Abnormal recognition method and device, computer equipment and storage medium

Abstract

The embodiment of the invention belongs to the field of financial science and technology, and relates to an anomaly recognition method and device, computer equipment and a storage medium, and the method comprises the following steps: obtaining historical electric energy load data, and carrying out the preprocessing of the historical electric energy load data, and obtaining standard electric energy load data; acquiring external factor data, and associating the external factor data with the standard electric energy load data to obtain an electric energy load data set; analyzing the electric energy load data set based on a time sequence analysis algorithm and a preset clustering algorithm, and constructing a power utilization mode model; performing electric energy load prediction based on the power utilization mode model and a preset regression algorithm to obtain predicted electric energy load information; acquiring real-time electric energy load information, and comparing the predicted electric energy load information with the real-time electric energy load information to obtain abnormal power consumption behavior information; and performing abnormal alarm according to the abnormal power utilization behavior information. According to the invention, the abnormal power consumption behavior can be effectively and accurately identified and timely alarmed.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, specifically to the field of financial technology, and in particular to an abnormality identification method, device, computer equipment and storage medium. Background Art

[0002] In the insurance industry, especially for power-related insurance products, the anomaly detection technology of power load data in smart grids is particularly important. The power load data in smart grids is complex and changeable. It is not only significantly affected by external factors such as weather and holidays, but also often faces abnormal situations such as equipment failure and power theft, which are directly related to the risk assessment and claims processing of insurance companies.

[0003] The load data in smart grids has obvious periodicity and trend characteristics, which makes it a major technical difficulty to accurately extract the characteristics of power consumption patterns. Insurance companies need to rely on these characteristics to identify potential power consumption anomalies and thus assess risks. However, abnormal situations can manifest in various forms, especially power theft, which is often more hidden and difficult to distinguish from normal fluctuations, which undoubtedly increases the difficulty of risk identification.

[0004] With the popularity of IoT devices such as smart meters, the scale and dimension of load data are increasing. How to achieve distributed computing and horizontal expansion of algorithms to cope with the processing needs of large-scale data has become another urgent problem to be solved. Insurance companies need to use advanced technical means to improve the accuracy and efficiency of anomaly detection, so as to more effectively manage risks and protect the interests of customers. Summary of the invention

[0005] The purpose of the embodiments of the present application is to propose an abnormality identification method, device, computer equipment and storage medium to solve the problem of being unable to effectively and accurately identify abnormal power usage behavior and issue an alarm in a timely manner.

[0006] In order to solve the above technical problems, the present application embodiment provides an abnormality identification method, which adopts the following technical solution:

[0007] Acquire historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data;

[0008] Acquire external factor data, and associate the external factor data with the standard electric energy load data to obtain an electric energy load data set;

[0009] Analyzing the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to construct an electric power consumption pattern model;

[0010] Performing electric energy load forecasting based on the power consumption pattern model and a preset regression algorithm to obtain forecasted electric energy load information;

[0011] Acquire real-time electric energy load information, compare the predicted electric energy load information with the real-time electric energy load information, and obtain abnormal power consumption behavior information;

[0012] An abnormal alarm is issued according to the abnormal electricity usage behavior information.

[0013] Furthermore, the step of acquiring external factor data, associating the external factor data with the standard electric energy load data, and obtaining an electric energy load data set specifically includes:

[0014] Extracting the external factor data from a database according to the historical power load data, wherein the external factor data includes weather data and holiday data;

[0015] Performing time stamp alignment on the weather data, the holiday data, and the standard electric energy load data;

[0016] The aligned standard electric energy load data is used as a main data set, and the aligned weather data and holiday data are added to the main data set to obtain the electric energy load data set.

[0017] Furthermore, the step of analyzing the electric energy load data set based on the time series analysis algorithm and the preset clustering algorithm to construct the power consumption mode model specifically includes:

[0018] Extracting time trend information and periodic characteristics from the electric energy load data set according to a time series analysis algorithm, and generating a time dimension electricity consumption pattern according to the time trend information and the periodic characteristics;

[0019] Perform user clustering on the electric energy load data set according to a preset clustering algorithm to obtain user electric energy load categories, and generate user-dimensional power consumption patterns according to the user electric energy load categories;

[0020] The power usage pattern model is constructed according to the time dimension power usage pattern and the user dimension power usage pattern.

[0021] Furthermore, the step of performing electric energy load forecasting based on the power consumption pattern model and the preset regression algorithm to obtain the predicted electric energy load information specifically includes:

[0022] Acquire current factor data, and fuse the current factor data with the power consumption pattern model to obtain a power consumption prediction model;

[0023] Performing electric energy load forecasting based on a preset regression algorithm and the power consumption forecasting model, and optimizing the power consumption forecasting model according to the forecasting result to obtain an optimized power consumption forecasting model;

[0024] The power load is predicted based on the optimized power consumption prediction model to obtain the predicted power load information.

[0025] Furthermore, the step of performing electric energy load forecasting based on the preset regression algorithm and the power consumption forecasting model, and optimizing the power consumption forecasting model according to the forecasting result to obtain the optimized power consumption forecasting model specifically includes:

[0026] Generate a corresponding predicted power load curve according to the prediction result;

[0027] Obtaining a preset rationality condition, and determining whether the predicted power load curve meets the preset rationality condition;

[0028] If the predicted electricity load curve does not meet the preset rationality condition, a prediction error is calculated for the predicted electricity load curve and the actual electricity load curve, and the algorithm parameters of the preset regression algorithm and the model parameters of the electricity consumption prediction model are adjusted according to the calculated prediction error to obtain the optimized electricity consumption prediction model.

