Hierarchical-based multi-dimensional electric quantity abnormity identification method, system and equipment and medium

Abstract

The invention belongs to the technical field of electric quantity data identification, and discloses a hierarchy-based multi-dimensional electric quantity abnormity identification method, system and device and a medium. The method comprises the steps of obtaining power consumption of a target region or a target industry in a preset time period and total power consumption of a power grid, and dividing a power consumption ratio based on the power consumption; acquiring a preset change range corresponding to each electric quantity abnormity type, identifying whether the electric quantity of the target region or the target industry deviates from each change range, if so, regarding the electric quantity as abnormity, and scoring and counting an abnormal integral; and dividing a priority level based on the power consumption ratio and the abnormal integral, and formulating a corresponding power abnormality response strategy. According to the method, through combination of hierarchical division and multi-dimensional detection, rapid and accurate identification of electric quantity abnormity is realized, the problems of information redundancy and low calculation efficiency possibly existing in a traditional method are avoided, the abnormity detection efficiency is improved, the detection accuracy is enhanced, and the method is suitable for popularization and application. And firm and powerful support is provided for safe operation of a power grid.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric quantity data identification, and specifically relates to a hierarchical-based multi-dimensional electric quantity anomaly identification method, system, equipment and medium. Background Art

[0002] With the rapid development of big data and machine learning technologies, the power industry has begun to use these advanced technologies to conduct in-depth mining and analysis of power data in order to improve the accuracy and reliability of power forecasts. By integrating and utilizing multi-dimensional information such as historical power data and meteorological data, we can gain a deeper understanding of the changing patterns of power characteristics in various industries and their influencing factors, thereby formulating more scientific power forecasting and allocation strategies.

[0003] However, in actual operation, power grid operation is often troubled by power anomalies, such as equipment failure, line aging, illegal power consumption, etc. These factors may cause abnormal fluctuations in power data. Power anomalies mainly include mutation anomalies, proportion anomalies, year-on-year anomalies, and data self-anomalies. These anomalies have a significant impact on the stability of the power grid, especially in cities or industries with a high proportion of power. Abnormal fluctuations may cause greater pressure on power dispatching. To solve this problem, Sun Yingtao et al. proposed a power data anomaly detection method based on multi-domain feature extraction. The method first analyzes the electrical parameter data through a multi-feature fusion algorithm to detect whether there is abnormal power consumption; then analyzes the power consumption pattern of the results suspected of abnormal power consumption, and finally determines whether there is abnormal power consumption. However, it is worth noting that since the nature and format of data in different dimensions may be different, data conversion and standardization processing will increase complexity. At the same time, multi-dimensional data fusion may lead to information redundancy and reduced computational efficiency. Summary of the invention

[0004] Based on the above-mentioned shortcomings and deficiencies in the prior art, one of the objects of the present invention is to solve at least one or more of the above-mentioned problems in the prior art. In other words, one of the objects of the present invention is to provide a hierarchical multi-dimensional power anomaly identification method, system, device and medium that meet one or more of the aforementioned needs, so as to meet the needs of efficient and accurate anomaly identification of power data, thereby achieving the purpose of improving the accuracy of power anomaly detection, reducing the complexity of data conversion and standardization processing, reducing information redundancy, and improving computing efficiency.

[0005] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a hierarchical multi-dimensional electricity anomaly identification method, comprising the steps of: S1, obtaining the electricity consumption of a target region or target industry and the total electricity consumption of a power grid within a preset time period, and dividing the electricity consumption proportion of the target region or target industry based on this, and the electricity consumption proportions are high proportion, medium proportion and low proportion respectively; S2, obtaining a preset variation range corresponding to each type of electricity anomaly, identifying whether the electricity consumption of the target region or target industry deviates from each of the variation ranges, if deviated, it is regarded as an abnormality and scored, and the abnormality points of the target region or target industry are counted, and the electricity anomaly types include sudden change anomalies and proportion anomalies; S3, based on the electricity consumption proportion and the abnormality points corresponding to the target region or target industry, dividing the priority level and formulating the corresponding electricity anomaly response strategy.

[0006] As a preferred solution, the priority level division and the formulation of corresponding power anomaly response strategies are specifically as follows: if the power consumption of the target area or target industry accounts for a high proportion, and the accumulated abnormal points are not less than three points, it is divided into a very high priority, recorded as level one, and scheduling is required; if the power consumption of the target area or target industry accounts for a high proportion, and the accumulated abnormal points are less than three points, it is divided into a high priority, recorded as level two, and close monitoring is required; if the power consumption of the target area or target industry accounts for a low proportion, and the accumulated abnormal points are not less than three points, it is divided into a medium priority, recorded as level three, and monitoring is required; if the power consumption of the target area or target industry accounts for a medium proportion, and the accumulated abnormal points are less than three points, it is divided into a low priority, recorded as level four, and monitoring is required; if the power consumption of the target area or target industry accounts for a low proportion, and the accumulated abnormal points are less than three points, it is divided into a secondary priority, recorded as level five.

