Data trend analysis system, method and medium for regional electricity users
By converting time-series data of electricity-consuming enterprises, generating target power time-series data, decomposing time series data, and calculating trends through bilateral detection, a trend chart of electricity-consuming enterprise data is generated, which solves the problem of accuracy in trend analysis of electricity-consuming enterprise data and enables accurate analysis and prediction of the power system.
Patent Information
- Application Number
- CN202410029627.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-08
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-01-08
AI Technical Summary
Existing data trend analysis technologies in the power industry are insufficient to discover new patterns and trends in the data of electricity-consuming enterprises, resulting in low analysis accuracy.
The system employs a power consumption enterprise time-series data conversion module, a target power time-series data generation module, a time series decomposition module, and a bilateral detection trend calculation module. Through data augmentation, anomaly detection, time series decomposition, and trend calculation, it generates trend charts of power consumption enterprise data for analysis.
It improves the accuracy of data trend analysis for electricity-consuming enterprises, enabling timely detection of abnormal events and faults in the power system, understanding of the basic characteristics and patterns of data, prediction of electricity demand and supply, and optimization of power system planning and dispatch.
Smart Images

Figure CN118014386B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data analysis, and in particular to a system, method and medium for analyzing data trends of regional electricity users. Background Art
[0002] With the rapid development of economy and society, the power industry has gradually become one of the key supporting industries for the country's economic development. The power industry has become one of the most important basic industries in the national economy. It is an indispensable energy supply guarantee for promoting the development of various fields such as industry, agriculture, service industry and information construction. With the continuous growth of electricity demand, the complexity and operation difficulty of the power system are also increasing. Therefore, it is necessary to conduct data trend analysis on the power industry to ensure the steady operation of the power industry.
[0003] Most of the existing data trend technologies in the power industry analyze the historical data of power users to find out the laws and trends. In practical applications, the analysis of historical data can only analyze and predict the known laws and trends, and it is difficult to discover new patterns and trends, which leads to overly one-sided analysis of data trends of power users, resulting in low accuracy in data trend analysis of power users. Summary of the invention
[0004] The present invention provides a system, method and medium for regional power consumption enterprise data trend analysis, the main purpose of which is to solve the problem of accuracy when performing power consumption enterprise data trend analysis.
[0005] To achieve the above-mentioned purpose, the present invention provides a data trend analysis system for regional power users, the system includes a power user time series data conversion module, a target power time series data generation module, a time series decomposition module, a bilateral detection trend calculation module and a power user data trend analysis module, wherein:
[0006] The power consumption enterprise time series data conversion module is used to obtain the power consumption enterprise data of the target area, perform data enhancement on the power consumption enterprise data to obtain the power consumption enterprise enhanced data, and convert the power consumption enterprise enhanced data into the power consumption enterprise time series data;
[0007] The target power time series data generating module is used to detect abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data by using a preset dual weight algorithm, and generate target power time series data corresponding to the power consuming enterprise time series data according to the abnormal time series data of the power consuming enterprise, wherein when the abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data is detected by using a preset dual weight algorithm, it is specifically used to:
[0008] Calculate the power data mean of the power consumption enterprise time series data; calculate the power weight of each time series data in the power consumption enterprise time series data according to the power data mean using the following dual weight algorithm:
[0009]
[0010] Among them, W ni is the power weight of the index data of the nth power-consuming enterprise at the i-th moment, x ni is the power value of the index data of the nth power-consuming enterprise at the i-th moment, B is the mean value of the power data, k ni is the position number of the index data of the nth power consumer at the i-th moment, A is the position number of the most recent time series data corresponding to the mean value of the power data, and e is the sequence length of the time series data of the power consumer;
[0011] Selecting the time series data whose power weight is less than a preset weight threshold as abnormal time series data;
[0012] Gathering the abnormal time series data into abnormal time series data of the power consuming enterprise;
[0013] The time series decomposition module is used to perform time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data;
[0014] The bilateral detection trend calculation module is used to generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics;
[0015] The power consumption enterprise data trend analysis module is used to generate a power data trend matrix based on the power time series data and the bilateral detection trend, generate a trend chart of the power consumption enterprise data based on the power data trend matrix, and perform trend analysis on the power consumption enterprise data using the trend chart.
[0016] Optionally, when the power consumption enterprise time series data conversion module performs data enhancement on the power consumption enterprise data to obtain the power consumption enterprise enhanced data, it is specifically used to:
[0017] Performing numerical quantization processing on the electricity-consuming enterprise data to obtain first electricity-consuming enterprise data;
[0018] Processing missing values on the first electricity consuming enterprise data to obtain second electricity consuming enterprise data;
[0019] Performing outlier processing on the second power consuming enterprise data to obtain third power consuming enterprise data;
[0020] The third electricity consuming enterprise data is standardized to obtain enhanced electricity consuming enterprise data.
[0021] Optionally, when converting the power consumption enterprise enhanced data into the power consumption enterprise time series data, the power consumption enterprise time series data conversion module is specifically used to:
[0022] Sorting the enhanced data of the power consuming enterprises in time order according to preset timestamps to obtain an initial data sequence of the power consuming enterprises;
[0023] The timestamps are matched one-to-one with the enhanced data of the power user enterprises in the initial power user enterprise data sequence to obtain time series corresponding data;
[0024] The time series corresponding data is converted into time series data of the electricity consuming enterprise according to the preset time granularity.
[0025] Optionally, when the time series decomposition module performs time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data, it is specifically used to:
[0026] The target power time series data is preliminarily decomposed using the following addition decomposition formula to obtain the target power time series decomposition data:
[0027] Y t =R t +Q t
[0028] Among them, Y t is the power observation value of the target power time series data at time t, R t is the power trend value of the target power time series data at time t, Q t is the power fluctuation value of the target power time series data at time t;
[0029] Performing a smoothing operation on the target power time series data according to a preset decomposition window to obtain target power smoothed time series data;
[0030] The trend value is determined according to the power observation value of the target power smoothing time series data and the power trend value of the target power time series decomposition data, wherein the trend value calculation formula is:
[0031]
[0032] Among them, R' t is the trend value at time t, Y t is the power observation value of the target power time series data at time t, is the initial fluctuation value at time t in the f+1th cycle, is the smoothed fluctuation value at time t in the f+1th cycle;
[0033] The trend value is numerically fitted to obtain trend component data of the target power time series data.
