Power grid anomaly analysis method and device, computer equipment and computer readable storage medium

Abstract

The application relates to a power grid anomaly analysis method and device, computer equipment and a computer readable storage medium. The method comprises the following steps: based on power grid time sequence data of a target power grid branch line, calculating power grid anomaly indexes corresponding to each sampling time; the power grid time sequence data comprises sampling data corresponding to each sampling time; identifying a power grid anomaly event occurring to the target power grid branch line according to the power grid anomaly indexes; determining an anomaly form of the target power grid branch line based on power grid anomaly events occurring at each sampling time within each time scale; the anomaly form comprises instantaneous disturbance anomaly, periodic anomaly or permanent anomaly. By accurately locating whether the anomaly form of the target power grid branch line belongs to instantaneous disturbance anomaly, periodic anomaly or permanent anomaly, false alarms or missed alarms are avoided, a solid data foundation is provided for subsequent differentiated operation and maintenance strategies, and the accuracy of power grid anomaly analysis and the safety and stability of power grid operation are improved.

Description

Technical Field

[0001] This application relates to the field of power grid technology, and in particular to a power grid anomaly analysis method, apparatus, computer equipment, and computer-readable storage medium. Background Technology

[0002] With the large-scale integration of distributed photovoltaic power, electric vehicle charging stations, and single-phase high-power residential / industrial loads into low-voltage distribution networks, grid anomalies such as heavy overload of low-voltage branch lines or three-phase imbalance may occur. These grid anomalies can be categorized into three types: transient, periodic, and permanent, each requiring different handling strategies.

[0003] However, current power grid anomaly analysis methods rely on power grid operation and maintenance data at a single time scale, which makes it impossible to distinguish anomaly patterns from multiple different time scales. This can easily lead to missed alarms and false alarms, thereby affecting the safe and stable operation of the power grid. Summary of the Invention

[0004] Therefore, it is necessary to provide a power grid anomaly analysis method, device, computer equipment, and computer-readable storage medium to address the aforementioned technical problems, so as to distinguish anomaly patterns from multiple different time scales, avoid missed alarms and false alarms, and achieve safe and stable operation of the power grid.

[0005] Firstly, this application provides a method for power grid anomaly analysis, wherein:

[0006] Based on the power grid time series data of the target power grid branch, calculate the power grid anomaly index corresponding to each sampling time; the power grid time series data includes sampling data corresponding to multiple sampling times respectively.

[0007] Based on the power grid anomaly indicators, identify the power grid anomaly events occurring on the target power grid branch line;

[0008] Based on the power grid anomaly events occurring at each sampling time within each time scale, the anomaly pattern of the target power grid branch is determined; the anomaly pattern includes transient disturbance anomaly, periodic anomaly, or permanent anomaly.

[0009] In one embodiment, each sampled data point includes a rated current value and a measured operating current value for each phase circuit branch; the power grid anomaly indicators include low-voltage branch overload indicators and / or three-phase imbalance indicators; the power grid anomaly indicators calculated based on the power grid time-series data of the target power grid branch at each sampling time include at least one of the following:

[0010] For each sampling time, determine the maximum value of the measured operating current, calculate the ratio of the maximum value to the rated current value, and obtain the low-voltage branch line overload index at the sampling time;

[0011] For each sampling time, the maximum, minimum, and average values ​​of the measured operating current are determined. The difference between the maximum and minimum values ​​is calculated, and the ratio of the difference to the average value is determined to obtain the three-phase imbalance index at the sampling time.

[0012] In one embodiment, determining the abnormal morphology of the target power grid branch based on the power grid anomaly events occurring at each of the sampling times within each time scale includes:

[0013] If at least one of the aforementioned power grid anomalies occurs within a single day, and the power grid anomalies occurring within a monthly, quarterly, or annual timescale do not meet the preset anomaly conditions, then the anomaly pattern of the target power grid branch is determined to be a transient disturbance anomaly.

[0014] In one embodiment, determining the abnormal morphology of the target power grid branch based on the power grid anomaly events occurring at each of the sampling times within each time scale includes:

[0015] If the abnormal power grid events occurring within a single day's time scale meet a preset regularity condition, the abnormal power grid events occurring within a monthly time scale meet a preset concentration condition, and the abnormal power grid events occurring within a quarterly and annual time scale meet a preset periodicity condition, then the abnormal pattern of the target power grid branch line is determined to be a periodic anomaly.

[0016] In one embodiment, determining the abnormal morphology of the target power grid branch based on the power grid anomaly events occurring at each of the sampling times within each time scale includes:

[0017] If the abnormal power grid events occurring within a single-day, monthly, quarterly, and annual timescale all meet the preset persistence conditions, then the abnormal morphology of the target power grid branch is determined to be a permanent abnormality.

[0018] In one embodiment, the method further includes:

[0019] Based on the described anomaly pattern and pre-acquired power grid operation and maintenance data, analyze the cause of the anomaly in the target power grid branch; and / or,

[0020] Based on the power grid time series data, the abnormal change trend of the target power grid branch line is deduced.

[0021] In one embodiment, the method further includes:

[0022] Based on the power grid anomaly indicators, the power grid anomaly events, and the anomaly patterns, the risk priority of the target power grid branch line is determined;

[0023] According to the risk priority, anomaly handling is performed on each of the target power grid branches in sequence.