[0029] Furthermore, the step of acquiring real-time electric energy load information, comparing the predicted electric energy load information with the real-time electric energy load information, and obtaining abnormal electricity consumption behavior information specifically includes:

[0030] Aligning the predicted electric energy load information and the real-time electric energy load information on a time axis to obtain an aligned load information set;

[0031] Performing time granularity regularization on the aligned load information set according to a dynamic time regularization algorithm to obtain a regularized load information set;

[0032] Fitting the regularized load information set according to a curve fitting algorithm to obtain a predicted electric energy load curve and a real-time electric energy load curve;

[0033] Calculating the time point deviation value between the predicted electric energy load curve and the real-time electric energy load curve to obtain an electric energy deviation value sequence;

[0034] Determining whether the electric energy deviation value sequence exceeds a preset deviation value threshold range;

[0035] If the electric energy deviation value sequence exceeds the preset deviation value threshold range, the corresponding time point in the electric energy deviation value sequence is marked as an abnormal power consumption point, and the abnormal power consumption points are clustered based on a density clustering algorithm. The clustered abnormal power consumption point clusters are determined as abnormal power consumption events, and the abnormal power consumption events are sorted to obtain the abnormal power consumption behavior information.

[0036] Furthermore, the abnormal power usage behavior information includes abnormal power usage events, and the step of performing abnormal alarm according to the abnormal power usage behavior information specifically includes:

[0037] Extracting abnormal duration characteristics and abnormal fluctuation amplitude characteristics from the abnormal power consumption event;

[0038] Determine the abnormal event level according to the abnormal duration characteristics and the abnormal fluctuation amplitude characteristics;

[0039] An abnormal alarm information table is extracted from a database, the abnormal power consumption event and the abnormal event level are mapped to the abnormal alarm information table, abnormal alarm information is obtained, and an abnormal alarm is performed according to the abnormal alarm information.

[0040] In order to solve the above technical problems, the embodiment of the present application further provides an abnormality identification device, which adopts the following technical solution:

[0041] A data processing module is used to obtain historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data;

[0042] A data association module, used for acquiring external factor data, and associating the external factor data with the standard electric energy load data to obtain an electric energy load data set;

[0043] A model building module, used to analyze the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to build an electric power consumption mode model;

[0044] A model prediction module, used to perform electric energy load prediction based on the power consumption mode model and a preset regression algorithm to obtain predicted electric energy load information;

[0045] An abnormality analysis module, used to obtain real-time electric energy load information, compare the predicted electric energy load information with the real-time electric energy load information, and obtain abnormal power consumption behavior information;

[0046] The abnormal alarm module is used to make an abnormal alarm according to the abnormal power consumption behavior information.

[0047] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0048] A computer device comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of any of the above-mentioned abnormality identification methods when executing the computer-readable instructions.

[0049] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0050] A computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the anomaly identification method described in any one of the above items are implemented.

[0051] Compared with the prior art, the embodiments of the present application have the following beneficial effects: the embodiments obtain historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data; obtain external factor data, associate the external factor data with the standard electric energy load data, and obtain an electric energy load data set; analyze the electric energy load data set based on the time series analysis algorithm and the preset clustering algorithm, and construct an electric energy consumption pattern model; predict the electric energy load based on the electric energy consumption pattern model and the preset regression algorithm, and obtain predicted electric energy load information; obtain real-time electric energy load information, compare the predicted electric energy load information with the real-time electric energy load information, and obtain abnormal electric energy consumption behavior information; and perform abnormal alarm according to the abnormal electric energy consumption behavior information. This effectively realizes accurate identification of abnormal electric energy consumption behavior and timely alarm. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0054] Figure 2 A flowchart of an embodiment of an abnormality identification method according to the present application;

[0055] Figure 3 yes Figure 2 A flowchart of a specific implementation of step S20;

[0056] Figure 4 yes Figure 2 A flowchart of a specific implementation of step S30;

[0057] Figure 5 yes Figure 2 A flowchart of a specific implementation of step S40;

[0058] Figure 6 yes Figure 5 A flowchart of a specific implementation of step S402;

[0059] Figure 7 yes Figure 2 A flowchart of a specific implementation of step S50;

[0060] Figure 8 yes Figure 2 A flowchart of a specific implementation of step S60;

[0061] Fig. 9 is a schematic structural diagram of an embodiment of an abnormality identification device according to the present application;

[0062] Fig.10 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0064] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor does it refer to non-related or alternative embodiments that are mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0065] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0066] like Figure 1As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0067] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0068] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to a laptop computer 1011, a tablet computer 1012 or a mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0069] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0070] It should be noted that the anomaly identification method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the anomaly identification device is generally arranged in the server / terminal device.