[0007] As a preferred solution, the mutation anomaly includes a single-point mutation anomaly, and obtaining a preset variation range corresponding to the single-point mutation anomaly includes the steps of: obtaining a preset sliding window size k; obtaining the current power consumption at the current time and recording it as the current value l i+k And intercept the first k data points to form a time series , the time series The median or mean is recorded as its statistic m i , to calculate the current value l i+k With the statistic m i The difference d i ; Use a double rolling aggregation transformer based on the sliding window and the time series , generate a new time series D and calculate its interquartile range IQR; based on the interquartile range IQR of time series D, set the variation range corresponding to the single point mutation anomaly as ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant.

[0008] As a preferred solution, the mutation anomaly also includes the cumulative fluctuation mutation anomaly, and obtaining the preset variation range corresponding to the cumulative fluctuation mutation anomaly includes the steps of: obtaining the historical seven-day electricity data of the target area or target industry; , and based on this calculate the percentage change in daily power , mean percentage change and percentage change standard deviation ; Based on the mean percentage change and percentage change standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The percentage change The calculation formula is , where For the seventh day in history The daily electricity For the Day's electricity.

[0009] As a preferred solution, the mutation anomaly also includes the cumulative fluctuation mutation anomaly, and obtaining the preset variation range corresponding to the cumulative fluctuation mutation anomaly also includes the steps of: obtaining the historical 30-day electricity data of the target area or target industry; , and calculate the average value of the historical 30-day electricity data and standard deviation ; Based on the mean and standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The historical thirty-day average power consumption and standard deviation The calculation formula is , , where For the first time in history Day's electricity.

[0010] As a preferred solution, the mutation anomaly also includes a horizontal mutation anomaly, and obtaining a preset range of variation corresponding to the horizontal mutation anomaly includes the steps of: obtaining a preset sliding window and its size k; forming a first sequence based on k data of the sliding window And get its adjacent sequence ; The first sequence The median or mean of ; The adjacent sequence The median or mean of Calculate the first sequence Statistics The adjacent sequence Statistics The difference ; Using a double rolling aggregation transformer based on the sliding window, the first sequence and the adjacent sequence , generate a new time series D and calculate its interquartile range IQR; based on the interquartile range IQR of time series D, set the range of change corresponding to the horizontal mutation anomaly to ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant.

[0011] As a preferred solution, obtaining a preset variation range corresponding to the abnormal proportion includes the following steps: assuming that the time series data ,in Indicates time The proportion of electricity consumption in the target area or target industry; Select the time The data of the past thirty days is used as a sliding window ; Calculate the sliding window the median of and interquartile range , set the change range corresponding to the abnormal proportion to ,in is a preset constant.

[0012] In a second aspect, the present invention provides a hierarchical multi-dimensional power anomaly identification system, characterized in that it is used to implement the multi-dimensional power anomaly identification method as described in the first aspect.

[0013] In a third aspect, the present invention provides an electronic device, wherein the computer device includes a memory, a processor, and a computer program, and when the computer program is executed by the processor, the multi-dimensional electrical quantity anomaly identification method as described in the first aspect is implemented.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the multi-dimensional electrical quantity anomaly identification method as described in the first aspect is implemented.

[0015] Compared with the prior art, the present invention has the following beneficial effects: First, the present invention realizes the deep classification of electricity usage through the fine division of electricity consumption proportion, ensures that higher priority is given to abnormal fluctuations in key areas and industries, and thus optimizes the resource allocation of abnormal detection. Secondly, the range of change corresponding to each type of electricity anomaly (sudden change anomaly, proportion anomaly, year-on-year anomaly, data self-anomaly) is preset, which can fully identify whether the electricity consumption of the target area or target industry deviates from the normal range. Once an anomaly is detected, it is immediately scored and the abnormal points are counted. This quantitative evaluation method helps to understand the degree of abnormality and the scope of influence more intuitively. Finally, based on the proportion of electricity consumption and abnormal points, the scheme can scientifically and reasonably divide the priority level, and formulate corresponding electricity abnormal response strategies accordingly. The formulation of this strategy helps to quickly respond to and handle electricity anomalies, ensure the stable operation of the power grid, reduce the pressure of power dispatching, and improve the overall safety of the power system. In summary, by combining hierarchical division and multi-dimensional detection, the present invention realizes the rapid and accurate identification of electricity anomalies, effectively avoiding the problems of information redundancy and low computational efficiency that may exist in traditional methods. This improvement not only significantly improves the efficiency of anomaly detection, but also greatly enhances the accuracy of detection, providing solid and strong support for the safe operation of the power grid.

