Data processing method, device and computer readable storage medium

Abstract

The present invention discloses a data processing method, device and computer-readable storage medium. The method includes: determining a multi-level label system including at least four levels, the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each label type in the multiple label types, the third-level label includes multiple data themes corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data themes; obtaining target power data; based on the multi-level label system, determining the target label of the target power data, the target label includes the corresponding labels included in the first-level label to the fourth-level label. The present invention solves the technical problem in the related art that there is a lack of a method for constructing effective user labels.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a data processing method, device and computer-readable storage medium. Background Art

[0002] Residents' electricity demand is characterized by diversification, personalization and rapidity. By carrying out user label layering and classification design, the service level of power companies can be effectively improved and customer satisfaction can be improved.

[0003] The tags constructed by the tag construction method in the related art have the problem of large tag granularity, which means that the tags obtained by the related art cannot reflect dynamic indicators, and the tag coverage dimensions are not comprehensive enough. That is, in the related art, there is a problem of lack of methods for constructing effective user tags.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a data processing method, device and computer-readable storage medium to at least solve the technical problem of the lack of a method for constructing effective user tags in the related art.

[0006] According to one aspect of an embodiment of the present invention, a data processing method is provided, characterized in that it includes: determining a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes a plurality of label types, the second-level label includes a plurality of data types corresponding to each of the plurality of label types, the third-level label includes a plurality of data topics corresponding to the plurality of data types, and the fourth-level label includes a data level corresponding to the plurality of data topics; acquiring target power data, wherein the target power data includes power data of a coal-to-electricity object, and the coal-to-electricity object includes an object that obtains energy from coal burning to electricity; based on the multi-level label system, determining a target label of the target power data, wherein the target label includes a target label type included in the first-level label, a target data type included in the target label type in the second-level label, a target data topic included in the target data type in the third level, and a target data level included in the target data topic in the fourth level.

[0007] Optionally, in the case that the power data of the target object includes multiple sub-power data, determining the target tag of the target power data based on the multi-level label system includes: acquiring the target tag of each sub-power data based on the multi-level label system for each sub-power data in the multiple sub-power data; and determining the target tag of the target power data based on the target tag of each sub-power data.

[0008] Optionally, in the case that the power data of the target object includes multiple sub-power data, determining the target tag of the target power data based on the multi-level label system includes: preprocessing the multiple sub-power data separately; integrating the multiple preprocessed sub-power data based on predetermined associated fields to obtain a target data set; and determining the target tag of the target power data based on the target data.

[0009] Optionally, in the case that the power data of the target object includes the total power consumption of the target object within a predetermined time period, determining the target tag of the target power data based on the multi-level label system and the power data of the target object includes: obtaining the power consumption level of the target object within the predetermined time period according to the total power consumption of the target object within the predetermined time period and a predetermined power consumption level threshold; determining the power consumption level to be a four-level label corresponding to the total power consumption of the target object within the predetermined time period, the four-level label corresponding to the total power consumption of the target object within the predetermined time period includes one of the following: high power, normal power, low power, and abnormal power consumption; determining the target tag of the target power data based on the power consumption level, and the third-level label, second-level label and first-level label corresponding to the power consumption level in the multi-level label system, the third-level label, second-level label and first-level label corresponding to the power consumption level respectively include power consumption level, power consumption level and statistical label.

[0010] Optionally, it also includes: obtaining the total electricity consumption of multiple predetermined objects within the predetermined time period, wherein the multiple predetermined objects include at least one of the following: an object geographically adjacent to the target object, and other objects other than the target object among multiple objects in the area where the target object is located; when the multiple predetermined objects are the multiple objects geographically adjacent to the target object, based on the comparison result of the total electricity consumption of the multiple predetermined objects within the predetermined time period and the total electricity consumption of the target object within the predetermined time period, correcting the electricity consumption level of the target object within the predetermined time period; when the multiple predetermined objects are other objects other than the target object among multiple objects in the area where the target object is located, based on the sorting result of the total electricity consumption of multiple objects in the area where the target object is located within the predetermined time period, correcting the electricity consumption level of the target object within the predetermined time period.

[0011] Optionally, when the power data of the target object includes behavior data of the target object within a predetermined time period, and the behavior data includes complaint behavior data, determining the target label of the target power data based on the multi-level label system and the power data of the target object includes: obtaining text data corresponding to the complaint behavior data according to the complaint behavior data; extracting complaint behavior object data contained in the text data; determining the complaint behavior object of the target object within the predetermined time period according to the complaint behavior object data contained in the text data; determining the complaint behavior object as a fourth-level label corresponding to the behavior data of the target object within the predetermined time period, and the four-level label corresponding to the behavior data of the target object within the predetermined time period includes one of the following: power outage, abnormality of electric energy meter, abnormality of collection equipment, refund of electricity fee, and failure of electrical appliance to be used normally; determining the target label of the target power data based on the complaint behavior object, and the third-level label, second-level label and first-level label corresponding to the complaint behavior object in the multi-level label system; the third-level label, second-level label and first-level label corresponding to the complaint behavior object include complaint subject, service demand behavior and mining label, respectively.

[0012] Optionally, in the case where the power data of the target object includes the power load distribution of the target object at multiple predetermined power consumption times within a predetermined time period, determining the target tag of the target power data based on the multi-level label system and the power data of the target object includes: according to the power load distribution and according to a predetermined power load anomaly determination method, obtaining the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level based on the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level as a four-level label corresponding to the power load distribution, the four-level label corresponding to the power load distribution including one of the following: high anomaly, relatively high anomaly, medium anomaly, general anomaly, normal; determining the target tag of the target power data based on the power load anomaly level, and the third-level label, second-level label and first-level label corresponding to the power load anomaly level in the multi-level label system, the third-level label, second-level label and first-level label corresponding to the power load anomaly level respectively including power load anomaly, anomaly information and mining label.

