: Traverse the power outage records (traverse power outage records with similar start and end times), count the frequency of power outages for each user, and take the highest frequency as the number K of the multi-user power outage events SJ after splitting (take the larger number when the frequencies are the same). Step S3.2: Sort the power outage records of N users with K power outage counts in ascending order by the start time of the power outage, and assign a sequence number to each record. Set user 1's power outage record as And the power outage record of user N is Each power outage record includes the start time and duration of the outage. The set of power outage records for different users corresponding to the same serial number X }, take the average of the power outage duration and power outage start time within the preset quantile range (20%~80%), and use them as the power outage duration and power outage start time of the power outage event SJ corresponding to the sequence number seq; Step S3.3: Merge the power outage records for users whose number of power outages is not equal to K. Merge each record with each sequence number (seq) of the power outage event (SJ) (i.e., among users with the same number of power outages, the average of the middle range for the start time and duration of the power outage). This involves merging the power outage records of different users corresponding to the same sequence number X. The power outage duration and start time are calculated and averaged within a preset quantile range (20%~80%). If the power outage duration deviates from the power outage event SJ of a certain sequence number seq by no more than a preset value (5%), it is included in the power outage event SJ of that sequence number seq. If there are multiple records that meet the conditions for the same user, the record with the closest start time to the power outage event is included in the power outage event SJ of that sequence number seq. Step S3.4: Add the final sequence numbers SJ to the SJ set.

[0022] Step S4 (Hazard Identification and Classification): For multiple-household power outages occurring only once for a single user, identify the hazard of repeated power outages within a user group, and classify the hazard events according to the outage scope. In Step S4, for a multiple-household power outage event SJ (including a user list, the start time, end time, and average outage duration for each user), if there are repeated outages within the same user group (user overlap ≥ 50% is considered the same user group), it is defined as a hazard and the corresponding outage event set YHSJ. The outage scope of YHSJ is then identified and classified, including multiple access points, single access point, multiple meter boxes, and single meter box. For multiple access points, the cause of the outage is identified. For transformer substations equipped with a main protection device (primary branch switch protection), determine whether the main protection device in the same substation has activated within the range of the power outage start time ± threshold. If it has, obtain the activation reason (leakage, overload, overcurrent) and mark the power outage reason (leakage, overload, overcurrent) of this transformer substation, and mark it as a switch-type hidden danger KGYH. If no main protection device has activated, mark it as a node-type breaking hidden danger JDYH. For transformer substations without a master warranty installed, the following information is provided in the (marketing system): repair report description, outage duration, percentage of affected users within the substation, and classification of lost load current after the outage. If the repair report text describes the switch or ZB as being greater than the threshold for the number of users in the distribution area (20%), it is judged as a switch-related fault. Among them: if the power outage duration is within the first time (within 2 minutes and similar), it is a leakage current switch hazard; if the power outage duration is the second time (more than 15 minutes and similar, and the loss of single-phase load after the power outage is greater than 100A), it is an overload switch hazard; if the repair report text describes that the switch or ZB is less than the threshold for the number of users in the distribution area (20%), it is identified as a node-related interruption hazard JDYH.

[0023] Step S5 (Hidden danger-related data retention and suspicious user confirmation): Retain key measurement point data according to the switch hidden danger type, and determine suspicious users of electric leakage; In step S5, for overload-type switch hidden dangers (classify first by power outage duration before calculation, or retain all without judgment), retain the station area load current TQmaxFH at the measurement point immediately before the overload power outage, and the current ZBmaxFH (close to the rated current of the switch) at the measurement point immediately before the overload tripping device of the main protection station area; for leakage-type switch hidden dangers (classify first by power outage duration before calculation, or retain all without judgment), obtain the zero-line and live-line difference LHC (live line measurement value - zero line measurement value) of all users under the station area at the measurement point immediately before the leakage power outage, and determine suspicious users; If it is a main protection station area, it is also necessary to obtain the leakage current action value DZZ at the tripping moment of the main protection, where: In the main protection station area, if there is a user satisfying DZZ / 2<LHC<DZZ, the user is determined as a suspicious user and the risk value is increased; otherwise, it is determined as three-phase electric leakage or line electric leakage; In a non-main protection station area, return the three users with the largest LHC at the measurement point immediately before the power outage, and if there is a user with LHC greater than the preset value (100mA), increase the risk value.

