A method for judging the authenticity of power outage events in distribution networks

Abstract

The present invention discloses a method for judging the authenticity of a power outage event in a distribution network, which relates to the technical field of power grid monitoring and management, and includes the following steps: collecting original data from different sources in the distribution network power supply system, including power outage reports, user feedback, meteorological information, power grid operation status data, and historical power outage records, and performing preprocessing operations on the collected original data to obtain a distribution information sequence table; extracting abnormal features related to the power outage event from the distribution information sequence table. The present invention can judge the authenticity of a power outage event in a distribution network in the first time by combining a real-time monitoring system with a risk assessment model, and integrates real-time distribution network status monitoring to make the judgment result more accurate. When a power outage event occurs, key data can be quickly captured, and the power outage risk coefficient can be instantly calculated through a risk assessment model, thereby quickly judging the authenticity and potential impact of the power outage event.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid monitoring and management, and in particular to a method for determining the authenticity of a power outage event in a distribution network. Background Art

[0002] With the continuous development of the power system and the gradual liberalization of the power market, users have increasingly higher requirements for power supply reliability and service quality. As an important part of the power system, the authenticity of power outages in the distribution network is directly related to whether the power supply company can grasp the power outage situation in a timely and accurate manner and take corresponding emergency repair measures to ensure the power demand of users. Frequent power outages will seriously affect the level of power supply reliability, resulting in large differences in the reliability levels of different regions and lines. If users access individual low-reliability lines, new weak nodes in the power grid will be formed, increasing frequent complaints and emergency repair work orders.

[0003] For example, the reliability power outage event collection integrity analysis and automatic completion method disclosed in Chinese patent publication number: CN108647328A includes building a unified analysis model for power outage events in the distribution network; aggregating the automatically collected discrete user power outage information to generate original power outage events; performing collection integrity analysis on the original power outage events to locate the power outage lines with incomplete collection and the users with missing collection; judging and identifying the users with real power outages among the missing users; automatically completing the power outage information for the missing users of the real power outage, and uploading the final power outage event to the reliability business system.

[0004] In the prior art, the missing users are located and searched, and shared data is used to determine whether the users are actually experiencing a power outage. The power outage information is automatically completed for the missing users who are actually experiencing a power outage. However, some users can restore power supply independently after a power outage, which will affect the accuracy of the judgment result. Combined with the interference of harsh environmental factors on power supply restoration, the risk identification of distribution network power outage events is further interfered with, and the accuracy of authenticity judgment of power outage events is further interfered with. Therefore, how to conduct a comprehensive analysis and judgment of power outage events and enhance the risk identification ability of distribution network power outage events to improve the accuracy of authenticity judgment is the problem we need to solve. To this end, a method for authenticity judgment of distribution network power outage events is proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a method for determining the authenticity of a power outage event in a distribution network, so as to solve the problems raised in the above-mentioned background technology.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for determining the authenticity of a power outage event in a distribution network comprises the following steps:

[0008] Step 1: collect raw data from different sources in the distribution network power supply system, including power outage reports, user feedback, meteorological information, power grid operation status data, historical power outage records, and perform pre-processing operations such as cleaning and sorting the collected raw data to obtain a distribution information sequence table;

[0009] Step 2: extract abnormal features related to the power outage event from the power distribution information sequence table, preliminarily screen out the power outage areas and equipment related to the power outage event by combining time-space correlation analysis, analyze the impact of the power outage event, and analyze the environmental interference trend of the power outage event area by combining meteorological information data;

[0010] Step 3: Identify the user's power supply behavior pattern, analyze historical power outage records, power usage habits and grid status, and evaluate the authenticity of the user's autonomous power supply restoration;

[0011] Step 4: Based on the analysis results of the impact of the power outage, the environmental interference trend, and the evaluation results of the user's autonomous power restoration, a risk assessment model is constructed using machine learning technology to analyze the risk of the power outage to further determine the authenticity of the power outage.

[0012] Step 5: Configure a real-time monitoring system to continuously track the distribution network status, generate a risk report for power outage events based on the output results of the risk assessment model, and feed back the risk report and judgment results to relevant personnel.

[0013] A further improvement of the technical solution of the present invention is that in step 1, the process of obtaining the power distribution information sequence table includes:

[0014] Step 101, collecting raw data from the power distribution system's power grid management system, customer service system, meteorological department, and historical database, wherein power grid operation status data, including voltage, current, power, equipment status, etc., are extracted from the power grid management system; user feedback and power outage reports, including report time, location, description, etc., are collected from the customer service system; meteorological information, including temperature, humidity, wind speed, rainfall, etc., is obtained from the meteorological department; historical power outage records and user power usage habit data, including power outage time, location, cause, duration, etc., are extracted from the historical database;

[0015] Step 102, integrating the raw data from different sources into a unified data set, adding a common timestamp to each record in the data set to facilitate subsequent analysis, and performing preprocessing operations on the raw data, including data cleaning and data conversion steps;

[0016] Step 103, classify and organize the data by time period, and establish a distribution information sequence table based on the organized data, wherein the distribution information sequence table includes but is not limited to the power outage event number, power outage time, power outage location, impact range, power outage cause, meteorological conditions, processing process, recovery time, user feedback summary and related equipment status, verify the organized distribution information sequence table to ensure the accuracy and completeness of the data, and store the verified distribution information sequence table and related data in the data warehouse.