[0034] Optionally, when generating power data statistics through the time series trend data of the power consuming enterprise, the bilateral detection trend calculation module is specifically used to:
[0035] The symbol value of the time series trend data of the electricity consuming enterprise is calculated by the following preset symbol function:
[0036]
[0037] Where H is the symbol value, y j is the time series data value of the index data of the nth power user in the time series data of the power user at the jth moment, y n(j+1) is the time series data value of the index data of the nth power consuming enterprise in the time series data of the power consuming enterprise at the j+1th moment;
[0038] Generate a symbol sequence corresponding to the time series trend data of the electricity consuming enterprise according to the symbol value;
[0039] Calculate the power data statistics of the power consumption enterprise's time series trend data according to the symbol sequence:
[0040]
[0041] Where S is the power data statistic, H d is the dth symbol value in the symbol sequence, and m is the sequence length of the symbol sequence.
[0042] Optionally, when calculating the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics, the bilateral detection trend calculation module is specifically used to:
[0043] Calculating the power data variance of the time series trend data of the power consuming enterprise according to the symbol sequence;
[0044] The bilateral test value of the time series data of the power consuming enterprise is calculated according to the power data statistics and the power data variance:
[0045]
[0046] Wherein, Z is the bilateral check value, S is the power data statistic, and D is the power data variance;
[0047] The bilateral detection trend of the time series data of the electricity consuming enterprise is determined according to the bilateral verification value.
[0048] Optionally, when the power consumption enterprise data trend analysis module generates a power data trend matrix according to the power time series data and the bilateral detection trend, it is specifically used to:
[0049] Classify the power time series data into power data types to obtain power time series category data;
[0050] extracting a bilateral category detection trend from the bilateral detection trend;
[0051] A power data trend matrix is generated according to the power time series category data and the bilateral category detection trend.
[0052] Optionally, when the power consumption enterprise data trend analysis module uses the trend chart to perform trend analysis on the power consumption enterprise data, it is specifically used to:
[0053] Determining the long-term trend of the electricity-consuming enterprise data according to the overall change trend of the trend graph;
[0054] Determine the periodic trend of the electricity-consuming enterprise data according to the peaks and valleys of the trend graph;
[0055] Determining the mutation trend of the electricity consuming enterprise data according to the mutation point of the trend graph;
[0056] The multi-dimensional trend changes of the electricity consuming enterprise data are determined through the long-term trend, the periodic trend and the sudden change trend.
[0057] In order to solve the above problems, the present invention also provides a method for analyzing data trends of regional power users, the method comprising:
[0058] Acquire the data of power consuming enterprises in the target area, perform data enhancement on the data of the power consuming enterprises to obtain enhanced data of the power consuming enterprises, and convert the enhanced data of the power consuming enterprises into time series data of the power consuming enterprises;
[0059] Detect abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise by using a preset double weight algorithm, and generate target power time series data corresponding to the time series data of the power consuming enterprise according to the abnormal time series data of the power consuming enterprise;
[0060] Performing time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data;
[0061] Generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics;
[0062] A power data trend matrix is generated according to the power time series data and the bilateral detection trend, a trend graph of the power consuming enterprise data is generated according to the power data trend matrix, and a trend analysis of the power consuming enterprise data is performed using the trend graph.
[0063] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned method for analyzing data trends of regional electricity users.
[0064] The embodiment of the present invention can timely discover abnormal events or faults in the power system by detecting abnormal conditions in the time series data of the power user; generate target power time series data according to the abnormal time series data, and understand the basic characteristics and laws of the power user data by analyzing the target power time series data; decompose the target power time series data into trend component data and other component data through time series decomposition technology, and the trend component data reflects the trend of long-term changes in the power system, which is conducive to judging the development trend of the power market and the growth trend of load demand; generate power user time series trend data according to the trend component data, and calculate the bilateral detection trend of the time series data of the power user through statistics, and intuitively display the development trend, periodic changes and other important information of the power user through trend charts and statistics; use trend charts to analyze the data of power users, which can help decision makers better understand the development dynamics of the power market, predict future power demand and supply, and optimize the planning and scheduling of the power system. Therefore, the system, method and medium for regional power user data trend analysis proposed by the present invention can solve the problem of low accuracy when performing power user data trend analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] Figure 1 A functional module diagram of a regional power-consuming enterprise data trend analysis system provided by an embodiment of the present invention;
[0066] Figure 2 A flowchart of an operating method of a regional electricity-consuming enterprise data trend analysis system provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION
[0067] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0068] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0069] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.
[0070] As used herein, the words "if" and "if" may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.
[0071] In addition, the step sequence in the following method embodiments is only an example and not a strict limitation.
[0072] In fact, the server-side device deployed for the regional power consumption enterprise data trend analysis system may be composed of one or more devices. The above-mentioned regional power consumption enterprise data trend analysis system can be implemented as: business instance, virtual machine, hardware device. For example, the regional power consumption enterprise data trend analysis system can be implemented as a business instance deployed on one or more devices in the cloud node. In simple terms, the regional power consumption enterprise data trend analysis system can be understood as a software deployed on the cloud node, which is used to provide a regional power consumption enterprise data trend analysis system for each user terminal. Alternatively, the regional power consumption enterprise data trend analysis system can also be implemented as a virtual machine deployed on one or more devices in the cloud node. The virtual machine is installed with application software for managing each user terminal. Alternatively, the regional power consumption enterprise data trend analysis system can also be implemented as a server consisting of many hardware devices of the same or different types, and one or more hardware devices are set to provide a regional power consumption enterprise data trend analysis system for each user terminal.
[0073] In terms of implementation, the data trend analysis system for regional power users and the user end are adapted to each other. That is, the data trend analysis system for regional power users is an application installed on the cloud service platform, and the user end is a client that establishes a communication connection with the application; or the data trend analysis system for regional power users is implemented as a website, and the user end is implemented as a web page; or the data trend analysis system for regional power users is implemented as a cloud service platform, and the user end is implemented as a small program in an instant messaging application.
[0074] Reference Figure 1 1 is a functional module diagram of a regional electricity-consuming enterprise data trend analysis system provided by an embodiment of the present invention.
[0075] The data trend analysis system 100 for regional power users of the present invention can be set in a cloud server. In terms of implementation, it can be used as one or more service devices, or it can be installed as an application on the cloud (such as a server of a mobile service operator, a server cluster, etc.), or it can be developed as a website. According to the functions implemented, the data trend analysis system 100 for regional power users can include a power user time series data conversion module 101, a target power time series data generation module 102, a time series decomposition module 103, a bilateral detection trend calculation module 104 and a power user data trend analysis module 105. The module described in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by a device processor and can complete fixed functions, which are stored in the memory of the device.