[0071] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0072] Continue to refer Figure 2 , shows a flow chart of an embodiment of a method for abnormality identification according to the present application. The abnormality identification method comprises the following steps:

[0073] Step S10, obtaining historical electric energy load data, preprocessing the historical electric energy load data, and obtaining standard electric energy load data;

[0074] In this embodiment, historical electric energy load data refers to the record of the amount of electric energy load consumed by the power system or a specific user in the past period of time. The historical electric energy load data can be obtained from sources such as smart grid systems, energy management systems, smart meters, etc., and stored in a database, and extracted from the database by a load data extraction identifier. The historical electric energy load data includes a timestamp (the time point at which the load data is recorded, such as date, hour, etc.), a load value (indicating the amount of electric energy load at a specific time point, usually in kilowatts (kW) or megawatts (MW)), and user information (including user ID, geographic location, electricity type, etc.). The preprocessing of historical electric energy load data includes invalid data removal: identifying and deleting missing value records, identifying and deleting duplicate value records, and identifying and deleting data records that exceed a reasonable range.

[0075] Step S20, obtaining external factor data, and associating the external factor data with the standard electric energy load data to obtain an electric energy load data set;

[0076] In this embodiment, the external factor data includes weather data and holiday data, wherein the weather data includes temperature, humidity, precipitation, wind speed, wind direction, air pressure, cloud cover, sunshine duration, etc. The holiday data includes holiday time, holiday length, holiday attributes, etc. By associating the external factor data with the standard power load data, an electric power load data set that effectively shows the impact of weather and holidays on the user's power load is obtained.

[0077] Step S30, analyzing the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to construct an electric power consumption mode model;

[0078] In this embodiment, the time series analysis algorithm is used to extract time trend information and periodic features from the electric energy load data set. The time series analysis algorithm can be implemented based on the ARIMA model. The ARIMA model is a differential integrated moving average autoregressive model. By fitting the ARIMA model, trend information and periodic features can be effectively extracted from the electric energy load data set. The preset clustering algorithm can adopt the K-means clustering algorithm. The K-means clustering algorithm is a partition-based clustering algorithm. The cluster of each data point is determined by calculating the distance from each data point in the electric energy load data set to the cluster center, and the cluster center and cluster allocation are continuously iterated and updated until the convergence condition is met. The power consumption pattern model includes typical power consumption patterns of different user categories in different time periods, as well as the statistical characteristics and change laws of each pattern.

[0079] Step S40, performing electric energy load forecasting based on the power consumption pattern model and a preset regression algorithm to obtain forecasted electric energy load information;

[0080] In this embodiment, the preset regression algorithm can adopt a support vector regression algorithm, which is a regression analysis method based on a support vector machine (SVM), which attempts to find an optimal hyperplane so that all training data points are as close to the plane as possible while allowing a certain error range (i.e., ε-insensitive loss function). By introducing a kernel function, the support vector regression algorithm can handle complex nonlinear relationships and map the input data to a high-dimensional space, thereby finding the optimal fitting model for prediction.

[0081] Step S50, acquiring real-time electric energy load information, comparing the predicted electric energy load information with the real-time electric energy load information, and obtaining abnormal power consumption behavior information;

[0082] In this embodiment, abnormal electricity usage behavior information refers to abnormal electricity usage detected based on comparison, and real-time power load information refers to the power load information obtained in the current period. By calculating the power deviation value between the real-time power load information and the predicted power load information, and then comparing the power deviation value with the preset deviation value threshold range, abnormal electricity usage behavior information can be effectively judged.

[0083] Step S60: issuing an abnormal alarm based on the abnormal electricity usage behavior information.

[0084] In this embodiment, corresponding abnormal alarm information is obtained based on the abnormal power usage event and the abnormal event level corresponding to the abnormal power usage behavior information. The abnormal alarm information is an alarm prompt information preset according to the type of abnormal power usage event and the corresponding level. The abnormal alarm information includes the corresponding alarm object and alarm text content, which can be configured and adjusted according to actual conditions.

[0085] This embodiment obtains historical power load data, pre-processes the historical power load data, and obtains standard power load data; obtains external factor data, associates the external factor data with the standard power load data, and obtains a power load data set; analyzes the power load data set based on a time series analysis algorithm and a preset clustering algorithm to construct a power consumption pattern model; predicts power load based on the power consumption pattern model and a preset regression algorithm to obtain predicted power load information; obtains real-time power load information, compares the predicted power load information with the real-time power load information to obtain abnormal power consumption behavior information; and performs abnormal alarms based on the abnormal power consumption behavior information. This effectively realizes accurate identification of abnormal power consumption behavior and timely alarms.

[0086] This embodiment can be applied to risk early warning systems of power-related insurance products or insurance businesses, so as to effectively identify and alarm users' abnormal power usage behaviors, thereby improving users' power usage safety.

[0087] refer to Figure 3 In some optional implementations of this embodiment, step S20 includes the following steps:

[0088] Step S201, extracting the external factor data from a database according to the historical power load data, wherein the external factor data includes weather data and holiday data;

[0089] In this embodiment, weather data and holiday data recorded in the corresponding time period are extracted from the database according to the timestamp information corresponding to the historical power load data, thereby effectively obtaining the external factor data. The external factor data can also be extracted from the database according to the factor data extraction identifier corresponding to the external factor data, wherein the factor data extraction identifier is a unique identifier corresponding to the external factor data, and the factor data extraction identifier can be a data ID.

[0090] Step S202, aligning the timestamps of the weather data, the holiday data, and the standard power load data;

[0091] In this embodiment, by searching for a field representing a timestamp (a combination of date and time fields or a separate date field) in the standard power load data, weather data, and holiday data, and then aligning the timestamps of the weather data, holiday data, and standard power load data to the same time granularity (the time interval of the data), the timestamps of the weather data, holiday data, and standard power load data are aligned.

