Electric energy metering data processing and analyzing method and application system thereof
By generating adversarial network expansion data, combining ARIMA and LSTM models for power metering data processing, using an isolated forest algorithm to identify abnormal patterns, and establishing a multi-dimensional risk assessment model, solving the problem of data singularity and insufficient risk quantification in the processing and analysis of electricity metering data, and achieving stable operation and efficient management of the power system.
Patent Information
- Application Number
- CN202510507199.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-22
- Publication Date
- 2025-07-18
AI Technical Summary
In the existing power metering data processing and analysis methods, the data source is single and the format is not uniform, resulting in one-sided analysis results. Time series analysis methods cannot take into account the linear and nonlinear characteristics of the data. Anomaly identification is difficult to accurately locate, lack of systematic risk quantification, and it is difficult to ensure the stable operation of the power system and efficient energy utilization.
Generative adversarial network (GAN) is used to expand data diversity, combine ARIMA and LSTM models for trend prediction, use isolated forest algorithm to identify abnormal power usage patterns, and establish a multi-dimensional risk assessment model, and form closed-loop management through intelligent early warning and strategy generation modules.
A comprehensive and accurate analysis of the power usage status has been achieved, timely warnings and effective response strategies have been provided, to ensure the stable operation of the power system, reduce the risk of economic losses and equipment damage, and improve energy utilization efficiency and management level.
Smart Images
Figure CN120338424A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of power metering data processing and analysis, and particularly relates to a power metering data processing and analysis method and its application system. Background Art
[0002] Power metering data is the record and measurement data of the relevant information of the electric energy consumed by power users within a certain period of time through power metering devices (such as smart meters, power sensors, etc.), covering parameters such as active power, reactive power, power factor, current, voltage, power, as well as power consumption time, metering point number, basic user information, etc., which can accurately reflect the power consumption situation of users and the operating status of the power system, and provide key basis for the electricity bill settlement, load analysis, power grid planning, energy management of power enterprises, and energy-saving optimization of users.
[0003] In the existing field of power metering data processing and analysis, at the data level, traditional methods rely on a single data source and the data formats are not unified, resulting in one-sided analysis results and restricting the comprehensiveness and accuracy of model training. On the other hand, in the prediction link, a single time series analysis method, such as only using ARIMA or moving average method, cannot take into account both the linear and non-linear characteristics of the data at the same time, resulting in poor prediction accuracy of the electricity usage trend. Moreover, in anomaly identification, traditional methods based on threshold judgment or simple statistical models are difficult to accurately locate abnormal electricity usage patterns in a complex electricity usage environment. In addition, most systems lack a systematic risk quantification system and cannot comprehensively evaluate potential losses considering multi-dimensional factors such as equipment value, abnormal duration, and impact on power grid stability, making it difficult to provide a comprehensive analysis of electricity usage, and thus difficult to timely and effectively respond to the operating risks of the power system and ensure the stable operation of the system and efficient utilization of energy. Therefore, it is necessary to improve this situation, and for this reason, a power metering data processing and analysis method and its application system are proposed. Summary of the Invention
[0004] The purpose of the present invention is to provide a power metering data processing and analysis method and its application system to solve the problems raised in the above background art.
[0005] To achieve the above purpose, the present invention adopts the following technical solutions: A power metering data processing and analysis system, comprising:
[0006] A client, which is used to load power metering data in different regions and different periods, display the data in a variety of visualization methods, and provide a user interaction interface to support users to customize query conditions;
[0007] Data access module, which is used to read the power metering data sources specified by the client, including the monitoring systems of smart meters and distributed generation equipment, parse and convert data in different formats and protocols, and unify the data format;
[0008] Data enhancement module, which expands the power data according to the output of the data access module, such as generating simulated data through similarity analysis of historical data, to increase the diversity and integrity of the data;
[0009] Trend analysis module, which is used to perform time series analysis on the data processed by the data enhancement module, and predict the long-term and short-term trends of power consumption, such as seasonal changes in electricity consumption, daily peak and trough trends;
[0010] Abnormal diagnosis module, which compares the real-time power data with the prediction results of the trend analysis module to identify abnormal power consumption patterns, such as sudden surges or drops in electricity consumption, abnormal power factor;
[0011] Risk quantification module, which evaluates the risks of the identified abnormal situations according to the results of the abnormal diagnosis module, and quantifies the possible economic losses and equipment damage risks;
[0012] Intelligent warning module, which, based on the evaluation results of the risk quantification module, when the risk reaches the preset threshold, timely pushes warning information to the client, and the warning methods include pop-up windows, text messages, and emails;
[0013] Strategy generation module, which generates corresponding countermeasures for different types of abnormal situations and risk levels, such as adjusting the power consumption plan, checking the equipment status;
[0014] Effect evaluation module, which is used to evaluate the effects after the implementation of the countermeasures proposed by the strategy generation module, and analyze whether the risks and abnormal situations are effectively reduced;
[0015] Result display module, which is used to display the results of trend analysis, abnormal diagnosis, risk quantification, strategy generation, and effect evaluation on the client interface, so as to facilitate users to comprehensively understand the power consumption situation.