[0017] A further improvement of the technical solution of the present invention is that in step 2, the analysis process of the impact degree of the power outage event and the environmental interference trend includes:

[0018] Step 201, filter out all power outage event records from the power distribution information sequence table, extract abnormal features of the power outage event for each power outage event record, including abnormal power outage time, abnormal power outage duration, abnormal expansion of power outage scope, and abnormal increase in the number of affected users, and perform preliminary analysis on the extracted abnormal features to identify corresponding abnormal patterns;

[0019] Step 202, according to the timestamps of the power outage events, analyze the time distribution of the power outage events, generate a time distribution chart of the power outage events, identify the time periods of frequent power outages, use GIS technology to mark the locations of the power outage events on the map, analyze the spatial distribution of the power outage events, identify the areas with frequent power outages and adjacent areas, perform superposition analysis on the results of the time distribution and the spatial distribution, identify the power outage areas that are both frequent in time and concentrated in space, and preliminarily screen out the areas with frequent power outages based on the results of the spatiotemporal correlation analysis;

[0020] Step 203, check the status records of relevant equipment in the frequent power outage area, including fault records and maintenance records, summarize and analyze the feedback from users in the frequent power outage area, and further screen out equipment directly related to the power outage event in combination with the equipment status and user feedback;

[0021] Step 204, based on the duration of the power outage event in the frequent power outage area, the number of users affected by the power outage, the severity score caused by the power outage, and the equipment failure status record of the power outage, a comprehensive analysis is performed to obtain a power outage event impact index to evaluate the impact of the power outage event;

[0022] Step 205, matching the meteorological information with the time point and location of the power outage event, obtaining meteorological data at the corresponding time point and location, analyzing the changing trend of the meteorological information over time, identifying the meteorological factors affecting the power outage event, calculating the environmental interference index based on the severity of the meteorological information and historical data, and quantifying the potential impact of the meteorological information on the power outage event;

[0023] Step 206 , combining the environmental interference index and the weather forecast data, evaluating the interference trend of the weather information change on the power outage event, so as to identify the high-risk time periods and areas.

[0024] A further improvement of the technical solution of the present invention is that the expression of the power outage event impact index is:

[0025]

[0026] Among them, S is the power outage event impact index, which represents the comprehensive impact score of the power outage event, T i is the duration of the ith power outage, T base is the baseline power outage duration, T max is the longest power outage duration during the observation period, F i is the number of users affected by the ith power outage, F max is the maximum number of users affected during the observation period, U i is the severity score of the user feedback of the i-th power outage, U base Score the baseline user feedback severity, R i is the equipment status record of the i-th power outage, and n is the total number of power outage events;

[0027] The expression of the environmental interference index is:

[0028]

[0029] Among them, EI is the environmental interference index, M j is the current value of the jth meteorological factor, M base is the baseline value of the jth meteorological factor, M max is the maximum value of the jth meteorological factor during the observation period, w j is the weight of the jth meteorological factor, indicating its importance in affecting the power outage event. α is the attenuation coefficient, which is used to adjust the attenuation speed of the historical impact HI. j is the historical impact index of the jth meteorological factor, and m is the total number of meteorological factors.

[0030] A further improvement of the technical solution of the present invention is that in step 3, the process of evaluating the authenticity of the user's autonomous power restoration includes:

[0031] Step 301, based on the power distribution information sequence table, extract historical power outage records and user's power usage habit data, wherein the historical power outage records include power outage time, location, duration, and cause, and the user's power usage habit data includes power consumption, peak power consumption period, and power equipment type;

[0032] Step 302, combining historical power outage records and user power usage habit data, analyzing the user's power usage behavior during historical power outage events, identifying user behavior patterns, and identifying typical behaviors of users using backup power and reducing power usage during power outages;

[0033] Step 303: Analyze the power supply demand of users under different circumstances in combination with historical power outage records and power consumption habit data, analyze the user's behavioral response in the power outage event, and evaluate the user's ability to restore power autonomously in combination with the capacity status of the user's backup power supply to determine whether the user has the conditions and ability to restore power autonomously;

[0034] Step 304, monitoring the operation status of the power grid, including load, power generation capacity, and power grid stability, and analyzing the current load distribution, redundancy and vulnerability status of the power grid, and evaluating the self-healing capability of the power grid and the response time when a fault occurs;

[0035] Step 305, applying regression analysis technology, combined with the analysis results of historical power outage records, power consumption habits and power grid status, calculates the power supply restoration index, analyzes the probability of users independently restoring power supply after a power outage, and comprehensively evaluates the authenticity and stability of users' autonomous power supply restoration.