[0092] Step S203: taking the aligned standard power load data as a main data set, and adding the aligned weather data and the holiday data to the main data set to obtain the power load data set.

[0093] In this embodiment, the step of adding weather data and holiday data to the main data set includes confirming field matching: ensuring that the fields in the weather data and holiday data (such as temperature, humidity, wind speed, holiday date, etc.) can correctly match the timestamp field in the main data set; merging data sets: using data processing tools to merge weather data and holiday data into the main data set based on timestamps. This step involves a "connect" or "merge" operation, in which timestamp is the key field for the connection; processing missing values: if there is missing weather data in the merged data set, it is filled using an interpolation method. If a time point is covered by multiple holidays at the same time, duplicate values ​​are removed.

[0094] This embodiment extracts the external factor data from the database according to the historical power load data, wherein the external factor data includes weather data and holiday data; performs timestamp alignment on the weather data, the holiday data, and the standard power load data; uses the aligned standard power load data as a main data set, and adds the aligned weather data and the holiday data to the main data set, thereby effectively obtaining an power load data set that associates the external factor data and the standard power load data, so as to facilitate the subsequent construction of a power consumption pattern model.

[0095] refer to Figure 4 In some optional implementations of this embodiment, step S30 includes the following steps:

[0096] Step S301, extracting time trend information and periodic characteristics from the electric energy load data set according to a time series analysis algorithm, and generating a time dimension power consumption pattern according to the time trend information and the periodic characteristics;

[0097] In this embodiment, the time series analysis algorithm adopts the ARIMA algorithm, which can be implemented based on the ARIMA model. By inputting the electric energy load data set into the pre-trained ARIMA model, time trend information (such as long-term growth or decline trend) and periodic characteristics (such as daily cycle, weekly cycle, monthly cycle or annual cycle) can be effectively extracted. The step of generating the time dimension power consumption pattern includes pattern definition: based on the extracted features, the time dimension power consumption pattern is defined, and the definition includes power consumption behavior patterns in different time periods (such as weekdays, weekends, holidays), and the impact of seasonal changes on power consumption. Pattern representation: The time dimension power consumption pattern can be represented by means of charts, mathematical models or statistical distributions.

[0098] Step S302, performing user clustering on the electric energy load data set according to a preset clustering algorithm to obtain user electric energy load categories, and generating user-dimensional power consumption patterns according to the user electric energy load categories;

[0099] In this embodiment, the preset clustering algorithm may adopt a K-means clustering algorithm, and the K-means clustering algorithm is used to cluster the users corresponding to the electric energy load data set, and the characteristics of the clustering may include the user's power consumption, power consumption time distribution, power consumption peak and valley characteristics, etc. After the clustering is completed, the user's power load category is obtained, and then the corresponding user-dimensional power consumption mode is generated according to the power consumption preferences and power consumption habits corresponding to different user power load categories.

[0100] Step S303: construct the power usage pattern model according to the time dimension power usage pattern and the user dimension power usage pattern.

[0101] In this embodiment, the power consumption pattern of the time dimension and the power consumption pattern of the user dimension are weighted and fused to obtain the power consumption pattern model. The weights can be assigned according to the importance of the time dimension and the user dimension in the power consumption pattern analysis, and then the data corresponding to the power consumption pattern of the time dimension and the power consumption pattern of the user dimension are standardized (normalized, standardized, converted or projected, etc.) to ensure that they are compared and fused on the same scale. Finally, according to the determined weight allocation, the power consumption pattern of the time dimension and the power consumption pattern of the user dimension are weighted and fused, so as to effectively construct the power consumption pattern model.

[0102] This embodiment extracts time trend information and periodic features from the power load data set according to a time series analysis algorithm, and generates a time dimension power consumption pattern according to the time trend information and the periodic features; performs user clustering on the power load data set according to a preset clustering algorithm to obtain user power load categories, and generates user dimension power consumption patterns according to the user power load categories; and constructs the power consumption pattern model according to the time dimension power consumption pattern and the user dimension power consumption pattern. Thus, an power consumption pattern model integrating multiple dimensional power consumption patterns is effectively obtained, which provides reliable data support for subsequent power load prediction.

[0103] refer to Figure 5 In some optional implementations of this embodiment, step S40 includes the following steps:

[0104] Step S401, obtaining current factor data, and fusing the current factor data with the power consumption pattern model to obtain a power consumption prediction model;

[0105] In this embodiment, the current factor data includes the weather data corresponding to the current time period and the holiday conditions corresponding to the current time period, and the current factor data can be extracted from the database according to the current time period and the corresponding data extraction identifier. The step of fusing the current factor data with the power consumption pattern model includes: converting the current factor data into features that can be used for the model, which can be achieved through operations such as data conversion, encoding, and normalization. The processed current factor data is used as an input feature and is fused with the power consumption pattern model, which can be achieved by expanding the input dimension of the power consumption pattern model, using feature splicing, or through the embedding layer of the deep learning model.

[0106] Step S402, performing electric energy load forecasting based on a preset regression algorithm and the power consumption forecasting model, and optimizing the power consumption forecasting model according to the forecasting result to obtain an optimized power consumption forecasting model;

[0107] In this embodiment, the preset regression algorithm adopts a support vector regression algorithm, and the power load in the future period is predicted by the support vector regression algorithm and the power consumption prediction model to obtain the prediction result based on the current factor data prediction. The prediction result of the current factor data prediction is then tested according to the preset rationality condition, so that when the prediction result of the current factor data prediction does not meet the preset rationality condition, the preset regression algorithm and the power consumption prediction model are optimized to obtain the optimized power consumption prediction model.