[0076] In the embodiment of the present invention, in the data trend analysis system for regional power users, each of the above modules can be implemented independently and called with other modules. The call here can be understood as a module that can connect to multiple modules of another type and provide corresponding services to the multiple modules connected to it. For example, the time series decomposition module can call the target power time series data generation module to obtain the information collected by the target power time series data generation module. Based on the above characteristics, in the data trend analysis system for regional power users provided by the embodiment of the present invention, the scope of application of the data trend analysis system architecture for regional power users can be adjusted by adding modules and directly calling them without modifying the program code, so as to achieve cluster-based horizontal expansion, so as to achieve the purpose of quickly and flexibly expanding the data trend analysis system for regional power users. In actual applications, the above modules can be set in the same device or different devices, or in virtual devices, such as service instances in cloud servers.
[0077] The following describes the various components and specific workflows of the regional power consumption enterprise data trend analysis system in conjunction with specific embodiments:
[0078] The power consumption enterprise time series data conversion module 101 is used to obtain power consumption enterprise data in a target area, perform data enhancement on the power consumption enterprise data to obtain power consumption enterprise enhanced data, and convert the power consumption enterprise enhanced data into power consumption enterprise time series data.
[0079] In an embodiment of the present invention, the electricity consuming enterprise data refers to data related to regional electricity consuming enterprises, including information on power generation, electricity consumption, electricity price, market size, voltage, grid frequency, etc., wherein the electricity consuming enterprise data of the target area can be obtained from a pre-stored storage area through a computer statement with a data capture function (such as Java statements, Python statements, etc.), wherein the storage area includes but is not limited to a database and a blockchain.
[0080] Furthermore, in order to obtain reliable power data for trend analysis of power consumption enterprise data, it is necessary to clean and process the power consumption enterprise data.
[0081] In the embodiment of the present invention, the power consumer enterprise enhanced data is the power consumer enterprise data after data preprocessing of the originally collected power consumer enterprise data to ensure the integrity of the power consumer enterprise data.
[0082] In the embodiment of the present invention, when the power consumption enterprise time series data conversion module 101 performs data enhancement on the power consumption enterprise data to obtain the power consumption enterprise enhanced data, it is specifically used to:
[0083] Performing numerical quantization processing on the electricity-consuming enterprise data to obtain first electricity-consuming enterprise data;
[0084] Processing missing values on the first electricity consuming enterprise data to obtain second electricity consuming enterprise data;
[0085] Performing outlier processing on the second power consuming enterprise data to obtain third power consuming enterprise data;
[0086] The third electricity consuming enterprise data is standardized to obtain enhanced electricity consuming enterprise data.
[0087] In detail, first, the data in the power consumer data that are not expressed numerically are quantified into power consumer data expressed numerically. For example, the market size in the power consumer data is not expressed numerically, then the market size needs to be converted into data expressed numerically. A large market size can be represented by a value of 1, and a small market size can be represented by a value of -1. For data in the power consumer data that are already expressed numerically, there is no need to perform numerical quantization processing. Then, the power consumer data after numerical quantization processing is used as the first power consumer data, thereby performing missing value, outlier and standardization processing on the first power consumer data.
[0088] Specifically, the pattern of missing data is identified through statistical methods or visualization tools (such as heat maps). If the proportion of missing data is very small, you can consider directly deleting the missing samples or features. For numerical data, that is, the data of the first power user can fill the missing values through interpolation methods (such as mean, median, linear interpolation, etc.); after the missing value processing of the first power user data, outlier processing is also required, that is, the outliers are detected by using statistical methods, box plots, etc. for the second power user data, and outliers that exceed a certain range are truncated or processed through conversion methods to alleviate the impact of outliers on the model. For skewed distribution data, you can use The median replaces the mean to reduce the impact of outliers. In addition, the duplicate value detection method can be used to find and remove duplicate records in the data set. For duplicate records with the same identification features but different other features, one of them can be retained and the other duplicate records can be deleted. Finally, the data of the third power user is standardized. That is, for numerical features, standardization (Z-score standardization) is used to convert the data into a standard normal distribution with a mean of 0 and a standard deviation of 1, and the feature values are scaled to a fixed range, usually [0,1]. This is very useful for algorithms that need to consider the weights between features (such as KNN), thereby obtaining enhanced data for power users.
[0089] Furthermore, after preprocessing the data of power users, in order to better discover the changing trends of the data of power users, it is necessary to time-series the data of power users. Time series data can better reflect the changing trends in the power system. Through time series data analysis, we can better understand the changing laws, trends and periodicities in power users, so as to better make predictions and decisions.
[0090] In the embodiment of the present invention, the electricity-consuming enterprise time series data refers to data related to electricity recorded in chronological order over a period of time.
[0091] In the embodiment of the present invention, when converting the power consumption enterprise enhanced data into the power consumption enterprise time series data, the power consumption enterprise time series data conversion module 101 is specifically used to:
[0092] Sorting the enhanced data of the power consuming enterprises in time order according to preset timestamps to obtain an initial data sequence of the power consuming enterprises;
[0093] The timestamps are matched one-to-one with the enhanced data of the power user enterprises in the initial power user enterprise data sequence to obtain time series corresponding data;
[0094] The time series corresponding data is converted into time series data of the electricity consuming enterprise according to the preset time granularity.
[0095] In detail, the timestamp refers to a field in the data used to indicate time, which may be a specific timestamp, such as a combination of date and time, or may be an indicator of a time interval, such as hourly, daily, etc.; the data is then sorted according to the timestamp to ensure that the data is arranged in chronological order, and the data is converted into a time series form, in which each time point corresponds to a row of data, ensuring that each time point has a corresponding observation value, thereby obtaining an initial power user enterprise data sequence; the sorted timestamps are matched one by one with the corresponding power user enterprise enhanced data to form time series corresponding data, which can ensure that each time point has corresponding power user enterprise enhanced data.
[0096] Specifically, the corresponding time series data are aggregated or interpolated according to the preset time granularity (for example, every hour, every day, or every month) to generate time series data for power users. The original corresponding time series data can be converted into more regular and easy-to-analyze time series data, which is convenient for subsequent modeling and analysis. Aggregation processing means that if the preset time granularity is larger than the time granularity of the original data, then multiple data points can be merged into one data point by aggregation, thereby reducing the amount of data and reducing the impact of data noise. Aggregation methods include but are not limited to average aggregation, maximum aggregation, and minimum aggregation. For example, if the time granularity of the original data is once per minute, and the preset time granularity is once per hour, then every 60 data points can be aggregated into one data point, and the average, maximum, or minimum value of these 60 data points is calculated as one hour of data. Interpolation processing means that if the preset time granularity is smaller than the time granularity of the original data, then new data points can be generated based on existing data points by interpolation, thereby increasing data density and improving data accuracy and reliability. Interpolation methods include but are not limited to linear interpolation, spline interpolation, and polynomial interpolation. For example, if the time granularity of the original data is once per hour, and the preset time granularity is once per minute, then linear interpolation can be used to insert 60 data points in each hour, making the data density higher and more accurate.