[0016] Further or more detailed beneficial effects will be described in detail in conjunction with specific examples in the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of the multi-dimensional power anomaly identification method provided in an embodiment of the present invention.

[0019] Figure 2 It is a schematic diagram of a flow chart of obtaining a preset variation range corresponding to the single point mutation abnormality provided by an embodiment of the present invention.

[0020] Figure 3 It is a flowchart of obtaining a preset range of changes corresponding to the abnormal level mutation provided by an embodiment of the present invention.

[0021] Figure 4 is a structural diagram of the electronic device provided in an embodiment of the present invention.

[0022] Figure Number: 400. Electronic equipment; 401, processor; 402, communication bus; 403, user interface; 404, network interface; 405, memory. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0024] In the following description, multiple embodiments of the present invention are provided, and different embodiments may be replaced or combined, so the present invention may also be considered to include all possible combinations of the same and / or different embodiments described. Therefore, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing one or more of A, B, C, and D, all other possible combinations, even though the embodiment may not be clearly described in the following text.

[0025] The following description provides examples and does not limit the scope, applicability or examples set forth in the claims. Changes may be made to the functions and arrangements of the elements described without departing from the scope of the present invention. Various processes or components may be appropriately omitted, substituted or added to each example. For example, the described method may be performed in an order different from the described order, and various steps may be added, omitted or combined. In addition, the features described in some examples may be combined in other examples.

[0026] In order to facilitate a better understanding of the embodiments of the present invention, before explaining the specific implementation modes of the present invention in detail, its application scenarios are first described.

[0027] The multidimensional power anomaly identification method described in the embodiments of this specification is applied to the power monitoring and management process of the smart grid. In these scenarios, the application of the multidimensional power anomaly identification method is intended to enable grid managers to identify power anomalies in each node or area in the grid in real time and accurately, so that measures can be taken quickly, such as adjusting power dispatch, repairing faulty equipment or optimizing the grid structure, to maintain the stable operation of the grid, reduce power losses, and improve the overall efficiency and safety of the power system.

[0028] The following is a brief explanation of the mutation anomaly, proportion anomaly, year-on-year anomaly, data self-anomaly, and double rolling aggregation converter involved in multiple embodiments of this specification: Sudden abnormality refers to a sudden and significant change in the power usage of a node or area in the power grid within a short period of time. This change is beyond the normal fluctuation range and may indicate equipment failure, illegal power use or grid structure problems.

[0029] An abnormal share means that the share of electricity used in different areas of the power grid or in different time periods is usually different. An abnormal share means that the share of electricity used in a certain area or time period is significantly different from the expected or historical data, which may reflect an unreasonable power grid structure, changes in electricity consumption habits, or energy distribution problems.

[0030] Year-on-year anomalies refer to significant differences in the power usage of a node or region in the current power grid compared to historical data in the same time period. This anomaly may be caused by factors such as changes in power demand, power grid transformation or policy adjustments, and requires in-depth analysis and adjustments by power grid managers.

[0031] Data self-abnormality refers to problems with the power grid power monitoring data itself, such as missing data, data errors, or data inconsistency. This abnormality may be caused by sensor failure, data transmission problems, or data processing algorithm errors, and grid managers need to check and repair them in a timely manner to ensure the accuracy and reliability of power monitoring data.

[0032] The double rolling aggregation transformer is an anomaly detector. Its detection principle is based on the electricity consumption of cities / industries / users. For each data point, a sliding window is established before it. Identification is performed by comparing the difference between the median in the sliding window and the data point after the window.

[0033] Embodiment 1: like Figure 1 As shown, the present embodiment provides a hierarchical multi-dimensional electricity anomaly identification method, including the steps of: S1, obtaining the electricity consumption of the target area or target industry and the total electricity consumption of the power grid within a preset time period, and dividing the electricity consumption proportion of the target area or target industry based on this, and the electricity consumption proportions are high proportion, medium proportion and low proportion respectively; S2, obtaining the preset variation range corresponding to each electricity anomaly type, identifying whether the electricity consumption of the target area or target industry deviates from each of the variation ranges, if deviated, it is regarded as abnormal and scored, and the abnormal points of the target area or target industry are counted, and the electricity anomaly types include sudden change anomalies and proportion anomalies; S3, based on the electricity consumption proportion and the abnormal points corresponding to the target area or target industry, dividing the priority level and formulating the corresponding electricity anomaly response strategy.

