WO2024255020A1 - 计量异常的分析方法及装置、存储介质、计算机设备 - Google Patents

计量异常的分析方法及装置、存储介质、计算机设备 Download PDF

Info

Publication number
WO2024255020A1
WO2024255020A1 PCT/CN2023/123294 CN2023123294W WO2024255020A1 WO 2024255020 A1 WO2024255020 A1 WO 2024255020A1 CN 2023123294 W CN2023123294 W CN 2023123294W WO 2024255020 A1 WO2024255020 A1 WO 2024255020A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
target
abnormal
measurement
analysis
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2023/123294
Other languages
English (en)
French (fr)
Inventor
张家铭
郑可
肖冀
成涛
徐鸿宇
邹波
万树伟
吕鑫科
陈冰凌
谭时顺
周峰
何珉
谢广成
刘拧滔
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Marketing Service Center of State Grid Chongqing Electric Power Co Ltd
State Grid Corp of China SGCC
Original Assignee
Marketing Service Center of State Grid Chongqing Electric Power Co Ltd
State Grid Corp of China SGCC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Marketing Service Center of State Grid Chongqing Electric Power Co Ltd, State Grid Corp of China SGCC filed Critical Marketing Service Center of State Grid Chongqing Electric Power Co Ltd
Priority to US18/578,470 priority Critical patent/US20240420157A1/en
Publication of WO2024255020A1 publication Critical patent/WO2024255020A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
    • G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/10—Pre-processing; Data cleansing
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply

Definitions

  • the present application relates to the technical field of data processing and analysis, for example, to an analysis method and device for measurement anomalies, a storage medium, and a computer device.
  • Metering anomalies are related to the fairness of the power market and the perception of customer power supply services. Ensuring the normal operation of power metering equipment is a very important part of the entire power marketing environment. The causes of abnormalities in power metering equipment are very complex, including business problems, equipment problems, special working conditions, operating environment problems, etc. If the abnormalities of power metering equipment are not clear and timely eliminated, major accidents will occur.
  • the measurement anomaly analysis method is to select some data items as key performance data from a large amount of measurement equipment operation information according to the degree of concern and importance, and then calculate indicators based on the key performance data, so as to determine the overall quality of the measurement equipment and measurement business based on the indicator value.
  • the analysis dimension is small, the verification logic is simple, and the accuracy of the judgment of the problem of multiple influencing factors is low, which leads to the recurrence, spread and transfer of measurement anomaly problems, and cannot play the role of reducing the cost of power operation and maintenance.
  • the present application provides a method and device for analyzing metering anomalies, a storage medium, and a computer device, which solves the problem of low accuracy in abnormal problem analysis due to small analysis dimensions and simple verification logic in the power metering anomaly analysis method.
  • the present application provides a method for analyzing measurement anomalies, comprising:
  • a device for analyzing measurement anomalies comprising:
  • a filtering rule determination module configured to obtain primary analysis data obtained through analysis by a source end system, and determine at least one data filtering rule based on the primary analysis data;
  • a data acquisition module configured to filter the monitoring data of the source system based on the data filtering rules to obtain target measurement abnormal data
  • a case determination module configured to compare and analyze the target measurement abnormal data with pre-configured abnormal case data, and determine at least one target case data from the abnormal case data;
  • the cause analysis module is configured to perform multidimensional cluster analysis on the target measurement abnormality data to obtain the aggregation degree of the target measurement abnormality data in multiple data dimensions; and analyze the cause of the measurement abnormality based on the target case data and the aggregation degree.
  • a storage medium stores at least one executable instruction, and the executable instruction enables a processor to execute the above-mentioned measurement abnormality analysis method.
  • a computer device comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;
  • the memory is configured to store at least one executable instruction, and the executable instruction enables the processor to execute the above-mentioned measurement abnormality analysis method.
  • FIG1 is a schematic flow chart of a method for analyzing measurement anomalies provided in an embodiment of the present application
  • FIG2 is a flow chart of another method for analyzing measurement anomalies provided in an embodiment of the present application.
  • FIG3 is a flow chart of another method for analyzing measurement anomalies provided in an embodiment of the present application.
  • FIG4 is a schematic diagram of the structure of a measurement anomaly analysis device provided in an embodiment of the present application.
  • FIG5 is a schematic diagram of the structure of another device for analyzing measurement anomalies provided in an embodiment of the present application.
  • FIG6 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application.
  • the present application embodiment provides an analysis method for a measuring device, as shown in FIG1 , the method comprising:
  • the current execution end obtains the preliminary analysis data obtained through the source end system analysis, wherein the source end system is used to characterize the system with functions such as online data monitoring of electric power metering equipment, inaccuracy monitoring of electric power metering equipment, and receiving customer complaints. It can collect the operating status data of all metering equipment in the system in real time, and can also dispatch work orders based on discrete functions, and find problems such as high abnormality rate of metering equipment and decreased customer service satisfaction rate in the monthly indicator measurement.
  • the preliminary analysis data is used to characterize the analysis data obtained by the system analysis function of the source end system, including nearly 100 key indicators such as acquisition success rate, electric energy meter inaccuracy, electric energy meter flow reversal, voltage phase failure, customer complaints, line loss exceeding the limit, load overcapacity, and sudden change in electricity.
  • the embodiment of the present application does not limit the type of preliminary analysis data. After the current execution end obtains the preliminary analysis data, it determines at least one data filtering rule based on the preliminary analysis data.
  • the key indicator with the indicator deviation problem is called the key indicator deviation item.
  • the data filtering rules it is necessary to filter the data based on the key indicator deviation item. If there is a deviation in the success rate of the electric energy meter collection, then based on the normal collection success, first determine the time period when the success rate of the electric energy meter collection is deviated, the metering equipment corresponding to the collection failure, and the detailed information related to the collection.
  • the detailed information includes the collection equipment file information (manufacturer, batch, model, specification and other static information), business information (operating time, collection and operation, on-site supplementary recording, on-site inspection, on-site inspection, assembly and disassembly, etc.), operation information that may affect the collection quality (voltage curve, current curve, terminal address, clock, battery undervoltage event, power-off power-on event, cover opening record, clock error and other dynamic information), etc.
  • the embodiment of the present application does not limit the type of detailed information.
  • the current execution end filters the collection-related detailed information involved in the key indicator deviation item of the above-mentioned electric energy meter collection success rate according to the time period when the deviation occurs and the metering equipment corresponding to the collection failure to determine the data filtering rules corresponding to the key indicator deviation item of the electric energy meter collection success rate, etc.
  • the embodiment of the present application does not limit the determination method.
  • the current execution end performs a filtering operation on the monitoring data of the source end system based on the data filtering rule to obtain the target metering abnormal data.
  • the data filtering rule corresponding to the key indicator deviation item of the electric energy meter acquisition success rate in step 101 when used, when performing the filtering operation, the detailed information related to the acquisition corresponding to the failed acquisition time period A of the electric energy meter 001 in the source end system, the detailed information related to the acquisition corresponding to the failed acquisition time period B of the electric energy meter 003 in the failed acquisition time period B, the detailed information related to the acquisition corresponding to the failed acquisition time period C of the electric energy meter 014 in the failed acquisition time period C, and the detailed information related to the failed acquisition time period C of the electric energy meter 036 in the failed acquisition time period C should be obtained.
  • the current execution end compares and analyzes the target measurement abnormal data with the pre-configured abnormal case data, and determines at least one target case data from the abnormal case data.
  • the abnormal case data includes abnormal data of multiple dimensions, and the current execution end compares the abnormal data of multiple dimensions based on similarity when performing comparative analysis, thereby determining at least one target case data.
  • the current execution end performs multi-dimensional cluster analysis on the target measurement abnormal data to obtain the aggregation degree of the target measurement abnormal data in multiple data dimensions.
  • the multi-dimensional cluster analysis can use statistical principles to perform quantitative statistics, and can also use machine learning models for cluster analysis, such as fuzzy clustering, multivariate regression, decision trees, support vector machines, etc.
  • the embodiment of the present application does not limit the method corresponding to the multi-dimensional cluster analysis.
  • the current execution end also analyzes the causes of measurement anomalies based on the target case data and the aggregation degree, such as analyzing the correlation between the causes of anomalies in the target case data and the attribute values on the data dimensions with high aggregation degrees, etc., which are not limited in the embodiment of the present application.
  • At least one data filtering rule is determined based on the primary analysis data, including:
  • the current execution end obtains the mapping relationship between the key indicator and the indicator association information, and determines the target indicator association information corresponding to the key indicator deviation item based on the mapping relationship.
  • the mapping relationship is pre-configured based on experience.
  • the target indicator association information that has a mapping relationship with the key indicator deviation item, the electric energy meter acquisition success rate deviation can be the electric energy meter file information (manufacturer, batch, model, specification and other static information), business information (operating time, acquisition and maintenance, on-site re-recording, on-site inspection, on-site inspection, installation and disassembly and other business time, personnel, conclusions, etc.), and operation information that may affect the electric energy meter acquisition quality (voltage curve, current curve, terminal address, clock, battery undervoltage event, power-off power-on event, cover opening record, clock error and other dynamic information).