[0013] Optionally, in the case that the power data of the target object includes the power consumption data distribution of the target object within a predetermined time period, the determination of the target label of the target power data based on the multi-level label system and the power data of the target object includes at least one of the following: according to a predetermined overcurrent abnormality determination method, obtaining the number of overcurrent abnormalities occurring in the target object in the power consumption data distribution, determining the overcurrent abnormality level according to the number of overcurrent abnormalities occurring, determining the overcurrent abnormality level as a fourth-level label corresponding to the power consumption data distribution, and determining the target label of the target power data based on the overcurrent abnormality level and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution in the multi-level label system; the four-level label corresponding to the power consumption data distribution includes one of the following: does not exist, exists once, exists multiple times, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution include overcurrent anomaly, abnormal information, and mining labels; according to a predetermined voltage anomaly determination method, determine whether the target object has voltage anomaly data in the power consumption data distribution, determine the voltage anomaly level according to whether there is voltage anomaly data, and determine the voltage anomaly level as the fourth-level label corresponding to the power consumption data distribution; based on the voltage anomaly level, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution in the multi-level label system, determine the target label of the target power data, and the four-level label corresponding to the power consumption data distribution includes one of the following: normal, abnormal, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution include voltage anomaly, abnormal information, and mining labels, respectively.

[0014] According to another aspect of an embodiment of the present invention, there is also provided a data processing device, comprising a first determination module, for determining a multi-level label system, wherein the multi-level label system comprises at least a first-level label, a second-level label, a third-level label and a fourth-level label, wherein the first-level label comprises a plurality of label types, the second-level label comprises a plurality of data types corresponding to each of the plurality of label types, the third-level label comprises a plurality of data subjects corresponding to the plurality of data types, and the fourth-level label comprises a data level corresponding to the plurality of data subjects; an acquisition module, for acquiring target power data, wherein the target power data comprises power data of a coal-to-electricity object, and the coal-to-electricity object comprises an object that obtains energy from coal burning to electricity; a second determination module, for determining a target label of the target power data based on the multi-level label system, wherein the target label comprises a target label type included in the first-level label, a target data type included in the target label type in the second-level label, a target data subject included in the target data type in the third level, and a target data level included in the target data subject in the fourth level.

[0015] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, characterized in that the storage medium includes a stored program, wherein when the program is run, the device where the storage medium is located is controlled to execute any one of the above-mentioned data processing methods.

[0016] In an embodiment of the present invention, a multi-level label system is determined, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes a plurality of label types, the second-level label includes a plurality of data types corresponding to each label type in the plurality of label types, the third-level label includes a plurality of data subjects corresponding to the plurality of data types, and the fourth-level label includes a data level corresponding to the plurality of data subjects; the target power data is obtained; based on the multi-level label system, a target label of the target power data is determined, wherein the target label includes a target label type included in the first-level label, a target data type included in the target label type in the second-level label, a target data subject included in the target data type in the third level, and a target data level included in the target data subject in the fourth level, thereby solving the technical problem of lack of a method for constructing effective user labels in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flow chart of an optional data processing method according to an embodiment of the present invention;

[0019] Figure 2 It is a framework diagram of an optional data processing device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme 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 only 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 should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] Example 1

[0023] According to an embodiment of the present invention, a method embodiment of a data processing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0024] Figure 1 is a data processing method according to an embodiment of the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0025] Step S102, determine a multi-level labeling system, wherein the multi-level labeling system includes at least first-level labels, second-level labels, third-level labels and fourth-level labels, the first-level labels include multiple label types, the second-level labels include multiple data types corresponding to each label type in the multiple label types, the third-level labels include multiple data themes corresponding to the multiple data types, and the fourth-level labels include data levels corresponding to the multiple data themes.

[0026] Step S104, obtaining target power data, wherein the target power data includes power data of coal-to-electricity conversion objects, and the coal-to-electricity conversion objects include objects that obtain energy from coal combustion and obtain energy from electricity;

[0027] Step S106, based on the multi-level label system, determine the target label of the target power data, wherein the target label includes the target label type included in the first level label, the target data type included in the target label type in the second level label, the target data subject included in the target data type in the third level, and the target data level included in the target data subject in the fourth level.

[0028] The customer label system obtained through the methods in the relevant technology is not suitable for users who have switched from coal to electricity. Directly using the label construction method in the relevant technology will cause the extracted labels to fail to fully reflect the electricity consumption characteristics of users who have switched from coal to electricity. Specifically, the labels obtained through the user label construction method in the relevant technology have the characteristics of large label granularity and insufficient fineness. In the actual construction process, due to the diversity of content or data, the labels extracted using the relevant technical methods cannot reflect dynamic indicators and the coverage dimensions are not comprehensive enough. In this optional embodiment, by determining a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each label type in the multiple label types, the third-level label includes multiple data topics corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data topics, the target power data is obtained, wherein the target power data includes power data of coal-to-electricity objects, and the coal-to-electricity objects include objects that obtain energy from coal burning to electricity; based on the multi-level label system, a target label for the target power data is determined, wherein the target label includes a target label type included in the first-level label, a target data type included in the target label type in the second-level label, a target data topic included in the target data type in the third level, and a target data level included in the target data topic in the fourth level. Target power data of various objects including coal-to-electricity objects are obtained, and target tags of the target power data are obtained according to the target power data and a multi-level tag system. The target tags include tag data of multiple levels, thereby solving the technical problem of lack of methods for constructing effective user tags in related technologies.

[0029] As an optional embodiment, in the case where the power data of the target object includes multiple sub-power data, the target tag of the target power data is determined based on a multi-level tag system, including: for each of the multiple sub-power data, respectively, the target tag of each sub-power data is obtained based on the multi-level tag system; based on the target tag of each sub-power data, the target tag of the target power data is determined. For multiple sub-power data, the target tag of each sub-power data is obtained respectively, and the target tag of the target power data is determined according to the target tag of each sub-power data. The target tag thus obtained is a multi-dimensional and multi-level tag, which can meet the requirements that the tag can reflect dynamic indicators and cover comprehensive dimensions in actual applications. The technical problem of the lack of methods for constructing effective user tags in related technologies is solved.

[0030] As an optional embodiment, when the power data of the target object includes multiple sub-power data, the target tag of the target power data is determined based on a multi-level tag system, including: pre-processing the multiple sub-power data respectively; integrating the multiple pre-processed sub-power data based on a predetermined associated field to obtain a target data set; and determining the target tag of the target power data based on the target data. In this way, multi-dimensional and multi-level tags can be obtained. In addition, by integrating multiple pre-processed sub-power data to obtain a target data set; and determining the target tag of the target power data based on the target data, the efficiency of generating the target tag is improved.