[0097] For example, the data of the power consumer collected at time point t1 is {A, B, C}, then the corresponding time series data is t1-{A, B, C}, where A can be the power consumption, B can be the power generation, and C can be the voltage. If the timestamp and the power consumer enhanced data do not completely correspond, interpolation or other processing methods are required to fill in the gaps, so as to generate the power consumer time series data. Finally, the generated power consumer time series data is stored in a fixed format to facilitate subsequent analysis and processing.
[0098] Furthermore, it is necessary to perform outlier detection processing on the abnormal data in the generated time series data of the power consuming enterprises to reduce the data false alarm rate of the power consuming enterprise data and reduce unnecessary interference and losses.
[0099] The target power time series data generating module 102 is used to detect abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data using a preset dual weight algorithm, and generate target power time series data corresponding to the power consuming enterprise time series data according to the abnormal time series data of the power consuming enterprise.
[0100] In the embodiment of the present invention, the voltage industry abnormal time series data refers to abnormal power data in the time series data of power consuming enterprises generated in real time, such as abnormal data on power consumption and abnormal data on power generation.
[0101] In the embodiment of the present invention, when the target power time series data generating module 102 detects abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data using a preset dual weight algorithm, it is specifically used to:
[0102] Calculate the power data mean of the time series data of the power consuming enterprise;
[0103] The power weight of each time series data in the time series data of the power consuming enterprise is calculated according to the mean value of the power data using the following dual weight algorithm:
[0104]
[0105] Among them, W ni is the power weight of the index data of the nth power-consuming enterprise at the i-th moment, x ni is the power value of the index data of the nth power-consuming enterprise at the i-th moment, B is the mean value of the power data, k ni is the position number of the index data of the nth power consumer at the i-th moment, A is the position number of the most recent time series data corresponding to the mean value of the power data, and e is the sequence length of the time series data of the power consumer;
[0106] Selecting the time series data whose power weight is less than a preset weight threshold as abnormal time series data;
[0107] The abnormal time series data are collected as the abnormal time series data of the power consuming enterprise.
[0108] Specifically, firstly, the mean value of the power data of the power consumption enterprise time series data is calculated. For example, the time series data corresponding to the power consumption index data in the power consumption enterprise time series data is (x 11 ,x 12 ,…,x 1i ,…,x 1t ), where x 1i is the power consumption at the i-th moment, x 1t is the power consumption at the tth moment; the time series data corresponding to the power generation index in the time series data of the power consuming enterprise is (x 21 ,x22 ,…,x 2i ,…,x 2t ), where x 2i is the power generation at the i-th moment, x 2t is the power generation at time t; the time series data of the power consumption enterprise is (x n1 ,x n2 ,…,x ni ,…,x nt ), where x ni The index data of the nth power consuming enterprise at the i-th moment is used to calculate the mean value of the power data corresponding to each time series data in the time series data of the power consuming enterprise, and then the power weight of each time series data is calculated according to the mean value of the power data.
[0109] Specifically, the dual weight algorithm combines information on time weight and distance weight, and can more comprehensively consider the abnormal situations of time series data, while reducing the impact of time window size on abnormal detection results, which can improve the accuracy of detection. In the dual weight algorithm, the former represents the robustness weight, and calculates the difference between the power value of the index data of the nth power consumer at the i-th moment and the mean of the time series data. The latter is the distance weight, which is based on the distance value between the index data of the nth power consumer at the i-th moment and the position number corresponding to the time series data closest to the mean of the power data. The weight is expressed by distance, and the power weight of each time series data can be calculated, and the data with a power weight less than the weight threshold is selected as abnormal time series data, and all abnormal time series data are collected as abnormal time series data of power consumers.
[0110] For example, the time series data corresponding to the power consumption index data in the time series data of the power consuming enterprise is (x 11 ,x 12 ,…,x 1i ,…,x 1t ), the average value of the power consumption index data is calculated as B1. In the double weight algorithm, when x ni When n=1 in , it represents the power consumption index data, so the robustness weight of each time series data is calculated according to the former in the double weight algorithm, and then its distance weight is calculated according to the latter. For example, the value of the power data mean B1 is closest to x in the time series data. 1i Data, if we calculate x 11 The distance weight is k ni It means x 11 The position number, that is, k 11 The position number of is 1, and x 1i The position number represented is A, so calculate x 11 to x 1i The distance between them is obtained, so as to obtain the distance weight, and finally, x 11The corresponding power weight is W 11 .
[0111] Furthermore, target power time series data corresponding to the power consuming enterprise time series data is generated based on the abnormal time series data of the power consuming enterprise, that is, the abnormal time series data of the power consuming enterprise is filtered out from the power consuming enterprise time series data, and the power consuming enterprise data after the abnormal data is filtered out is used as the target power time series data, which can ensure that there is no abnormal data in the target power time series data and ensure accurate analysis of subsequent power consuming enterprise data.
[0112] Furthermore, by analyzing the trend components of the target power time series data, we can better understand the long-term trend of the data and make more accurate predictions about future development trends.
[0113] The time series decomposition module 103 is used to perform time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data.
[0114] In the embodiment of the present invention, the trend component data refers to a part extracted from the target power time series data by a time series decomposition method, indicating the long-term trend in the data and representing the change of the data on a longer time scale.
[0115] In the embodiment of the present invention, when the time series decomposition module 103 performs time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data, it is specifically used to:
[0116] The target power time series data is preliminarily decomposed using the following addition decomposition formula to obtain the target power time series decomposition data:
[0117] Y t =R t +Q t
[0118] Among them, Y t is the power observation value of the target power time series data at time t, R t is the power trend value of the target power time series data at time t, Q t is the power fluctuation value of the target power time series data at time t;
[0119] Performing a smoothing operation on the target power time series data according to a preset decomposition window to obtain target power smoothed time series data;
[0120] The trend value is determined according to the power observation value of the target power smoothing time series data and the power trend value of the target power time series decomposition data, wherein the trend value calculation formula is:
[0121]
[0122] Among them, R' t is the trend value at time t, Y t is the power observation value of the target power time series data at time t, is the initial fluctuation value at time t in the f+1th cycle, is the smoothed fluctuation value at time t in the f+1th cycle;
[0123] The trend value is numerically fitted to obtain trend component data of the target power time series data.