[0016] Preferably, the data access module adopts data middleware technology to achieve efficient access to multiple data sources and data format conversion, and supports fast transmission of real-time data.
[0017] Preferably, the data enhancement module uses a generative adversarial network (GAN) to generate simulated power data, and improves the richness and generalization ability of the data by mixing with real data.
[0018] Preferably, the trend analysis module uses a method combining ARIMA (Autoregressive Integrated Moving Average Model) and LSTM (Long Short-Term Memory Network) for time series prediction to improve the accuracy of prediction.
[0019] Preferably, the anomaly diagnosis module uses the Isolation Forest algorithm to identify abnormal electricity consumption patterns and determines whether a data point is an outlier by calculating its degree of isolation.
[0020] Preferably, the risk quantification module establishes a risk assessment model, comprehensively considering factors such as the value of electrical equipment, the duration of anomalies, and the impact on grid stability, to quantify the risks brought by abnormal situations.
[0021] Preferably, the intelligent warning module classifies and pushes different types of warning information according to the priorities and preferences set by the user to ensure that the user can obtain key information in a timely manner.
[0022] Preferably, the strategy generation module uses a combination of an expert system and machine learning algorithms to generate optimal response strategies based on historical data and real-time situations.
[0023] A method for processing and analyzing electricity metering data, applied to the electricity metering data processing and analysis system according to claims 1-8, includes the following steps:
[0024] S1. Load electricity metering data from different regions and different time periods through the client, and use the data access module to parse and convert data in different formats and protocols to unify the data format.
[0025] S2. Use the data augmentation module to expand the electricity data to generate simulated data and increase the diversity and integrity of the data.
[0026] S3. Conduct time series analysis on the processed data, and use a method combining ARIMA and LSTM to predict the long-term and short-term trends of electricity consumption.
[0027] S4. Compare the real-time electricity data with the prediction results, and use the Isolation Forest algorithm to identify abnormal electricity consumption patterns.
[0028] S5. According to the identified abnormal situations, establish a risk assessment model, and comprehensively consider various factors to quantify possible economic losses, equipment damage risks, etc.
[0029] S6. When the risk reaches the preset threshold, push warning information to the client through various methods according to the priorities and preferences set by the user.
[0030] S7. For different types of abnormal situations and risk levels, use a combination of expert systems and machine learning algorithms to generate corresponding coping strategies;
[0031] S8. Evaluate the effect after the implementation of the coping strategy and analyze whether it effectively reduces the occurrence of risks and abnormal situations;
[0032] S9. Display the results of trend analysis, abnormal diagnosis, risk quantification, strategy generation, and effect evaluation on the client interface to complete the dynamic analysis and intelligent early warning of electricity metering data.