[0108] Step S403: perform electric energy load forecasting based on the optimized electricity consumption forecasting model to obtain the forecasted electric energy load information.

[0109] In this embodiment, the power load forecast for a period of time in the future is performed based on the optimized power consumption forecast model, so as to effectively obtain the predicted power load information predicted based on the current factor data.

[0110] This embodiment obtains a page template extraction identifier, extracts the predefined page template from the database according to the page template extraction identifier; generates an initial adaptation page based on the predefined page template, the business data characteristics, and the device portrait information based on the template engine generator; determines whether the initial adaptation page meets the device compatibility requirements, and uses the initial adaptation page that meets the device compatibility requirements as the device adaptation page. In this way, a device adaptation page that matches and is compatible with the user's target device object is effectively obtained to facilitate subsequent optimization processing.

[0111] Continue to refer Figure 6 In some optional implementations of this embodiment, step S402 includes the following steps:

[0112] Step S4021, generating a corresponding predicted power load curve according to the prediction result;

[0113] In this embodiment, the prediction results are arranged in chronological order, and the load prediction values ​​at each time point are connected to generate a predicted power load curve. A chart tool can be used to connect the load prediction values ​​at each time point using a line chart (broken line chart) with the timestamp as the horizontal axis (X axis) and the load prediction value as the vertical axis (Y axis) to form a predicted power load curve.

[0114] Step S4022, obtaining a preset rationality condition, and determining whether the predicted power load curve meets the preset rationality condition;

[0115] In this embodiment, the preset rationality conditions include the smoothness of the load curve, the rationality of the peak and valley values, the continuity of load changes, etc. By comparing the smoothness of the predicted power load curve, the peak and valley values, and the continuity of changes with the preset rationality conditions, it is determined whether the predicted power load curve meets the preset rationality conditions.

[0116] Step S4023: if the predicted power load curve meets the preset rationality condition, the power consumption prediction model is used as a power consumption prediction model that does not need to be optimized to perform power load prediction;

[0117] In this embodiment, when the predicted power load curve meets the preset rationality condition, it means that the prediction result of the power consumption prediction model is valid. At this time, the power consumption prediction model can be directly used to perform power load prediction.

[0118] Step S4024: If the predicted power load curve does not meet the preset rationality condition, a prediction error is calculated for the predicted power load curve and the actual power load curve, and the algorithm parameters of the preset regression algorithm and the model parameters of the power consumption prediction model are adjusted according to the calculated prediction error to obtain the optimized power consumption prediction model.

[0119] In this embodiment, the difference between the load prediction value of the predicted power load curve at each time point and the actual load value of the actual power load curve at each time point is calculated to obtain the calculated prediction error. Then, the algorithm parameters of the preset regression algorithm, such as the learning rate, regularization parameter, etc., are adjusted according to the prediction error to improve the prediction performance of the algorithm. At the same time, the model parameters of the power consumption prediction model, such as feature selection, model structure, etc., can also be adjusted to improve the accuracy and generalization ability of the model, so as to obtain the optimized power consumption prediction model.

[0120] This embodiment generates a corresponding predicted power load curve according to the prediction result; obtains a preset rationality condition, and determines whether the predicted power load curve meets the preset rationality condition; if the predicted power load curve meets the preset rationality condition, the power consumption prediction model is used as a power consumption prediction model that does not need to be optimized to predict the power load; if the predicted power load curve does not meet the preset rationality condition, the prediction error of the predicted power load curve and the actual power load curve is calculated, and the algorithm parameters of the preset regression algorithm and the model parameters of the power consumption prediction model are adjusted according to the calculated prediction error to obtain the optimized power consumption prediction model. This effectively realizes the optimization of the power consumption prediction model to improve the accuracy of the power load prediction.

[0121] Continue to refer Figure 7 In some optional implementations of this embodiment, step S50 includes the following steps:

[0122] Step S501, aligning the predicted electric energy load information and the real-time electric energy load information on a time axis to obtain an aligned load information set;

[0123] In this embodiment, the predicted electric energy load information and the real-time electric energy load information are matched in terms of timestamps to ensure alignment at the same time point, thereby obtaining an aligned load information set after time axis alignment is completed.

[0124] Step S502, performing time granularity regularization on the aligned load information set according to a dynamic time regularization algorithm to obtain a regularized load information set;

[0125] In this embodiment, the time granularity of the aligned load information set is warped by a dynamic time warping (DTW) algorithm, thereby effectively finding the optimal alignment between the two time series, even if they have different speeds or forms at different time points, and the time granularity of the two sets of data in the aligned load information set is warped to the same time interval, so that the two sets of data in the aligned load information set are more matched on the time axis.

[0126] Step S503, fitting the regularized load information set according to a curve fitting algorithm to obtain a predicted electric energy load curve and a real-time electric energy load curve;

[0127] In this embodiment, the curve fitting algorithm can use polynomial fitting to fit the regularized predicted power load information and real-time power load information. The polynomial fitting algorithm uses the least squares method to minimize the error between the fitting curve and the actual data points, thereby effectively obtaining the predicted power load curve and the real-time power load curve that represent the regular load information set in the form of a curve.