[0031] In an optional embodiment, when the power data of the target object includes the total power consumption of the target object within a predetermined time period, based on the multi-level label system and the power data of the target object, the target label of the target power data is determined, including: according to the total power consumption of the target object within the predetermined time period and the predetermined power consumption level threshold, the power consumption level of the target object within the predetermined time period is obtained; the power consumption level is determined to be a four-level label corresponding to the total power consumption of the target object within the predetermined time period, and the four-level label corresponding to the total power consumption of the target object within the predetermined time period includes one of the following: high power, general power, low power, abnormal power consumption (or low zero power); based on the power consumption level, and the third-level label, second-level label and first-level label corresponding to the power consumption level in the multi-level label system, the target label of the target power data is determined, and the third-level label, second-level label and first-level label corresponding to the power consumption level respectively include power consumption level, power consumption level, and statistical label. Thus, when the power data of the target object includes the total power consumption of the target object within the predetermined time period, a multi-dimensional multi-layer label for the total power consumption can be obtained.

[0032] In an optional embodiment, it also includes: obtaining the total power consumption of multiple predetermined objects within a predetermined time period, wherein the multiple predetermined objects include at least one of the following: an object that is geographically adjacent to the target object, and other objects other than the target object among multiple objects in the area where the target object is located; in the case where the multiple predetermined objects are multiple objects that are geographically adjacent to the target object, according to the comparison result of the total power consumption of the multiple predetermined objects within the predetermined time period and the total power consumption of the target object within the predetermined time period, the power consumption level of the target object within the predetermined time period is corrected; in the case where the multiple predetermined objects are other objects other than the target object among multiple objects in the area where the target object is located, according to the sorting result of the total power consumption of multiple objects in the area where the target object is located within the predetermined time period, the power consumption level of the target object within the predetermined time period is corrected. Through the correction, an accurate power consumption level can be obtained, the accuracy of the generated target tag is improved, and the dynamic update of the target tag is realized.

[0033] In an optional embodiment, when the power data of the target object includes the behavior data of the target object within a predetermined time period, and the behavior data includes complaint behavior data, the target tag of the target power data is determined based on the multi-level label system and the power data of the target object, including: obtaining text data corresponding to the complaint behavior data according to the complaint behavior data; extracting the complaint behavior object data contained in the text data; determining the complaint behavior object of the target object within the predetermined time period according to the complaint behavior object data contained in the text data; determining the complaint behavior object as a fourth-level label corresponding to the behavior data of the target object within the predetermined time period, and the four-level label corresponding to the behavior data of the target object within the predetermined time period includes one of the following: power outage, abnormality of electric energy meter, abnormality of collection equipment, refund of electricity fee, and inability of electrical appliances to be used normally; determining the target tag of the target power data based on the complaint behavior object, and the third-level label, second-level label and first-level label corresponding to the complaint behavior object in the multi-level label system; the third-level label, second-level label and first-level label corresponding to the complaint behavior object include complaint subject, service demand behavior and mining label, respectively. Therefore, when the power data of the target object includes behavior data of the target object within a predetermined time period, and the behavior data includes complaint behavior data, multi-dimensional and multi-layer labels for the behavior data of the target object within the predetermined time period can be obtained.

[0034] In an optional embodiment, when the power data of the target object includes the power load distribution of the target object at multiple predetermined power consumption times within a predetermined time period, the target tag of the target power data is determined based on the multi-level label system and the power data of the target object, including: according to the power load distribution, according to a predetermined power load anomaly determination method, obtaining the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level based on the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level as a four-level label corresponding to the power load distribution, the four-level label corresponding to the power load distribution including one of the following: high anomaly, relatively high anomaly, medium anomaly, general anomaly, normal; determining the target tag of the target power data based on the power load anomaly level, and the third-level label, second-level label and first-level label corresponding to the power load anomaly level in the multi-level label system, the third-level label, second-level label and first-level label corresponding to the power load anomaly level respectively include power load anomaly, anomaly information and mining label. Therefore, when the power data of the target object includes the power load distribution of the target object at multiple scheduled power consumption times within a predetermined time period, multi-dimensional and multi-layer labels for the power load distribution of the target object at multiple scheduled power consumption times within a predetermined time period can be obtained.

[0035] In an optional embodiment, when the power data of the target object includes the distribution of power consumption data of the target object within a predetermined time period, the target label of the target power data is determined based on the multi-level label system and the power data of the target object, including at least one of the following: according to a predetermined overcurrent abnormality determination method, the number of overcurrent abnormalities occurring in the power consumption data distribution of the target object is obtained, the overcurrent abnormality level is determined according to the number of overcurrent abnormalities occurring, the overcurrent abnormality level is determined as a four-level label corresponding to the power consumption data distribution, and the target label of the target power data is determined based on the overcurrent abnormality level and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution in the multi-level label system; the four-level label corresponding to the power consumption data distribution includes one of the following: does not exist, exists once, There are multiple times, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution include overcurrent anomaly, abnormal information, and mining labels; according to the predetermined voltage anomaly determination method, determine whether there is voltage anomaly data in the power consumption data distribution of the target object, determine the voltage anomaly level according to whether there is voltage anomaly data, and determine the voltage anomaly level as the fourth-level label corresponding to the power consumption data distribution; based on the voltage anomaly level, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution in the multi-level label system, determine the target label of the target power data, and the fourth-level label corresponding to the power consumption data distribution includes one of the following: normal, abnormal, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution include voltage anomaly, abnormal information, and mining labels respectively. Therefore, in the case where the power data of the target object includes the power consumption data distribution of the target object within a predetermined time period, a multi-dimensional and multi-layer label for the power consumption data distribution of the target object within the predetermined time period can be obtained.

[0036] According to the above embodiment and optional embodiment, the present invention provides an optional implementation mode. In this optional implementation mode, a data processing method is provided to realize the construction of user tags. In this optional implementation mode, the target object is a coal-to-electricity user as an example for explanation. It should be noted that the coal-to-electricity user is equivalent to the coal-to-electricity object in the above embodiment, and the coal-to-electricity user includes users who change from obtaining energy from coal burning to using electricity.

[0037] It should be understood that the demand for electricity consumption by residents is characterized by diversification, personalization, and rapidity. As an important means of electricity marketing, the "Internet + power supply service" model needs to carry out user label classification and layering design, carry out label design from a practical and effective perspective, explore customer needs, and integrate different label dimensions to provide customers with energy analysis and electricity strategy guidance, thereby effectively improving the service level of power companies and improving customer satisfaction. The customer label system obtained according to the method in the relevant technology is not suitable for users who switch from coal to electricity. Directly using the label construction method in the relevant technology will cause the extracted labels to fail to fully reflect the electricity consumption characteristics of users who switch from coal to electricity. Specifically, the labels obtained by the user label construction method in the relevant technology have the characteristics of large label granularity and insufficient fineness. In the actual construction process, due to the diversity of content or data, the labels extracted using the relevant technical methods cannot reflect dynamic indicators, the coverage dimensions are not comprehensive, and the rationality of the threshold division of label system indicators also needs to be demonstrated. That is, in the relevant technology, there is a lack of methods for constructing effective user labels.