[0034] It is understandable that according to the difference in the contribution of the electricity share of various cities and industries to the overall electricity demand, the electricity share stratification model is adopted to divide the target cities or target industries into three levels: high share, medium share and low share. This division method can more intuitively reflect the potential impact of different regions or industries on the fluctuation of the province's power load, and ensure that abnormal fluctuations in key regions and industries are given higher priority. Specifically, cities or industries with a high proportion of electricity play a leading role in the operation of the power grid. When they experience abnormal fluctuations, it may lead to an imbalance in the load of the power grid in the province. Therefore, cities / industries with a high proportion of electricity will be placed under key monitoring, while cities / industries with a low proportion of electricity will be classified as secondary monitoring objects to optimize the allocation of resources for anomaly detection.

[0035] Specifically, the present embodiment provides a preferred implementation method, and the priority level division and the formulation of corresponding power anomaly response strategies are as follows: if the power consumption of the target area or target industry accounts for a high proportion, and the accumulated abnormal points are not less than three points, it is divided into a very high priority, recorded as level one, and scheduling is required; if the power consumption of the target area or target industry accounts for a high proportion, and the accumulated abnormal points are less than three points, it is divided into a high priority, recorded as level two, and close monitoring is required; if the power consumption of the target area or target industry accounts for a low proportion, and the accumulated abnormal points are not less than three points, it is divided into a medium priority, recorded as level three, and monitoring is required; if the power consumption of the target area or target industry accounts for a medium proportion, and the accumulated abnormal points are less than three points, it is divided into a low priority, recorded as level four, and monitoring is required; if the power consumption of the target area or target industry accounts for a low proportion, and the accumulated abnormal points are less than three points, it is divided into a secondary priority, recorded as level five.

[0036] Specifically, Figure 2 The embodiment shown provides a preferred implementation method, wherein the mutation anomaly includes a single-point mutation anomaly, and obtaining a preset variation range corresponding to the single-point mutation anomaly includes the steps of: obtaining a preset sliding window size k; obtaining the current power consumption at the current time and recording it as the current value l i+k And intercept the first k data points to form a time series , the time series The median or mean is recorded as its statistic m i , to calculate the current value l i+k With the statistic m i The difference d i ; Use a double rolling aggregation transformer based on the sliding window and the time series , generate a new time series D and calculate its interquartile range IQR; based on the interquartile range IQR of time series D, set the variation range corresponding to the single point mutation anomaly as ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant. It can be understood that the use of the double rolling aggregation transformer to compare each time series value with the statistics generated in the previous window helps to smooth the time series and capture potential abnormal signals. At the same time, this type of anomaly refers to the degree of deviation of the sequence data point from the data trend over a period of time in the past, that is, the data point is not accidentally deviated, but shows a persistent abnormal behavior, so this method is effective in capturing persistent anomalies in time series (relative to short-term anomalies).

[0037] It is understandable that when analyzing abnormal fluctuations in electricity consumption, we first need to identify the reasons for the dramatic changes in electricity consumption. These changes often exceed the normal fluctuation range and may be caused by industrial production fluctuations, emergencies, or seasonal demand changes. When quantifying abnormal fluctuations and capturing abnormal points in the data, we can accurately identify abnormal fluctuations in electricity consumption based on historical data, periodic laws, and real-time fluctuation characteristics.

[0038] Specifically, short-term changes in electricity volume are often affected by temporary factors, such as equipment failures, weather changes, or emergencies. Since the absolute electricity volume may not change much in the short term, but its relative change may be very obvious, the fluctuation percentage can more intuitively capture these short-term fluctuations. For example, if the electricity volume on a certain day suddenly rises or falls sharply, even if the actual electricity volume does not change much, the percentage change relative to the previous few days may be very significant, thus reflecting an abnormality. This embodiment provides a preferred implementation method for short-term fluctuations, and the mutation anomaly also includes a cumulative fluctuation mutation anomaly. Obtaining a preset range of changes corresponding to the cumulative fluctuation mutation anomaly includes the steps of: obtaining historical seven-day electricity volume data for the target area or target industry. , and based on this calculate the percentage change in daily power , mean percentage change and percentage change standard deviation ; Based on the mean percentage change and percentage change standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The percentage change The calculation formula is , where For the seventh day in history The daily electricity For the Day's electricity.