[0024] Secondly, this application also provides a power grid anomaly analysis device, the device comprising:

[0025] The indicator calculation module is used to calculate the power grid anomaly indicators corresponding to each sampling time based on the power grid time series data of the target power grid branch; the power grid time series data includes sampling data corresponding to multiple sampling times respectively.

[0026] An anomaly identification module is used to identify power grid anomaly events occurring on the target power grid branch line based on the power grid anomaly indicators.

[0027] The morphology analysis module is used to determine the abnormal morphology of the target power grid branch based on the power grid abnormal events that occur at each of the sampling times within each time scale; the abnormal morphology includes transient disturbance abnormalities, periodic abnormalities, or permanent abnormalities.

[0028] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps described in any of the above method embodiments.

[0029] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps described in any of the above method embodiments.

[0030] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps described in any of the above method embodiments.

[0031] The aforementioned power grid anomaly analysis method, device, computer equipment, and computer-readable storage medium can calculate the power grid anomaly indicators corresponding to each sampling moment using the power grid time-series data of the target power grid branch line. This allows for the accurate identification of power grid anomaly events. Furthermore, statistical analysis of the occurring power grid anomaly events is performed at different time scales to accurately determine whether the anomaly of the target power grid branch line is a transient disturbance anomaly, a periodic anomaly, or a permanent anomaly. This significantly reduces false alarms caused by sudden data changes at a single moment and missed alarms caused by persistent minor exceedances that are not captured by single-point thresholds. This provides a solid data foundation for the subsequent development of differentiated operation and maintenance strategies, thereby significantly improving the accuracy of power grid anomaly analysis and the safety and stability of power grid operation. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 Here is a flowchart of a power grid anomaly analysis method in one embodiment;

[0034] Figure 2 This is a flowchart illustrating the steps of a power grid anomaly analysis method in one embodiment;

[0035] Figure 3 This is a structural block diagram of a power grid anomaly analysis device in one embodiment;

[0036] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0038] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0039] The power grid anomaly analysis method provided in this application can be applied to servers with data computing and processing capabilities. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud servers, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The server can be a backend server for the aforementioned target application, used to provide backend services to the client of the target application.

[0040] Specifically, such as Figure 1 As shown, this power grid anomaly analysis method can implement the following process:

[0041] Step 101: Based on the power grid time series data of the target power grid branch line, calculate the power grid anomaly index corresponding to each sampling time; the power grid time series data includes sampling data corresponding to multiple sampling times.

[0042] The power grid time series data includes multiple sample data arranged in time sequence. Each sample data has a corresponding sampling time. These sample data are continuously collected and uploaded by smart meters, branch monitoring terminals and other data acquisition devices deployed in the low-voltage distribution network according to a preset sampling period (such as one sampling point every 15 minutes).

[0043] After receiving the power grid time-series data, the server calculates the power grid anomaly indicators for each sampling time according to preset power grid anomaly indicator calculation rules. Specifically, the power grid anomaly indicators include at least the load factor and the three-phase current imbalance. Through the above calculations, the power grid time-series data is transformed into quantifiable power grid anomaly indicators, providing a unified basis for subsequent anomaly event identification.

[0044] Step 102: Identify the abnormal power grid events occurring on the target power grid branch line based on the power grid anomaly indicators.

[0045] The server compares the power grid anomaly indicators calculated at each sampling time with preset anomaly judgment thresholds. When a power grid anomaly indicator at a certain sampling time exceeds its corresponding anomaly judgment threshold, it is identified as a power grid anomaly event at that time, and the type and occurrence time of the power grid anomaly event are recorded simultaneously; otherwise, if it does not exceed the threshold, it is considered that the operation is normal at that time. Through point-by-point comparison and judgment process, it is possible to determine whether a power grid anomaly event has occurred on the target power grid branch at each sampling time.

[0046] For example, grid anomaly indicators may include load rate. When the load rate at a certain sampling time is greater than 80%, a grid anomaly event of overload anomaly is determined to have occurred; when it is greater than 100%, a grid anomaly event of overload anomaly is determined to have occurred. Grid anomaly indicators may also include three-phase current imbalance indicators. When the three-phase current imbalance at a certain sampling time is greater than 15%, a grid anomaly event of three-phase imbalance anomaly is determined to have occurred.

[0047] Step 103: Based on the abnormal power grid events that occur at each sampling time within each time scale, determine the abnormal morphology of the target power grid branch line; the abnormal morphology includes transient disturbance anomalies, periodic anomalies, or permanent anomalies.

[0048] The server can perform statistical analysis on discrete power grid anomalies at multiple preset time scales. It can statistically analyze the number, distribution, and clustering of power grid anomalies occurring at each sampling moment within each time scale, and then determine the anomaly pattern of the target power grid branch based on the distribution differences or correlations of the anomalies. Each time scale contains several sampling moments, and the time scale can include, but is not limited to, daily, monthly, quarterly, and annual time scales.

[0049] For example, if an anomalous event occurs only on a single-day timescale and does not meet the anomalous conditions on a longer timescale, it is determined to be a transient disturbance anomaly; if an anomalous event exhibits regular repetition on both the single-day and monthly timescales and matches a known cycle, but does not have year-round continuity on the annual timescale, it is determined to be a periodic anomaly; if an anomalous event occurs continuously on the daily, monthly, quarterly, and annual timescales and meets its respective anomalous conditions, it is determined to be a permanent anomaly.