[0038] In view of this, this optional implementation method determines a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each label type in the multiple label types, the third-level label includes multiple data topics corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data topics; obtains the target power data; based on the multi-level label system, determines the target label of the target power data, wherein the target label includes the target label type included in the first-level label, the target data type included in the target label type in the second-level label, the target data topic included in the target data type in the third level, and the target data level included in the target data topic in the fourth level, thereby solving the technical problem of the lack of methods for constructing effective user labels in related technologies. The following is a detailed description.

[0039] In this optional embodiment, by obtaining the label dimension of the target object, a label classification hierarchical framework is formed; the power data involved in the label dimension is obtained, and the power data includes multiple sub-power data; the label of the target object is generated according to the multiple sub-power data, wherein the label includes sub-labels corresponding to the multiple sub-power data. The present invention solves the technical problem of the lack of a method for constructing effective user labels in the related art. The following is a detailed description.

[0040] In this alternative embodiment, by conducting an analysis of user requirements and the company's service business, label dimensions are screened to form a label classification hierarchy framework; the data is preprocessed, and basic labels are obtained using direct extraction and timestamp calculation methods; quantitative analysis such as data scaling, data transformation, and statistical analysis is combined with qualitative analysis to extract statistical labels such as the proportion of valley electricity consumption, and the percentile is used as the interval threshold for the minimum label granularity calculation; methods such as Isolation Forest, One-Class-SVM, DBCSAN clustering, k-means clustering, text keyword extraction, and topic models are used to extract mining labels such as abnormal electricity load, abnormal voltage, electricity consumption preferences, and complaint topics, so as to solve the problems of insufficient automation and intelligence in the construction of existing coal-to-electricity user labels and the long time for generating user labels.

[0041] The data processing method provided in this alternative embodiment includes the following steps:

[0042] Step S1, conduct an analysis of the business relationship between user requirements and the company's service business, sort out the existing data fields of power data, screen out important first-level and second-level classification dimensions, and fix them as labels. Combining the label refinement principle, each label is decomposed into the smallest granularity, and the label is refined to a depth of 4 levels.

[0043] Step S11, starting from the perspectives of user requirements and company service management, conduct an analysis of users' diverse and personalized electricity consumption requirements and the company's diversified energy efficiency service business, explore the relationship between them, draw on the label frameworks of existing industries, and use methods such as comparative analysis, empirical analysis, and literature review to screen out important first-level and second-level classification dimensions that meet the needs of coal-to-electricity users, and fix them as labels. The first-level label classification is divided into basic labels, statistical labels, and mining labels from the perspective of label generation. The second-level labels of basic labels include basic information, and the second-level labels of statistical labels include electricity consumption level, electricity consumption analysis, etc. The second-level labels of mining labels include abnormal information, service demand behavior, electricity consumption behavior, etc.

[0044] Step S12, based on the first-level and second-level classification dimensions, use methods such as data distribution and descriptive statistics to analyze the distribution of data such as electricity consumption, charging, current voltage, and basic information. Using the MECE analysis method layer by layer, each label is decomposed into the smallest granularity, the label is refined to a depth of 4 levels, and the division indicators and standards for the smallest granularity are determined to form a complete label classification hierarchy framework. For example, for the label of the proportion of valley electricity consumption level in the electricity consumption level of statistical labels, it is divided into three categories A, B, and C according to the proportion of valley electricity consumption. The proportion standard of A is [proportion >= a1, proportion < a2), where a1 and a2 are predetermined proportion thresholds.

[0045] Step S2, data preprocessing. Clean the data involved in the coal-to-electricity tag, such as user information, electricity consumption, electricity charges, current, voltage, load, peak and valley electricity consumption, charges, complaint text, etc., to find missing, abnormal, or converted data, and form the data set required before the tag calculation application;

[0046] Data preprocessing objects include non-numeric data and numerical data, and the processing steps generally include: data cleaning, data transformation, data description and data integration.

[0047] Step 2 specifically includes the following steps:

[0048] Step 211: extract the known data such as power consumption and charges according to the determined label dimension, and formulate data governance rules for data of different numerical types according to data quality analysis and reference to data preprocessing steps;

[0049] Step 212: According to the data governance rules formulated in step 21, data problems such as missing, abnormal, and multiple devices are cleaned up, and integrated to form multi-dimensional and multi-type data sets covering electricity consumption, charges, current, voltage, complaint texts, etc.

[0050] In an optional embodiment, the data is preprocessed by the following method:

[0051] Step S221, converting the known power data into a data set in k×n format, where k is the sum of the number of each user multiplied by the corresponding time range, and n is a power-related variable. At the same time, the required fields are extracted from the charging, current and voltage, and basic information data tables to form a new data set;

[0052] Step S222: according to the data set of step S221, big data technology is used to analyze the data distribution trend and quality issues, sort out the anomalies, missing, duplication, multi-device problems, etc. in the data, and compile a data quality report;

[0053] Step S223, according to the data quality report and in combination with the data preprocessing step, set the processing methods for non-numeric type, missing numeric type, abnormality, multiple devices and other problems, and set the limit threshold of extreme value to formulate data governance rules;

[0054] According to the data governance rules of step S223, the data is processed taking into account the consistency of data in each system.

[0055] Step S3, based on the cloud-edge collaborative coal-to-electricity user monitoring data analysis and evaluation platform, compile analysis application modules such as neighborhood electricity consumption comparison and regional user electricity consumption level ranking to conduct energy-saving analysis and electricity consumption strategy analysis. According to the analysis results, new labels for electricity consumption in the same region are added to the sorted labels to further improve the label dimensions, and summarize reasonable electricity consumption suggestions or label descriptions.

[0056] Step S31: By collecting the user data obtained from the terminal, the data governance module and the energy-saving analysis and electricity consumption strategy analysis module are applied to perform neighborhood electricity consumption comparison, regional user electricity consumption level ranking, inter-regional user electricity consumption behavior analysis, heating electricity consumption analysis of the same type and brand equipment, etc., to obtain user electricity consumption behavior analysis trends, energy-saving electricity consumption strategies and carbon emission rankings; or, through scripting, neighborhood electricity consumption comparison, regional user electricity consumption level ranking and other analysis application modules are compiled to perform neighborhood electricity consumption comparison, etc.