[0039] Specifically, in the medium term, changes in electricity consumption are more affected by macro factors, such as changes in production cycles, fluctuations in seasonal demand, etc. Compared to the short term, using specific electricity values ​​can better capture these changes, because medium-term electricity consumption changes tend to reflect overall trends. For example, as the seasons change, electricity demand increases or decreases, and medium-term electricity values ​​can better indicate long-term trends and avoid the impact of single-day fluctuations on the results. This embodiment provides a preferred implementation method for medium-term fluctuations, and the mutation anomaly also includes a cumulative fluctuation mutation anomaly, obtaining a preset range of changes corresponding to the cumulative fluctuation mutation anomaly, and also includes the steps of: obtaining historical thirty-day electricity consumption data for the target area or target industry. , and calculate the average value of the historical 30-day electricity data and standard deviation ; Based on the mean and standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The historical thirty-day average power consumption and standard deviation The calculation formula is , , where For the first time in history Day's electricity.

[0040] It is understandable that for the use of standard deviation multiples, short-term fluctuations often contain higher uncertainty and suddenness, so using 2 standard deviations can more sensitively capture abnormal fluctuations. 2 standard deviations cover about 95% of the data, so when the power change exceeds this range, short-term abnormalities can be identified more quickly. This sensitivity is suitable for scenarios with rapid response, such as when the power load suddenly increases. Medium-term changes are usually more stable and trendy, so using 3 standard deviations (covering 99.7% of the data) to determine abnormalities can more accurately exclude accidental fluctuations and focus on continuous deviations from the trend. 3 standard deviations can prevent false alarms caused by single emergencies, making abnormality identification more robust and reliable, and suitable for macro trend analysis.

[0041] Based on the above two determination strategies provided by this embodiment for the cumulative fluctuation anomaly, the following three scenarios will be formed. First, short-term abnormality but medium-term non-abnormality: This indicates that the power has experienced sudden and drastic fluctuations in a short period of time, but from a longer-term perspective, this fluctuation is still within the normal range of the historical 30 days, which means that this change is temporary and has no lasting impact on the overall trend. It is usually caused by a one-time event, such as equipment failure, climate change, or sudden increase in production demand. Second, short-term non-abnormality but medium-term abnormality: This means that although the power on the current date has fluctuated relatively smoothly in the short term and has not changed significantly, the power has shown significant deviations relative to the historical mean and standard deviation of the past 30 days. This situation usually indicates that the power has gradually accumulated deviations over a long period of time, and short-term fluctuations are not enough to reflect these changes. It may be related to seasonal demand, production structure adjustment, or the gradual implementation of policies. Third, both short-term and medium-term abnormalities: This shows that the power has not only experienced drastic fluctuations in the short term, but also that this fluctuation has caused the power on the current date to exceed the mean range of the historical 30 days. In other words, the sharp fluctuations in the short term are also a direct result of changes in medium-term trends, indicating that electricity demand is showing abnormal conditions in both the short and medium term.

[0042] Specifically, Figure 3 The embodiment shown provides a preferred implementation, wherein the mutation anomaly also includes a horizontal mutation anomaly, and obtaining a preset range of variation corresponding to the horizontal mutation anomaly includes the steps of: obtaining a preset sliding window and its size k; forming a first sequence based on k data of the sliding window And get its adjacent sequence ; The first sequence The median or mean of ; The adjacent sequence The median or mean of Calculate the first sequence Statistics The adjacent sequence Statistics The difference ; Using a double rolling aggregation transformer based on the sliding window, the first sequence and the adjacent sequence , generate a new time series D and calculate its interquartile range IQR; based on the interquartile range IQR of time series D, set the range of change corresponding to the horizontal mutation anomaly to ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant.

[0043] More specifically, mutation anomalies also include periodic mutation anomalies, which refer to the phenomenon of drastic fluctuations in electricity caused by external periodic factors within a specific time period. Its triggering factors cover a wide range of fields, including climate change, holiday effects, industry production cycles, and natural disasters. The typical manifestation of this anomaly is that there is a significant sudden increase or decrease in electricity during certain seasons or specific events. In addition, during major holidays, industrial electricity consumption plummets due to factory shutdowns, while residential electricity consumption increases, resulting in drastic fluctuations in electricity in the short term. Therefore, the range of changes corresponding to periodic mutation anomalies can be flexibly set according to actual conditions. This usually requires consideration of historical data, seasonal factors, holiday arrangements, and industry production cycles, etc., in order to formulate a range of electricity changes that is both in line with actual conditions and can effectively identify anomalies. Through such a setting, periodic mutation anomalies can be more accurately identified and responded to, providing strong support for the stable operation of the power grid and power dispatching.