[0024] Step S6 (Risk assessment of node-type breaking hidden dangers): Calculate the risk value of node-type breaking hidden dangers; in step S6, for the node-type breaking hidden danger JDYH, by calculating relevant information (taking the user group of the power outage event SJ as the research object, for each of the 96 daily measurement points, taking the sum of the currents of the group as the current, and the average value of the voltages as the voltage, to study the changes of voltage and current; the calculated relevant information includes resistance, resistance change at high power moments, heating power, resistance constancy, correlation between power outage moments and high load, and correlation between power outage moments and temperature), the risk value of the node-type breaking hidden danger JDYH is determined, and the specific implementation includes the following steps: Step S6.1: Take the power outage user set of the node-type breaking hidden danger JDYH, take the average measured voltage DY of the user set in the week before the power outage, and simultaneously calculate the total current DL of JDYH; Step S6.2: By calculating the voltage change DYBH (DYBH=|DY of the previous measurement point - DY of the next measurement point|) under the condition of severe DL fluctuation DLBH (the difference between the currents of two adjacent measurement points is greater than 20% of the maximum current, that is, DLBH=|DL of the previous measurement point - DL of the next measurement point| (DLBH>topDL*20%)), select the events with reverse voltage and current changes, calculate the hidden danger resistance DZ=DLBH / DYBH, and obtain the average resistance PJDZ of all events; Step S6.3: Identify whether the potential resistance increases during high current events. In the event of severe current fluctuation (i.e., severe fluctuation DLBH situation), take the value of the maximum current at the measuring point (the maximum value of the two measuring points before and after a week event can be taken as the top 5% of all events) to calculate the potential resistance DDLDZ. If DDLDZ > average resistance PJDZ multiplied by the threshold (30%), the potential risk is increased (resistance increase caused by heat). Step S6.4: If there is a high load moment (DL>top5DL average*0.5) in the first 2 measuring points of the power outage time, the proportion of power outage times that meet the conditions is greater than the threshold (70%), and the hidden danger risk value is increased; Step S6.5: If the temperature at the time of the power outage is located in the lowest preset range of the temperature measurement point (i.e., the lowest 10% range) in the past month, the proportion of power outages that meet the conditions is greater than the threshold (70%), thus increasing the risk value of potential hazards.

[0025] Step S7: Quantify the risk values ​​of the categorized multi-household hazards (quantified according to dimensions such as outage cause, number of leakage detections, total outage duration, user scale, work order status, outage recurrence after work orders, and JDYH risk value of node disconnection hazards). This can be done using the analytic hierarchy process (AHP). Risk value = dimension * weight. Leakage, overload, and JDYH node disconnection hazards have special weights, while the others are shared weights. Risk value = Common weight + Type-specific weight: Total weight = Σ Number of power outages * weight 1 + Total power outage duration * weight 2 + Number of households experiencing power outages * weight 3 + Number of work orders * weight 4 + Number of power outages after a work order * weight 5.

[0026] The selection criteria for each weight are as follows: Weight 1: Based on the probability of repeated occurrences of group-related hazards in low-voltage distribution areas, the more power outages there are, the higher the likelihood that the hazard persists and has not been eradicated. Combining historical fault statistics and on-site operation and maintenance experience, the contribution of the number of power outages to the hazard risk is used as the main criterion for judgment, and corresponding weight values ​​are assigned.

[0027] Weight 2: The total outage duration reflects the difficulty of fault recovery and the severity of potential hazards. The longer the duration, the more complex the fault, the more difficult the handling, and the greater the impact on power supply reliability. The risk contribution weight of the total outage duration is determined based on the actual emergency repair data of the power supply station and the degree of power outage impact.

[0028] Weight 3: The scale of users affected by a power outage represents the level of group risk. The more households affected by a power outage, the greater the social impact and the pressure on power supply services. The weight corresponding to the number of households affected by a power outage is determined according to the user base of the low-voltage distribution area, the power supply reliability assessment requirements, and the classification standards for group incidents.

[0029] Weight 4: The number of repair work orders and emergency repair requests submitted through 95598 reflects user perception and the degree of fault visibility. The more work orders there are, the more significant the impact on user electricity supply and the stronger the user's demands. Based on user sensitivity and emergency repair priority requirements, the number of work orders is assigned a corresponding weight.

[0030] Weight 5: The number of times the power goes out again after the hidden danger has been dealt with represents the effectiveness of the defect eradication. The more times the problem recurs, the more likely the original hidden danger has not been completely eliminated and there is a risk of repeated power outages. This is a key indicator for determining whether it is necessary to strengthen monitoring and proactive repair.