[0033] In summary, compared with the prior art, the beneficial effects of the present invention are as follows: The electricity metering data processing and analysis method and its application system of the present invention utilize a generative adversarial network (GAN) to expand data diversity, combine the ARIMA and LSTM models to improve the accuracy of trend prediction, accurately identify abnormal electricity consumption patterns through the isolation forest algorithm, establish a multi-dimensional risk assessment model to quantify potential losses, and finally form a closed-loop management relying on the intelligent early warning and strategy generation module, providing users with a comprehensive and accurate analysis of electricity usage conditions, timely warning and providing effective coping strategies, ensuring the stable operation of the power system, reducing the risk of economic losses and equipment damage, and improving energy utilization efficiency and management level. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 is the process structure diagram of the present invention;
[0035] Figure 2 is the system structure diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0036] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0037] Please refer to Figure 1-2 , the present invention provides a technical solution:
[0038] An electricity metering data processing and analysis system, comprising:
[0039] A client, which is used to load electricity metering data in different regions and different periods, display the data in a variety of visual ways, and provide a user interaction interface to support users to customize query conditions;
[0040] Data access module, which is used to read the power metering data sources specified by the client, including the monitoring systems of smart meters and distributed generation devices, parse and convert data in different formats and protocols, and unify the data format;
[0041] Data enhancement module, which expands the power data according to the output of the data access module, such as generating simulated data through similarity analysis of historical data, to increase the diversity and integrity of the data;
[0042] Trend analysis module, which is used to perform time series analysis on the data processed by the data enhancement module, and predict the long-term and short-term trends of power consumption, such as seasonal changes in electricity consumption and daily peak and trough trends;
[0043] Abnormal diagnosis module, which compares the real-time power data with the prediction results of the trend analysis module to identify abnormal power consumption patterns, such as sudden surges or drops in electricity consumption and abnormal power factor;
[0044] Risk quantification module, which assesses the risks of the identified abnormal situations according to the results of the abnormal diagnosis module, and quantifies the possible economic losses and equipment damage risks;
[0045] Intelligent warning module, which, based on the evaluation results of the risk quantification module, when the risk reaches the preset threshold, timely pushes warning information to the client, and the warning methods include pop-up windows, text messages, and emails;
[0046] Strategy generation module, which generates corresponding countermeasures for different types of abnormal situations and risk levels, such as adjusting the power consumption plan and checking the equipment status;
[0047] Effect evaluation module, which is used to evaluate the effects after the implementation of the countermeasures proposed by the strategy generation module, and analyze whether the risks and abnormal situations have been effectively reduced;
[0048] Result display module, which is used to display the results of trend analysis, abnormal diagnosis, risk quantification, strategy generation, and effect evaluation on the client interface, so as to facilitate users to comprehensively understand the power consumption situation.
[0049] The data access module adopts data middleware technology to achieve efficient access to multiple data sources and data format conversion, and supports fast transmission of real-time data.
[0050] The data enhancement module uses a generative adversarial network (GAN) to generate simulated power data, and improves the richness and generalization ability of the data by mixing with real data.
[0051] The trend analysis module uses a method combining ARIMA (Autoregressive Integrated Moving Average) and LSTM (Long Short-Term Memory Network) for time series prediction to improve the accuracy of prediction.
[0052] The anomaly diagnosis module uses the Isolation Forest algorithm to identify abnormal electricity consumption patterns and determines whether a data point is an outlier by calculating its degree of isolation.
[0053] The risk quantification module establishes a risk assessment model, comprehensively considering factors such as the value of electrical equipment, the duration of anomalies, and the impact on grid stability, to quantify the risks brought by abnormal situations.
[0054] The intelligent warning module classifies and pushes different types of warning information according to the priorities and preferences set by users to ensure that users can obtain key information in a timely manner.
[0055] The strategy generation module uses a combination of an expert system and machine learning algorithms to generate optimal response strategies based on historical data and real-time situations.
[0056] A method for processing and analyzing electrical energy measurement data, applied to the electrical energy measurement data processing and analysis system described in claims 1-8, includes the following steps:
[0057] S1, Load electrical energy measurement data from different regions and at different times through the client, and use the data access module to parse and convert data in different formats and protocols to unify the data format;
[0058] S2, Use the data augmentation module to expand the electrical energy data to generate simulated data, increasing the diversity and integrity of the data;
[0059] S3, Conduct time series analysis on the processed data, and use a method combining ARIMA and LSTM to predict the long-term and short-term trends of electricity consumption;
[0060] S4, Compare the real-time electrical energy data with the prediction results, and use the Isolation Forest algorithm to identify abnormal electricity consumption patterns;
[0061] S5, According to the identified abnormal situations, establish a risk assessment model, comprehensively considering various factors to quantify possible economic losses, equipment damage risks, etc.;
[0062] S6, When the risk reaches the preset threshold, push warning information to the client through various methods according to the priorities and preferences set by users;
[0063] S7, For different types of abnormal situations and risk levels, use a combination of an expert system and machine learning algorithms to generate corresponding response strategies;
[0064] S8. Evaluate the effects after the implementation of the response strategies, and analyze whether the risks and abnormal situations have been effectively reduced;
[0065] S9. Display the results of trend analysis, abnormal diagnosis, risk quantification, strategy generation, and effect evaluation on the client interface to complete the dynamic analysis and intelligent early warning of power metering data.