[0050] As can be seen from the above, the solution provided in this application can calculate the power grid anomaly indicators corresponding to each sampling time using the power grid time series data of the target power grid branch, thereby accurately identifying power grid anomaly events. Furthermore, statistical analysis of the power grid anomaly events occurring at different time scales can accurately determine whether the anomaly of the target power grid branch is a transient disturbance anomaly, a periodic anomaly, or a permanent anomaly. This significantly reduces false alarms caused by sudden changes in data at a single moment, as well as missed alarms caused by continuous minor exceedances not being captured by single-point thresholds. This provides a solid data foundation for the subsequent formulation of differentiated operation and maintenance strategies, thereby significantly improving the accuracy of power grid anomaly analysis and the safety and stability of power grid operation.

[0051] In an exemplary embodiment, each sampled data includes a rated current value and a measured operating current value for each phase circuit branch; the power grid anomaly indicators include low-voltage branch overload indicators and / or three-phase imbalance indicators; step 101 includes at least one of the following:

[0052] For each sampling time, determine the maximum value of the measured operating current, calculate the ratio of the maximum value to the rated current value, and obtain the low-voltage branch line overload index at the sampling time;

[0053] For each sampling time, determine the maximum, minimum, and average values ​​of the measured operating current, calculate the difference between the maximum and minimum values, and determine the ratio of the difference to the average value to obtain the three-phase unbalance index at the sampling time.

[0054] Specifically, the rated current value is the rated current carrying capacity of the cables or switches used in the target low-voltage branch line, which is an inherent attribute parameter of the line; the measured operating current value of each phase circuit branch, including the actual operating current value of phase A, phase B and phase C of the line at the current sampling time, is synchronously collected and uploaded by the front-end acquisition device according to the preset sampling period.

[0055] Power grid anomaly indicators include at least one of the following: low-voltage branch line overload indicator and three-phase imbalance indicator. The low-voltage branch line overload indicator is used to characterize whether the current load pressure of the line is close to or exceeds its design capacity, while the three-phase imbalance indicator is used to characterize the degree of amplitude deviation between the three-phase currents.

[0056] When calculating the heavy overload index of a low-voltage branch line, firstly, the maximum value is selected from the measured operating current values ​​of phases A, B, and C. Then, the ratio of this maximum value to the rated current value is calculated, and this ratio is expressed as a percentage. The result is the heavy overload index of the low-voltage branch line at that sampling time. This index reflects the proportion of the maximum current load borne by the line at that moment to its rated current carrying capacity; the higher the value, the greater the pressure the line bears. This index numerically represents the ratio of the current maximum phase current to the rated current. For example, a ratio of 0.85 indicates that the maximum phase current is 85% of the rated current, and a ratio of 1.10 indicates that the maximum phase current is 110% of the rated current. The calculation process of the above-mentioned heavy overload index of the low-voltage branch line can be expressed as follows:

[0057]

[0058] in, For low-voltage branch line heavy overload indicators; This represents the maximum value of the measured operating current. This is the rated current value.

[0059] When calculating the three-phase unbalance index, the maximum and minimum values ​​are selected from the measured operating current values ​​of phases A, B, and C collected at each sampling time, and the arithmetic mean of the three phase current values ​​is calculated. Then, the difference between the maximum and minimum values ​​is calculated, and this difference is divided by the average of the three-phase currents. The resulting ratio is expressed as a percentage, and the result is the three-phase unbalance index at that sampling time. This index reflects the degree of amplitude deviation between the three-phase currents; a higher value indicates a more unbalanced three-phase load distribution and a greater risk of zero-sequence current borne by the neutral line. The calculation process for the three-phase current unbalance can be expressed as follows:

[0060]

[0061] in, This refers to the three-phase current imbalance. The maximum value among the measured operating current values ​​of phases A, B, and C; This is the minimum value among the measured operating current values ​​of phases A, B, and C; This represents the average of the measured operating current values ​​for phases A, B, and C.

[0062] By standardizing the calculation of two types of power grid anomaly indicators, the degree of anomaly of line overload and three-phase imbalance at each sampling time can be stably and accurately quantified. This provides a unified and standardized quantitative judgment basis for subsequent power grid anomaly event identification and multi-scale anomaly pattern classification, reduces the data dependence and computational complexity of the indicator calculation process, and ensures the objectivity and accuracy of subsequent anomaly analysis results.

[0063] In one exemplary embodiment, step 103 includes:

[0064] If at least one power grid anomaly occurs within a single day, and the power grid anomalies occurring within a monthly, quarterly, or annual timescale do not meet the preset anomaly conditions, then the anomaly pattern of the target power grid branch is determined to be an instantaneous disturbance anomaly.

[0065] Specifically, after completing the statistical analysis at each time scale, the server first determines whether at least one power grid anomaly event occurred on the target power grid branch within a single day. If at least one sampling moment within a single day is marked as a power grid anomaly event, it indicates that the line experienced a short-term exceedance on that day.

[0066] Then, the server can determine whether the statistical results meet the corresponding preset anomaly conditions on the monthly, quarterly, and annual time scales. If they do not meet the conditions, that is, if the proportion of abnormal days to total days on the monthly time scale does not reach the monthly time scale threshold, the total number of abnormal days or abnormal months on the quarterly time scale does not reach the quarterly time scale threshold, or the abnormal event does not show a continuous distribution characteristic throughout the year or the total number of abnormal days on the annual time scale does not reach the annual time scale threshold, it indicates that although the abnormal event occurred on a single day, it does not have statistical significance or regular repetitive characteristics on a longer time scale.