[0124] In detail, the target power time series data is first represented as a trend component and a fluctuation component, and the long-term trend change of the power consumption enterprise data can be obtained. Then the trend component is completely separated from the target power time series and the fluctuation component to obtain the trend component data. First, the target power time series data is decomposed into trend components R by adding a wig decomposition formula. t and the fluctuation component Q t , where the power trend component R t Represents the long-term trend change, the power fluctuation component Q t Reveals the short-term fluctuations, Y t It reflects the actual electricity consumption or generation. By analyzing the power trend value, the long-term change trend of the power system can be determined. By analyzing the power fluctuation value, the stability and reliability of the power system can be evaluated. At the same time, the complexity and dynamics of power time series data can be better understood.
[0125] Specifically, according to the preset decomposition window, the target power time series data can be smoothed to obtain the target power smoothed time series data. The smoothing operation can remove short-term fluctuations and noise, making the data smoother and more stable, so as to better capture the long-term trend. The smoothing method can adopt the moving average method, exponential smoothing method, etc. Among them, the moving average method is a commonly used smoothing method. It realizes smoothing by calculating the mean of the data in a certain time window. Assuming that the preset decomposition window size is n, a suitable initial point is selected as the starting point. For each moment t, the average value of the data from tn to t is calculated as the target power smoothed time series data at moment t. Repeat the above steps until all moments are processed. Through such a smoothing operation, the smoothed target power time series data can be obtained to better understand the long-term trend and characteristics of the power system. The smoothed data is usually more stable, which is conducive to reducing the impact of noise and outliers and improving the reliability and interpretability of the data.
[0126] Furthermore, the trend value can be determined by the power observation value of the target power smoothed time series data and the power fluctuation value of the target power time series decomposition data. Then, the power fluctuation value at each moment is removed from the target power time series data, and the smoothed fluctuation value is added. By adding the smoothed fluctuation value, the trend characteristics in the data can be better highlighted. The smoothed fluctuation value usually retains the trend component, highlights the long-term change trend and periodic characteristics of the data, and makes these trends more obvious and observable. That is, the trend value of all moments can be obtained by continuously looping the power fluctuation value of the initial decomposition, and then the decomposed trend value is numerically fitted to obtain the trend component data of the target power time series data, such as according to the trend value R' t Calculate the time series corresponding to each moment t, for example, from t=1 to t=N, and fit the time series through polynomial fitting and exponential fitting methods. After numerical fitting, a fitting curve can be obtained, and then the data corresponding to each moment in the fitting curve can be determined as trend component data.
[0127] Furthermore, the decomposed trend component data can be expressed in a time series, and trend forecasting can be performed, that is, the development trend of future electricity users can be speculated.
[0128] The bilateral detection trend calculation module 104 is used to generate time series trend data of power consuming enterprises according to the trend component data, generate power data statistics through the time series trend data of power consuming enterprises, and calculate the bilateral detection trend of the time series data of the power consuming enterprises according to the power data statistics.
[0129] In an embodiment of the present invention, the trend component data are generated into time series trend data of the power consuming enterprise in the order of timestamps, and the trend component data decomposed from each target power time series data are used to generate the time series data of the power consuming enterprise, that is, the trend component data are reorganized in chronological order to obtain the time series trend data of the power consuming enterprise.
[0130] Furthermore, by conducting data statistics based on the time series trend data of electricity users, a large amount of data can be integrated into several simple statistics, such as mean value, standard deviation, etc., which helps to improve the readability and comprehensibility of the data, and also facilitates the comparison and analysis of the data. For example, statistics such as the mean value and standard deviation of electricity demand can be used to predict the growth trend and fluctuation range of future electricity demand.
[0131] In the embodiment of the present invention, the power data statistics are used to judge the significance of the change trend of the power consumption enterprise data based on the time series trend data of the power consumption enterprise.
[0132] In the embodiment of the present invention, when the bilateral detection trend calculation module 104 generates power data statistics through the time series trend data of the power consuming enterprise, it is specifically used to:
[0133] The symbol value of the time series trend data of the electricity consuming enterprise is calculated by the following preset symbol function:
[0134]
[0135] Where H is the symbol value, y j is the time series data value of the index data of the nth power user in the time series data of the power user at the jth moment, y n(j+1) is the time series data value of the index data of the nth power consuming enterprise in the time series data of the power consuming enterprise at the j+1th moment;
[0136] Generate a symbol sequence corresponding to the time series trend data of the electricity consuming enterprise according to the symbol value;
[0137] Calculate the power data statistics of the power consumption enterprise's time series trend data according to the symbol sequence:
[0138]
[0139] Where S is the power data statistic, H d is the dth symbol value in the symbol sequence, and m is the sequence length of the symbol sequence.
[0140] In detail, firstly, the symbol value of the time series trend data of the power consumption enterprise is calculated by the symbol function, and the difference between each adjacent data point of the time series data of the power consumption enterprise can be converted into the corresponding symbol value, so as to help analyze the trend and characteristics of the power consumption enterprise. By calculating the difference between adjacent data points and converting them into symbol values, the trend of the power consumption enterprise can be observed more intuitively. For example, the sequence of the time series trend data of the power consumption enterprise is expressed as (y n1 ,y n2 ,…,y ni ,…,y nt ), the difference between adjacent data points is calculated respectively. If the difference is greater than zero, the symbol value between adjacent data points is determined to be -1. If the difference is equal to zero, the symbol value between adjacent data points is determined to be 0. If the difference is less than zero, the symbol value between adjacent data points is determined to be -1. Then, a symbol sequence is generated according to the symbol value of each adjacent time series data, such as y n1 -y n2 >0,y n2 -y n3 <0,y nj -y n(j+1) =0, etc., then the symbol sequence may be (1,-1,…,0,…,1), and various statistics can be calculated based on the symbol sequence to comprehensively evaluate the trends and characteristics of the electricity consuming enterprise data.
[0141] Specifically, the power data statistics of the time series trend data of the power consuming enterprises are calculated according to the symbol sequence, and the average value of all symbol values in the symbol sequence is calculated as the power data statistics. In addition, the standard deviation and volatility can also be calculated according to the symbol sequence, where the volatility refers to the standard deviation of the symbol sequence divided by the average value. Therefore, the overall characteristics and volatility of the time series data of the power consuming enterprises can be better understood according to the power data statistics.