[0066] Working principle:
[0067] Data collection and preparation
[0068] Data loading
[0069] After the user starts the client of the power metering data dynamic analysis and intelligent early warning system, they enter a simple and intuitive operation interface. On this interface, the user can select and load power metering data from different regions and different time periods through various interaction methods. These data sources are extensive, covering various regions such as urban commercial areas, residential areas, and industrial development zones, as well as different time periods such as daily, weekly, and monthly. In addition, the client also provides a rich custom query function. The user can not only accurately enter the start and end times through the time range box, but also use the area filtering drop-down menu to select a specific power consumption area to obtain data that meets their own needs.
[0070] Data access and conversion
[0071] The data access module, as the connection hub between the system and external data sources, monitors the user's operation instructions on the client in real time. Once it receives a data loading instruction, this module quickly establishes connections with data sources such as smart meters and the monitoring systems of distributed generation equipment. Considering the huge differences in data formats and protocols among different devices, the data access module adopts data middleware technology. Through built-in multiple parsers, it parses and converts data in different formats and protocols. Taking the Modbus protocol data commonly used by smart meters as an example, the data access module can accurately parse it and convert it into the unified standard format of the system, including unifying the time data into a standardized timestamp and performing unit conversion on data such as power, voltage, and current to ensure that the data can be smoothly processed by the subsequent modules of the system. At the same time, the data middleware technology can optimize the data transmission path to ensure that real-time data is quickly and stably transmitted into the system and stored in the central database.
[0072] Data enhancement
[0073] The data enhancement module extracts the data processed by the data access module from the central database and uses the generative adversarial network (GAN) to expand the data. GAN consists of a generator and a discriminator. The generator simulates and generates new electric energy data by learning from historical data. These data cover electricity consumption in different scenarios, such as peak electricity consumption during high temperature periods in summer and fluctuations in electricity consumption during holidays. The discriminator identifies the data generated by the generator and continuously encourages the generator to improve the quality of the generated data. After multiple iterative training, the data enhancement module mixes the generated simulated data with the real data, effectively solving the problems of insufficient data volume or uneven data distribution, providing a richer and more comprehensive data set for subsequent analysis models, and significantly improving the generalization ability of the model in complex scenarios.
[0074] Data analysis and prediction
[0075] Trend Analysis
[0076] The trend analysis module obtains the data set processed by the data enhancement module from the central database, and uses a combination of ARIMA (autoregressive integrated moving average model) and LSTM (long short-term memory network) to perform time series analysis. When performing analysis, the ARIMA model can effectively capture the linear characteristics and seasonal changes in the data. For example, by analyzing historical data, the cyclical change trend of monthly electricity consumption can be determined, while the LSTM model has excellent capture capabilities for long-term dependencies and nonlinear characteristics in the data due to its unique gating mechanism, and can accurately identify abnormal fluctuations in electricity consumption caused by sudden equipment failures. By combining the advantages of these two models, the trend analysis module can not only predict the long-term trend of electricity use, such as the annual growth or decline in electricity consumption, but also accurately predict the daily peak and valley periods of electricity consumption, providing strong support for the dispatching and management of the power system.
[0077] Anomaly Detection and Risk Assessment
[0078] Abnormal diagnosis
[0079] The anomaly diagnosis module collects electric energy data in real time and compares it with the prediction results obtained by the trend analysis module. During the comparison process, the anomaly diagnosis module uses the isolation forest algorithm to identify abnormal power consumption patterns. The isolation forest algorithm randomly divides the data space by constructing multiple random binary trees. When the path length of a data point in the tree deviates significantly from other data points, the algorithm considers that this data point has a high degree of isolation and determines it as an outlier. For example, when the real-time collected power consumption surges or decreases sharply in a short period of time and exceeds the prediction range, or the power factor fluctuates abnormally, the anomaly diagnosis module can quickly identify these abnormal power consumption patterns and transmit the relevant information to the risk quantification module.