[0067] In this situation, the server determines that the abnormality of the target power grid branch is a transient disturbance anomaly, indicating that the anomaly is caused by an occasional, short-term temporary load impact or transient power grid interference, and is not due to a structural or persistent defect in the line itself.

[0068] In this way, occasional, short-term temporary load disturbances can be accurately distinguished from periodic, regular hidden dangers and permanent hard faults, avoiding the misjudgment of short-term daily fluctuations as structural faults and triggering invalid alarms, effectively reducing the false alarm rate, thereby improving the accuracy of abnormal pattern judgment and the reliability of operation and maintenance decisions.

[0069] In one exemplary embodiment, step 103 includes:

[0070] If the abnormal power grid events occurring within a single day's time scale meet the preset regularity conditions, the abnormal power grid events occurring within a monthly time scale meet the preset concentration conditions, and the abnormal power grid events occurring within a quarterly and annual time scale meet the preset periodicity conditions, then the abnormal pattern of the target power grid branch line is determined to be a periodic anomaly.

[0071] Specifically, the first step is to determine whether the abnormal power grid events occurring on the target power grid branch within a single day's timescale meet a preset regularity condition. This preset regularity condition means that the abnormal event repeats itself within the same or similar time periods over multiple natural days, rather than occurring sporadically on a single isolated date. For example, the server records the specific time periods of abnormal events each day. If the line experiences excessive load during the evening peak hours for several consecutive workdays, while remaining normal during other times, then it is determined to meet the regularity condition on a single day's timescale. This condition is designed to capture repetitive anomalies with stable intraday fluctuation patterns, distinguishing them from randomly occurring transient disturbances.

[0072] Based on this, the server further determines whether the abnormal power grid events occurring within the monthly timescale meet a preset concentration condition. The preset concentration condition refers to the distribution of abnormal events within the month exhibiting significant temporal clustering; that is, abnormal days are concentrated in specific periods of the month (such as the first or last week of the month), or the frequency of abnormal days is not uniformly distributed along the timeline but concentrated within a continuous time interval. The server calculates the dispersion of abnormal day distribution by statistically analyzing the occurrence of abnormal events on each natural day of the month. If the distribution of abnormal days on the monthly timeline meets the preset concentration criteria (such as abnormal days concentrated within 7 consecutive days, or the proportion of abnormal days to the total number of abnormal days in the month exceeding a preset percentage), then the concentration condition on the monthly timescale is deemed met. This condition aims to distinguish whether anomalies are concentrated in a specific period of a month with high frequency or are evenly distributed throughout the month.

[0073] Furthermore, the server can determine whether power grid anomalies occurring on quarterly and annual timescales meet preset periodicity conditions. These preset periodicity conditions refer to a verifiable temporal match between the peak periods of anomalies and known periodic patterns. Specifically, the server retrieves a preset periodic feature library, which includes at least the time intervals of seasonal changes, statutory holidays, and typical industrial and commercial production cycles (such as quarterly peak production periods in manufacturing). The server compares the quarter or month in which the anomalies occur with the time intervals in the aforementioned periodic feature library. If the peak periods of anomalies fall within or largely overlap with any of these periodic intervals, and this match is repeated in consecutive years, then the periodicity conditions on the quarterly and annual scales are met. This condition aims to identify the inherent correlation between anomalies and seasonal temperature changes, holiday load fluctuations, and production scheduling patterns.

[0074] When all three conditions mentioned above are met—namely, when the daily timescale exhibits regular repetition, the monthly timescale shows periodic concentration, and the quarterly and annual timescales match known cycles—the server determines that the abnormal pattern of the target power grid branch is a periodic anomaly. This indicates that the anomaly is a recurring hidden danger, and its cause is closely related to medium- and long-term factors such as seasonal electricity consumption, holiday load changes, or cyclical industrial production scheduling. In this way, periodic anomalies can be accurately distinguished from random, transient disturbances and year-round permanent faults, achieving precise identification of recurring hidden dangers.

[0075] In one exemplary embodiment, step 103 includes:

[0076] If the abnormal power grid events occurring within a single day, monthly, quarterly, and annual time scale all meet the preset persistence conditions, then the abnormal morphology of the target power grid branch line is determined to be a permanent abnormality.

[0077] Specifically, after completing statistical analysis at each time scale, the server determines whether the abnormal power grid events occurring on the target power grid branch line at the daily, monthly, quarterly, and annual time scales all meet preset persistence conditions. The preset persistence conditions refer to the following: within the corresponding time scale, the frequency or proportion of abnormal power grid events reaches or exceeds the preset persistence threshold for that scale, and the abnormal distribution does not exhibit significant intermittent or periodic regularity, but rather shows continuous existence or high-frequency recurrence on the time axis. The server performs the determination for each scale one by one, and only when all four scales meet the persistence conditions is a permanent anomaly determination triggered.

[0078] For example, on a daily timescale, the server statistically analyzes the occurrence of power grid anomalies at each sampling time point of the day. The preset persistence condition on the daily timescale is as follows: multiple sampling times within a day are marked as power grid anomalies, and the proportion of the number of anomaly sampling points to the total number of sampling points for the day exceeds the preset daily persistence threshold. For instance, if the number of sampling points per day is 96 and the preset persistence threshold is 50%, then at least 48 sampling times must show anomalies for the day to be considered to meet the persistence condition on the daily timescale. This condition ensures that the anomalies of the day are not sporadic and transient, but rather substantial exceedances over a relatively long period.