[0036] A further improvement of the technical solution of the present invention is that the expression of the power supply restoration index is:

[0037]

[0038] Among them, PR is the power supply restoration index, T i is the duration of the ith power outage, T avg is the average duration of all power outage events, n is the total number of power outage events, D i is the power consumption during the i-th power outage, D avg is the average power consumption of all power outage events, w i is the weight of the ith power outage, β and δ are adjustment indexes used to balance the impact of different parts in the formula.

[0039] A further improvement of the technical solution of the present invention is that in step 4, the process of constructing the risk assessment model includes:

[0040] Step 401, extracting data related to the impact of power outage events, environmental interference trends, and user autonomous power restoration assessment based on the power distribution information sequence table, and performing feature extraction on the acquired data, selecting features associated with power outage risks, including power outage frequency, duration, impact range, user feedback, and meteorological conditions, to obtain a risk feature set;

[0041] Step 402, dividing the risk feature set into a training set and a test set, using the training set data in combination with the neural network model to build a risk assessment model, and using the test set data to verify the trained risk assessment model to assess the risk of the power outage event;

[0042] Step 403, based on the evaluation results of the risk evaluation model, and in combination with the power outage event impact index, the environmental interference index and the power supply restoration index, a power outage risk coefficient is calculated, and the authenticity of the power outage event is further analyzed using the trend of the power outage risk coefficient and the impact caused by the power outage event;

[0043] Step 404, based on the calculation result of the power outage risk coefficient and the impact and severity of the power outage event, the risk level of the power outage event is divided into low risk level, medium risk level and high risk level, and a corresponding risk threshold is matched for each risk level.

[0044] A further improvement of the technical solution of the present invention is that the expression of the power outage risk coefficient is:

[0045]

[0046] Among them, F is the power outage risk factor, which reflects the overall risk of a power outage event; S is the power outage event impact index, which reflects the severity of the power outage event; EI is the environmental interference index, which quantifies the potential impact of meteorological conditions on power outage events; PR is the power supply restoration index, which reflects the probability that users will restore power supply autonomously after a power outage; and w S 、w EI 、w PR They are the weight coefficients of power outage impact index, environmental interference index and power supply restoration index, which are used to adjust the relative importance of different indexes in risk assessment.

[0047] A further improvement of the technical solution of the present invention is that: the multiple risk levels correspond to multiple risk thresholds, wherein the risk thresholds include an upper threshold and a lower threshold;

[0048] The multiple risk levels and the multiple risk thresholds satisfy the following relationship:

[0049] Low risk level 0≤F <F M ;

[0050] Medium risk level F M ≤F <F H ;

[0051] High risk level F H ≤F≤1;

[0052] Among them, F is the power outage risk factor, F Mis the lower threshold corresponding to the medium risk level and the upper threshold corresponding to the low risk level, F H is the lower threshold corresponding to the high risk level and the upper threshold corresponding to the medium risk level, F M =0.33, F H =0.66.

[0053] A further improvement of the technical solution of the present invention is that in step 5, the process of generating the power outage event risk report includes:

[0054] Step 501, configure a real-time monitoring system, and integrate the trained risk assessment model into the real-time monitoring system to obtain various real-time data in the distribution network, including power outage reports, user feedback, meteorological information, and power grid operation status data;

[0055] Step 502, based on the real-time monitoring system, continuously tracking the distribution network status, inputting the collected real-time data into the risk assessment model, calculating the power outage risk coefficient, and generating a preliminary risk assessment result to assess the risk level of the power outage event;

[0056] Step 503, based on the risk assessment results, automatically generate a risk report of the power outage event, the risk report includes the risk level, the impact of the power outage event, and the recommended response measures, and distribute the risk report and the judgment results to relevant personnel;

[0057] Step 504, collect feedback from relevant personnel on risk reports and judgment results, further analyze the effects and problems of the risk assessment model in application, and iteratively optimize the real-time monitoring system, risk assessment model, and risk report based on the feedback to improve the accuracy and practicality of the system.

[0058] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art:

[0059] 1. The present invention provides a method for judging the authenticity of a distribution network power outage event. By combining a real-time monitoring system with a risk assessment model, the authenticity of a distribution network power outage event can be judged in the first time, and real-time distribution network status monitoring is integrated to make the judgment result more accurate. When a power outage event occurs, key data can be quickly captured, and the power outage risk coefficient can be instantly calculated through a risk assessment model, thereby quickly judging the authenticity and potential impact of the power outage event.