[0080] Risk quantification
[0081] Based on the detection results of the anomaly diagnosis module, the risk quantification module establishes a comprehensive risk assessment model. This model comprehensively considers multiple factors and quantitatively evaluates the potential economic losses and equipment damage risks brought by abnormal situations. During the evaluation process, the risk quantification module first determines the potential loss value of the equipment according to the type, specification, and market value of the electrical equipment. Secondly, by analyzing the duration of the anomaly, it evaluates the degree of its impact on the power grid stability. For example, long-term overload operation may cause the transformer to overheat and be damaged. In addition, the model also considers the chain reaction of abnormal situations on surrounding equipment and the power grid. For example, voltage fluctuations may affect other electrical equipment in adjacent areas. Through the comprehensive analysis of these factors, the risk quantification module can accurately calculate the risk level of abnormal situations and the possible economic losses.
[0082] Early warning and response
[0083] Intelligent early warning
[0084] The intelligent early warning module receives the evaluation results of the risk quantification module in real time. When the risk reaches the preset threshold, the intelligent early warning module classifies the early warning information according to the priorities and preferences set by the user on the client side. For abnormal situations in important equipment or key areas, the system will give priority to pushing early warning information. At the same time, the intelligent early warning module supports multiple early warning push methods, including pop-up reminders on the client side, SMS notifications, email sending, etc., to ensure that users can obtain key information in a timely manner so as to take corresponding measures.
[0085] Strategy generation
[0086] The strategy generation module generates response strategies for different types of abnormal situations and risk levels by combining an expert system and machine learning algorithms. A large amount of professional knowledge and historical experience in the power field are stored in the expert system, which can quickly generate preliminary response strategies according to the type of anomaly and risk level. For example, when it is detected that the power consumption in a certain area is too high, the expert system may suggest adjusting the power consumption plan in that area and optimizing the equipment operation time. The machine learning algorithm optimizes and adjusts the strategies generated by the expert system through learning a large amount of historical data and real-time situations. For example, through the analysis of historical data, the machine learning algorithm can predict the implementation effects of different response strategies, so as to select the optimal response plan.
[0087] Effect evaluation
[0088] The effect evaluation module is responsible for evaluating the effect after the implementation of the coping strategies proposed by the strategy generation module. This module judges the effectiveness of the strategies by collecting power data after the implementation of the strategies and analyzing indicators such as whether the risk level has decreased and whether the occurrence frequency of abnormal situations has decreased. At the same time, the effect evaluation module will also collect the feedback from users on the implementation effect of the strategies, such as whether there are any operational difficulties during the implementation of the strategies. According to the evaluation results and user feedback, the effect evaluation module provides a basis for subsequent strategy optimization, continuously improving the system's ability to cope with abnormal situations.
[0089] Result display
[0090] The result display module presents the results of trend analysis, anomaly diagnosis, risk quantification, strategy generation, and effect evaluation in an intuitive visual manner on the client interface. In terms of interface design, the system uses a variety of visual elements, such as line charts, bar charts, pie charts, etc., to display different types of data. For example, the long-term and short-term trends of electricity consumption are displayed through line charts, enabling users to clearly understand the changes in electricity consumption. The risk levels in different regions are compared through bar charts, facilitating users to quickly locate high-risk regions. The proportion of different types of abnormal situations is shown through pie charts, helping users comprehensively grasp the distribution of abnormal situations. In addition, users can view detailed data information and strategy implementation situations through interactive operations, thereby making reasonable decisions and achieving effective management of the power system.