[0142] Furthermore, by performing bilateral detection trend analysis on power data statistics, we can derive the trend direction of the power users' time series data and the significance of the trend, which is more comprehensive and accurate than simply observing the changes in a single data point.
[0143] In the embodiment of the present invention, the bilateral detection trend refers to an increasing trend, a decreasing trend or a stabilization trend of the power consumption enterprise data based on the analysis of the power consumption enterprise time series data.
[0144] In the embodiment of the present invention, when the bilateral detection trend calculation module 104 calculates the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics, it is specifically used to:
[0145] Calculating the power data variance of the time series trend data of the power consuming enterprise according to the symbol sequence;
[0146] The bilateral test value of the time series data of the power consuming enterprise is calculated according to the power data statistics and the power data variance:
[0147]
[0148] Wherein, Z is the bilateral check value, S is the power data statistic, and D is the power data variance;
[0149] The bilateral detection trend of the time series data of the electricity consuming enterprise is determined according to the bilateral verification value.
[0150] In detail, the variance in the symbol sequence is counted, that is, the variance of all symbol values is calculated as the variance of the power data. The variance is a statistic that measures the degree of data dispersion. By using the symbol sequence to calculate the variance of the time series trend data of the power consumer, the degree of dispersion of the data in the trend can be understood. The larger the variance, the higher the volatility of the data in the trend, and vice versa. The bilateral test value of the time series data of the power consumer is then calculated based on the power data statistics and the power data variance.
[0151] Specifically, the bilateral test value is calculated according to the power data statistic and the power data variance. First, the positive or negative of the power data statistic S is determined, and then the power data statistic S is substituted into different formulas according to the positive or negative of the power data statistic S to obtain different test values. Then, the bilateral detection trend of the time series data of the power consumer is determined according to the test value. The bilateral test value is a statistic used to determine whether the data is significant. If the bilateral test value is greater than zero, it indicates that the data has significant differences, indicating that the sequence has an increasing trend; if the bilateral test value is less than zero, it indicates that the data has significant differences, indicating that the sequence has a decreasing trend; if the bilateral test value is equal to zero, it indicates that the data has no significant differences. In the analysis of time series data of power consumers, the bilateral test value can determine whether the trend of the data is significant.
[0152] Furthermore, by arranging the power data in a matrix in chronological order, the changing trend of the power data can be seen more intuitively, making it easier to discover patterns and anomalies in the data.
[0153] The power consumption enterprise data trend analysis module 105 is used to generate a power data trend matrix based on the power time series data and the bilateral detection trend, generate a trend chart of the power consumption enterprise data based on the power data trend matrix, and perform trend analysis on the power consumption enterprise data using the trend chart.
[0154] In the embodiment of the present invention, the power data trend matrix refers to using the power time series category data as the row dimension and the bilateral category detection trend as the column dimension, and each power time series category data corresponds to a bilateral category detection trend.
[0155] In the embodiment of the present invention, when the power consumption enterprise data trend analysis module 105 generates the power data trend matrix according to the power time series data and the bilateral detection trend, it is specifically used to:
[0156] Classify the power time series data into power data types to obtain power time series category data;
[0157] extracting a bilateral category detection trend from the bilateral detection trend;
[0158] A power data trend matrix is generated according to the power time series category data and the bilateral category detection trend.
[0159] In detail, the power time series data is divided according to the power data type. For example, the power data type includes power consumption, power generation, voltage, power price, etc., and the power time series category data can be obtained. For example, the power time series category data corresponding to the power consumption index is (y 11 ,y 12 ,…,y 1i ,…,y 1t), the power time series category data corresponding to the power generation index is (y 21 ,y 22 ,…,y 2i ,…,y 2t ), the power time series category data corresponding to the voltage index is (y 31 ,y 32 ,…,y 3i ,…,y 3t ), and determine the bilateral category detection trend based on the bilateral detection value calculated based on the power time series category data corresponding to each category, so as to take the power time series category data as the row dimension and the bilateral category detection trend as the column dimension, thereby generating a power data trend matrix.
[0160] For example, if the bilateral category detection trend corresponding to the power consumption indicator is an upward trend, the bilateral category detection trend corresponding to the power generation indicator is an upward trend, and the bilateral category detection trend corresponding to the voltage indicator is a downward trend, then the power data trend matrix is
[0161] Furthermore, the corresponding time series data and bilateral detection trends in the power data trend matrix are used to construct a trend chart of the power consumer data, and the corresponding time series data in the power time series category data are used to generate a trend chart, with the horizontal axis representing time and the vertical axis representing the value of the power consumer data. Various tools and software can be used to draw the trend chart.
[0162] Furthermore, by analyzing trend graphs of electricity-using enterprise data, long-term trends and patterns can be identified, which helps to understand the overall development trends of electricity demand, supply and consumption, and provide guidance for future planning and decision-making.
[0163] In the embodiment of the present invention, when the power consumption enterprise data trend analysis module 105 uses the trend chart to perform trend analysis on the power consumption enterprise data, it is specifically used to:
[0164] Determining the long-term trend of the electricity-consuming enterprise data according to the overall change trend of the trend graph;
[0165] Determine the periodic trend of the electricity-consuming enterprise data according to the peaks and valleys of the trend graph;
[0166] Determining the mutation trend of the electricity consuming enterprise data according to the mutation point of the trend graph;
[0167] The multi-dimensional trend changes of the electricity consuming enterprise data are determined through the long-term trend, the periodic trend and the sudden change trend.
[0168] In detail, by observing the overall change trend of the trend chart, such as linear growth, exponential growth or cyclical fluctuations, the long-term trend of the data of the power user can be determined. The long-term trend is to observe the overall change trend of the data, whether it shows an upward, downward or stable trend; by observing the peaks and troughs in the trend chart, that is, periodic fluctuations, the cyclical trend of the data of the power user can be determined, which helps to identify the cyclical patterns of electricity demand and supply in specific seasons, months or days of the week. Based on the cyclical trend, power dispatching and resource planning can be optimized to meet peak demand and avoid supply shortages. The seasonal trend is to find seasonal changes in the data, such as periodic seasonal peaks and troughs; by observing the mutation points in the trend chart, that is, abnormal and sudden data changes, the mutation trend of the data of the power user can be determined. The mutation points may be caused by emergencies, equipment failures, policy adjustments and other factors. Early detection and identification of these mutation trends can allow corresponding measures to respond to changes and mitigate potential risks.