[0128] Step S504, calculating the time point deviation value between the predicted electric energy load curve and the real-time electric energy load curve to obtain an electric energy deviation value sequence;

[0129] In this embodiment, by aligning the time points of the predicted power load curve and the real-time power load curve, and then calculating the difference between the predicted load value of the predicted power load curve and the actual load value of the real-time power load curve at the aligned time point, the difference can be calculated according to the Euclidean distance measurement method, so as to effectively quantify the degree of deviation and obtain the time point deviation value corresponding to each time point. After obtaining the time point deviation value, the time point deviation value is sorted in the order of the time points to obtain the power deviation value sequence.

[0130] Step S505, determining whether the electric energy deviation value sequence exceeds a preset deviation value threshold range;

[0131] In this embodiment, the time point deviation value corresponding to each time point in the electric energy deviation value sequence is compared with the preset deviation value threshold range one by one, so as to identify the time point at which the time point deviation value is abnormal. The above-mentioned time point deviation value can be expressed as a specific load value, or as a percentage value based on the predicted load value. When expressed as a percentage value, the preset deviation value threshold range is 50%, which can be adjusted accordingly according to the actual situation.

[0132] Step S506: if the electric energy deviation value sequence exceeds the preset deviation value threshold range, the corresponding time point in the electric energy deviation value sequence is marked as an abnormal power consumption point, and the abnormal power consumption point is clustered based on a density clustering algorithm, and the abnormal power consumption point cluster obtained by clustering is determined as an abnormal power consumption event, and the abnormal power consumption event is sorted to obtain the abnormal power consumption behavior information;

[0133] In this embodiment, time point marking can be achieved by adding the corresponding time points in the electric energy deviation value sequence to the power consumption abnormal point record table. The density clustering algorithm can adopt the DBSCAN clustering algorithm. The DBSCAN clustering algorithm can recognize clusters of any shape and does not require the number of clusters to be specified in advance. By taking the power consumption abnormal points as input data and applying the density clustering algorithm for clustering, one or more power consumption abnormal point clusters are obtained, and each cluster represents a potential power consumption abnormal event.

[0134] Step S507: if the electric energy deviation value sequence does not exceed the preset deviation value threshold range, then the corresponding time point in the electric energy deviation value sequence is marked as a normal power consumption point.

[0135] In this embodiment, the time point marking can be achieved by adding the corresponding time point in the electric energy deviation value sequence to the normal power consumption point record table.

[0136] In this embodiment, the predicted electric energy load information and the real-time electric energy load information are aligned on the time axis to obtain an aligned load information set; the aligned load information set is time-granularly regularized according to a dynamic time regularization algorithm to obtain a regularized load information set; the regularized load information set is fitted according to a curve fitting algorithm to obtain a predicted electric energy load curve and a real-time electric energy load curve; the time point deviation value of the predicted electric energy load curve and the real-time electric energy load curve is calculated to obtain an electric energy deviation value sequence; it is determined whether the electric energy deviation value sequence exceeds a preset deviation value threshold range; if the electric energy deviation value sequence exceeds the preset deviation value threshold range, the corresponding time point in the electric energy deviation value sequence is marked as an abnormal power consumption point, and the abnormal power consumption point is clustered based on a density clustering algorithm, the abnormal power consumption point cluster obtained by clustering is determined as an abnormal power consumption event, and the abnormal power consumption event is sorted to obtain the abnormal power consumption behavior information; if the electric energy deviation value sequence does not exceed the preset deviation value threshold range, the corresponding time point in the electric energy deviation value sequence is marked as a normal power consumption point. This can effectively identify abnormal power consumption in real-time power load information and obtain accurate and effective abnormal power consumption behavior information.

[0137] Continue to refer Figure 8 In some optional implementations of this embodiment, step S60 includes the following steps:

[0138] Step S601, extracting abnormal duration characteristics and abnormal fluctuation amplitude characteristics from the abnormal power consumption event;

[0139] In this embodiment, the abnormal duration characteristic can be obtained by calculating the difference between the start time and the end time of the abnormal event, and the abnormal fluctuation amplitude characteristic can be obtained by calculating the difference between the maximum and minimum values ​​of its load deviation (or the difference in absolute values).

[0140] Step S602, determining the abnormal event level according to the abnormal duration characteristic and the abnormal fluctuation amplitude characteristic;

[0141] In this embodiment, different grading standards can be set according to the combination of abnormal duration and abnormal fluctuation amplitude. For example, the following grades can be set: Grade 1 (minor): the abnormal duration is short and the fluctuation amplitude is small. Grade 2 (general): the abnormal duration is long or the fluctuation amplitude is large, but it does not reach the serious level. Grade 3 (severe): the abnormal duration is long and the fluctuation amplitude is large, which may have a significant impact on the safety of the power grid or the power consumption of users. Conditional statements or rule engines can be used to determine the level of each abnormal power consumption event based on the extracted feature values.

[0142] Step S603, extracting an abnormal alarm information table from a database, mapping the abnormal power consumption event and the abnormal event level to the abnormal alarm information table, obtaining abnormal alarm information, and performing an abnormal alarm according to the abnormal alarm information.

[0143] In this embodiment, the abnormal alarm information table is a preset information table recording abnormal alarm information, and the abnormal alarm information is set corresponding to the abnormal power consumption event and the abnormal event level. The abnormal alarm can be generated by filling the fields corresponding to the abnormal power consumption event and the abnormal event level into the alarm template through a preset alarm template, and the abnormal alarm information is sent to the corresponding email address to realize the alarm.