[0091] S1: Data loading and format conversion. The user operates on the client to select electricity metering data for different regions and time periods. These data sources include smart meters, distributed generation equipment monitoring systems, etc. The client can display the data in visual charts such as line charts, bar charts, and heat maps, facilitating users to intuitively grasp the general situation of the data. At the same time, the user can customize query conditions according to their own needs, such as specifying a time range and filtering specific regions;
[0092] The data access module uses data middleware technology to read the data from the specified data sources on the client in real time. Due to differences in data protocols (such as Modbus, IEC 61850) and formats (such as CSV, JSON) of different devices, the data access module uses parsers to convert this heterogeneous data into a unified format, perform standardized timestamp processing and unit conversion, and then store the processed data in the central database;
[0093] S2: Data augmentation and simulation generation. The data enhancement module obtains data from the central database and conducts similarity analysis on historical data based on the generative adversarial network (GAN). Through this technology, simulated electricity consumption data is generated, such as data for scenarios like simulated holiday electricity peaks and electricity fluctuations under extreme weather;
[0094] Mix the generated simulation data with the real data to construct an extended dataset, increasing the diversity and richness of the data and effectively improving the generalization ability of the subsequent model in complex scenarios;
[0095] S3: The multi-model fusion prediction power consumption trend analysis module obtains the enhanced dataset and performs time series analysis on the data using a hybrid model that combines ARIMA (Autoregressive Integrated Moving Average Model) and LSTM (Long Short-Term Memory Network);
[0096] The ARIMA model is responsible for capturing the linear trends in the data, such as identifying seasonal cyclic changes in power consumption; the LSTM model focuses on learning the non-linear features of the data, such as sudden power consumption fluctuations;
[0097] Through the collaborative work of these two models, the long-term trends of power consumption, such as annual power consumption growth, and short-term trends, such as daily peak and valley power consumption periods, are output;
[0098] S4: Abnormal power consumption pattern recognition. The abnormal diagnosis module compares the real-time collected power data with the prediction results obtained by the trend analysis module;
[0099] Using the Isolation Forest algorithm, by randomly dividing the data space and calculating the isolation degree of data points in the space, that is, the path length, when the isolation degree of the data point exceeds the normal range, it is identified as an abnormal point, such as sudden sharp increases or decreases in power consumption, abnormal power factor, etc.;
[0100] S5: Multi-dimensional risk quantification. The risk quantification module constructs a risk assessment model based on the results of the abnormal diagnosis module, and comprehensively considers various factors during the assessment process;
[0101] According to the equipment type (transformer, generator, etc.) and service life, calculate the potential loss of the equipment, quantify the impact of the abnormality on the power grid stability through the duration of the abnormality, such as long-term overload may cause equipment damage, and evaluate the chain reaction of the abnormality on adjacent areas, such as the impact of voltage fluctuations on surrounding equipment;
[0102] Finally, the risk level is output, divided into three levels: low, medium, and high, and the possible economic losses are estimated;
[0103] S6: Personalized warning push. The intelligent warning module receives the evaluation results of the risk quantification module. When the risk reaches the preset threshold, it classifies according to the priorities set by the user on the client side, such as equipment importance, regional criticality, and push preferences, such as push time, push method, etc.;
[0104] Through methods such as client pop-up windows, text messages, and emails, send warning messages to users. The warning content includes the type of abnormality, risk level, and targeted recommended measures;
[0105] S7: The multi - method fusion generates countermeasures. The countermeasure generation module combines an expert system and machine learning algorithms to generate countermeasures;
[0106] Based on the historical case library, the expert system generates preliminary countermeasures, such as adjusting the power consumption plan and arranging equipment inspections;
[0107] The machine learning algorithm dynamically optimizes the strategy according to real - time data, predicts future risk trends, and adjusts the strategy in advance;
[0108] The generated strategies include adjusting load distribution, starting standby power supplies, arranging emergency repairs, etc.;
[0109] S8: Evaluation of the implementation effect of the strategy. The effect evaluation module conducts a quantitative evaluation of the effect after the implementation of the strategy.
[0110] Compare indicators such as the risk index, abnormal occurrence rate, and economic losses before and after the implementation of the strategy. At the same time, collect users' evaluations of the effectiveness of the strategy, such as whether the problem is solved in a timely manner and whether the risk is reduced;
[0111] Generate an evaluation report based on the evaluation and feedback information, providing a basis for subsequent optimization of the strategy model;
[0112] S9: Display results and closed - loop optimization. The result display module centrally displays various information on the client interface, including long - term and short - term power consumption trend charts, time axes and type statistics of abnormal events, risk level distribution maps and economic loss estimates, lists of countermeasures and their execution status, and comparison data before and after the implementation of the strategy.
[0113] Based on the displayed results, users adjust the strategy, forming a closed - loop management mode of "analysis - warning - strategy - evaluation", continuously optimizing the system performance, enhancing the stability of the power system operation, and reducing economic losses and equipment risks.
[0114] As mentioned above, it is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and all should be covered within the protection scope of the present invention.