  • the target indicator association information that has a mapping relationship with the key indicator deviation item, the electric energy meter inaccuracy deviation can be the electric energy meter file information (manufacturer, batch, model, specification, accuracy, etc.).
  • the embodiments of the present application do not limit the type of target indicator associated information.
  • the current execution end obtains the time period in which the key indicator deviation item is generated, and determines the data filtering rule based on the time period in which the key indicator deviation item is generated and the target indicator association information. For example, when the key indicator deviation item is the deviation of the success rate of electric energy meter acquisition, the target indicator association information of the corresponding electric energy meter in the time period in which the above deviation is generated is determined as the data filtering rule; when the key indicator deviation item is the inaccuracy deviation of the electric energy meter, the target indicator association information of the corresponding electric energy meter in the time period in which the above deviation is generated is determined as the data filtering rule, etc.
  • the embodiment of the present application does not limit the method of determining the data filtering rule based on the time period and the target indicator association information.
  • the method further includes:
  • the current execution end performs data logic verification processing on the target metering abnormal data, deletes the target metering abnormal data that does not conform to the data logic in the target metering abnormal data, and realizes data cleaning in terms of data logic.
  • the current execution end can also perform data consistency verification processing on the target metering abnormal data, deletes the target metering abnormal data that does not conform to the data consistency in the target metering abnormal data, and realizes data cleaning in terms of data consistency.
  • target measurement abnormal data is compared and analyzed with pre-configured abnormal case data, and at least one target case data is determined from the abnormal case data, including:
  • Acquire case attribute information of the abnormal case data match the case attribute information with the to-be-matched attribute information of the target measurement abnormal data, and determine the target attribute information from the to-be-matched attribute information; perform similarity calculation based on the attribute value corresponding to the target attribute information and the attribute value corresponding to the case attribute information, and determine the abnormal case data corresponding to the case attribute information with a similarity higher than a threshold as the target case data.
  • the current execution end obtains case attribute information of abnormal case data, including the name of abnormal metering equipment, abnormal index, abnormal duration, abnormal impact range, etc.
  • the type of example attribute information is not limited.
  • the current execution end uses multiple attribute information of the target metering abnormal data as the attribute information to be matched, matches the case attribute information with the attribute information to be matched, and determines the target attribute information from the attribute information to be matched. For example, the name of the abnormal metering equipment, abnormal index, abnormal duration, and abnormal impact range that are successfully matched from the numerous attribute information of the target metering abnormal data are determined as the target attribute information.
  • the current execution end performs a degree of calculation based on the attribute value corresponding to the target attribute information and the attribute value corresponding to the case attribute information, such as the attribute value corresponding to the name of the abnormal metering device is the electric energy meter, the attribute value corresponding to the abnormal indicator is the acquisition success rate value, the attribute value corresponding to the abnormal duration is the duration (such as 5 minutes, 1 hour, 30 minutes, etc.), the attribute value corresponding to the abnormal impact range is the size of the impact area (such as the number of abnormal electric energy meters, the number of areas with abnormal electric energy meters, etc.), etc.
  • the embodiment of the present application does not limit the measurement method of the attribute value.
  • the current execution end determines the abnormal case data corresponding to the case attribute information with a similarity higher than the similarity threshold as the target case data.
  • the threshold is a numerical value pre-set according to needs.
  • a multidimensional cluster analysis is performed on the target measurement abnormal data to obtain the aggregation degree of the target measurement abnormal data in multiple data dimensions, including:
  • the current execution end obtains the attribute value corresponding to a single attribute information in the target metering abnormal data, and uses the cluster analysis model to perform cluster analysis on the attribute value to obtain the degree of aggregation on a single data dimension.
  • the attribute value corresponding to the attribute information of the manufacturer ⁇ manufacturer A, manufacturer B, manufacturer A, manufacturer C, manufacturer B, ... ⁇ is obtained for cluster analysis to obtain the degree of aggregation of manufacturer A, the degree of aggregation of manufacturer B, the degree of aggregation of manufacturer C, etc.; for another example, the attribute value corresponding to the attribute information of the installation position ⁇ position 1, position 3, position 1, position 1, position 2, position 3, position 3, ... ⁇ is obtained for cluster analysis to obtain the degree of aggregation of position 1, the degree of aggregation of position 2, the degree of aggregation of position 3, etc.
  • the current execution end combines the above-obtained aggregation degrees on a single data dimension to obtain aggregation degrees on multiple data dimensions.
  • the above-mentioned aggregation degrees can be used to analyze batch abnormalities of metering equipment, such as the deviation of the inaccuracy rate of electric energy meters is mainly due to the products of manufacturer A; the deviation of the collection success rate of electric energy meters mainly occurs at installation location 3, etc.
  • the method further includes:
  • the current execution end sorts the aggregation degrees in descending order, and obtains the attribute value with the highest aggregation degree in at least one data dimension as the attribute value to be analyzed. For example, when the aggregation degrees of electric energy meter manufacturer A, manufacturer B, and manufacturer C in the above embodiment are 74, 8, and 11 respectively, manufacturer A can be determined as the attribute value to be analyzed; when the aggregation degrees of electric energy meter installation position 1, installation position 2, and installation position 3 in the above embodiment are 9, 7, and 77 respectively, installation position 3 can be determined as the attribute value to be analyzed, etc.
  • the number of data dimensions to be sorted is not limited.
  • the degree of aggregation on only one data dimension shows the degree of aggregation characteristics
  • only the degree of aggregation on this data dimension can be sorted;
  • multiple condition constraints, weighted statistics or conditional analysis can be used to sort the degrees of aggregation on multiple data dimensions to form a combination scheme of attribute values to be analyzed.
  • the current execution end obtains the source end system backtracking rule associated with the attribute value to be analyzed, wherein the source end system backtracking rule is used to characterize the method of using the control variable method to perform source end system abnormality backtracking on the attribute value to be analyzed.
  • the actual cause of the metering abnormality is the process problem of the installer, but because of the objective law of procurement and supply, there is a batch of electric energy meters generated by the same manufacturer, the same specifications, the same production batch, and even the geographical location of the installation is relatively close, then the determined attribute value to be analyzed contains the attribute values of the same manufacturer, the same specifications, the same batch, and the same installation location.
  • the associated source end system backtracking rule can be to obtain the monitoring data of electric energy meters with the same manufacturer, the same production batch, and the same specifications, but installed in different power units, to determine whether there is an abnormality or an increase in the abnormality rate; or to obtain the monitoring data of electric energy meters with the same installation location, the same manufacturer, the same specifications, but different installation times, to determine whether there is an abnormality or an increase in the abnormality rate, etc.
  • the embodiment of the present application does not limit the content of the source end system backtracking rule.
  • the current execution end determines whether the attribute value to be analyzed is the cause of the abnormality based on the source end system backtracking rule, and deletes the attribute value to be analyzed corresponding to the cause of the abnormality.
  • the source end system backtracking rule of step 302 is used to obtain the monitoring data of the electric energy meters of the same manufacturer, the same production batch, and the same specification, but installed in different power units.
  • the attribute value corresponding to the installation location is deleted, and the cause of the abnormality of the electric energy meter caused by the installation location is excluded, that is, the attribute value corresponding to the installation location is a non-abnormal cause; if the source system backtracking rule of step 302 is adopted to obtain the monitoring data of the electric energy meters with the same installation location, the same manufacturer, the same specifications, but different installation times, and it is determined that there is no abnormality in the monitoring data of the electric energy meters with different installation times, the attribute value corresponding to the installation time is deleted, and the cause of the abnormality of the electric energy meter caused by the installation time is excluded, that is, the attribute value corresponding to the installation time is a non-abnormal cause, etc.
  • the method further includes:
  • the measurement anomaly is risk graded according to a preset classification and grading strategy to obtain the measurement anomaly risk level; when the measurement anomaly risk level exceeds the risk level threshold, a language model is used to generate risk warning information; and the risk warning information is sent to the relevant operation and maintenance party through a preset risk warning interface.