[0057] Step S32: According to the analysis results of step S31, based on the original label framework, a new label for electricity consumption in the same area is added, and the original label dimension is corrected, the existing coal-to-electricity label dimension is continuously improved, and reasonable electricity consumption suggestions are summarized.

[0058] Step S4, label extraction, according to the label dimensions refined in step S1, integrate qualitative and quantitative analysis, data mining technology to obtain indicators such as valley electricity proportion, use isolation forest, One-Class-SVM and other algorithms to classify abnormalities of voltage, current and load, use K-means clustering analysis algorithm to divide electricity consumption preferences for indicators such as peak electricity proportion, use DBSCAN adaptive clustering to build a user segmentation model, integrate Jieba word segmentation, TF-IDF (Term Frequency-Inverse Document Frequency) keyword extraction and LDA (Latent Dirichlet Allocation) topic model to carry out complaint topic analysis, form label extraction and construction from multiple dimensions such as electricity consumption, current and voltage, load, complaint text, etc., and use percentiles to calculate thresholds such as valley electricity proportion level, and obtain the specific division thresholds of each interval of the corresponding minimum granularity label;

[0059] Label extraction and construction: Based on the label dimensions extracted in step S1, we integrate qualitative and quantitative analysis, data mining technology, k-means clustering analysis, and DBSCAN adaptive clustering methods to extract and construct labels from multiple dimensions such as electricity consumption, current and voltage, and charges. We also use percentiles to calculate the threshold value to obtain the specific division standard of the minimum granularity label.

[0060] DBSCAN is a density-based clustering method, whose basic assumption is that the degree of compactness determines the number and position of categories. The distance between samples of the same category is closer, and the distance between samples of different categories is farther. By classifying closely connected samples into one category, a cluster category is obtained, and by classifying each group of closely connected samples into different categories, the final clustering effect is achieved.

[0061] K-means algorithm is a clustering algorithm. Clustering is based on the principle of similarity, which divides data objects with high similarity into the same cluster and divides data objects with high dissimilarity into different clusters. The main steps are as follows:

[0062] (1) Randomly select the initial centers of K clusters;

[0063] (2) For any sample point, find its Euclidean distance to the K cluster centers, and classify the sample point into the cluster with the smallest distance. Repeat this process n times.

[0064]

[0065] Among them, ρ represents the Euclidean distance between the point (x1, ..., z1) and the point (x2, ..., z2), x2 represents the value corresponding to the first variable in the point (x2, ..., z2), x1 represents the value corresponding to the first variable in the point (x1, ..., z1), z2 represents the value corresponding to the last variable in the point (x2, ..., z2), and z1 represents the value corresponding to the last variable in the point (x1, ..., z1).

[0066] (3) In each iteration, the center point (centroid) of each cluster is updated using methods such as mean;

[0067]

[0068] Among them, cj represents the jth class, |cj| represents the number of samples in the class, and x represents the high-dimensional samples.

[0069] (4) After iteratively updating the K cluster centers according to steps (2) and (3), if the position point changes very little (a threshold can be set), it is considered that a stable state has been reached and the iteration ends. Different colors can be selected for different cluster blocks and cluster centers.

[0070] The One-Class-SVM algorithm is to find a hypersphere or hyperplane and circle the positive examples in the sample. The prediction process is based on this hyperplane to make decisions, and the samples in the circle are considered to be positive samples.

[0071] Assume that the parameters of the generated hypersphere are the center o and the corresponding hypersphere radius r>0, the hypersphere volume V(r) is minimized, and the center o is a linear combination of the support rows; similar to the traditional SVM method, it can be required that all training data points x i The distance to the center is strictly less than r. However, a slack variable ζi with a penalty coefficient C is constructed at the same time, and the optimization problem is as follows:

[0072]

[0073] ||xi -o||≤r+ξ i , i=1,2,3,…,m

[0074] ξ i ≥0, i=1, 2, …, m

[0075] Here, m represents the sample size.

[0076] After solving the Lagrangian duality, we can determine whether the new data point z is inside the hypersphere. If the distance from z to the center is less than or equal to the radius r, it is not an outlier. If it is outside the hypersphere, it is an outlier.

[0077] The isolation forest algorithm mainly targets outliers in continuous structured data. The premise of its use is to define outliers as outliers that are easily isolated. Outliers that are easily isolated can be understood as points that are sparsely distributed and far away from high-density groups. It should be understood that from a statistical point of view, in the data space, if there are only sparsely distributed points in an area, it means that the probability of data points falling in this area is very low, so the points in these areas can be considered abnormal. For data with a small proportion of abnormal data in the total sample size and a large difference between the characteristic values ​​of abnormal points and normal points, the isolation forest algorithm can better select outliers.

[0078] TF-IDF is a statistical method to evaluate the importance of a word to a document set or one of the documents in a corpus. The importance of a word increases in direct proportion to the number of times it appears in a document, but decreases in inverse proportion to the frequency of its appearance in the corpus. The main steps include: calculating word frequency; calculating inverse document frequency; calculating TF-IDF. The LDA topic model can give the topic in a text in the form of probability distribution, so that after analyzing some documents to extract their topics or distribution, topic clustering or text classification can be performed based on the topics or distribution.

[0079] Step 41: Based on the data set of step S2, with the help of box plots, data distribution and other methods, analyze the current data of each period with a frequency of 15 minutes from the perspectives of normal, continuously low, and sudden increase, explore the abnormal distribution of current, and use the LOF algorithm to judge the abnormality of the current in each time period, count the number of abnormalities, classify the overcurrent abnormalities according to the statistical number, and make corresponding suggestions.

[0080] Step 42: Based on the voltage data set of step S2, obtain normal samples and abnormal samples according to the voltage anomaly judgment conditions. The normal samples are used as the training set. The first 1000 normal samples (the amount of abnormal sample data) are taken and combined with the abnormal samples to form a new data set. The One-Class-SVM algorithm is used to build an abnormal classification model for training and verification. The new data set is used to perform a generalization performance test on the model to test the classification accuracy of the model. The model is continuously optimized to obtain the optimal model, and the model is used to determine whether the voltage is abnormal.