Claims
1. An electric energy metering data processing and analysis system, characterized in that Including: A client, which is used to load the power metering data of different regions and different time periods, display the data in various visualization methods, and provide a user interaction interface to support users to customize query conditions; A data access module, which is used to read the power metering data sources specified by the client, including the monitoring systems of smart meters and distributed generation equipment, parse and convert data in different formats and protocols, and unify the data format; A data enhancement module, which expands the power data according to the output of the data access module, such as generating simulation data through the similarity analysis of historical data to increase the diversity and integrity of the data; A trend analysis module, which is used to perform time series analysis on the data processed by the data enhancement module to predict the long-term and short-term trends of power consumption, such as seasonal changes in electricity consumption and daily peak and trough trends; An anomaly diagnosis module, which compares the real-time power data with the prediction results of the trend analysis module to identify abnormal power consumption patterns, such as sudden surges or drops in electricity consumption and abnormal power factor; A risk quantification module, which conducts a risk assessment on the identified abnormal situations according to the results of the anomaly diagnosis module and quantifies the possible economic losses and equipment damage risks; An intelligent warning module, which, based on the evaluation results of the risk quantification module, when the risk reaches the preset threshold, timely pushes warning information to the client, and the warning methods include pop-up windows, text messages, and emails; A strategy generation module, which generates corresponding coping strategies for different types of abnormal situations and risk levels, such as adjusting the power consumption plan and checking the equipment status; An effect evaluation module, which is used to evaluate the effects after the implementation of the coping strategies proposed by the strategy generation module and analyze whether the risks and abnormal situations have been effectively reduced; A result display module, which is used to display the results of trend analysis, anomaly diagnosis, risk quantification, strategy generation, and effect evaluation on the client interface to facilitate users to comprehensively understand the power consumption situation.
2. The power consumption metering data processing and analysis system according to claim 1, wherein The data access module adopts data middleware technology to achieve efficient access to multiple data sources and data format conversion, and supports the rapid transmission of real-time data.
3. As described in claim 2, a power metering data processing and analysis system, wherein the data enhancement module uses a generative adversarial network (GAN) to generate simulated power data, and improves the richness and generalization ability of the data by mixing with real data.
4. The power consumption metering data processing and analysis system according to claim 1, wherein The trend analysis module uses a method combining ARIMA (Autoregressive Integrated Moving Average Model) and LSTM (Long Short-Term Memory Network) for time series prediction to improve the prediction accuracy.
5. The power consumption data processing and analysis system according to claim 1, wherein The anomaly diagnosis module uses the Isolation Forest algorithm to identify abnormal power consumption patterns and determines whether a data point is an outlier by calculating the degree of isolation of the data point.
6. The power consumption data processing and analysis system according to claim 1, wherein The risk quantification module establishes a risk assessment model, comprehensively considers factors such as the value of power consumption equipment, the duration of anomalies, and the impact on grid stability, and quantifies the risks brought by abnormal situations.
7. The power consumption data processing and analysis system according to claim 1, wherein The intelligent early warning module classifies and pushes different types of early warning information according to the priorities and preferences set by the user, ensuring that the user can obtain key information in a timely manner.
8. The power consumption data processing and analysis system according to claim 1, wherein The policy generation module uses a combination of expert systems and machine learning algorithms to generate optimal response policies based on historical data and real-time situations.
9. A method for processing and analyzing electric energy metering data, characterized in that Applied to the electric energy metering data processing and analysis system described in claims 1-8, it includes the following steps: S1. Load the electric energy metering data of different regions and different time periods through the client, and use the data access module to parse and convert data of different formats and protocols to unify the data format. S2. Use the data enhancement module to expand the electric energy data to generate simulated data, increasing the diversity and integrity of the data. S3. Conduct time series analysis on the processed data, and use a combination of ARIMA and LSTM methods to predict the long-term and short-term trends of electric energy use. S4. Compare the real-time electric energy data with the prediction results, and use the isolation forest algorithm to identify abnormal electricity consumption patterns. S5. According to the identified abnormal situations, establish a risk assessment model, comprehensively consider various factors to quantify the possible economic losses, equipment damage risks, etc. S6. When the risk reaches the preset threshold, push early warning information to the client through various methods according to the priorities and preferences set by the user. S7. For different types of abnormal situations and risk levels, use a combination of expert systems and machine learning algorithms to generate corresponding response policies. S8. Evaluate the effect after the implementation of the response policy, and analyze whether it effectively reduces the occurrence of risks and abnormal situations. S9. Display the results of trend analysis, abnormal diagnosis, risk quantification, policy generation, and effect evaluation on the client interface to complete the dynamic analysis and intelligent early warning of electric energy metering data.