[0079] On a monthly timescale, the server counts the number of days marked as abnormal within a month and calculates the proportion of these days to the total number of days in the month. The preset persistence condition at the monthly scale is that this proportion exceeds a preset monthly persistence threshold (e.g., greater than 50%), meaning that more than half of the days in the month experience abnormal power grid events, and these abnormal days are relatively evenly distributed throughout the month, rather than being concentrated in a specific period. This condition ensures that anomalies have a persistent characteristic at the monthly level, rather than occurring only within a short period of the month.

[0080] On a quarterly timescale, the server counts the total number of abnormal days or months within each month of the quarter and calculates their proportion to the total number of days or months in the quarter. The preset persistence condition at the quarterly scale is as follows: if this proportion exceeds a preset quarterly persistence threshold (e.g., greater than 50%), meaning that more than half of the days or months in the quarter have power grid anomalies. This condition ensures that anomalies have a sustained distribution across months throughout the quarter, rather than being concentrated in a single month.

[0081] On an annual timescale, the server counts the total number of abnormal days in each month of the year and calculates its proportion of the total number of days in the year. The preset persistence condition at the annual scale is as follows: if this proportion exceeds the preset annual persistence threshold (e.g., greater than 40%), it means that a significant portion of the days in the year are in an abnormal state, and the distribution of anomalies across months does not exhibit significant seasonal gaps or periodic intermittent declines, but rather shows a persistent characteristic across quarters and seasons. This condition ensures that the anomalies are long-term, persistent problems at the annual level, rather than existing only in specific seasons or specific production cycles.

[0082] When all four time scales mentioned above meet their respective preset persistence conditions, the server determines that the abnormal morphology of the target power grid branch is a permanent anomaly. This indicates that the anomaly is a persistent problem caused by inherent defects in the line itself or unreasonable structural configuration. Its causes are closely related to permanent factors such as undersized line diameter, severely uneven three-phase load distribution, wiring errors, and equipment aging. In this way, permanent fault-type anomalies can be accurately distinguished from transient disturbances and periodic fluctuations, avoiding misjudgments caused by short-term data fluctuations or seasonal high loads.

[0083] In one exemplary embodiment, the method further includes:

[0084] Based on the abnormal patterns and pre-acquired power grid operation and maintenance data, analyze the causes of the abnormalities in the target power grid branch lines; and / or, based on power grid time series data, deduce the abnormal change trends of the target power grid branch lines.

[0085] Specifically, based on the determined anomaly type and pre-acquired power grid operation and maintenance data, the server automatically invokes differentiated source tracing analysis logic matching the anomaly type to analyze the cause of the anomaly in the target power grid branch. The pre-acquired power grid operation and maintenance data includes at least equipment ledger data (such as line diameter, commissioning year, switch model, assigned transformer area, and power user type), environmental meteorological data (such as daily average temperature and extreme weather records), distributed energy access data (such as photovoltaic and charging pile access capacity and operating periods), and historical fault maintenance logs. After the source tracing analysis is completed, the server saves the analyzed root cause of the anomaly and corresponding supporting data, forming a complete source tracing result set containing the anomaly type, anomaly type, root cause, and supporting data.

[0086] For example, for transient disturbance-type anomalies, the server retrieves short-term load data, user fault reporting logs, and real-time meteorological data during the period in which the anomaly occurred and cross-compares them to identify whether it is caused by factors such as the activation of temporary high-power equipment, temporary power supply at the construction site, or short-term voltage fluctuations.

[0087] In response to periodic anomalies, the server integrates monthly and quarterly load change curves, full-cycle meteorological time-series data, transformer area user attributes, and holiday operation records. It then performs time-series matching between periods of high anomaly occurrence and seasonal temperature changes, factory production schedules, residential electricity consumption patterns, and statutory holidays to identify whether the anomalies are caused by seasonal load fluctuations or regular factors such as industrial and commercial periodic production.

[0088] For permanent anomalies, the server retrieves basic line ledgers, three-phase current steady-state operation data, distributed energy access information, and maintenance and fault records from previous years to conduct a comprehensive analysis: For heavy overload issues, by comparing the line's rated current carrying capacity with the long-term average load and considering the equipment's years of operation, it determines whether there are hard issues such as insufficient wire diameter, line aging, or overall overload; for three-phase imbalance issues, by checking the distribution of three-phase users, it determines whether there are structural issues such as uneven distribution of single-phase load, incorrect three-phase wiring, or concentrated single-phase access of distributed energy, and combines historical fault data to determine whether the anomaly is a recurring fault.

[0089] In an exemplary embodiment, the server can also use a preset time-series extrapolation algorithm based on the power grid time-series data to extrapolate the future trends of abnormal indicators of the target power grid branch. Specifically, the server takes the historical load rate and three-phase current imbalance of the line in the multi-period time-series dataset as input, extracts its change trajectory at daily, monthly, quarterly, and annual time scales, fits the evolution law of the abnormal indicators of the line over time, and extrapolates forward for a preset period (such as the next 1 to 3 months) to predict the possible trend of abnormal indicators of the line within the preset period.

[0090] For example, based on the extrapolation results, the server categorizes abnormal trends into three types: continuously deteriorating, basically stable, and self-resolving. If the extrapolated curve shows that the abnormal indicator will continue to rise in the future, it is marked as continuously deteriorating; if the abnormal indicator fluctuates slightly around the current level in the future, without a significant upward or downward trend, it is marked as basically stable; if the abnormal indicator shows a gradual decline and tends towards the normal range in the future, it is marked as self-resolving. After completing the trend extrapolation, the server persistently stores the extrapolation results, along with anomaly details, classification conclusions, source tracing conclusions, and a governance priority list, in the backend database and pushes a complete analysis report to the frontend for maintenance personnel to refer to.