[0081] Step 43: Based on the power load data of step S2, analyze the load change and distribution characteristics of different time periods within a period of time, preliminarily determine the occurrence of abnormally low or high load characteristics, use the isolation forest algorithm to analyze and classify the power load anomaly in each time period, statistically summarize the number of anomalies in different time periods, and classify the power load anomaly by dividing the number of times;

[0082] Step 44: Use the trend chart to analyze the change trend of electricity consumption during peak, flat and valley periods, and use the K-means clustering algorithm to classify electricity consumption preferences based on indicators such as valley electricity consumption proportion, peak electricity consumption proportion, flat electricity consumption proportion, and electricity consumption, and summarize the characteristics of each type of electricity consumption preference;

[0083] Step 45: Conduct user electricity consumption behavior analysis, explore electricity consumption patterns, combine the energy-saving analysis and electricity consumption strategy analysis results of step S3, use quantitative analysis methods such as statistical analysis, data conversion, and data scaling, combined with qualitative analysis methods, and extract basic labels and statistical labels after secondary processing. Basic labels include user number and first transformation time in basic information, and statistical labels include valley electricity proportion level, heating electricity consumption level per unit area, subsidized electricity fee, electricity consumption trend in electricity consumption analysis, etc., and write percentile calculation module in Python to calculate the interval threshold of minimum granularity label for indicators such as valley electricity proportion level and heating electricity consumption level per unit area.

[0084] Percentiles can characterize the order in which a set of n observations are arranged in terms of numerical magnitude. For example, the value at the p% position is called the pth percentile, and observations with the same numerical value have the same percentage.

[0085]

[0086] Among them, PR j represents percentile, j represents the jth user, N represents the total number of users, and nj represents the number of users whose value of a certain indicator is smaller than that of user j. Users with the same value of a certain indicator have the same corresponding percentage.

[0087] Step 46: Comprehensively analyze the electricity consumption behavior data of users in different dimensions, use hierarchical clustering R-type clustering to select important features from a large number of features as the user segmentation index, integrate DBSCAN adaptive clustering and spectral clustering to build a user segmentation model, explore customers of different values, and label customers with insightful labels.

[0088] Step 47: Use Jieba word segmentation and stop word dictionary to perform pre-processing on the complaint samples related to "coal to electricity" electricity consumption, such as word segmentation, stop word removal, and irrelevant content removal. Use the professional word library construction method based on information entropy to obtain some long phrases with clear customer demand semantics. Use the TF-IDF statistical algorithm to extract keywords related to "coal to electricity", infer the distribution of 5 complaint topics based on the LDA topic model, and make suggestions based on the characteristics of the complaint topics.

[0089] Step 48: Based on the aforementioned label extraction method, in accordance with the label hierarchy framework of step S1, three categories of basic labels, statistical labels and mining labels are formed, and they correspond one by one to the 2nd, 3rd and 4th level label dimensions, forming a classified and graded, three-dimensional "coal to electricity" label framework.

[0090] Step 49: Solidify and automate all steps to form a complete application module that can automatically and quickly complete data processing, analysis, and label calculation and extraction.

[0091] Example 2

[0092] According to an embodiment of the present invention, a device for implementing the above data processing method is also provided. Figure 2 2 is a framework diagram of an optional data processing device according to an embodiment of the present invention, the device includes: a first determination module 202, an acquisition module 204, and a second determination module 206. The following is a detailed description.

[0093] The first determination module 202 is used to determine a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, wherein the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each of the multiple label types, the third-level label includes multiple data themes corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data themes; the acquisition module 204 is connected to the above-mentioned first determination module 202, and is used to acquire target power data, wherein the target power data includes power data of coal-to-electricity objects, and the coal-to-electricity objects include objects that obtain energy from coal burning to electricity; the second determination module 206 is connected to the above-mentioned acquisition module 204, and is used to determine the target label of the target power data based on the multi-level label system, wherein the target label includes the target label type included in the first-level label, the target data type included in the target label type in the second-level label, the target data theme included in the target data type in the third level, and the target data level included in the target data theme in the fourth level.

[0094] It should be noted here that the above-mentioned first determination module 202, acquisition module 204 and second determination module 206 correspond to steps S102 to S106 in Example 1, and the three modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the contents disclosed in the above-mentioned Example 1.

[0095] The embodiment of the present invention further provides a computer-readable storage medium. Optionally, in this embodiment, the non-volatile storage medium can be used to store the program code executed by the data processing method provided in the first embodiment.

[0096] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0097] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: determining a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each label type in the multiple label types, the third-level label includes multiple data topics corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data topics; obtaining target power data, wherein the target power data includes power data of coal-to-electricity objects, and the coal-to-electricity objects include objects that obtain energy from coal burning to electricity; based on the multi-level label system, determining a target label for the target power data, wherein the target label includes a target label type included in the first-level label, a target data type included in the target label type in the second-level label, a target data topic included in the target data type in the third level, and a target data level included in the target data topic in the fourth level.

[0098] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: in the case where the target power data includes multiple sub-power data, based on the multi-level label system, determining the target tag of the target power data, including: for each sub-power data in the multiple sub-power data, respectively, obtaining the target tag of each sub-power data based on the multi-level label system; determining the target tag of the target power data based on the target tag of each sub-power data.

[0099] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: when the target power data includes multiple sub-power data, based on a multi-level label system, determining a target label of the target power data, including: preprocessing the multiple sub-power data separately; integrating the multiple preprocessed sub-power data based on predetermined associated fields to obtain a target data set; and determining a target label of the power data based on the target data.

[0100] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for performing the following steps: in a case where the target power data includes the total power consumption of the target object within a predetermined time period, based on the multi-level label system and the target power data, determining the target label of the target power data, including: obtaining the power consumption level of the target object within the predetermined time period according to the total power consumption of the target object within the predetermined time period and a predetermined power consumption level threshold; determining the power consumption level to be a four-level label corresponding to the total power consumption of the target object within the predetermined time period, the four-level label corresponding to the total power consumption of the target object within the predetermined time period including one of the following: high power, normal power, low power, and abnormal power consumption; determining the target label of the target power data based on the power consumption level, and the third-level label, the second-level label and the first-level label corresponding to the power consumption level in the multi-level label system, the third-level label, the second-level label and the first-level label corresponding to the power consumption level respectively including power consumption level, power consumption level, and statistical label.