[0044] Specifically, this embodiment provides a preferred implementation method for obtaining a preset range of changes corresponding to the abnormal proportion, including the steps of: assuming that the time series data ,in Indicates time The proportion of electricity consumption in the target area or target industry; Select the time The data of the past thirty days is used as a sliding window ; Calculate the sliding window the median of and interquartile range , set the change range corresponding to the abnormal proportion to ,in is a preset constant. It is understandable that the selection of 30 days as a window is mainly to capture the medium- and long-term cyclical characteristics of the electricity share data. The electricity share in the power industry is usually relatively stable. The 30-day time span can smooth short-term fluctuations, reduce the impact of accidental events on the data, and ensure the reliability of the detection results. In addition, the parameter c=10 is used as a multiple of the IQR (interquartile range) to avoid false alarms due to small normal fluctuations, and only those changes that significantly deviate from the normal range are identified as abnormal. This configuration can not only effectively adapt to the characteristics of large fluctuations in power data, but also improve the accuracy of detection.

[0045] More specifically, the type of power anomaly also includes year-on-year anomalies. It is understandable that the year-on-year anomaly determination for power consumption in cities is mainly divided into holiday / normal day strategies and strategies for different seasons, such as the high temperature period in summer and the cooling period in winter. Affected by extreme weather in some years, the year-on-year changes will be large. Therefore, different thresholds can be set for summer, winter, autumn, and spring to flexibly determine year-on-year anomalies. At the same time, in terms of holidays, especially the Spring Festival period, there are differences in the period when the Spring Festival occurs each year. Therefore, when it was the Spring Festival period last year but not this year, the year-on-year power growth rate often changes significantly. To deal with such situations, a Spring Festival label can be added for anomaly determination, and a higher threshold can be given to the year-on-year change range for such situations to prevent misjudgment. This embodiment provides a preferred implementation for year-on-year anomalies, as follows: First, flexible thresholds are set based on the electricity consumption characteristics of different seasons. In summer and winter, electricity consumption fluctuates greatly due to high and low temperatures, so a higher year-on-year abnormal threshold (such as ±40%) is set. In spring and autumn, the climate is relatively mild, and electricity consumption fluctuates less, so the threshold can be set to ±20%.

[0046] The year-on-year calculation formula is: .

[0047] Secondly, when comparing electricity consumption during holidays (especially during the Spring Festival) and normal days, different abnormality determination thresholds are set. During the Spring Festival, electricity consumption fluctuates greatly, so a higher threshold (such as ±40%) is set. On normal working days, electricity consumption fluctuates less, so the threshold can be set to ±20%.

[0048] The year-on-year calculation formula is: .

[0049] Third, we introduce the Spring Festival tag. For periods that were Spring Festival last year but not this year, we add the Spring Festival tag for special treatment. Such situations are prone to dramatic year-on-year fluctuations in electricity consumption, so we set the year-on-year abnormality threshold to a higher level (such as ±100%) to prevent misjudgment. For other periods, we continue to use the year-on-year abnormality threshold for normal days (such as ±30%).

[0050] More specifically, the power abnormality type also includes data self-abnormality. This embodiment provides a preferred implementation for data self-abnormality, which is as follows: First, missing data can be directly analyzed using simple statistical methods to count the location and number of missing values.

[0051] Secondly, if the capacity exceeds the limit, the historical electricity consumption data of users or cities is collected, especially the past overcapacity situations. The historical overcapacity ratio can be calculated. For daily overcapacity ratio and monthly overcapacity frequency: , , Occasional overcapacity is acceptable in the power system. Therefore, when formulating the overcapacity abnormality judgment strategy, a more reasonable judgment can be made based on the historical overcapacity situation. The frequency of overcapacity is calculated based on historical data. For example, a user may have overcapacity in 10% of the time periods in a year. In this case, an allowable overcapacity frequency threshold (such as 15%) is set. When the overcapacity frequency exceeds this ratio, the system determines that the user or city has overcapacity abnormality. The frequency threshold can be dynamically adjusted according to the historical performance of different users or cities. For example, if the historical overcapacity frequency of a user has always been less than 5%, the overcapacity frequency threshold can be set to 7%-10%; while the historical overcapacity frequency of a user is high, a higher threshold can be set, and the specific threshold is adjusted within a reasonable range. Analyze the maximum value of historical overcapacity, for example, a certain city has historically had an overcapacity ratio of up to 25%. Based on this historical maximum value, set an allowable overcapacity range, such as the allowable overcapacity ratio not exceeding 80%-90% of the historical maximum value. For some cities, if the overcapacity phenomenon occurs more frequently during extreme weather or special periods, the tolerance for the overcapacity ratio can be adjusted during these specific periods. For example, during high temperatures in summer, the overcapacity ratio can be set to 90% of the historical maximum value; while in other seasons, the threshold can be set to 80%.