[0144] This embodiment extracts abnormal duration characteristics and abnormal fluctuation amplitude characteristics from the abnormal power consumption event; determines the abnormal event level according to the abnormal duration characteristics and the abnormal fluctuation amplitude characteristics; extracts the abnormal alarm information table from the database, maps the abnormal power consumption event and the abnormal event level to the abnormal alarm information table, obtains abnormal alarm information, and performs abnormal alarm according to the abnormal alarm information. Thus, timely alarm of abnormal power consumption behavior is effectively realized.

[0145] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions 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-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0146] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0147] Further references Fig. 9 , as a response to the above Figure 1In order to realize the method shown in the figure, the present application provides an embodiment of an abnormality identification device, and the device embodiment is Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0148] like Fig. 9 As shown, the abnormality identification device 700 described in this embodiment includes: a data processing module 701, a data association module 702, a model construction module 703, a model prediction module 704, an abnormality analysis module 705, and an abnormality alarm module 706. Among them:

[0149] The data processing module 701 is used to obtain historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data;

[0150] The data association module 702 is used to obtain external factor data, and associate the external factor data with the standard electric energy load data to obtain an electric energy load data set;

[0151] A model building module 703 is used to analyze the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to build an electric power consumption mode model;

[0152] A model prediction module 704 is used to perform electric energy load prediction based on the power consumption mode model and a preset regression algorithm to obtain predicted electric energy load information;

[0153] The abnormality analysis module 705 is used to obtain real-time power load information, compare the predicted power load information with the real-time power load information, and obtain abnormal power consumption behavior information;

[0154] The abnormal alarm module 706 is used to generate an abnormal alarm based on the abnormal power usage behavior information.

[0155] By adopting the above-mentioned abnormality identification device, this embodiment can obtain historical power load data, pre-process the historical power load data, and obtain standard power load data; obtain external factor data, associate the external factor data with the standard power load data, and obtain a power load data set; analyze the power load data set based on the time series analysis algorithm and the preset clustering algorithm to construct a power consumption pattern model; predict the power load based on the power consumption pattern model and the preset regression algorithm to obtain predicted power load information; obtain real-time power load information, compare the predicted power load information with the real-time power load information to obtain abnormal power consumption behavior information; and make abnormal alarms based on the abnormal power consumption behavior information. This effectively realizes accurate identification of abnormal power consumption behavior and timely alarms.

[0156] To solve the above technical problems, the present application also provides a computer device. Fig.10 , Fig.10 This is a basic structural block diagram of the computer device in this embodiment.

[0157] The computer device 8 includes a memory 81, a processor 82, and a network interface 83 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 8 with components 81-83, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0158] The computer device may be a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with the user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0159] The memory 81 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 81 can be an internal storage unit of the computer device 8, such as a hard disk or memory of the computer device 8. In other embodiments, the memory 81 can also be an external storage device of the computer device 8, such as a plug-in hard disk equipped on the computer device 8, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Of course, the memory 81 can also include both the internal storage unit of the computer device 8 and its external storage device. In this embodiment, the memory 81 is generally used to store the operating system and various application software installed on the computer device 8, such as computer-readable instructions of the abnormality identification method, etc. In addition, the memory 81 can also be used to temporarily store various types of data that have been output or are to be output.

[0160] The processor 82 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor 82 is generally used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to run computer-readable instructions or process data stored in the memory 81, such as computer-readable instructions for running the abnormality identification method.

[0161] The network interface 83 may include a wireless network interface or a wired network interface. The network interface 83 is generally used to establish a communication connection between the computer device 8 and other electronic devices.

[0162] By adopting the above-mentioned computer equipment, this embodiment can obtain historical power load data, pre-process the historical power load data, and obtain standard power load data; obtain external factor data, associate the external factor data with the standard power load data, and obtain a power load data set; analyze the power load data set based on the time series analysis algorithm and the preset clustering algorithm to construct a power consumption pattern model; predict the power load based on the power consumption pattern model and the preset regression algorithm to obtain predicted power load information; obtain real-time power load information, compare the predicted power load information with the real-time power load information to obtain abnormal power consumption behavior information; and make abnormal alarms based on the abnormal power consumption behavior information. This effectively realizes accurate identification of abnormal power consumption behavior and timely alarms.

[0163] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned abnormality identification method.

[0164] By adopting the above-mentioned computer-readable storage medium, this embodiment can obtain historical power load data, pre-process the historical power load data, and obtain standard power load data; obtain external factor data, associate the external factor data with the standard power load data, and obtain a power load data set; analyze the power load data set based on the time series analysis algorithm and the preset clustering algorithm to construct a power consumption pattern model; predict the power load based on the power consumption pattern model and the preset regression algorithm to obtain predicted power load information; obtain real-time power load information, compare the predicted power load information with the real-time power load information to obtain abnormal power consumption behavior information; and make abnormal alarms based on the abnormal power consumption behavior information. This effectively realizes accurate identification of abnormal power consumption behavior and timely alarms.

[0165] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0166] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A method for identifying anomalies, characterized in that: The steps include: Acquire historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data; Acquire external factor data, and associate the external factor data with the standard electric energy load data to obtain an electric energy load data set; Analyzing the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to construct an electric power consumption pattern model; Performing electric energy load forecasting based on the power consumption pattern model and a preset regression algorithm to obtain forecasted electric energy load information; Acquire real-time electric energy load information, compare the predicted electric energy load information with the real-time electric energy load information, and obtain abnormal power consumption behavior information; An abnormal alarm is issued according to the abnormal electricity usage behavior information.