[0060] 2. The present invention provides a method for determining the authenticity of a power outage event in a distribution network. By comprehensively analyzing historical power outage records, user power usage habits, environmental interference trends, and the user's ability to restore power independently, the accuracy and efficiency of responding to power outage events can be significantly improved. Through a real-time monitoring system and a risk assessment model, the authenticity of a power outage event and the risk of a power outage can be quickly identified, thereby reducing unnecessary waste of resources. In addition, the response priority can be dynamically adjusted according to the trend and severity of the power outage risk factor, thereby improving the overall efficiency of handling power outage events. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0062] Figure 1 is a flow chart of the method of the present invention;

[0063] Figure 2 The flowchart of the analysis of the impact degree of the power outage event and the environmental interference trend of the present invention;

[0064] Figure 3 A flow chart for evaluating the authenticity of user autonomous power restoration in the present invention;

[0065] Figure 4 The figure is a flow chart for constructing the risk assessment model of the present invention. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] Embodiment 1, as Figure 1-Figure 3 As shown, the present invention provides a method for determining the authenticity of a power outage event in a distribution network, comprising the following steps:

[0068] Step 1: Collect raw data from different sources in the distribution network power supply system, including power outage reports, user feedback, meteorological information, power grid operation status data, and historical power outage records, and perform pre-processing operations such as cleaning and sorting on the collected raw data to obtain a distribution information sequence table, collect raw data from the power grid management system, customer service system, meteorological department, and historical database of the distribution network power supply system, wherein the power grid operation status data, including voltage, current, power, equipment status, etc., are extracted from the power grid management system, user feedback and power outage reports, including report time, location, description, etc., are collected from the customer service system, meteorological information, including temperature, humidity, wind speed, rainfall, etc., is obtained from the meteorological department, historical power outage records and user electricity usage habit data, including power outage time, location, cause, duration, etc., are extracted from the historical database, and raw data from different sources are integrated into a unified In the data set, a common timestamp is added to each record in the data set to facilitate subsequent analysis, and the original data is preprocessed, including data cleaning and data conversion steps, wherein missing values ​​are identified and processed through data cleaning, duplicate records are identified and deleted, and obvious errors or abnormal values ​​are identified and corrected, the cleaned data are converted into a unified format and unit through data conversion, the data are classified and sorted by time period, and a distribution information sequence table is established based on the sorted data, wherein the distribution information sequence table includes but is not limited to the power outage event number, power outage time, power outage location, impact range, power outage cause, meteorological conditions, processing process, recovery time, user feedback summary and related equipment status, the sorted distribution information sequence table is verified to ensure the accuracy and completeness of the data, and the verified distribution information sequence table and related data are stored in the data warehouse;

[0069] Step 2, extract the abnormal features related to the power outage event from the power distribution information sequence table, combine the time and space correlation analysis to preliminarily screen out the power outage areas and equipment related to the power outage event, analyze the impact of the power outage event, and analyze the environmental interference trend of the power outage event area in combination with meteorological information data, screen out all the records of power outage events from the power distribution information sequence table, and for each power outage event record, extract the abnormal features of the power outage event, including abnormal power outage time, abnormal power outage duration, abnormal expansion of the power outage range, and abnormal increase in the number of affected users, and perform a preliminary analysis on the extracted abnormal features to identify the corresponding abnormal patterns, analyze the time distribution of the power outage event according to the timestamp of the power outage event, generate a time distribution chart of the power outage event, identify the time period of frequent power outages, use GIS technology to mark the location of the power outage event on the map, analyze the spatial distribution of the power outage event, identify the areas and adjacent areas with frequent power outages, and overlay the results of the time distribution and spatial distribution to identify the power outage areas that are both frequent in time and concentrated in space. Based on the results of spatiotemporal correlation analysis, the areas with frequent power outages are preliminarily screened out, and the status records of relevant equipment in the areas with frequent power outages are checked, including fault records and maintenance records. The feedback from users in the areas with frequent power outages is summarized and analyzed. Combined with the equipment status and user feedback, the equipment directly related to the power outage events is further screened out. Based on the duration of the power outage events in the areas with frequent power outages, the number of users affected by the power outages, the severity score caused by the power outages, and the equipment failure status records of the power outages, a comprehensive analysis is performed to obtain the power outage event impact index, and the impact of the power outage events is evaluated. The meteorological information is matched with the time and location of the power outage events, and the meteorological data of the corresponding time and location are obtained. The changing trend of meteorological information over time is analyzed to identify the meteorological factors affecting the power outage events. According to the severity of the meteorological information and historical data, the environmental interference index is calculated to quantify the potential impact of meteorological information on the power outage events. Combined with the environmental interference index and meteorological forecast data, the interference trend of meteorological information changes on the power outage events is evaluated to identify high-risk time periods and areas.