  • the current execution end classifies the risk of the metering anomaly based on the cause of the metering anomaly and in accordance with a preset classification and grading strategy, and obtains the risk level of the metering anomaly.
  • the metering anomaly risk is divided into level one, level two, level three, etc. according to factors such as the potential risk range, the degree of harmfulness, and the size of the impact range.
  • the embodiment of the present application does not limit the method of dividing the metering anomaly risk level.
  • the current execution end compares the metering anomaly risk level with the set risk level threshold. When the metering anomaly risk level exceeds the risk level threshold, a language model is used to generate risk warning information.
  • a risk warning information is generated.
  • the embodiment of the present application does not limit the set risk level threshold.
  • the current execution end obtains a pre-configured risk warning interface and sends the risk warning information to the relevant operation and maintenance party through the risk warning interface.
  • the risk warning interface is used to characterize the network interface of a peripheral communication unit that completes data communication in a wired or wireless manner.
  • the peripheral communication unit is responsible for the necessary transcoding of warning information and control instructions, and adopts reasonable communication methods to transmit them to system servers, monitoring screens, acquisition terminals, energy meters, operation terminals, handheld devices, staff mobile phones and other external devices, and ensure the quality of data transmission.
  • the transmission channel involves the application of public networks or external network devices, data desensitization or encryption processing should also be performed as needed.
  • risk warning can also be carried out by system prompts, screen pop-ups, emails, phone calls, etc.
  • the current execution end can also access the risk information for system storage and temporary storage, and include it in the abnormal case database after the cause of the problem is verified.
  • the embodiment of the present application provides a method for analyzing metering anomalies.
  • the present application obtains primary analysis data obtained by analyzing the source end system, and determines at least one data filtering rule based on the primary analysis data; performs filtering operation on the monitoring data of the source end system based on the data filtering rule to obtain target metering anomaly data; compares and analyzes the target metering anomaly data with pre-configured abnormal case data, and determines at least one target case data from the abnormal case data; and compares the target metering anomaly data with the pre-configured abnormal case data.
  • the target metering anomaly data is subjected to multi-dimensional cluster analysis to obtain the aggregation degree of the target metering anomaly data in multiple data dimensions; and the cause of the metering anomaly is analyzed based on the target case data and the aggregation degree, thereby realizing the cause analysis of the power metering anomaly.
  • the present application utilizes the primary analysis data obtained by calculating the indicators in the relevant technology to obtain source data of more dimensions, which not only expands the analysis dimension but also reduces the amount of data acquired from the source system.
  • the present application adopts a method that combines case matching and cluster analysis on the source system data to improve the accuracy of the analysis and judgment of metering anomaly problems under the joint action of multiple factors.
  • an embodiment of the present application provides a measurement abnormality analysis device, as shown in FIG. 4 , the device includes:
  • the filtering rule determination module 41 is configured to obtain the primary analysis data obtained by the source system analysis, and determine at least one data filtering rule based on the primary analysis data;
  • the data acquisition module 42 is configured to filter the monitoring data of the source system based on the data filtering rule to obtain the target measurement abnormality data;
  • the case determination module 43 is configured to compare and analyze the target measurement abnormality data with the pre-configured abnormal case data, and determine at least one target case data from the abnormal case data;
  • the cause analysis module 44 is configured to perform multidimensional clustering analysis on the target measurement abnormality data to obtain the aggregation degree of the target measurement abnormality data in multiple data dimensions; and analyze the cause of measurement abnormality based on the target case data and the aggregation degree.
  • the primary analysis data includes at least one key indicator deviation item; the filtering rule determination module 41 is also configured as:
  • the device further includes a data cleaning module 45 configured to:
  • the case determination module 43 is further configured to:
  • Acquire case attribute information of the abnormal case data match the case attribute information with the to-be-matched attribute information of the target measurement abnormal data, and determine the target attribute information from the to-be-matched attribute information; perform similarity calculation based on the attribute value corresponding to the target attribute information and the attribute value corresponding to the case attribute information, and determine the abnormal case data corresponding to the case attribute information with a similarity higher than a threshold as the target case data.
  • the cause analysis module 44 is further configured to:
  • the device further includes an abnormality backtracking module 46, which is configured as follows:
  • the aggregation degrees on at least one data dimension are sorted in descending order, and the attribute value to be analyzed with the highest aggregation degree on each data dimension is obtained; a source-end system backtracking rule associated with the attribute value to be analyzed is obtained, wherein the source-end system backtracking rule is used to characterize a method of performing source-end system abnormality backtracking on the attribute value to be analyzed by using a control variable method; based on the source-end system backtracking rule, whether the attribute value to be analyzed is an abnormal cause is judged, and the attribute value to be analyzed corresponding to a non-abnormal cause is deleted.
  • the device further includes an abnormal warning module 47, which is configured as follows:
  • the measurement anomaly is risk graded according to a preset classification and grading strategy to obtain the measurement anomaly risk level; when the measurement anomaly risk level exceeds the risk level threshold, a language model is used to generate risk warning information; and the risk warning information is sent to the relevant operation and maintenance party through a preset risk warning interface.
  • the embodiment of the present application provides an analysis device for metering anomalies.
  • the present application obtains the primary analysis data obtained by the source end system analysis, and determines at least one data filtering rule based on the primary analysis data; performs filtering operation on the monitoring data of the source end system based on the data filtering rule to obtain target metering anomaly data; compares and analyzes the target metering anomaly data with the pre-configured abnormal case data, and determines at least one target case data from the abnormal case data; performs multidimensional clustering analysis on the target metering anomaly data to obtain the aggregation degree of the target metering anomaly data in multiple data dimensions; and analyzes the cause of metering anomalies based on the target case data and the aggregation degree, thereby realizing the cause analysis of power metering anomalies.
  • the present application uses the primary analysis data obtained by the indicator calculation in the related art to obtain source end data with more dimensions, which not only expands the analysis dimension but also reduces the amount of data obtained from the source end system.
  • the present application adopts a method combining case matching and cluster analysis on the source end system data, which improves the accuracy of the analysis and judgment of metering anomaly problems under the joint action of multiple factors.
  • a storage medium stores at least one executable instruction, and the computer executable instruction can execute the measurement anomaly analysis method in any of the above method embodiments.
  • FIG6 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application.
  • the embodiment of the present application does not limit the implementation of the computer device.
  • the computer device may include: a processor (processor) 602 , a communication interface (Communications Interface) 604 , a memory (memory) 606 , and a communication bus 608 .
  • processor processor
  • Communication interface Communication interface
  • memory memory
  • communication bus 608
  • the processor 602 , the communication interface 604 , and the memory 606 communicate with each other via a communication bus 608 .
  • the communication interface 604 is configured to communicate with network elements of other devices such as clients or other servers.
  • the processor 602 is configured to execute the program 610 and can execute the relevant steps of the above-mentioned measurement anomaly analysis method.
  • the program 610 may include program codes including computer operation instructions.
  • Processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.
  • the one or more processors included in the computer device may be processors of the same type, such as one or more CPUs; or may be processors of different types, such as one or more CPUs and one or more ASICs.
  • the memory 606 is configured to store the program 610.
  • the memory 606 may include a high-speed random access memory (RAM) memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage.
  • RAM random access memory
  • non-volatile memory such as at least one disk storage.
  • the program 610 may be used to cause the processor 602 to perform the following operations:
  • the above-mentioned multiple modules or multiple steps of the present application can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into multiple integrated circuit modules respectively, or multiple modules or steps therein can be made into a single integrated circuit module for implementation.
  • the present application is not limited to any specific combination of hardware and software.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • General Business, Economics & Management (AREA)
  • Water Supply & Treatment (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Primary Health Care (AREA)
  • Marketing (AREA)
  • Human Resources & Organizations (AREA)
  • Probability & Statistics with Applications (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