[0101] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: obtaining the total power consumption of multiple predetermined objects within a predetermined time period, wherein the multiple predetermined objects include at least one of the following: an object geographically adjacent to the target object, and other objects other than the target object among multiple objects in the area where the target object is located; in a case where the multiple predetermined objects are multiple objects geographically adjacent to the target object, based on a comparison result of the total power consumption of the multiple predetermined objects within the predetermined time period and the total power consumption of the target object within the predetermined time period, correcting the power consumption level of the target object within the predetermined time period; in a case where the multiple predetermined objects are other objects other than the target object among multiple objects in the area where the target object is located, correcting the power consumption level of the target object within the predetermined time period based on a sorting result of the total power consumption of the multiple objects in the area where the target object is located within the predetermined time period.

[0102] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: when the target power data includes behavior data of the target object within a predetermined time period, and the behavior data includes complaint behavior data, determining the target tag of the target power data based on the multi-level label system and the target power data of the target object, including: obtaining text data corresponding to the complaint behavior data based on the complaint behavior data; extracting complaint behavior object data contained in the text data; determining the complaint behavior object of the target object within the predetermined time period based on the complaint behavior object data contained in the text data; determining the complaint behavior object as a fourth-level label corresponding to the behavior data of the target object within the predetermined time period, the four-level label corresponding to the behavior data of the target object within the predetermined time period including one of the following: power outage, abnormality of electric energy meter, abnormality of collection equipment, refund of electricity fee, and inability of electrical appliances to be used normally; determining the target tag of the target power data based on the complaint behavior object, and the third-level label, second-level label and first-level label corresponding to the complaint behavior object in the multi-level label system; the third-level label, second-level label and first-level label corresponding to the complaint behavior object respectively include complaint subject, service demand behavior and mining label.

[0103] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: when the power data of the target object includes the power load distribution of the target object at multiple predetermined power consumption times within a predetermined time period, based on the multi-level label system and the power data of the target object, determining the target label of the target power data, including: according to the power load distribution, according to a predetermined power load anomaly determination method, obtaining the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level based on the number of power load anomalies occurring in the power data distribution of the target object; determining the power load anomaly level as a four-level label corresponding to the power load distribution, the four-level label corresponding to the power load distribution including one of the following: high anomaly, relatively high anomaly, medium anomaly, general anomaly, normal; determining the target label of the target power data based on the power load anomaly level, and the third-level label, the second-level label and the first-level label corresponding to the power load anomaly level in the multi-level label system, the third-level label, the second-level label and the first-level label corresponding to the power load anomaly level respectively including power load anomaly, anomaly information and a mining label.

[0104] Optionally, in this embodiment, the non-volatile storage medium is configured to store program code for executing the following steps: in a case where the power data of the target object includes the distribution of power consumption data of the target object within a predetermined time period, based on the multi-level label system and the power data of the target object, determining the target label of the target power data, including at least one of the following: according to a predetermined overcurrent abnormality determination method, obtaining the number of overcurrent abnormalities occurring in the power consumption data distribution of the target object, determining the overcurrent abnormality level according to the number of overcurrent abnormalities occurring, determining the overcurrent abnormality level as a fourth-level label corresponding to the power consumption data distribution, and determining the target label of the target power data based on the overcurrent abnormality level and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution in the multi-level label system; the fourth-level label corresponding to the power consumption data distribution includes: The method comprises one of the following: does not exist, exists once, exists multiple times, and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution include overcurrent anomaly, abnormal information, and mining labels; according to a predetermined voltage anomaly determination method, determine whether there is voltage anomaly data of the target object in the power consumption data distribution, determine the voltage anomaly level according to whether there is voltage anomaly data, and determine the voltage anomaly level as the fourth-level label corresponding to the power consumption data distribution; based on the voltage anomaly level, and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution in the multi-level label system, determine the target label of the target power data, and the fourth-level label corresponding to the power consumption data distribution includes one of the following: normal, abnormal, and the third-level label, the second-level label and the first-level label corresponding to the power consumption data distribution include voltage anomaly, abnormal information, and mining labels, respectively.

[0105] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0106] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0107] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0108] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0109] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or 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 software functional units.

[0110] If the integrated unit 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. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0111] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A data processing method, characterized in that: include: Determine a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes multiple label types, the second-level label includes multiple data types corresponding to each label type in the multiple label types, the third-level label includes multiple data topics corresponding to the multiple data types, and the fourth-level label includes data levels corresponding to the multiple data topics; Acquire target power data, wherein the target power data includes power data of coal-to-electricity conversion targets, and the coal-to-electricity conversion targets include targets that obtain energy from coal combustion and change to obtain energy from electricity; Based on the multi-level label system, determine the target label of the target power data, wherein the target label includes the target label type included in the first level label, the target data type included in the target label type in the second level label, the target data subject included in the target data type in the third level, and the target data level included in the target data subject in the fourth level; Wherein, in a case where the power data of the target object includes the distribution of power consumption data of the target object within a predetermined time period, determining the target tag of the target power data based on the multi-level tag system and the power data of the target object includes at least one of the following: According to a predetermined overcurrent anomaly determination method, the number of overcurrent anomalies occurring in the target object in the power consumption data distribution is obtained, the overcurrent anomaly level is determined according to the number of overcurrent anomalies occurring, the overcurrent anomaly level is determined as a four-level label corresponding to the power consumption data distribution, and based on the overcurrent anomaly level, and the three-level label, the two-level label and the one-level label corresponding to the power consumption data distribution in the multi-level label system, the target label of the target power data is determined; the four-level label corresponding to the power consumption data distribution includes one of the following: does not exist, exists once, exists multiple times, and the three-level label, the two-level label and the one-level label corresponding to the power consumption data distribution include overcurrent anomaly, abnormal information, and mining label; According to a predetermined voltage anomaly determination method, determine whether there is voltage anomaly data for the target object in the electricity consumption data distribution, determine the voltage anomaly level according to whether there is voltage anomaly data, and determine the voltage anomaly level as a four-level label corresponding to the electricity consumption data distribution; based on the voltage anomaly level, and the third-level label, second-level label and first-level label corresponding to the electricity consumption data distribution in the multi-level label system, determine the target label of the target power data, the four-level label corresponding to the electricity consumption data distribution includes one of the following: normal, abnormal, the third-level label, second-level label and first-level label corresponding to the electricity consumption data distribution include voltage anomaly, abnormal information, and mining label, respectively.

2. The method according to claim 1, characterized in that In the case where the target power data includes a plurality of sub-power data, determining the target tag of the target power data based on the multi-level tag system includes: For each of the plurality of sub-power data, respectively, obtaining a target tag for each sub-power data based on the multi-level tag system; Based on the target tag of each sub-power data, a target tag of the target power data is determined.