[0031] Type-specific weight: Node interruption risk JDYH = Σ resistance * weight 1 + resistance change rate at high power * weight 2 + heat generation power (I * I * R) * weight 3 + resistance standard deviation * weight 4 + power outage rate after high load * weight 5 + power outage rate at low temperature * weight 6.

[0032] The potential for a switch to trip due to leakage current is calculated as follows: LDKGYH = Σ leakage current magnitude (total measured zero-sequence current magnitude LX) * weight 7 + similarity between user's zero-live difference LHC and leakage current 1 - ((LX-LHC) / max(LX,LHC)) * weight 8.

[0033] Switch overload hazard GFHKGYH=Σ Total maximum phase current before power outage / Rated current * weight 9||max(Transformer single-phase current before power outage - Transformer single-phase current after power outage) * weight 10.

[0034] The selection criteria for each weight are as follows: Weight 1: Based on the fact that the contact resistance at the potential fault point directly reflects the degree of joint oxidation and poor contact, the greater the resistance, the higher the risk of interruption. The weight is determined according to the contribution of contact resistance to the occurrence of the fault.

[0035] Weight 2: Based on the increase in resistance under high power, which characterizes thermal instability, the faster the resistance increases with power, the more significant the risk of overheating and burnout. The weight is determined according to the degree of influence of resistance change on the deterioration of potential hazards.

[0036] Weight 3: Based on heating power I 2 R reflects the long-term heat deterioration trend of potential hazards. The greater the heat output, the higher the probability of equipment burn-out and instantaneous interruption. Weights are assigned according to the severity of heat deterioration.

[0037] Weight 4: The resistance fluctuation amplitude reflects the degree of instability of the contact state. The larger the standard deviation, the less reliable the joint contact. The weight is determined according to the contribution of resistance fluctuation to the risk of breakage.

[0038] Weight 5: Based on the characteristics of cold contraction interruption after high load ends, the higher the proportion of power outages after high load, the more it conforms to the typical fault law of node thermal expansion and contraction. The weight is determined according to the feature matching degree.

[0039] Weight 6: Based on the physical characteristics that low temperature environment exacerbates the risk of joint cold shrinkage and breakage, the higher the proportion of power outages under low temperature, the more typical the hidden danger, and the weight is determined according to the contribution of low temperature to the fault.

[0040] Weight 7: The severity of leakage is directly reflected by the magnitude of the total zero-sequence current. The closer the current is to the trip value, the more significant the leakage fault. The weight is determined according to the contribution of leakage intensity to the tripping risk.

[0041] Weight 8: The closer the user's zero-current difference is to the total leakage current, the more likely it is to be a source of leakage fault. The higher the similarity, the higher the accuracy of the location. The weight is determined according to the reliability of the fault source matching.

[0042] Weight 9: The degree of overload severity is reflected by the extent to which the current exceeds the rated value. The higher the limit is exceeded, the stronger the inevitability of overload tripping. The weight is determined according to the degree of overload.

[0043] Weight 10: The actual operation of the switch is reflected by the load drop before and after the power outage. The larger the drop, the more it confirms the overload trip. The weight is determined according to the confirmation degree of overload characteristics.

[0044] The above calculations are averaged for multiple power outages.

[0045] Step S8: Strengthen monitoring and proactive response, specifically implemented through the following steps: Step S8.1: Include the top 20% of quantified risk values ​​that are greater than the preset value, or the hidden dangers that are the last of the multiple power outage events corresponding to the hidden dangers and are the last of the 95598 work order or safety domain emergency repair ticket, into the enhanced monitoring, with a monitoring period of 1 month; Step S8.2: Monitor the overall protection action and user power outage records through the real-time measurement center. If a power outage event occurs with a similar user range (overlapping with the user range corresponding to the hidden danger by ≥50%), push the information to the corresponding power supply station's emergency repair manager via SMS and app (iGuowang). The information includes the hidden danger type, shared weight (number of power outages, total power outage duration, scale of power outage users, number of work orders, and power outage recurrence after the work order), unique weight of the hidden danger type (node ​​disconnection hidden danger (JDYH risk value)), switch leakage hidden danger (the magnitude of leakage current and the degree of matching of leakage current), switch overload hidden danger (the magnitude of the overload current in the overall protection area), and the load reduction after overload in non-overload areas (to facilitate the operation personnel to handle according to priority and quickly locate the problem). Step S8.3: For overload hazards, if the total protection current value of the corresponding overload switch in the main protection area exceeds ZBmaxFH after a preset number of times (3 times or more), or if the current in the non-main protection area exceeds TQmaxFH after a preset number of times (more than 3 times), monitoring shall be stopped.

[0046] It is worth mentioning that the technical features such as the middle platform involved in this patent application should be regarded as prior art. The specific structure, working principle, and possible control methods and spatial arrangement of these technical features can be adopted using conventional choices in the field, and should not be regarded as the inventive point of this patent. This patent will not be further elaborated in detail.

[0047] For those skilled in the art, modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the protection scope of this invention.

Claims

1. A method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, characterized in that, Includes the following steps: Step S1: Collect relevant data and categorize historical low-voltage user power outage records by transformer area. After removing relevant power outage records, cleaned power outage records are obtained. Step S2: Sort the cleaned power outage records in ascending order of time sequence, and merge the power outage events according to the start and end time deviation and the threshold. Step S3: Deduplicatize and split the multi-household power outage events. After splitting, each user appears only once in each multi-household power outage event. Step S4: For multiple power outages that occur only once for a single user, identify potential risks of repeated power outages for user groups, and classify potential risks according to the scope of the power outage. Step S5: Record key measurement point data according to the type of switch hazard to identify users suspected of leakage; Step S6: Calculate the risk value of node-type breakage hazards; In step S6, for the node-type breakage hazard JDYH, the risk value of the node-type breakage hazard JDYH is determined by calculating relevant information. The specific implementation is as follows: Step S6.1: Take the set of power outage users of node-type interruption hazard JDYH, take the average voltage DY of the measurement points of the user set one week before the power outage, and calculate the total current DL of JDYH at the same time. Step S6.2: Calculate the voltage change DYBH under the condition of DL drastic fluctuation DLBH, take the event of reverse voltage and current change, calculate the hidden danger resistance DZ=DYBH / DLBH, and obtain the average resistance PJDZ of all events. Step S6.3: Identify whether the potential resistance increases during high current events. In events with drastic current fluctuations, calculate the potential resistance DDLDZ by taking the maximum current value at the measurement point. If DDLDZ > average resistance PJDZ multiplied by the threshold, the potential risk increases. Step S6.4: If there are high load moments among the two measuring points before the power outage, and the proportion of power outage moments that meet the conditions is greater than the threshold, the potential risk value will be increased. Step S6.5: If the temperature at the time of power outage is located in the lowest preset range of the temperature measuring point, the proportion of power outage times that meet the conditions is greater than the threshold, thus increasing the potential risk value. Step S7: Quantify the risk values ​​of the classified multi-household hazards; Step S8: Enhance monitoring and proactive response.

2. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 1, is characterized in that... The relevant data for step S1 includes data from the power data platform, safety monitoring domain, real-time measurement center, meteorological center, PMS ledger, and low-voltage distribution area data in the power supply service command system. Low-voltage distribution area data includes power outage records, measurement data, branch main protection action data, and user repair work order data.

3. The method for identifying and quantifying low-voltage transformer area group hazards based on multi-source data from a central data platform, as described in claim 2, is characterized in that... The relevant power outage records for step S1 include power outage records corresponding to renovation work in the power security domain, power outage records corresponding to meter replacement or business expansion work in the marketing system, power outage records corresponding to power outage events in the master meter terminal, and power outage records corresponding to planned power outage work in the marketing system or power supply service command system.

4. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 3, is characterized in that... In step S2, the power outage start time deviation Spc and power outage end time deviation Tpc between each power outage record and the previous record are calculated. A threshold Tyz is set. If Spc ≤ Tyz and Tpc ≤ Tyz, it is marked as the same power outage event; otherwise, it is marked as the start of a new event. Multiple power outage event sets are generated within the transformer area, resulting in power outage event SJ, where: Power outage start time deviation Spc = current power outage start time minus the previous power outage start time; Power outage end time deviation Tpc = current power outage end time minus the previous power outage end time.

5. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 4, is characterized in that... In step S3, for a power outage event SJ containing power outage records of multiple households, if the same user exists, it is defined as an event to be split SJM and subjected to splitting processing, and after splitting, each user appears only once in each multi-household power outage event SJ, and the specific implementation of the splitting processing is as follows: Step S3.1: Traverse the power outage records, count the occurrence frequency of the number of power outages of each user, and take the number with the highest frequency as the number K of multi-household power outage events SJ after splitting; Step S3.2: Sort the power outage records of N users with K power outage counts in ascending order by the start time of the power outage, and assign a sequence number to each record. Set user 1's power outage record as And the power outage record of user N is Each power outage record includes the start time and duration of the outage. The set of power outage records for different users corresponding to the same serial number X The average value of the outage duration and outage start time within the preset quantile range is used as the outage duration and outage start time of the outage event SJ corresponding to the sequence number seq. Step S3.3: Perform merging processing on the power outage records where the number of power outages of the user is not equal to K, compare each record with the power outage duration and power outage start time of each power outage event SJ with sequence number seq, if the deviation between the power outage duration and the power outage event SJ with a certain sequence number seq does not exceed a preset value, it is included in the power outage event SJ with that sequence number seq, and when there are multiple records of the same user that meet the conditions, the record closest to the start time of the power outage event is taken and included in the power outage event SJ with that sequence number seq; Step S3.4: Add the finally formed SJs of all sequence numbers to the SJ set.

6. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 5, is characterized in that... In step S4, for a multi-household power outage event SJ in which a single user appears only once, if there are repeated power outages of the same user group, it is defined as a hidden danger and the corresponding power outage event set YHSJ, and power outage scope identification and classification are performed on YHSJ, where the power outage scope identification and classification includes multi-access point, single-access point, multi-meter box and single-meter box, and power outage cause identification is performed for multi-access points: For a distribution area installed with a main protection device, determine whether there is a main protection operation of the same distribution area within the range of power outage start time ± threshold, if it exists, obtain the operation cause and mark the power outage cause of this SJ, and mark it as a switch-type hidden danger KGYH, if there is no main protection operation, mark it as a node-type disconnection hidden danger JDYH; For a distribution area not installed with a main protection device, classification is performed through repair text description, power outage duration, the proportion ZB of the number of affected user groups in the distribution area, and the lost load current after power outage, wherein: If the repair text description refers to a switch or ZB is greater than the user number threshold of the distribution area, it is determined as a switch-type fault, wherein: it is further determined that a power outage duration of the first time is a leakage-type switch hidden danger; a power outage duration of the second time is an overload-type switch hidden danger; if the repair text description does not refer to a switch or ZB is less than the user number threshold of the distribution area, it is identified as a node-type disconnection hidden danger JDYH.

7. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 6, is characterized in that... In step S5, for the overload-type switch hidden danger, the distribution area load current TQmaxFH at the measuring point immediately before the overload power outage, and the current ZBmaxFH at the measuring point immediately before the overload tripping device in the distribution area with main protection are retained; For the leakage-type switch hidden danger, obtain the neutral-live line difference LHC of all users under the distribution area at the measuring point immediately before the leakage power outage and determine suspicious users; If it is a distribution area with main protection, it is also necessary to obtain the leakage current operation value DZZ at the moment of main protection tripping, wherein: In the distribution area with main protection, if there exists a user satisfying DZZ / 2<LHC<DZZ, the user is determined as a suspicious user and the risk value is increased, otherwise it is determined as three-phase leakage or line leakage; In non-main protection areas, return to the three users with the highest LHC at the measurement point before the power outage. If there are users with LHC > preset value, the risk value is increased.

8. The method for identifying and quantifying potential hazards in low-voltage distribution areas based on multi-source data from a central data platform, as described in claim 7, is characterized in that... Step S8 is specifically implemented as follows: Step S8.1: Include the hidden dangers whose quantified risk values ​​are greater than the preset values, or whose last power outage event among multiple events is a 95598 work order or a safety domain emergency repair ticket, into the enhanced monitoring system. The monitoring period is 1 month. Step S8.2: Monitor the operation of the main protection system and user power outage records through the real-time measurement center. If a power outage event occurs in a similar user range, push information to the corresponding power supply station emergency repair manager via SMS and app. The information includes the type of hazard, shared weight, exclusive weight of the hazard type, switch leakage hazard, switch overload hazard, and the load reduction after overload in non-main protection areas. Step S8.3: For overload hazards, when the total protection current value of the corresponding overload switch in the main protection area exceeds ZBmaxFH after a preset number of times, or when the current in the non-main protection area exceeds TQmaxFH after a preset number of times, monitoring shall be stopped.

Images