计量异常的分析方法及装置、存储介质、计算机设备。计量异常的分析方法包括：获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则(101)；基于所述数据过滤规则对源端系统的滤监测数据进行过滤操作，得到目标计量异常数据(102)；将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据(103)；对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因(104)。

Description

计量异常的分析方法及装置、存储介质、计算机设备
本申请要求在2023年06月16日提交中国专利局、申请号为202310720828.3的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及数据处理及分析技术领域，例如涉及计量异常的分析方法及装置、存储介质、计算机设备。
背景技术
计量异常事关电力市场公平公正和客户供电服务感知，保证电力计量设备正常运行是整个电力营销环境中非常重要的一环。电力计量设备的异常成因十分复杂，包括业务问题、设备问题、特殊工况问题、运行环境问题等，若对电力计量设备的异常排除不清楚、不及时，则会造成重大事故。
计量异常分析方法为根据关注程度和重要性，从大量计量设备运行信息中，选取部分数据项作为关键业绩数据，再基于关键业绩数据进行指标计算，从而基于指标值确定计量设备和计量业务的整体质量。但是，采用上述分析方法时分析维度小、校验逻辑简单，对多影响因素共同作用问题的研判准确性低，从而导致计量异常问题复发、扩散和转移，无法起到缩减电力运维成本的效果。
发明内容
本申请提供计量异常的分析方法及装置、存储介质、计算机设备，解决了电力计量异常分析方法由于分析维度小、校验逻辑简单导致的异常问题研判准确性低的问题。
本申请一个方面，提供了一种计量异常的分析方法，包括：
获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；
基于所述数据过滤规则对所述源端系统的监测数据进行过滤操作，得到目标计量异常数据；
将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；
对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据 在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
本申请另一个方面，提供了一种计量异常的分析装置，包括：
过滤规则确定模块，设置为获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；
数据获取模块，设置为基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据；
案例确定模块，设置为将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；
成因分析模块，设置为对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
本申请的又一方面，提供了一种存储介质，所述存储介质中存储有至少一可执行指令，所述可执行指令使处理器执行上述的计量异常的分析方法。
本申请再一个方面，提供了一种计算机设备，包括处理器、存储器、通信接口和通信总线，所述处理器、所述存储器和所述通信接口通过所述通信总线完成相互间的通信；
所述存储器设置为存放至少一可执行指令，所述可执行指令使所述处理器执行上述的计量异常的分析方法。
附图说明
图1是本申请实施例提供的一种计量异常的分析方法的流程示意图；
图2是本申请实施例提供的另一种计量异常的分析方法的流程示意图；
图3是本申请实施例提供的又一种计量异常的分析方法的流程示意图；
图4是本申请实施例提供的一种计量异常的分析装置的结构示意图；
图5是本申请实施例提供的又一种计量异常的分析装置的结构示意图；
图6是本申请实施例提供的一种计算机设备的结构示意图。
具体实施方式
下面将参照附图描述本申请的示例性实施例。虽然附图中显示了本申请的示例性实施例，然而可以以多种形式实现本申请。提供这些实施例是为了理解本 申请。
本申请实施例提供了一种计量设备的分析方法，如图1所示，该方法包括：
101、获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则。
本申请实施例中，当前执行端获取经源端系统分析得到的初阶分析数据，其中，源端系统用于表征具有电力计量设备的在线数据监测、电力计量设备的失准监测、接收客户投诉等功能的系统，可以实时采集系统内所有计量设备的运行状态数据，还可以基于离散的功能派发工单，并在每月指标测算中发现计量设备异常发生率偏高、客户服务满意率下降等问题。其中，初阶分析数据用于表征源端系统的系统分析功能得到的分析数据，包括采集成功率、电能表失准、电能表潮流反向、电压断相、客户投诉、线损越限、负荷超容、电量突变等近百项关键指标，本申请实施例对初阶分析数据的类型不做限定。当前执行端得到初阶分析数据后，基于初阶分析数据确定至少一条数据过滤规则。
由于初阶分析数据可以反应一部分计量设备的异常问题，即至少一个关键指标发生指标偏差问题，将发生指标偏差问题的关键指标称为关键指标偏差项。在确定数据过滤规则时，需要基于关键指标偏差项进行数据过滤。如电能表采集成功率出现了偏差，那就基于正常的采集成功率先确定电能表采集成功率出现偏差的时间段、采集失败对应的计量设备以及和采集相关的明细信息，明细信息包括采集设备档案信息(厂家、批次、型号、规约等静态信息)、业务信息(运行时间、采集运维、现场补录、现场巡视、现场检验、装拆等业务的时间、人员、结论等)、可能影响采集质量的运行信息(电压曲线、电流曲线、终端地址、时钟、电池欠压事件、停电上电事件、开盖记录、时钟误差等动态信息)等，本申请实施例对明细信息的类型不做限定。当前执行端将上述电能表采集成功率关键指标偏差项所涉及的采集相关的明细信息，按出现偏差的时间段和采集失败对应的计量设备进行数据过滤确定与电能表采集成功率关键指标偏差项对应的数据过滤规则等，本申请实施例对确定方式不做限定。
102、基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据。
本申请实施例中，当前执行端基于数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据。如采用步骤101中与电能表采集成功率关键指标偏差项对应的数据过滤规则，进行过滤操作时，应获取源端系统中采集失败电能表001在采集失败时间段A对应的与采集相关的明细信息、采集失败电能表003在采集失败时间段B对应的与采集相关的明细信息、采集失败电能表014在采集失败时间段C对应的与采集相关的明细信息、采集失败电能表036 在采集失败时间段D对应的与采集相关的明细信息、采集失败电能表102在采集失败时间段E对应的与采集相关的明细信息等。其中，与采集相关的明细信息参见步骤101。
103、将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据。
本申请实施例中，当前执行端将目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据。其中，异常案例数据包括多个维度的异常数据，当前执行端在进行对比分析时基于多个维度的异常数据相似度进行对比，从而确定出至少一个目标案例数据。
104、对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
本申请实施例中，当前执行端对目标计量异常数据进行多维度聚类分析，得到目标计量异常数据在多个数据维度上的聚合度。其中，多维聚类分析可以采用统计学原理进行数量统计，还可以采用机器学习模型进行聚类分析，如模糊聚类、多元回归、决策树、支持向量机等，本申请实施例对多维聚类分析对应的方法不做限定。当前执行端还基于目标案例数据与聚合度分析计量异常成因，如分析目标案例数据中的异常成因与聚合度高的数据维度上的属性值之间的相关性等，本申请实施例不做限定。
一实施例中，为了获取与计量异常相关的数据，过滤掉与计量异常无关的数据，基于所述初阶分析数据确定至少一条数据过滤规则，包括：
获取关键指标与指标关联信息的映射关系，并基于所述映射关系确定与所述关键指标偏差项对应的目标指标关联信息；获取产生所述关键指标偏差项的时间段，并基于所述时间段和所述目标指标关联信息确定所述数据过滤规则。
本申请实施例中，当前执行端获取关键指标与指标关联信息的映射关系，并基于映射关系确定与关键指标偏差项对应的目标指标关联信息。其中，映射关系为根据经验预先配置的，如与关键指标偏差项电能表采集成功率偏差具有映射关系的目标指标关联信息可以是电能表档案信息(厂家、批次、型号、规格等静态信息)、业务信息(运行时间、采集运维、现场补录、现场巡视、现场检验、装拆等业务的时间、人员、结论等)、可能影响电能表采集质量的运行信息(电压曲线、电流曲线、终端地址、时钟、电池欠压事件、停电上电事件、开盖记录、时钟误差等动态信息)等。又例如，与关键指标偏差项电能表失准偏差具有映射关系的目标指标关联信息可以是电能表档案信息(厂家、批次、型号、规格、准 确度等级、检定误差等静态信息)、业务信息(历次现场检验误差、安装时间、安装位置、运行环境、近期现场业务的内容和人员、近期状态评价结论等等)、可能影响运行误差的运行信息(电压、电流、过载、缺相、负荷不平衡、特殊功率因素、开盖记录等)等，本申请实施例对目标指标关联信息的类型不做限定。
本申请实施例中，当前执行端获取产生关键指标偏差项的时间段，并基于产生关键指标偏差项的时间段和目标指标关联信息确定所述数据过滤规则。如当关键指标偏差项为电能表采集成功率偏差时，则将产生上述偏差的时间段内相应电能表的目标指标关联信息确定为数据过滤规则；当关键指标偏差项为电能表失准偏差时，则将产生上述偏差的时间段内相应电能表的目标指标关联信息确定为数据过滤规则等，本申请实施例对基于时间段和目标指标关联信息确定数据过滤规则的方式不做限定。
一实施例中，为了使得获取的源端系统监测数据在逻辑和数据一致性上更加准确，便于后期数据分析，在基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据之后，所述方法还包括：
对所述目标计量异常数据进行数据逻辑核验处理，删除所述目标计量异常数据中不符合数据逻辑的所述目标计量异常数据；和/或，对所述目标计量异常数据进行数据一致性核验处理，删除所述目标计量异常数据中不符合数据一致性的所述目标计量异常数据。
本申请实施例中，当前执行端对目标计量异常数据进行数据逻辑核验处理，删除目标计量异常数据中不符合数据逻辑的目标计量异常数据，实现了在数据逻辑上对数据进行清洗。此外，当前执行端还可以对目标计量异常数据进行数据一致性核验处理，删除目标计量异常数据中不符合数据一致性的目标计量异常数据，实现了在数据一致性上对数据进行清洗。
一实施例中，为了支持类案检索，提高类案检索的可参考性，将目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据，包括：
获取所述异常案例数据的案例属性信息，并将所述案例属性信息与所述目标计量异常数据的待匹配属性信息进行匹配处理，从所述待匹配属性信息中确定目标属性信息；基于所述目标属性信息对应的属性值和所述案例属性信息对应的属性值进行相似度计算，将相似度高于阈值的案例属性信息对应的异常案例数据确定为所述目标案例数据。
本申请实施例中，当前执行端获取异常案例数据的案例属性信息，包括异常计量设备名称、异常指标、异常持续时间、异常影响范围等，本申请实施例对案 例属性信息的类型不做限定。当前执行端将目标计量异常数据的多个属性信息作为待匹配属性信息，将案例属性信息与待匹配属性信息进行匹配处理，从待匹配属性信息中确定目标属性信息。如从目标计量异常数据的众多属性信息中将匹配成功的异常计量设备名称、异常指标、异常持续时间、异常影响范围确定为目标属性信息等。
本申请实施例中，当前执行端基于目标属性信息对应的属性值和案例属性信息对应的属性值进行相度计算，如异常计量设备名称对应的属性值为电能表、异常指标对应的属性值为采集成功率值、异常持续时间对应的属性值为持续时长(如5分钟、1小时、30分钟等)、异常影响范围对应的属性值为影响区域的大小(如异常电能表数量、电能表异常的片区数量等)等，本申请实施例对属性值的计量方式不做限定。当前执行端将相似度高于相似度阈值的案例属性信息对应的异常案例数据确定为目标案例数据。其中，该阈值为根据需要预先设定的数值。