[0169] Specifically, by integrating the analysis results of long-term trends, cyclical trends and mutation trends, the multi-dimensional trend changes of power users' data can be determined, which can provide a more comprehensive perspective to understand the change patterns and trends of power users on different time scales. Using trend charts to perform trend analysis on power users' data can help decision makers better understand long-term trends, cyclical changes, mutation points, and comprehensively analyze trend changes in multiple dimensions, thereby guiding the management, planning and decision-making of power users.
[0170] The embodiment of the present invention can timely discover abnormal events or faults in the power system by detecting abnormal conditions in the time series data of the power user; generate target power time series data according to the abnormal time series data, and understand the basic characteristics and laws of the power user data by analyzing the target power time series data; decompose the target power time series data into trend component data and other component data through time series decomposition technology, and the trend component data reflects the trend of long-term changes in the power system, which is conducive to judging the development trend of the power market and the growth trend of load demand; generate power user time series trend data according to the trend component data, and calculate the bilateral detection trend of the time series data of the power user through statistics, and intuitively display the development trend, periodic changes and other important information of the power user through trend charts and statistics; use trend charts to analyze the data of power users, which can help decision makers better understand the development dynamics of the power market, predict future power demand and supply, and optimize the planning and scheduling of the power system. Therefore, the system, method and medium for regional power user data trend analysis proposed by the present invention can solve the problem of low accuracy when performing power user data trend analysis.
[0171] Reference Figure 2FIG. 1 is a flow chart of an operation method of a regional power consumption enterprise data trend analysis system provided by an embodiment of the present invention. In this embodiment, the operation method of the regional power consumption enterprise data trend analysis system includes:
[0172] S1. Acquire the data of power users in the target area, perform data enhancement on the data of the power users to obtain enhanced data of the power users, and convert the enhanced data of the power users into time series data of the power users;
[0173] S2. Detect abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise by using a preset dual weight algorithm, and generate target power time series data corresponding to the time series data of the power consuming enterprise according to the abnormal time series data of the power consuming enterprise;
[0174] S3, performing time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data;
[0175] S4, generating time series trend data of the power consuming enterprise according to the trend component data, generating power data statistics through the time series trend data of the power consuming enterprise, and calculating the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics;
[0176] S5. Generate a power data trend matrix based on the power time series data and the bilateral detection trend, generate a trend chart of the power consuming enterprise data based on the power data trend matrix, and perform trend analysis on the power consuming enterprise data using the trend chart.
[0177] The device for implementing the operating method of the regional power consumption enterprise data trend analysis system may include a processor, a memory, a communication bus and a communication interface, and may also include a computer program stored in the memory and executable on the processor, such as a regional power consumption enterprise data trend analysis system program.
[0178] The system program for regional power-consuming enterprise data trend analysis stored in the memory of the device is a combination of multiple instructions, and when executed in the processor, it can achieve:
[0179] Acquire the data of power consuming enterprises in the target area, perform data enhancement on the data of the power consuming enterprises to obtain enhanced data of the power consuming enterprises, and convert the enhanced data of the power consuming enterprises into time series data of the power consuming enterprises;
[0180] Detect abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise by using a preset double weight algorithm, and generate target power time series data corresponding to the time series data of the power consuming enterprise according to the abnormal time series data of the power consuming enterprise;
[0181] Performing time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data;
[0182] Generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics;
[0183] A power data trend matrix is generated according to the power time series data and the bilateral detection trend, a trend graph of the power consuming enterprise data is generated according to the power data trend matrix, and a trend analysis of the power consuming enterprise data is performed using the trend graph.
[0184] Specifically, the specific implementation method of the processor for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.
[0185] Furthermore, if the module / unit integrated in the device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).
[0186] The present invention further provides a computer-readable storage medium, wherein the medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement:
[0187] Acquire the data of power consuming enterprises in the target area, perform data enhancement on the data of the power consuming enterprises to obtain enhanced data of the power consuming enterprises, and convert the enhanced data of the power consuming enterprises into time series data of the power consuming enterprises;
[0188] Detect abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise by using a preset double weight algorithm, and generate target power time series data corresponding to the time series data of the power consuming enterprise according to the abnormal time series data of the power consuming enterprise;
[0189] Performing time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data;
[0190] Generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics;
[0191] A power data trend matrix is generated according to the power time series data and the bilateral detection trend, a trend graph of the power consuming enterprise data is generated according to the power data trend matrix, and a trend analysis of the power consuming enterprise data is performed using the trend graph.
[0192] In the several embodiments provided by the present invention, it should be understood that the disclosed media, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.
[0193] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0194] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0195] The above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A data trend analysis system for regional electricity users, characterized in that: The system includes a power consumption enterprise time series data conversion module, a target power time series data generation module, a time series decomposition module, a bilateral detection trend calculation module and a power consumption enterprise data trend analysis module, wherein: The power consumption enterprise time series data conversion module is used to obtain the power consumption enterprise data of the target area, perform data enhancement on the power consumption enterprise data to obtain the power consumption enterprise enhanced data, and convert the power consumption enterprise enhanced data into the power consumption enterprise time series data; The target power time series data generating module is used to detect abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data by using a preset dual weight algorithm, and generate target power time series data corresponding to the power consuming enterprise time series data according to the abnormal time series data of the power consuming enterprise, wherein when the abnormal time series data of the power consuming enterprise of the power consuming enterprise time series data is detected by using a preset dual weight algorithm, it is specifically used to: Calculate the power data mean of the power consumption enterprise time series data; calculate the power weight of each time series data in the power consumption enterprise time series data according to the power data mean using the following dual weight algorithm: Among them, W ni is the power weight of the index data of the nth power-consuming enterprise at the i-th moment, x ni is the power value of the index data of the nth power-consuming enterprise at the i-th moment, B is the mean value of the power data, k ni is the position number of the index data of the nth power user at the i-th moment, A is the position number of the most recent time series data corresponding to the mean value of the power data, and e is the sequence length of the time series data of the power user; Selecting the time series data whose power weight is less than a preset weight threshold as abnormal time series data; Gathering the abnormal time series data into abnormal time series data of the power consuming enterprise; The time series decomposition module is used to perform time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data; The bilateral detection trend calculation module is used to generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics; The power consumption enterprise data trend analysis module is used to generate a power data trend matrix based on the power time series data and the bilateral detection trend, generate a trend chart of the power consumption enterprise data based on the power data trend matrix, and perform trend analysis on the power consumption enterprise data using the trend chart.