[0052] Thirdly, the frequency of overload is determined in a short-term window (30 days) based on the historical frequency of overload during the same period to determine an allowable frequency threshold. For example, if the frequency of overload of a user in a week is 10% in history, this value is used to determine whether an abnormality has occurred. This determination method takes into account the reasonable range of historical data and allows a certain degree of overload.

[0053] Fourth, the excess capacity is adjusted dynamically. According to the historical maximum excess capacity ratio, the excess capacity thresholds for different seasons or situations are set. For example, during high temperatures in summer, the allowed excess capacity ratio can reach 90% of the historical maximum value; while in the off-season, the threshold can be lowered to 80%. For example, a higher excess capacity ratio can be allowed in July and August in summer, while in other months, the excess capacity ratio can be controlled at a lower level based on historical data.

[0054] Fifth, continuous overcapacity time determination. Setting of consecutive days: When the overcapacity phenomenon of a user or city occurs for more than 3 consecutive days, it should be determined as abnormal even if it does not reach 80%-90% of the historical maximum overcapacity ratio.

[0055] Embodiment 2: This embodiment provides a hierarchical multi-dimensional power anomaly identification system, which is used to implement the multi-dimensional power anomaly identification method as described in the first embodiment.

[0056] Embodiment three: like Figure 4 As shown, this embodiment provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.

[0057] The communication bus can be used to realize the connection and communication among the above-mentioned components.

[0058] The user interface may include buttons, and the optional user interface may also include a standard wired interface or a wireless interface.

[0059] The network interface may include but is not limited to a Bluetooth module, an NFC module, a Wi-Fi module, etc.

[0060] Among them, the processor may include one or more processing cores. The processor uses various interfaces and lines to connect the various parts of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory. Optionally, the processor can be implemented in at least one hardware form of DSP, FPGA, and PLA. The processor can integrate one or a combination of CPU, GPU, modem, etc. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display; the modem is used to handle wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor, but may be implemented separately through a chip.

[0061] Among them, the memory may include RAM or ROM. Optionally, the memory includes a non-transitory computer-readable medium. The memory can be used to store instructions, programs, codes, code sets or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory may also be at least one storage device located away from the aforementioned processor. The memory as a computer storage medium may include an operating system, a network communication module, a user interface module, and an identification application. The processor may be used to call the identification application stored in the memory and execute the steps of the multi-dimensional power anomaly identification method mentioned in the above-mentioned embodiment.

[0062] Embodiment 4: This embodiment provides a computer-readable storage medium, which stores instructions. When the computer-readable storage medium is executed on a computer or a processor, the computer or the processor executes the above-mentioned Figure 1 One or more steps in the illustrated embodiment. If the components of the electronic device described above are implemented in the form of software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0063] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of this specification is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state drive (SSD)).

[0064] A person of ordinary skill in the art can understand that all or part of the processes in the method of the first embodiment can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes. In the absence of conflict, the technical features in this embodiment and the implementation scheme can be combined arbitrarily.

[0065] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0066] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0067] The above is only an exemplary embodiment of the present invention and cannot be used to limit the scope of the present invention. That is, any equivalent changes and modifications made according to the teachings of the present invention are still within the scope of the present invention. After considering the specification and practicing the disclosure here, it will be easy for those skilled in the art to think of the implementation scheme of the present invention. The present invention is intended to cover any modification, use or adaptation of the present invention, which follows the general principles of the present invention and includes common knowledge or customary technical means in the art that are not recorded in the present invention. The description and examples are only regarded as exemplary, and the scope and spirit of the present invention are defined by the claims.

Claims

1. A hierarchical multi-dimensional power anomaly identification method, characterized in that: Includes steps: S1. Obtain the power consumption of the target region or target industry and the total power consumption of the power grid within a preset time period, and divide the power consumption proportion of the target region or target industry into high proportion, medium proportion and low proportion in sequence; S2. Obtain preset change ranges corresponding to each type of power anomaly, identify whether the power consumption of the target area or target industry deviates from each of the change ranges, and if deviated, it is considered abnormal and scored, and the abnormal points of the target area or target industry are counted. The power anomaly types include mutation anomaly and proportion anomaly; S3. Based on the electricity consumption proportion and the abnormal score corresponding to the target region or target industry, the priority levels are divided and corresponding electricity abnormality response strategies are formulated.