[0070] Furthermore, the expression of the power outage event impact index is:

[0071]

[0072] Among them, S is the power outage event impact index, which represents the comprehensive impact score of the power outage event, T i is the duration of the ith power outage, T base is the baseline power outage duration, T max is the longest power outage duration during the observation period, F i is the number of users affected by the ith power outage, F max is the maximum number of users affected during the observation period, U i is the severity score of the user feedback of the i-th power outage, Ubase Score the baseline user feedback severity, R i is the equipment status record of the i-th power outage, n is the total number of power outage events, R i is the device status, 0 indicates failure and 1 indicates normal. As the duration of the power outage, the number of affected users, the severity of user feedback, and the deterioration of the device status, S increases, indicating that the comprehensive impact of the power outage event is greater;

[0073] The expression of environmental interference index is:

[0074]

[0075] Among them, EI is the environmental interference index, M j is the current value of the jth meteorological factor, M base is the baseline value of the jth meteorological factor, M max is the maximum value of the jth meteorological factor during the observation period, w j is the weight of the jth meteorological factor, indicating its importance in affecting the power outage event. α is the attenuation coefficient, which is used to adjust the attenuation speed of the historical impact HI. j is the historical impact index of the jth meteorological factor, ranging from 0 to 1, where 1 indicates that the meteorological factor has a significant impact on power outage events in history, and m is the total number of meteorological factors. As the meteorological factor value deviates from the baseline value, the weight increases, and the historical impact index increases, the EI value increases, indicating that the meteorological conditions have a greater potential impact on power outage events;

[0076] Step 3: Identify the user's power supply behavior pattern, analyze historical power outage records, power consumption habits and power grid status, evaluate the authenticity of the user's self-restored power supply, and extract historical power outage records and user's power consumption habit data based on the power distribution information sequence table, where the historical power outage records include the power outage time, location, duration, and cause, and the user's power consumption habit data includes power consumption, peak power consumption period, and power equipment type. Combine the historical power outage records and the user's power consumption habit data to analyze the user's power consumption behavior in historical power outage events, identify user behavior patterns, and identify the user's typical behavior of using backup power and reducing power consumption during power outages. Combine the historical power outage records and power consumption habit data to analyze the user's power consumption behavior in different power outage events. The power supply demand in the event of a power outage is analyzed, and the user's behavioral response to the power outage is analyzed. Combined with the capacity status of the user's backup power supply, the user's ability to restore power supply autonomously is evaluated to determine whether the user has the conditions and ability to restore power supply autonomously. The operation status of the power grid is monitored, including load, power generation capacity, and grid stability. The load distribution, redundancy, and vulnerability status of the current power grid are analyzed. The self-healing ability of the power grid and the response time in the event of a fault are evaluated. Regression analysis technology is applied to combine historical power outage records, power consumption habits, and the analysis results of the power grid status to calculate the power supply restoration index, analyze the probability of users autonomously restoring power supply after a power outage, and comprehensively evaluate the authenticity and stability of users' autonomous power supply restoration;

[0077] Furthermore, the expression of the power supply restoration index is:

[0078]

[0079] Among them, PR is the power supply restoration index, T i is the duration of the ith power outage, T avg is the average duration of all power outage events, n is the total number of power outage events, D i is the power consumption during the i-th power outage, D avg is the average power consumption of all power outage events, w i is the weight of the ith power outage, β and δ are adjustment indexes used to balance the impact of different parts in the formula, It means that the deviation of the power outage duration from the average duration is adjusted using an exponential function, which decreases rapidly as the power outage duration increases. It represents the cube of the ratio of power consumption to average power consumption, and is used to emphasize the deviation of power consumption. The value of PR is between 0 and 1, where 0 indicates almost no possibility of autonomous power restoration, and 1 indicates a very high probability of autonomous power restoration. As the power outage duration increases and the power consumption decreases, the PR value decreases, indicating that the possibility of autonomous power restoration decreases. On the contrary, if the power outage duration is short and the power consumption is close to the average level, the PR value increases, indicating that the possibility of autonomous power restoration increases.

[0080] Step 4: Based on the analysis results of the impact of the power outage, the environmental interference trend, and the evaluation results of the user's autonomous power restoration, a risk assessment model is constructed using machine learning technology to analyze the risk of the power outage to further determine the authenticity of the power outage.

[0081] Step 5: Configure a real-time monitoring system to continuously track the distribution network status, generate a risk report for power outage events based on the output results of the risk assessment model, and feed back the risk report and judgment results to relevant personnel.