3. The method according to claim 1, characterized in that In the case where the target power data includes a plurality of sub-power data, determining the target tag of the target power data based on the multi-level tag system includes: Preprocessing the plurality of sub-power data respectively; Integrate the plurality of preprocessed sub-power data based on predetermined associated fields to obtain a target data set; Based on the target data, a target tag of the power data is determined.

4. The method according to claim 1, characterized in that: In a case where the target power data includes the total power consumption of the target object within a predetermined time period, determining the target tag of the target power data based on the multi-level tag system and the target power data includes: According to the total power consumption of the target object in the predetermined time period and the predetermined power consumption level threshold, obtaining the power consumption level of the target object in the predetermined time period; Determining the power consumption level as a four-level label corresponding to the total power consumption of the target object within a predetermined time period, wherein the four-level label corresponding to the total power consumption of the target object within the predetermined time period includes one of the following: high power, normal power, low power, and abnormal power consumption; Based on the electricity consumption level, and the third-level label, second-level label and first-level label corresponding to the electricity consumption level in the multi-level label system, the target label of the target power data is determined, and the third-level label, second-level label and first-level label corresponding to the electricity consumption level respectively include electricity consumption level, electricity consumption level and statistical label.

5. The method according to claim 4, characterized in that Also includes: Obtaining the total amount of electricity consumption of a plurality of predetermined objects within the predetermined time period, wherein the plurality of predetermined objects include at least one of the following: an object geographically adjacent to the target object, and other objects other than the target object among a plurality of objects in the area where the target object is located; In a case where the plurality of predetermined objects are the plurality of objects that are geographically adjacent to the target object, according to a comparison result of the total power consumption of the plurality of predetermined objects in the predetermined time period and the total power consumption of the target object in the predetermined time period, the power consumption level of the target object in the predetermined time period is corrected; In the case where the multiple predetermined objects are objects other than the target object among the multiple objects in the area where the target object is located, the power consumption level of the target object in the predetermined time period is corrected according to the sorting result of the total power consumption of the multiple objects in the area where the target object is located in the predetermined time period.

6. The method according to claim 1, characterized in that In a case where the target power data includes behavior data of a target object within a predetermined time period, and the behavior data includes complaint behavior data, determining a target tag of the target power data based on the multi-level tag system and the target power data of the target object includes: According to the complaint behavior data, obtaining text data corresponding to the complaint behavior data; Extracting the complaint behavior object data contained in the text data; Determining the complaint behavior object of the target object within a predetermined time period according to the complaint behavior object data contained in the text data; Determine that the complaint behavior object is a four-level label corresponding to the behavior data of the target object within a predetermined time period, and the four-level label corresponding to the behavior data of the target object within a predetermined time period includes one of the following: power outage, abnormality of electric energy meter, abnormality of collection equipment, refund of electricity fee, and failure of electrical appliance to be used normally; Based on the complaint behavior object, and the third-level label, second-level label and first-level label corresponding to the complaint behavior object in the multi-level label system, the target label of the target power data is determined; the third-level label, second-level label and first-level label corresponding to the complaint behavior object respectively include complaint subject, service demand behavior and mining label.

7. The method according to claim 1, characterized in that In a case where the power data of the target object includes power load distribution of the target object at multiple predetermined power consumption times within a predetermined time period, determining a target tag of the target power data based on the multi-level tag system and the power data of the target object includes: According to the power load distribution, and according to a predetermined power load anomaly determination method, obtaining the number of times the power load anomaly of the target object occurs in the power load data distribution; Determining a power load abnormality level based on the number of times the power load of the target object is abnormal in the power load data distribution; Determining the power load abnormality level as a four-level label corresponding to the power load distribution, wherein the four-level label corresponding to the power load distribution includes one of the following: high abnormality, relatively high abnormality, medium abnormality, general abnormality, and normal; Based on the power load abnormality level, and the third-level label, second-level label and first-level label corresponding to the power load abnormality level in the multi-level label system, the target label of the target power data is determined, and the third-level label, second-level label and first-level label corresponding to the power load abnormality level respectively include power load abnormality, abnormal information and mining label.

8. A data processing device, characterized in that: include: A first determination module is used to determine a multi-level label system, wherein the multi-level label system includes at least a first-level label, a second-level label, a third-level label and a fourth-level label, the first-level label includes a plurality of label types, the second-level label includes a plurality of data types corresponding to each label type in the plurality of label types, the third-level label includes a plurality of data themes corresponding to the plurality of data types, and the fourth-level label includes data levels corresponding to the plurality of data themes; An acquisition module, used for acquiring target power data, wherein the target power data includes power data of a coal-to-electricity conversion object, and the coal-to-electricity conversion object includes an object that obtains energy from coal combustion and changes to obtain energy from electricity; A second determination module is configured to determine a target tag of the target power data based on the multi-level tag system, wherein the target tag includes a target tag type included in the first-level tag, a target data type included in the target tag type in the second-level tag, a target data subject included in the target data type in the third level, and a target data level included in the target data subject in the fourth level; Among them, the second determination module is also used to obtain the number of overcurrent anomalies that occur in the target object in the power consumption data distribution according to a predetermined overcurrent anomaly determination method, determine the overcurrent anomaly level according to the number of overcurrent anomalies, determine the overcurrent anomaly level as a four-level label corresponding to the power consumption data distribution, and determine the target label of the target power data based on the overcurrent anomaly level and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution in the multi-level label system; the four-level label corresponding to the power consumption data distribution includes one of the following: does not exist, exists once, exists multiple times, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution include overcurrent anomaly, abnormal information, and mining labels; according to a predetermined voltage anomaly determination method, determine whether there is voltage anomaly data in the power consumption data distribution of the target object, determine the voltage anomaly level according to whether there is voltage anomaly data, and determine the voltage anomaly level as a four-level label corresponding to the power consumption data distribution; based on the voltage anomaly level, and the third-level label, second-level label and first-level label corresponding to the power consumption data distribution in the multi-level label system, determine the target label of the target power data, the four-level label corresponding to the power consumption data distribution includes one of the following: normal, abnormal, the third-level label, second-level label and first-level label corresponding to the power consumption data distribution respectively include voltage anomaly, abnormal information, and mining labels.

9. A computer-readable storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the data processing method according to any one of claims 1 to 7.

Images