2. The data trend analysis system for regional power users according to claim 1, characterized in that: When the power-consuming enterprise time series data conversion module performs data enhancement on the power-consuming enterprise data to obtain the power-consuming enterprise enhanced data, it is specifically used to: Performing numerical quantization processing on the electricity-consuming enterprise data to obtain first electricity-consuming enterprise data; Processing missing values on the first electricity consuming enterprise data to obtain second electricity consuming enterprise data; Performing outlier processing on the second power consuming enterprise data to obtain third power consuming enterprise data; The third electricity consuming enterprise data is standardized to obtain enhanced electricity consuming enterprise data.
3. The data trend analysis system for regional power users according to claim 1, characterized in that: When converting the power consumption enterprise enhanced data into the power consumption enterprise time series data, the power consumption enterprise time series data conversion module is specifically used to: Sorting the enhanced data of the power consuming enterprises in time order according to preset timestamps to obtain an initial data sequence of the power consuming enterprises; The timestamps are matched one-to-one with the enhanced data of the power user enterprises in the initial power user enterprise data sequence to obtain time series corresponding data; The time series corresponding data is converted into time series data of the electricity consuming enterprise according to the preset time granularity.
4. The regional power-consuming enterprise data trend analysis system according to claim 1, characterized in that: When the time series decomposition module performs time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data, it is specifically used to: The target power time series data is preliminarily decomposed using the following addition decomposition formula to obtain the target power time series decomposition data: Y t =R t +Q t Among them, Y t is the power observation value of the target power time series data at time t, R t is the power trend value of the target power time series data at time t, Q t is the power fluctuation value of the target power time series data at time t; Performing a smoothing operation on the target power time series data according to a preset decomposition window to obtain target power smoothed time series data; The trend value is determined according to the power observation value of the target power smoothing time series data and the power trend value of the target power time series decomposition data, wherein the trend value calculation formula is: Among them, R ′ t is the trend value at time t, Y t is the power observation value of the target power time series data at time t, is the initial fluctuation value at time t in the f+1th cycle, is the smoothed fluctuation value at time t in the f+1th cycle; The trend value is numerically fitted to obtain trend component data of the target power time series data.
5. The regional power-consuming enterprise data trend analysis system according to claim 1, characterized in that: When the bilateral detection trend calculation module generates power data statistics through the time series trend data of the power consuming enterprise, it is specifically used to: The symbol value of the time series trend data of the electricity consuming enterprise is calculated by the following preset symbol function: Where H is the symbol value, y nj is the time series data value of the index data of the nth power user in the time series data of the power user at the jth moment, y n(j+1) is the time series data value of the index data of the nth power consuming enterprise in the time series data of the power consuming enterprise at the j+1th moment; Generate a symbol sequence corresponding to the time series trend data of the electricity consuming enterprise according to the symbol value; Calculate the power data statistics of the power consumption enterprise's time series trend data according to the symbol sequence: Where S is the power data statistic, H d is the dth symbol value in the symbol sequence, and m is the sequence length of the symbol sequence.
6. The data trend analysis system for regional power users as claimed in claim 5, characterized in that: When the bilateral detection trend calculation module calculates the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics, it is specifically used to: Calculating the power data variance of the time series trend data of the power consuming enterprise according to the symbol sequence; The bilateral test value of the time series data of the power consuming enterprise is calculated according to the power data statistics and the power data variance: Wherein, Z is the bilateral test value, S is the power data statistic, and D is the power data variance; The bilateral detection trend of the time series data of the electricity consuming enterprise is determined according to the bilateral verification value.
7. The regional power-consuming enterprise data trend analysis system according to claim 1, characterized in that: When the power consumption enterprise data trend analysis module generates a power data trend matrix according to the power time series data and the bilateral detection trend, it is specifically used to: Classify the power time series data into power data types to obtain power time series category data; extracting a bilateral category detection trend from the bilateral detection trend; A power data trend matrix is generated according to the power time series category data and the bilateral category detection trend.
8. The regional power-consuming enterprise data trend analysis system according to claim 1, characterized in that: When the power consumption enterprise data trend analysis module uses the trend chart to perform trend analysis on the power consumption enterprise data, it is specifically used to: Determining the long-term trend of the electricity-consuming enterprise data according to the overall change trend of the trend graph; Determine the periodic trend of the electricity-consuming enterprise data according to the peaks and valleys of the trend graph; Determining the mutation trend of the electricity consuming enterprise data according to the mutation point of the trend graph; The multi-dimensional trend changes of the electricity consuming enterprise data are determined through the long-term trend, the periodic trend and the sudden change trend.
9. A method for operating a regional power-consuming enterprise data trend analysis system, characterized in that: The method comprises: Acquire the data of power consuming enterprises in the target area, perform data enhancement on the data of the power consuming enterprises to obtain enhanced data of the power consuming enterprises, and convert the enhanced data of the power consuming enterprises into time series data of the power consuming enterprises; The abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise is detected by using a preset dual weight algorithm, and the target power time series data corresponding to the time series data of the power consuming enterprise is generated according to the abnormal time series data of the power consuming enterprise, wherein the abnormal time series data of the power consuming enterprise of the time series data of the power consuming enterprise is detected by using a preset dual weight algorithm, which is specifically used for: Calculate the power data mean of the power consumption enterprise time series data; calculate the power weight of each time series data in the power consumption enterprise time series data according to the power data mean using the following dual weight algorithm: Among them, W ni is the power weight of the index data of the nth power-consuming enterprise at the i-th moment, x ni is the power value of the index data of the nth power-consuming enterprise at the i-th moment, B is the mean value of the power data, k ni is the position number of the index data of the nth power user at the i-th moment, A is the position number of the most recent time series data corresponding to the mean value of the power data, and e is the sequence length of the time series data of the power user; Selecting the time series data whose power weight is less than a preset weight threshold as abnormal time series data; Gathering the abnormal time series data into abnormal time series data of the power consuming enterprise; Performing time series decomposition on the target power time series data according to a preset decomposition window to obtain trend component data of the target power time series data; Generate time series trend data of the power consuming enterprise according to the trend component data, generate power data statistics through the time series trend data of the power consuming enterprise, and calculate the bilateral detection trend of the time series data of the power consuming enterprise according to the power data statistics; A power data trend matrix is generated according to the power time series data and the bilateral detection trend, a trend graph of the power consuming enterprise data is generated according to the power data trend matrix, and a trend analysis of the power consuming enterprise data is performed using the trend graph.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, the operating method of the regional power consumption enterprise data trend analysis system as claimed in claim 9 is implemented.