一实施例中，为了从多个维度上对目标计量异常数据进行深入分析，如图2所示，对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度，包括：
201、获取所述目标计量异常数据中单个属性信息对应的属性值。
202、采用聚类分析模型对所述属性值进行聚类分析，得到单个数据维度上的聚合度。
203、将所述单个数据维度上的聚合度进行合并，得到多个数据维度上的聚合度。
本申请实施例中，当前执行端获取目标计量异常数据中单个属性信息对应的属性值，采用聚类分析模型对属性值进行聚类分析，得到单个数据维度上的聚合度。如，获取厂家这一个属性信息对应的属性值{厂家A、厂家B、厂家A、厂家C、厂家B、......}进行聚类分析，得到厂家A的聚合度、厂家B的聚合度、厂家C的聚合度等；又如，获取安装位置这一属性信息对应的属性值{位置1、位置3、位置1、位置1、位置2、位置3、位置3、......}进行聚类分析，得到位置1的聚合度、位置2的聚合度、位置3的聚合度等。
当前执行端将上述得到的单个数据维度上的聚合度进行合并，得到多个数据维度上的聚合度。通过上述聚合度可以分析计量设备的批量异常情况，如，电能表失准率偏差主要为生产厂家A的产品；电能表采集成功率偏差主要发生在安装位置3上等。
一实施例中，为了避免幸存者误差，即避免分析的异常原因不符合实际异常 情况，提高计量异常分析的准确率，如图3所示，在对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度之后，所述方法还包括：
301、按由高到低的顺序对至少一个数据维度上的聚合度分别进行排序处理，并获取每个数据维度上聚合度最高的待分析属性值。
本申请实施例中，当前执行端按由高到低的顺序对聚合度进行排序处理，并分别获取至少一个数据维度上聚合度最高的属性值作为待分析属性值，如，当上述实施例中电能表生产厂家A的聚合度、厂家B的聚合度、厂家C的聚合度分别为74、8、11时，可以将厂家A确定为待分析属性值；当上述实施例中电能表安装位置1的聚合度、安装位置2的聚合度、安装位置3的聚合度分别为9、7、77时，可以将安装位置3确定为待分析属性值等。
本实施例中，对进行排序处理的数据维度的数量不作限定。在仅有一个数据维度上的聚合度表现出聚合度特征的情况下，则可以仅对该数据维度上的聚合度排序；在有多个数据维度上的聚合度均表现出聚合度特征的情况下，则可以采用多条件约束、加权统计或者条件分析等方式，对多个数据维度上的聚合度进行排序处理，形成待分析属性值的组合方案。
302、获取与所述待分析属性值相关联的源端系统回溯规则。
本申请实施例中，当前执行端获取与待分析属性值相关联的源端系统回溯规则，其中，源端系统回溯规则用于表征采用控制变量法对所述待分析属性值进行源端系统异常回溯的方法。如，计量异常的实际原因是安装人员的工艺问题，但是因为采购供货的客观规律，有一批电能表表生成厂家相同、规格相同、生产批次相同，甚至安装的地理位置都比较接近，则确定的待分析属性值中包含相同厂家、相同规格、相同批次、相同安装位置的属性值。那么，相关联的源端系统回溯规则可以是获取相同生产厂家、相同生产批次、相同规格，但安装在不同电力单位中的电能表的监测数据，判断有没有异常或者异常率上升；或者获取相同安装位置、相同厂家、相同规格，但安装时间不同的电能表监测数据，判断有没有异常或者异常率上升等，本申请实施例对源端系统回溯规则的内容不做限定。
303、基于所述源端系统回溯规则对所述待分析属性值是否为异常成因进行判断，并删除非异常成因对应的待分析属性值。
本申请实施例中，当前执行端基于源端系统回溯规则对待分析属性值是否为异常成因进行判断，并删除异常成因对应的待分析属性值。如采用步骤302中获取相同生产厂家、相同生产批次、相同规格，但安装在不同电力单位中的电能表的监测数据的源端系统回溯规则，经判断安装在不同电力单位中的电能表无 异常，则将安装位置对应的属性值删除，排除安装位置导致电能表异常的成因，即安装位置对应的属性值为非异常成因；如采用步骤302中获取相同安装位置、相同厂家、相同规格，但安装时间不同的电能表监测数据的源端系统回溯规则，经判断安装时间不同的电能表监测数据无异常，则将安装时间对应的属性值删除，排除安装时间导致电能表异常的成因，即安装时间对应的属性值为非异常成因等。
一实施例中，为了快速且准确地向相关运维方发送预警信息，在基于所述目标案例数据与所述聚合度分析计量异常成因之后，所述方法还包括：
基于所述计量异常成因，按照预置的分类分级策略对计量异常进行风险定级，得到计量异常风险级别；当所述计量异常风险级别超出风险级别阈值，采用语言模型生成风险预警信息；通过预置的风险预警接口，将所述风险预警信息发送至相关运维方。
本申请实施例中，当前执行端基于计量异常成因，按照预置的分类分级策略对计量异常进行风险定级，得到计量异常风险级别。如根据潜在风险范围、危害性高低、影响范围大小等因素将计量异常风险划分为一级、二级、三级等，本申请实施例对划分计量异常风险级别的方式不做限定。当前执行端对计量异常风险级别与设定的风险级别阈值进行比较，当计量异常风险级别超出风险级别阈值时，采用语言模型生成风险预警信息。如当计量异常风险级别为三级时超出风险级别阈值，生成风险预警信息等，本申请实施例对设定的风险级别阈值不做限定。当前执行端获取预先配置的风险预警接口，并通过风险预警接口将风险预警信息发送至相关运维方。其中，风险预警接口用于表征以有线或无线方式完成数据通信的外围通信单元的网络接口。外围通讯单元负责将预警信息和控制指令进行必要的转码，并采取合理的通讯方式传输至系统服务器、监控屏幕、采集终端、电能表、作业终端、掌机、工作人员手机等外部设备，并保证数据传输质量。传输信道涉及公共网络或外网设备应用时，还应按需进行数据脱敏或加密处理。
除了采用上述实施例中的风险预警方式以外，还可以采用系统提示、屏幕弹窗、发送邮件、拨打电话等方式进行风险预警。当前执行端除了发送预警信息以外，还可以将风险信息接入系统存储并暂存，待问题成因核实后纳入异常案例数据库。
本申请实施例提供了一种计量异常的分析方法，与相关技术相比，本申请通过获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据；将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；对所述 目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因，实现了对电力计量异常的成因分析。本申请利用相关技术中指标计算得到的初阶分析数据，获取更多维度的源端数据，不仅扩大了分析维度还减小了对源端系统的数据获取量。本申请采用对源端系统数据进行案例匹配和聚类分析相结合的方法，提高了多因素共同作用下计量异常问题的研判准确性。
作为对上述图1所示方法的实现，本申请实施例提供了一种计量异常的分析装置，如图4所示，该装置包括：
过滤规则确定模块41，设置为获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；数据获取模块42，设置为基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据；案例确定模块43，设置为将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；成因分析模块44，设置为对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
所述初阶分析数据包括至少一个关键指标偏差项；所述过滤规则确定模块41还设置为：
获取关键指标与指标关联信息的映射关系，并基于所述映射关系确定与所述关键指标偏差项对应的目标指标关联信息；获取产生所述关键指标偏差项的时间段，并基于所述时间段和所述目标指标关联信息确定所述数据过滤规则。
如图5所示，所述装置还包括数据清洗模块45，设置为：
对所述目标计量异常数据进行数据逻辑核验处理，删除所述目标计量异常数据中不符合数据逻辑的所述目标计量异常数据；和/或，对所述目标计量异常数据进行数据一致性核验处理，删除所述目标计量异常数据中不符合数据一致性的所述目标计量异常数据。
所述案例确定模块43还设置为：
获取所述异常案例数据的案例属性信息，并将所述案例属性信息与所述目标计量异常数据的待匹配属性信息进行匹配处理，从所述待匹配属性信息中确定目标属性信息；基于所述目标属性信息对应的属性值和所述案例属性信息对应的属性值进行相似度计算，将相似度高于阈值的案例属性信息对应的异常案例数据确定为所述目标案例数据。
所述成因分析模块44还设置为：
获取所述目标计量异常数据中单个属性信息对应的属性值；采用聚类分析模型对所述属性值进行聚类分析，得到单个数据维度上的聚合度；将所述单个数据维度上的聚合度进行合并，得到多个数据维度上的聚合度。
如图5所示，所述装置还包括异常回溯模块46，设置为：
按由高到低的顺序对至少一个数据维度上的聚合度分别进行排序处理，并获取每个数据维度上聚合度最高的待分析属性值；获取与所述待分析属性值相关联的源端系统回溯规则，其中，所述源端系统回溯规则用于表征采用控制变量法对所述待分析属性值进行源端系统异常回溯的方法；基于所述源端系统回溯规则对所述待分析属性值是否为异常成因进行判断，并删除非异常成因对应的待分析属性值。
如图5所示，所述装置还包括异常预警模块47，设置为：
基于所述计量异常成因，按照预置的分类分级策略对计量异常进行风险定级，得到计量异常风险级别；当所述计量异常风险级别超出风险级别阈值，采用语言模型生成风险预警信息；通过预置的风险预警接口，将所述风险预警信息发送至相关运维方。
本申请实施例提供了一种计量异常的分析装置，与相关技术相比，本申请通过获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据；将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因，实现了对电力计量异常的成因分析。本申请利用相关技术中指标计算得到的初阶分析数据，获取更多维度的源端数据，不仅扩大了分析维度还减小了对源端系统的数据获取量。本申请采用对源端系统数据进行案例匹配和聚类分析相结合的方法，提高了多因素共同作用下计量异常问题的研判准确性。
根据本申请一个实施例提供了一种存储介质，所述存储介质存储有至少一可执行指令，该计算机可执行指令可执行上述任意方法实施例中的计量异常的分析方法。
图6为本申请实施例提供的一种计算机设备的结构示意图，本申请实施例并不对计算机设备的实现做限定。
如图6所示，该计算机设备可以包括：处理器(processor)602、通信接口(Communications Interface)604、存储器(memory)606、以及通信总线608。
处理器602、通信接口604、以及存储器606通过通信总线608完成相互间的通信。
通信接口604，设置为与其它设备比如客户端或其它服务器等的网元通信。
处理器602，设置为执行程序610，可以执行上述计量异常的分析方法的相关步骤。
程序610可以包括程序代码，该程序代码包括计算机操作指令。
处理器602可能是中央处理器(Central Processing Unit，CPU)，或者是特定集成电路(Application Specific Integrated Circuit,ASIC)，或者是被配置成实施本申请实施例的一个或多个集成电路。计算机设备包括的一个或多个处理器，可以是同一类型的处理器，如一个或多个CPU；也可以是不同类型的处理器，如一个或多个CPU以及一个或多个ASIC。
存储器606，设置为存放程序610。存储器606可能包含高速随机存储器(Random Access Memory，RAM)存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。
程序610可以用于使得处理器602执行以下操作：
获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；基于所述数据过滤规则对源端系统的监测数据进行过滤操作，得到目标计量异常数据；将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
本领域的技术人员应该明白，上述的本申请的多个模块或多个步骤可以用通用的计算装置来实现，它们可以集中在单个的计算装置上，或者分布在多个计算装置所组成的网络上，它们可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，并且在一些情况下，可以以不同于此处的顺序执行所示出或描述的步骤，或者将它们分别制作成多个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本申请不限制于任何特定的硬件和软件结合。