[0091] By further expanding the analysis of abnormal causes and the deduction of abnormal change trends based on the judgment of abnormal patterns, maintenance personnel not only know what kind of abnormality exists now, but also know why the abnormality is caused and how it will develop in the future. This allows them to proactively arrange pre-maintenance before the abnormality worsens, eliminating the accident in the bud and truly realizing the transformation from passive alarm to proactive early warning.

[0092] In one exemplary embodiment, the method further includes:

[0093] Based on power grid anomaly indicators, power grid anomaly events, and anomaly patterns, determine the risk priority of the target power grid branch line;

[0094] According to risk priority, anomalies are handled sequentially for each target power grid branch.

[0095] Specifically, the server can invoke a pre-defined multi-dimensional risk scoring model to quantitatively score each low-voltage branch line exhibiting anomalies from multiple evaluation dimensions. These evaluation dimensions include at least the anomaly type, anomaly severity, duration, impact range, and recurrence frequency. The server then weights and sums the scores from each dimension according to pre-defined weights to obtain the total risk score for the line.

[0096] In the anomaly type dimension, which has a maximum score of 30 points, the server is scored based on the anomaly form determined in step 103: 30 points for a permanent anomaly, 18 points for a periodic anomaly, and 5 points for a transient disturbance anomaly. Permanent anomalies receive the highest score because they are caused by inherent defects in the line itself and require priority handling; transient disturbance anomalies receive the lowest score because they are occasional and temporary events with relatively low urgency for maintenance.

[0097] In the anomaly severity dimension, which has a maximum score of 25 points, the server assigns a score based on the specific values ​​of load rate and three-phase current imbalance calculated in step 101. The anomaly severity scoring rule is as follows: the higher the load rate or three-phase imbalance, the greater the electrical stress currently borne by the line, and the more serious the impact on equipment safety and power supply quality, resulting in a higher score. Specifically, the server uses the more severe exceedance of either the load rate or the three-phase imbalance as the scoring basis, mapping the measured exceedance value to a score between 0 and 25 points according to a preset severity grading scoring table. The greater the exceedance, the higher the score, thus quantifying the immediate hazard of the anomaly to the safe operation of the line.

[0098] In the duration dimension, which has a maximum score of 20 points, the server assigns scores based on the distribution of power grid anomalies across various time scales as described in step 102. The duration scoring rule is as follows: the more time scales an anomaly spans—that is, extending from a single-day time scale to a monthly time scale, and then to a quarterly time scale—the longer the anomaly persists, the more deeply entrenched it becomes, and the greater the threat to the long-term safe operation of the line, resulting in a higher score. For example, transient disturbances that only occur at a single-day time scale score low in this dimension; periodic anomalies that span both single-day and monthly time scales score medium in this dimension; and permanent anomalies that span four scales—daily, monthly, quarterly, and yearly—score the highest in this dimension. The server counts the actual number of time scales covered by the anomaly and maps the number of covered scales to a score between 0 and 20 points according to a preset duration segmentation scoring table.

[0099] In the impact scope dimension, which has a maximum score of 15 points, the server assigns a score based on the number and type of users supplied by the line, as determined by pre-acquired equipment ledger data. The impact scope scoring rules are as follows: the more users supplied, the more end-users affected when an anomaly occurs, and the greater the social impact, the higher the score. Additionally, if the load supplied by the line involves important loads such as industrial production and public welfare (e.g., hospitals, schools, residential elevators, and water supply facilities), extra points are awarded on top of the user quantity score. The server comprehensively considers the range of user quantity and user type weights for the line, calculating a score between 0 and 15 points according to a pre-defined impact scope scoring table to quantify the actual breadth of the anomaly's impact on the user side.

[0100] In the recurrence frequency dimension, with a maximum score of 10, the server assigns a score based on the historical occurrence records of similar anomalies on the line from pre-acquired historical fault operation and maintenance data. The scoring rule for recurrence frequency is: the more times the same type of anomaly event recurs in the same historical period, the more likely the anomaly is to recur, and simple temporary handling is insufficient to eradicate it, requiring systematic rectification; the corresponding score is higher. The server counts the total number of similar anomalies occurring on the line in the past year or a preset historical period, and maps the number of recurrences to a score between 0 and 10 according to a preset recurrence frequency tiered scoring table.

[0101] Subsequently, the server sorts all abnormal lines in descending order of their total risk score and assigns corresponding governance levels based on the score range. In an exemplary embodiment, the governance levels include three levels: Emergency, Important, and General. The Emergency level corresponds to the highest range of total risk scores, indicating that the anomaly requires immediate governance; the Important level corresponds to the middle range of total risk scores, indicating that the anomaly requires governance in the near future; and the General level corresponds to the lowest range of total risk scores, indicating that the anomaly can be included in the regular maintenance plan for governance. After completing the above scoring and sorting, the server generates a standardized governance priority list, which includes at least the following fields: line number, corresponding transformer area, anomaly type, anomaly morphology, total risk score, governance level, and suggested processing timeframe.

[0102] Furthermore, anomalies can be handled sequentially for each target power grid branch line according to risk priority. The server pushes the aforementioned governance priority list to the front-end client, allowing maintenance personnel to carry out on-site anomaly handling work in the order of priority in the list. After logging into the front-end client, maintenance personnel can view the list of anomaly lines to be handled by default in order of governance level from high to low and total risk score from large to small, and arrange on-site investigation and governance for each one.

[0103] By constructing a multi-dimensional risk scoring model, it is ensured that high-risk, high-impact permanent faults and periodic hidden dangers with high recurrence frequency can be prioritized for treatment. This enables the system to continuously improve the accuracy of analysis and the effectiveness of treatment recommendations during long-term operation, thereby effectively enhancing the level of refined management of power distribution network operation and maintenance.

[0104] like Figure 2 As shown below, the power grid anomaly analysis method provided in this application is described through a specific implementation. The method includes the following steps:

[0105] Step S1: Data Acquisition and Preprocessing. The server synchronizes multi-source data, including full-cycle raw operation data, equipment ledger data, environmental meteorological data, and fault operation and maintenance data, from multiple external business systems according to preset interfaces and acquisition rules. Then, the raw multi-source data is cleaned, including standardized processing procedures such as filling in missing values, filtering outliers, and normalization, thereby eliminating invalid and dirty data, unifying data format and units, and generating power grid time series data that can be used for subsequent calculations.

[0106] Step S2: Multi-timescale anomaly classification and identification. After receiving the power grid time series data output in step S1, the server batch calculates the two core power grid anomaly indicators, namely the load rate and three-phase current imbalance, for each low-voltage branch line. Then, it combines the four time scales of daily, monthly, quarterly, and yearly to carry out linkage judgment, distinguish the anomaly form, and complete the marking. This solves the problem of missed alarms and false alarms that are easy to occur in single time scale judgment.

[0107] Among them, algorithms for multi-timescale linkage anomaly classification include:

[0108] 1) Transient disturbance anomalies: The indicators exceed the standard only during certain periods of a single day, while the monthly, quarterly, and annual statistics are all normal. The causes are mostly temporary high-power load start-ups and shutdowns, and transient grid interference, which are low-risk temporary anomalies.

[0109] 2) Periodic anomalies: Daily indicators regularly exceed the standard, monthly indicators show high incidence at fixed times, and quarterly / annual indicators are strongly correlated with seasons, holidays, and production cycles (such as summer air conditioning load and seasonal factory production), which are medium- to long-term regular hidden dangers.

[0110] 3) Permanent anomalies: Indicators consistently exceed standards across daily, monthly, quarterly, and yearly time scales, with no significant differences over time periods. The causes are mostly insufficient line diameter, incorrect three-phase wiring, equipment aging, and unreasonable load configuration in the transformer area, which are high-risk hard faults.

[0111] Step S3: Multi-factor joint tracing. The server reads the anomaly type data marked in step S2, and automatically matches differentiated tracing analysis logic according to three different anomaly classifications: transient disturbance anomaly, periodic anomaly, and permanent anomaly. It combines multi-dimensional factors such as equipment ledger, user attributes, meteorological information, distributed energy operation data, and historical maintenance records to carry out correlation analysis, automatically locate the root cause of the anomaly, and simultaneously retain the tracing evidence.

[0112] For example, the source tracing logic for different anomaly types includes:

[0113] 1) Instantaneous disturbance anomaly: Retrieve short-term load data, user fault reporting logs, and real-time meteorological data during the period of the anomaly and cross-compare them to identify the causes, such as temporary activation of high-power equipment in a single household, temporary power supply at construction sites, and short-term voltage fluctuations.

[0114] 2) Periodic anomalies: Integrate monthly and quarterly load curves, meteorological data, and user attributes to compare the matching relationship between high-incidence periods of anomalies and seasonal temperature changes, factory production schedules, residential electricity consumption patterns, and statutory holidays, and distinguish between two core causes: seasonal load fluctuations and cyclical production in industry and commerce.

[0115] 3) Permanent anomalies: Conduct a comprehensive assessment using basic ledgers, distributed energy information, and maintenance records: For heavy overload issues, compare the rated current carrying capacity of the line with the long-term average load, and determine whether there are problems such as insufficient wire diameter, line aging, or overall overload in combination with the equipment's service life; For three-phase imbalance issues, check the distribution of three-phase users to determine whether there are problems such as uneven distribution of single-phase load, incorrect three-phase wiring, or concentrated single-phase access of distributed energy, and at the same time, combine historical data to determine the recurrence of faults.

[0116] Step S4: Multi-dimensional risk quantification scoring and governance priority ranking. Based on the source tracing result dataset, the preset multi-dimensional risk scoring model is called to quantify and score each low-voltage branch line with anomalies. The scores are combined to complete the descending order and the corresponding governance level is divided. This helps operation and maintenance personnel quickly distinguish the severity and urgency of hidden dangers, rationally allocate operation and maintenance resources, and finally output a standardized hidden danger governance priority list.

[0117] Step S5: Trend extrapolation and result output. Based on multi-period time-series datasets, a time-series extrapolation algorithm is used to extrapolate the trend of indicator changes over the next 1-3 months for the sorted abnormal lines, distinguishing between three development trends: continuous deterioration, basic stability, and self-mitigation. Subsequently, all data, including anomaly details, classification results, source tracing conclusions, priority lists, and trend predictions, are persistently stored in the backend database, while the analysis results are pushed to the front-end page, supporting online viewing, multi-condition filtering, and export of Excel-format reports by maintenance personnel. After maintenance personnel complete on-site hazard management, they send the maintenance logs back to the system, update the historical fault database, iteratively optimize the source tracing and analysis model, and form a complete data closed loop.

[0118] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0119] Based on the same inventive concept, this application also provides a power grid anomaly analysis device for implementing the power grid anomaly analysis method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more power grid anomaly analysis device embodiments provided below can be found in the limitations of the power grid anomaly analysis method described above, and will not be repeated here.

[0120] In one exemplary embodiment, such as Figure 3 As shown, a power grid anomaly analysis device is provided, comprising:

[0121] The indicator calculation module 201 is used to calculate the power grid anomaly indicators corresponding to each sampling time based on the power grid time series data of the target power grid branch line; the power grid time series data includes sampling data corresponding to multiple sampling times respectively.

[0122] The anomaly identification module 202 is used to identify power grid anomaly events occurring on the target power grid branch line based on the power grid anomaly indicators.

[0123] The morphology analysis module 203 is used to determine the abnormal morphology of the target power grid branch based on the power grid abnormal events that occur at each of the sampling times within each time scale; the abnormal morphology includes transient disturbance abnormalities, periodic abnormalities, or permanent abnormalities.

[0124] Each module in the aforementioned power grid anomaly analysis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0125] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores operating computer programs and a database. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The database stores measurement data and / or location information. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a power grid anomaly analysis method.

[0126] Those skilled in the art can understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0127] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps described in any of the above method embodiments.

[0128] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps described in any of the above method embodiments.

[0129] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps described in any of the above method embodiments.

[0130] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0131] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0132] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0133] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the appended claims.

Claims

1. A method for power grid anomaly analysis, characterized in that, The method includes: Based on the power grid time series data of the target power grid branch, calculate the power grid anomaly index corresponding to each sampling time; the power grid time series data includes sampling data corresponding to multiple sampling times respectively. Based on the power grid anomaly indicators, identify the power grid anomaly events occurring on the target power grid branch line; Based on the abnormal power grid events that occur at each sampling time within each time scale, the abnormal morphology of the target power grid branch is determined; the abnormal morphology includes transient disturbance abnormalities, periodic abnormalities, or permanent abnormalities.

2. The method according to claim 1, characterized in that, Each sampled data point includes the rated current value and the measured operating current value of each phase circuit branch; the power grid anomaly indicators include the low-voltage branch line overload indicator and / or the three-phase imbalance indicator; the power grid anomaly indicators calculated based on the power grid time series data of the target power grid branch line at each sampling time include at least one of the following: For each sampling time, determine the maximum value of the measured operating current, calculate the ratio of the maximum value to the rated current value, and obtain the low-voltage branch line overload index at the sampling time; For each sampling time, the maximum, minimum, and average values ​​of the measured operating current are determined. The difference between the maximum and minimum values ​​is calculated, and the ratio of the difference to the average value is determined to obtain the three-phase imbalance index at the sampling time.

3. The method according to claim 1, characterized in that, The determination of the abnormal morphology of the target power grid branch line based on the power grid abnormal events occurring at each sampling time within each time scale includes: If at least one of the aforementioned power grid anomalies occurs within a single day, and the aforementioned power grid anomalies occurring within a monthly, quarterly, or annual timescale do not meet the preset anomaly conditions, then the anomaly pattern of the target power grid branch is determined to be a transient disturbance anomaly.

4. The method according to claim 1, characterized in that, The determination of the abnormal morphology of the target power grid branch line based on the power grid abnormal events occurring at each sampling time within each time scale includes: If the abnormal power grid events occurring within a single day's time scale meet a preset regularity condition, the abnormal power grid events occurring within a monthly time scale meet a preset concentration condition, and the abnormal power grid events occurring within a quarterly and annual time scale meet a preset periodicity condition, then the abnormal pattern of the target power grid branch line is determined to be a periodic anomaly.

5. The method according to claim 1, characterized in that, The determination of the abnormal morphology of the target power grid branch line based on the power grid abnormal events occurring at each sampling time within each time scale includes: If the abnormal power grid events occurring within a single-day, monthly, quarterly, and annual timescale all meet the preset persistence conditions, then the abnormal morphology of the target power grid branch is determined to be a permanent abnormality.

6. The method according to claim 1, characterized in that, The method further includes: Based on the abnormal pattern and the pre-acquired power grid operation and maintenance data, analyze the cause of the abnormality of the target power grid branch line; and / or, based on the power grid time series data, deduce the abnormal change trend of the target power grid branch line.

7. The method according to claim 1, characterized in that, The method further includes: Based on the power grid anomaly indicators, the power grid anomaly events, and the anomaly patterns, the risk priority of the target power grid branch line is determined; According to the risk priority, anomaly handling is performed on each of the target power grid branches in sequence.

8. A power grid anomaly analysis device, characterized in that, The device includes: The indicator calculation module is used to calculate the power grid anomaly indicators corresponding to each sampling time based on the power grid time series data of the target power grid branch; the power grid time series data includes sampling data corresponding to multiple sampling times respectively. An anomaly identification module is used to identify power grid anomaly events occurring on the target power grid branch line based on the power grid anomaly indicators. The morphology analysis module is used to determine the abnormal morphology of the target power grid branch based on the power grid abnormal events that occur at each of the sampling times within each time scale; the abnormal morphology includes transient disturbance abnormalities, periodic abnormalities, or permanent abnormalities.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images