2. The abnormality identification method according to claim 1, characterized in that: The step of acquiring the external factor data, associating the external factor data with the standard electric energy load data, and obtaining the electric energy load data set specifically includes: Extracting the external factor data from a database according to the historical power load data, wherein the external factor data includes weather data and holiday data; Performing time stamp alignment on the weather data, the holiday data, and the standard electric energy load data; The aligned standard electric energy load data is used as a main data set, and the aligned weather data and holiday data are added to the main data set to obtain the electric energy load data set.

3. The abnormality identification method according to claim 1, characterized in that: The step of analyzing the electric energy load data set based on the time series analysis algorithm and the preset clustering algorithm to construct the power consumption mode model specifically includes: Extracting time trend information and periodic characteristics from the electric energy load data set according to a time series analysis algorithm, and generating a time dimension electricity consumption pattern according to the time trend information and the periodic characteristics; Perform user clustering on the electric energy load data set according to a preset clustering algorithm to obtain user electric energy load categories, and generate user-dimensional power consumption patterns according to the user electric energy load categories; The power usage pattern model is constructed according to the time dimension power usage pattern and the user dimension power usage pattern.

4. The abnormality identification method according to claim 1, characterized in that: The step of performing electric energy load prediction based on the power consumption mode model and the preset regression algorithm to obtain predicted electric energy load information specifically includes: Acquire current factor data, and fuse the current factor data with the power consumption pattern model to obtain a power consumption prediction model; Performing electric energy load forecasting based on a preset regression algorithm and the power consumption forecasting model, and optimizing the power consumption forecasting model according to the forecasting result to obtain an optimized power consumption forecasting model; The power load is predicted based on the optimized power consumption prediction model to obtain the predicted power load information.

5. The abnormality identification method according to claim 4, characterized in that: The step of performing electric energy load forecasting based on the preset regression algorithm and the power consumption forecasting model, and optimizing the power consumption forecasting model according to the forecasting result to obtain the optimized power consumption forecasting model specifically includes: Generate a corresponding predicted electricity load curve according to the prediction result; Obtaining a preset rationality condition, and determining whether the predicted power load curve meets the preset rationality condition; If the predicted electricity load curve does not meet the preset rationality condition, a prediction error is calculated for the predicted electricity load curve and the actual electricity load curve, and the algorithm parameters of the preset regression algorithm and the model parameters of the electricity consumption prediction model are adjusted according to the calculated prediction error to obtain the optimized electricity consumption prediction model.

6. The abnormality identification method according to claim 1, characterized in that: The step of acquiring real-time electric energy load information, comparing the predicted electric energy load information with the real-time electric energy load information, and obtaining abnormal electricity consumption behavior information specifically includes: Aligning the predicted electric energy load information and the real-time electric energy load information on a time axis to obtain an aligned load information set; Performing time granularity regularization on the aligned load information set according to a dynamic time regularization algorithm to obtain a regularized load information set; Fitting the regularized load information set according to a curve fitting algorithm to obtain a predicted electric energy load curve and a real-time electric energy load curve; Calculating the time point deviation value between the predicted electric energy load curve and the real-time electric energy load curve to obtain an electric energy deviation value sequence; Determining whether the electric energy deviation value sequence exceeds a preset deviation value threshold range; If the electric energy deviation value sequence exceeds the preset deviation value threshold range, the corresponding time point in the electric energy deviation value sequence is marked as an abnormal power consumption point, and the abnormal power consumption points are clustered based on a density clustering algorithm. The clustered abnormal power consumption point clusters are determined as abnormal power consumption events, and the abnormal power consumption events are sorted to obtain the abnormal power consumption behavior information.

7. The abnormality identification method according to claim 1, characterized in that: The abnormal power usage behavior information includes abnormal power usage events, and the step of performing abnormal alarm according to the abnormal power usage behavior information specifically includes: Extracting abnormal duration characteristics and abnormal fluctuation amplitude characteristics from the abnormal power consumption event; Determine the abnormal event level according to the abnormal duration characteristics and the abnormal fluctuation amplitude characteristics; An abnormal alarm information table is extracted from a database, the abnormal power consumption event and the abnormal event level are mapped to the abnormal alarm information table, abnormal alarm information is obtained, and an abnormal alarm is performed according to the abnormal alarm information.

8. An abnormality identification device, characterized in that: include: A data processing module is used to obtain historical electric energy load data, pre-process the historical electric energy load data, and obtain standard electric energy load data; A data association module, used for acquiring external factor data, and associating the external factor data with the standard electric energy load data to obtain an electric energy load data set; A model building module, used to analyze the electric energy load data set based on a time series analysis algorithm and a preset clustering algorithm to build an electric power consumption mode model; A model prediction module, used to perform electric energy load prediction based on the power consumption mode model and a preset regression algorithm to obtain predicted electric energy load information; An abnormality analysis module, used to obtain real-time electric energy load information, compare the predicted electric energy load information with the real-time electric energy load information, and obtain abnormal power consumption behavior information; The abnormal alarm module is used to make an abnormal alarm according to the abnormal power consumption behavior information.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the abnormality identification method according to any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the abnormality identification method according to any one of claims 1 to 7 are implemented.

Images