Claims (10)

  1. 一种计量异常的分析方法，包括：
    获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；
    基于所述数据过滤规则对所述源端系统的监测数据进行过滤操作，得到目标计量异常数据；
    将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；
    对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
  2. 根据权利要求1所述的方法，其中，所述初阶分析数据包括至少一个关键指标偏差项；
    所述基于所述初阶分析数据确定至少一条数据过滤规则，包括：
    获取关键指标与指标关联信息的映射关系，并基于所述映射关系确定与所述关键指标偏差项对应的目标指标关联信息；
    获取产生所述关键指标偏差项的时间段，并基于所述时间段和所述目标指标关联信息确定所述数据过滤规则。
  3. 根据权利要求1所述的方法，在所述基于所述数据过滤规则对所述源端系统的监测数据进行过滤操作，得到目标计量异常数据之后，还包括以下至少之一：
    对所述目标计量异常数据进行数据逻辑核验处理，删除所述目标计量异常数据中不符合数据逻辑的目标计量异常数据；
    对所述目标计量异常数据进行数据一致性核验处理，删除所述目标计量异常数据中不符合数据一致性的目标计量异常数据。
  4. 根据权利要求1所述的方法，其中，所述将目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据，包括：
    获取所述异常案例数据的案例属性信息，并将所述案例属性信息与所述目标计量异常数据的待匹配属性信息进行匹配处理，从所述待匹配属性信息中确定目标属性信息；
    基于所述目标属性信息对应的属性值和所述案例属性信息对应的属性值进 行相似度计算，将相似度高于阈值的案例属性信息对应的异常案例数据确定为所述目标案例数据。
  5. 根据权利要求1所述的方法，其中，所述对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度，包括：
    获取所述目标计量异常数据中单个属性信息对应的属性值；
    采用聚类分析模型对所述属性值进行聚类分析，得到所述单个数据维度上的聚合度；
    将单个数据维度上的聚合度进行合并，得到多个数据维度上的聚合度。
  6. 根据权利要求1所述的方法，在所述对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度之后，还包括：
    按由高到低的顺序对至少一个数据维度上的聚合度分别进行排序处理，并获取每个数据维度上聚合度最高的待分析属性值；
    获取与所述待分析属性值相关联的源端系统回溯规则，其中，所述源端系统回溯规则用于表征采用控制变量法对所述待分析属性值进行源端系统异常回溯的方法；
    基于所述源端系统回溯规则对所述待分析属性值是否为异常成因进行判断，并删除非异常成因对应的待分析属性值。
  7. 根据权利要求1～6任一项所述的方法，在所述基于所述目标案例数据与所述聚合度分析计量异常成因之后，还包括：
    基于所述计量异常成因，按照预置的分类分级策略对计量异常进行风险定级，得到计量异常风险级别；
    在所述计量异常风险级别超出风险级别阈值的情况下，采用语言模型生成风险预警信息；
    通过预置的风险预警接口，将所述风险预警信息发送至运维方。
  8. 一种计量异常的分析装置，包括：
    过滤规则确定模块，设置为获取经源端系统分析得到的初阶分析数据，并基于所述初阶分析数据确定至少一条数据过滤规则；
    数据获取模块，设置为基于所述数据过滤规则对所述源端系统的监测数据进行过滤操作，得到目标计量异常数据；
    案例确定模块，设置为将所述目标计量异常数据与预先配置的异常案例数据进行对比分析，从所述异常案例数据中确定至少一个目标案例数据；
    成因分析模块，设置为对所述目标计量异常数据进行多维聚类分析，得到所述目标计量异常数据在多个数据维度上的聚合度；并基于所述目标案例数据与所述聚合度分析计量异常成因。
  9. 一种存储介质，所述存储介质中存储有至少一可执行指令，所述可执行指令执行如权利要求1-7中任一项所述的计量异常的分析方法。
  10. 一种计算机设备，包括处理器、存储器、通信接口和通信总线，所述处理器、所述存储器和所述通信接口通过所述通信总线完成相互间的通信；
    所述存储器设置为存放至少一可执行指令，所述可执行指令使所述处理器执行如权利要求1-7中任一项所述的计量异常的分析方法。
PCT/CN2023/123294 2023-06-16 2023-10-08 计量异常的分析方法及装置、存储介质、计算机设备 Ceased WO2024255020A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/578,470 US20240420157A1 (en) 2023-06-16 2023-10-08 Metering abnormality analysis method and apparatus, storage medium, and computer device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202310720828.3 2023-06-16
CN202310720828.3A CN116756684A (zh) 2023-06-16 2023-06-16 一种计量异常的分析方法及装置、存储介质、计算机设备