[0082] Embodiment 2, as Figure 4 As shown, based on Example 1, the present invention provides a technical solution: Preferably, in step 4, the process of constructing the risk assessment model includes:

[0083] Based on the power distribution information sequence table, data related to the impact degree of the power outage event, the trend of environmental interference, and the user's autonomous power supply restoration assessment are extracted, and features are extracted from the acquired data. Features associated with the power outage risk are selected, including power outage frequency, duration, impact range, user feedback, and meteorological conditions, to obtain a risk feature set, and the risk feature set is divided into a training set and a test set. The risk assessment model is constructed using the training set data in combination with the neural network model, and the trained risk assessment model is verified using the test set data to assess the risk of the power outage event. Based on the assessment results of the risk assessment model, and combined with the power outage event impact index, environmental interference index, and power supply restoration index, the power outage risk coefficient is calculated. The authenticity of the power outage event is further analyzed using the trend of the power outage risk coefficient and the impact caused by the power outage event. According to the calculation results of the power outage risk coefficient and the impact and severity caused by the power outage event, the risk level of the power outage event is divided into low risk level, medium risk level, and high risk level, and the corresponding risk threshold is matched for each risk level.

[0084] Furthermore, the expression of power outage risk coefficient is:

[0085]

[0086] Among them, F is the power outage risk factor, which reflects the overall risk of a power outage event; S is the power outage event impact index, which reflects the severity of the power outage event; EI is the environmental interference index, which quantifies the potential impact of meteorological conditions on power outage events; PR is the power supply restoration index, which reflects the probability that users will restore power supply autonomously after a power outage; and w S 、w EI 、w PRare the weight coefficients of the power outage impact index, environmental interference index, and power supply restoration index, respectively, which are used to adjust the relative importance of different indexes in risk assessment. log(1+S) means using a logarithmic function to process the power outage impact index to reduce the impact of extreme values ​​and keep the value stable. Indicates that the power supply restoration index is adjusted using the Sigmoid function and limited between 0 and 1 to reflect the relative ability of users to restore power supply. The value of F is between 0 and 1, where 0 indicates no risk and 1 indicates extremely high risk. As S increases, EI improves, and PR decreases, the F value increases, indicating a higher risk level of the power outage event.

[0087] Furthermore, the multiple risk levels correspond to multiple risk thresholds, wherein the risk threshold includes an upper threshold and a lower threshold;

[0088] Multiple risk levels and multiple risk thresholds satisfy the following relationship:

[0089] Low risk level 0≤F <F M ; A power outage risk factor within this range indicates that the risk of a power outage event is low, the possible impact is small, and the self-healing ability of the power grid or the user's ability to restore power supply independently is strong;

[0090] Medium risk level F M ≤F <F H ; A power outage risk factor within this range indicates that the risk of a power outage event is medium. Certain intervention measures may be required, such as monitoring, early warning, or preparation of emergency resources;

[0091] High risk level F H ≤F≤1; A power outage risk factor within this range indicates a high risk of a power outage event. Immediate action is required, such as strengthening monitoring, preparing emergency response plans, or implementing preventive maintenance;

[0092] Among them, F is the power outage risk factor, F M is the lower threshold corresponding to the medium risk level and the upper threshold corresponding to the low risk level, F H is the lower threshold corresponding to the high risk level and the upper threshold corresponding to the medium risk level, F M =0.33, F H =0.66;

[0093] In step 5, the process of generating the power outage event risk report includes:

[0094] Configure a real-time monitoring system and integrate the trained risk assessment model into the real-time monitoring system to obtain various real-time data in the distribution network, including power outage reports, user feedback, meteorological information, and power grid operation status data. Based on the real-time monitoring system, the distribution network status is continuously tracked, and the collected real-time data is input into the risk assessment model to calculate the power outage risk coefficient, generate preliminary risk assessment results, evaluate the risk level of the power outage event, and automatically generate a risk report for the power outage event based on the risk assessment results. The risk report includes the risk level, the impact of the power outage event, and the recommended response measures. The risk report and judgment results are distributed to relevant personnel, and feedback from relevant personnel on the risk report and judgment results is collected. The effect and problems of the risk assessment model in the application are further analyzed. Based on the feedback, the real-time monitoring system, risk assessment model, and risk report are iteratively optimized to improve the accuracy and practicality of the system.

[0095] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for determining the authenticity of a power outage event in a distribution network, characterized in that: The following steps are involved: Step 1: collect raw data from different sources in the distribution network power supply system, including power outage reports, user feedback, meteorological information, power grid operation status data, and historical power outage records, and pre-process the collected raw data to obtain a distribution information sequence table; Step 2: extract abnormal features related to the power outage event from the power distribution information sequence table, preliminarily screen out the power outage areas and equipment related to the power outage event by combining time-space correlation analysis, analyze the impact of the power outage event, and analyze the environmental interference trend of the power outage event area by combining meteorological information data; Step 3: Identify the user's power supply behavior pattern, analyze historical power outage records, power usage habits and grid status, and evaluate the authenticity of the user's autonomous power supply restoration; Step 4: Based on the analysis results of the impact of the power outage, the environmental interference trend, and the evaluation results of the user's autonomous power restoration, a risk assessment model is constructed using machine learning technology to analyze the risk of the power outage to further determine the authenticity of the power outage. Step 5: Configure a real-time monitoring system to continuously track the distribution network status, generate a risk report for power outage events based on the output results of the risk assessment model, and feed back the risk report and judgment results to relevant personnel.

2. A method for determining the authenticity of a power outage event in a distribution network according to claim 1, characterized in that: In step 1, the process of obtaining the power distribution information sequence table includes: Step 101, collecting raw data from the power distribution system's power grid management system, customer service system, meteorological department, and historical database, wherein power grid operation status data is extracted from the power grid management system, user feedback and power outage reports are collected from the customer service system, meteorological information is obtained from the meteorological department, and historical power outage records and user power usage habit data are extracted from the historical database; Step 102, integrating the raw data from different sources into a unified data set, adding a common timestamp to each record in the data set, and performing preprocessing operations on the raw data, including data cleaning and data conversion steps; Step 103, classify and organize the data by time period, and establish a power distribution information sequence table based on the organized data, verify the organized power distribution information sequence table, and store the verified power distribution information sequence table and related data in the data warehouse.

3. A method for determining the authenticity of a power outage event in a distribution network according to claim 2, characterized in that: In step 2, the analysis process of the impact of the power outage event and the environmental interference trend includes: Step 201, filter out all power outage event records from the power distribution information sequence table, extract abnormal features of the power outage event for each power outage event record, including abnormal power outage time, abnormal power outage duration, abnormal expansion of power outage scope, and abnormal increase in the number of affected users, and perform preliminary analysis on the extracted abnormal features to identify corresponding abnormal patterns; Step 202, according to the timestamps of the power outage events, analyze the time distribution of the power outage events, generate a time distribution chart of the power outage events, identify the time periods of frequent power outages, use GIS technology to mark the locations of the power outage events on the map, analyze the spatial distribution of the power outage events, identify the areas with frequent power outages and adjacent areas, perform superposition analysis on the results of the time distribution and the spatial distribution, identify the power outage areas that are both frequent in time and concentrated in space, and preliminarily screen out the areas with frequent power outages based on the results of the spatiotemporal correlation analysis; Step 203, check the status records of relevant equipment in the frequent power outage area, including fault records and maintenance records, summarize and analyze the feedback from users in the frequent power outage area, and further screen out equipment directly related to the power outage event in combination with the equipment status and user feedback; Step 204, based on the duration of the power outage event in the frequent power outage area, the number of users affected by the power outage, the severity score caused by the power outage, and the equipment failure status record of the power outage, a comprehensive analysis is performed to obtain a power outage event impact index to evaluate the impact of the power outage event; Step 205, matching the meteorological information with the time point and location of the power outage event, obtaining meteorological data at the corresponding time point and location, analyzing the changing trend of the meteorological information over time, identifying the meteorological factors affecting the power outage event, calculating the environmental interference index based on the severity of the meteorological information and historical data, and quantifying the potential impact of the meteorological information on the power outage event; Step 206 , combining the environmental interference index and the weather forecast data, evaluating the interference trend of the weather information change on the power outage event, so as to identify the high-risk time periods and areas.

4. A method for determining the authenticity of a power outage event in a distribution network according to claim 3, characterized in that: The expression of the power outage event impact index is: Among them, S is the impact index of power outage event, T i is the duration of the ith power outage, T base is the baseline power outage duration, T max is the longest power outage duration during the observation period, F i is the number of users affected by the ith power outage, F max is the maximum number of users affected during the observation period, U i is the severity score of the user feedback of the i-th power outage, U base Score the baseline user feedback severity, R i is the equipment status record of the i-th power outage, and n is the total number of power outage events; The expression of the environmental interference index is: Among them, EI is the environmental interference index, M j is the current value of the jth meteorological factor, M base is the baseline value of the jth meteorological factor, M max is the maximum value of the jth meteorological factor during the observation period, w j is the weight of the jth meteorological factor, α is the attenuation coefficient, HI j is the historical impact index of the jth meteorological factor, and m is the total number of meteorological factors.

5. A method for determining the authenticity of a power outage event in a distribution network according to claim 4, characterized in that: In step 3, the process of evaluating the authenticity of the user's autonomous power restoration includes: Step 301, based on the power distribution information sequence table, extract historical power outage records and user's power usage habit data, wherein the historical power outage records include power outage time, location, duration, and cause, and the user's power usage habit data includes power consumption, peak power consumption period, and power equipment type; Step 302, combining historical power outage records and user power usage habit data, analyzing the user's power usage behavior during historical power outage events, identifying user behavior patterns, and identifying typical behaviors of users using backup power and reducing power usage during power outages; Step 303: Analyze the power supply demand of users under different circumstances in combination with historical power outage records and power consumption habit data, analyze the user's behavioral response in the power outage event, and evaluate the user's ability to restore power autonomously in combination with the capacity status of the user's backup power supply to determine whether the user has the conditions and ability to restore power autonomously; Step 304, monitoring the operation status of the power grid, including load, power generation capacity, and power grid stability, and analyzing the current load distribution, redundancy and vulnerability status of the power grid, and evaluating the self-healing capability of the power grid and the response time when a fault occurs; Step 305, applying regression analysis technology, combined with the analysis results of historical power outage records, power consumption habits and power grid status, calculates the power supply restoration index, analyzes the probability of users independently restoring power supply after a power outage, and comprehensively evaluates the authenticity and stability of users' autonomous power supply restoration.

6. A method for determining the authenticity of a power outage event in a distribution network according to claim 5, characterized in that: The expression of the power restoration index is: Among them, PR is the power restoration index, T i is the duration of the ith power outage, T avg is the average duration of all power outage events, n is the total number of power outage events, D i is the power consumption during the i-th power outage, D avg is the average power consumption of all power outage events, w i is the weight of the ith power outage, and β and δ are adjustment indexes.

7. A method for determining the authenticity of a power outage event in a distribution network according to claim 6, characterized in that: In step 4, the process of constructing the risk assessment model includes: Step 401, extracting data related to the impact of power outage events, environmental interference trends, and user autonomous power restoration assessment based on the power distribution information sequence table, and performing feature extraction on the acquired data, selecting features associated with power outage risks, including power outage frequency, duration, impact range, user feedback, and meteorological conditions, to obtain a risk feature set; Step 402, dividing the risk feature set into a training set and a test set, using the training set data in combination with the neural network model to build a risk assessment model, and using the test set data to verify the trained risk assessment model to assess the risk of the power outage event; Step 403, based on the evaluation results of the risk evaluation model, and in combination with the power outage event impact index, the environmental interference index and the power supply restoration index, a power outage risk coefficient is calculated, and the authenticity of the power outage event is further analyzed using the trend of the power outage risk coefficient and the impact caused by the power outage event; Step 404, based on the calculation result of the power outage risk coefficient and the impact and severity of the power outage event, the risk level of the power outage event is divided into low risk level, medium risk level and high risk level, and a corresponding risk threshold is matched for each risk level.

8. A method for determining the authenticity of a power outage event in a distribution network according to claim 7, characterized in that: The expression of the power outage risk coefficient is: Among them, F is the power outage risk factor, S is the power outage event impact index, EI is the environmental interference index, PR is the power supply recovery index, and w S 、w EI 、w PR They are the weight coefficients of power outage impact index, environmental interference index and power supply restoration index respectively.

9. A method for determining the authenticity of a power outage event in a distribution network according to claim 8, characterized in that: The plurality of risk levels correspond to the plurality of risk thresholds, wherein the risk thresholds include an upper threshold and a lower threshold; The multiple risk levels and the multiple risk thresholds satisfy the following relationship: Low risk level 0≤F <F M ; Medium risk level F M ≤F <F H ; High risk level F H ≤F≤1; Among them, F is the power outage risk factor, F M is the lower threshold corresponding to the medium risk level and the upper threshold corresponding to the low risk level, F H is the lower threshold corresponding to the high risk level and the upper threshold corresponding to the medium risk level, F M =0.33, F H =0.

66.

10. A method for determining the authenticity of a power outage event in a distribution network according to claim 9, characterized in that: In step 5, the process of generating the power outage event risk report includes: Step 501, configure a real-time monitoring system, and integrate the trained risk assessment model into the real-time monitoring system to obtain various real-time data in the distribution network, including power outage reports, user feedback, meteorological information, and power grid operation status data; Step 502, based on the real-time monitoring system, continuously tracking the distribution network status, inputting the collected real-time data into the risk assessment model, calculating the power outage risk coefficient, and generating a preliminary risk assessment result to assess the risk level of the power outage event; Step 503, based on the risk assessment results, automatically generate a risk report of the power outage event, the risk report includes the risk level, the impact of the power outage event, and the recommended response measures, and distribute the risk report and the judgment results to relevant personnel; Step 504, collect feedback from relevant personnel on the risk report and judgment results, further analyze the effect and existing problems of the risk assessment model in the application, and iteratively optimize the real-time monitoring system, risk assessment model, and risk report based on the feedback.

Images