2. The hierarchical multi-dimensional power anomaly identification method according to claim 1 is characterized in that: The division of priority levels and formulation of corresponding power abnormality response strategies are specifically as follows: If the target area or target industry accounts for a high proportion of electricity consumption and the accumulated abnormal points are not less than three points, it will be classified as a very high priority, recorded as level one, and dispatching is required; If the target region or target industry accounts for a high proportion of electricity consumption and the accumulated abnormal points are less than three points, it will be classified as high priority, recorded as level 2 and require close monitoring; If the target area or target industry has a low electricity consumption ratio and the accumulated abnormal points are not less than three points, it will be classified as medium priority, recorded as level three and need to be monitored; If the target region or target industry accounts for a medium proportion of electricity consumption and the accumulated abnormal points are less than three points, it will be classified as low priority, recorded as level 4 and require monitoring; If the electricity consumption of the target area or target industry accounts for a low proportion and the accumulated abnormal points are less than three points, it will be classified as the second priority and recorded as level five.

3. The hierarchical multi-dimensional power anomaly identification method according to claim 2 is characterized in that: The mutation abnormality includes a single point mutation abnormality, and obtaining a preset variation range corresponding to the single point mutation abnormality includes the steps of: Get the preset sliding window size k; Get the current power consumption and record it as current value l i+k And intercept the first k data points to form a time series , the time series The median or mean is recorded as its statistic m i , to calculate the current value l i+k With the statistic m i The difference d i ; Use a dual rolling aggregation transformer based on the sliding window and the time series , generate a new time series D and calculate its interquartile range IQR; The interquartile range IQR setting based on the time series D corresponds to the variation range of the single point mutation abnormality: ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant.

4. The hierarchical multi-dimensional power anomaly identification method according to claim 3 is characterized in that: The mutation anomaly also includes a cumulative fluctuation mutation anomaly, and obtaining a preset variation range corresponding to the cumulative fluctuation mutation anomaly includes the steps of: Obtain historical seven-day electricity consumption data for the target region or target industry , and based on this calculate the percentage change in daily power , mean percentage change and percentage change standard deviation ; Based on the mean percentage change and percentage change standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The percentage change The calculation formula is , In the formula, For the seventh day in history The daily electricity For the Day's electricity.

5. The hierarchical multi-dimensional power anomaly identification method according to claim 3 is characterized in that: The mutation anomaly also includes a cumulative fluctuation mutation anomaly, and obtaining a preset variation range corresponding to the cumulative fluctuation mutation anomaly also includes the steps of: Obtain the historical 30-day electricity consumption data for the target region or target industry , and calculate the average value of the historical 30-day electricity data and standard deviation ; Based on the mean and standard deviation , set the variation range corresponding to the cumulative fluctuation mutation anomaly to ; The historical thirty-day average power consumption and standard deviation The calculation formula is ， ， In the formula, For the first time in thirty days of history Day's electricity.

6. A hierarchical multi-dimensional power anomaly identification method according to claim 4 or 5, characterized in that: The mutation anomaly also includes a horizontal mutation anomaly, and obtaining a preset variation range corresponding to the horizontal mutation anomaly includes the steps of: Get the preset sliding window and its size k; The k data based on the sliding window form a first sequence And get its adjacent sequence ; The first sequence The median or mean of ; The adjacent sequence The median or mean of ; Calculate the first sequence Statistics The adjacent sequence Statistics The difference ; A dual rolling aggregate transformer is used to transform the first sequence based on the sliding window. and the adjacent sequence , generate a new time series D and calculate its interquartile range IQR; Based on the interquartile range (IQR) of the time series D, the range of variation corresponding to the abnormal horizontal mutation is set to ,in is the first quartile of time series D, is the third quartile of time series D, is a preset constant.

7. The hierarchical multi-dimensional power anomaly identification method according to claim 6 is characterized in that: Obtaining a preset variation range corresponding to the abnormal proportion includes the following steps: Assuming time series data ,in Indicates time The proportion of electricity consumption in the target region or target industry; Select time The data of the past thirty days is used as a sliding window ; Calculate the sliding window the median of and interquartile range , set the change range corresponding to the abnormal proportion to ,in is a preset constant.

8. A hierarchical multi-dimensional power anomaly identification system, characterized in that: Used to implement the multi-dimensional electrical quantity anomaly identification method as described in any one of claims 1 to 7.

9. A computer device, comprising a memory, a processor and a computer program, characterized in that: When the computer program is executed by a processor, the multi-dimensional electrical quantity anomaly identification method as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the multi-dimensional electrical quantity anomaly identification method as described in any one of claims 1 to 7 is implemented.

Images