Publications (1)

Publication Number Publication Date
WO2024255020A1 true WO2024255020A1 (zh) 2024-12-19

Family

ID=87952807

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/123294 Ceased WO2024255020A1 (zh) 2023-06-16 2023-10-08 计量异常的分析方法及装置、存储介质、计算机设备

Country Status (2)

Country Link
CN (1) CN116756684A (zh)
WO (1) WO2024255020A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119379425A (zh) * 2024-12-31 2025-01-28 杭银消费金融股份有限公司 一种用于逾期指标变化的归因分析方法及系统
CN119886934A (zh) * 2024-12-26 2025-04-25 深圳供电局有限公司 一种对于电能表表码录入异常的检测方法及系统
CN120067879A (zh) * 2025-04-29 2025-05-30 环德(福建)环保科技有限公司 污水处理设备的多源数据融合与远程诊断方法及系统
CN120064522A (zh) * 2025-04-29 2025-05-30 国网福建省电力有限公司电力科学研究院 一种实验室仪器数据自动采集与分析方法及系统
CN121502619A (zh) * 2026-01-13 2026-02-10 江阴长仪集团有限公司 基于电能表大数据的电能表风险自动识别方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116756684A (zh) * 2023-06-16 2023-09-15 国网重庆市电力公司营销服务中心 一种计量异常的分析方法及装置、存储介质、计算机设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109523116A (zh) * 2018-10-11 2019-03-26 平安科技（深圳）有限公司 企业风险分析方法、装置、计算机设备和存储介质
CN113556358A (zh) * 2021-07-30 2021-10-26 平安普惠企业管理有限公司 异常流量数据检测方法、装置、设备及存储介质
CN114066265A (zh) * 2021-11-18 2022-02-18 广东电网有限责任公司广州供电局 一种计量指标异常分析方法、装置、设备和介质
CN114416410A (zh) * 2021-12-31 2022-04-29 联想(北京)有限公司 一种异常分析方法、设备和计算机可读存储介质
CN116756684A (zh) * 2023-06-16 2023-09-15 国网重庆市电力公司营销服务中心 一种计量异常的分析方法及装置、存储介质、计算机设备

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106161138A (zh) * 2016-06-17 2016-11-23 贵州电网有限责任公司贵阳供电局 一种智能自动计量方法及装置
CN111475680A (zh) * 2020-03-27 2020-07-31 深圳壹账通智能科技有限公司 检测异常高密子图的方法、装置、设备及存储介质
CN112084220B (zh) * 2020-09-21 2024-05-14 国网重庆市电力公司营销服务中心 一种电能计量装置异常诊断方法、装置及可读存储介质
CN113283620B (zh) * 2021-06-29 2024-07-16 中国平安人寿保险股份有限公司 基于人工智能的运维方法、装置、设备及存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109523116A (zh) * 2018-10-11 2019-03-26 平安科技（深圳）有限公司 企业风险分析方法、装置、计算机设备和存储介质
CN113556358A (zh) * 2021-07-30 2021-10-26 平安普惠企业管理有限公司 异常流量数据检测方法、装置、设备及存储介质
CN114066265A (zh) * 2021-11-18 2022-02-18 广东电网有限责任公司广州供电局 一种计量指标异常分析方法、装置、设备和介质
CN114416410A (zh) * 2021-12-31 2022-04-29 联想(北京)有限公司 一种异常分析方法、设备和计算机可读存储介质
CN116756684A (zh) * 2023-06-16 2023-09-15 国网重庆市电力公司营销服务中心 一种计量异常的分析方法及装置、存储介质、计算机设备

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN119886934A (zh) * 2024-12-26 2025-04-25 深圳供电局有限公司 一种对于电能表表码录入异常的检测方法及系统
CN119379425A (zh) * 2024-12-31 2025-01-28 杭银消费金融股份有限公司 一种用于逾期指标变化的归因分析方法及系统
CN120067879A (zh) * 2025-04-29 2025-05-30 环德(福建)环保科技有限公司 污水处理设备的多源数据融合与远程诊断方法及系统
CN120064522A (zh) * 2025-04-29 2025-05-30 国网福建省电力有限公司电力科学研究院 一种实验室仪器数据自动采集与分析方法及系统
CN121502619A (zh) * 2026-01-13 2026-02-10 江阴长仪集团有限公司 基于电能表大数据的电能表风险自动识别方法及系统

Also Published As

Publication number Publication date
CN116756684A (zh) 2023-09-15

Similar Documents

Publication Publication Date Title
WO2024255020A1 (zh) 计量异常的分析方法及装置、存储介质、计算机设备
CN114298863B (zh) 一种智能抄表终端的数据采集方法及系统
CN110908883B (zh) 用户画像数据监控方法、系统、设备及存储介质
US20210319375A1 (en) Churn prediction in a broadband network
CN118761745B (zh) 应用于企业oa协同工作流程优化方法
US20150371163A1 (en) Churn prediction in a broadband network
CN113360722B (zh) 一种基于多维数据图谱的故障根因定位方法及系统
CN114514141A (zh) 充电站监测方法和设备
CN108063699B (zh) 网络性能监控方法、装置、电子设备、存储介质
CN106126391A (zh) 系统监控方法和装置
US10680902B2 (en) Virtual agents for facilitation of network based storage reporting
CN105184886A (zh) 一种云数据中心智能巡检系统及方法
CN107704387B (zh) 用于系统预警的方法、装置、电子设备及计算机可读介质
CN114443437B (zh) 告警根因输出方法、装置、设备、介质和程序产品
CN111984442A (zh) 计算机集群系统的异常检测方法及装置、存储介质
CN108390793A (zh) 一种分析系统稳定性的方法及装置
WO2024067358A1 (zh) 仓库管理系统的效率分析方法、系统及计算机设备
CN119441068B (zh) 一种实时数据测试方法、装置、计算机设备以及存储介质
CN119473445A (zh) Flink流处理任务的生成方法及电子设备、存储介质、计算机程序产品
CN113139729A (zh) 电网运行风险分析方法、系统、终端及存储介质
CN119250499A (zh) 电力系统告警信息处理方法、装置、电子设备及存储介质
US20240420157A1 (en) Metering abnormality analysis method and apparatus, storage medium, and computer device
CN117311295A (zh) 基于无线网络设备的生产质量提升方法及系统
CN110941255B (zh) 故障信息管理系统
CN116714469A (zh) 充电桩健康监测方法、装置、终端及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23941244

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE