Electric meter box, electrical load prediction method and computer readable storage medium

Abstract

The invention discloses an electric meter box, an electrical load prediction method and a computer readable storage medium, and relates to the technical field of electric meter boxes, and the electric meter box comprises a box body which is internally provided with a plurality of meter positions for installing electric energy meters; a plurality of acquisition modules which are coupled to the incoming line loops of the meter positions in a one-to-one correspondence manner and are used for acquiring the real-time power of the meter positions; the main control module is arranged in the box body, is electrically connected with each acquisition module, and is used for receiving the real-time power acquired by each acquisition module; according to the first real-time power set of each meter position received in a preset time window, analyzing user electricity utilization rule information corresponding to each meter position; and according to the user electricity utilization rule information corresponding to each meter position and a second real-time power set of each meter position received in the current time window, predicting an electricity utilization load demand value of each meter position in a future time window after the current moment.

Description

Technical Field

[0001] This application relates to the field of meter box technology, and in particular to an meter box, a method for predicting electricity load, and a computer-readable storage medium. Background Technology

[0002] Electricity meter boxes are the infrastructure connecting the power grid to end users. Currently, the widely used multi-meter meter boxes (such as centralized residential meter boxes) mainly focus on electricity metering and overload protection, and are essentially passive electrical connection devices.

[0003] With the development of smart grids, higher demands are placed on the refined management and load forecasting of distribution networks. Currently, load forecasting relies heavily on data from the main distribution meter or historical big data analysis from the master station system. This fails to achieve refined perception of the real-time electricity consumption behavior of each individual meter position (i.e., a single user). Consequently, the power grid dispatch center struggles to accurately capture changes in the electricity consumption habits of micro-users, resulting in poor accuracy in load forecasting. This leads to a mismatch between power generation and demand, affecting the stability of power grid operation and causing a waste of electrical resources. Summary of the Invention

[0004] The main purpose of this application is to provide an electricity meter box, an electricity load prediction method, and a computer-readable storage medium, which aims to improve the accuracy of electricity load prediction, so as to ensure the stability of power grid operation and avoid the waste of electrical energy resources to a certain extent.

[0005] This application provides an electricity meter box, including: The enclosure contains multiple mounting positions for installing electricity meters; Multiple acquisition modules are coupled one-to-one with the incoming line circuit of each meter position to acquire the real-time power of each meter position. A main control module, located inside the enclosure and electrically connected to each of the acquisition modules, is used for: Receive the real-time power acquired by each of the acquisition modules; Based on the first real-time power set received for each meter position within a preset time window, analyze the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set received for each meter position within the current time window, predict the electricity load demand value of each meter position in the future time window after the current moment; The acquisition module includes: A current sampling unit is coupled to the incoming circuit of a meter position in the enclosure and is used to collect the load current on the incoming circuit of the meter position. The metering unit has its input terminal electrically connected to the current sampling unit and its output terminal electrically connected to the main control module. It is used to calculate the real-time power of the meter position based on the load current and transmit the calculated real-time power to the main control module.

[0006] In one embodiment, the meter box further includes: A communication module, which is electrically connected to the main control module, is used to exchange data with the power grid master station and / or distribution substation. The main control module is also used to send the predicted power load demand value of each meter position and the received real-time power of each meter position to the power grid master station and / or distribution substation through the communication module.

[0007] Furthermore, to achieve the above objectives, this application also provides an electricity load forecasting method, applied to the meter box described above, the method comprising: Obtain the real-time power of each meter position in the meter box; Based on the first real-time power set of each meter position obtained within the preset time window, analyze the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained within the current time window, predict the electricity load demand value of each meter position in the future time window after the current moment; The step of obtaining the real-time power of each meter position in the meter box includes: Obtain the load current on the incoming circuit of each of the aforementioned positions; The real-time power of each meter position is calculated based on the load current of each meter position.

[0008] In one embodiment, the step of analyzing the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set obtained within a preset time window includes: For any of the aforementioned positions, a power curve for the position within the preset time window is generated based on each real-time power in the first real-time power set of the position and the acquisition time corresponding to each real-time power. The power curve is analyzed to extract the user electricity consumption pattern information corresponding to the meter position; wherein, the user electricity consumption pattern information includes at least the daily peak load period, the weekly load periodic fluctuation characteristics, and the seasonal electricity consumption trend.

[0009] In one embodiment, before the step of analyzing the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set obtained within a preset time window, the method further includes: Data cleaning processing is performed on the first real-time power set of each of the aforementioned positions; Based on the first real-time power set of each meter position after data cleaning and processing, analyze the user electricity consumption pattern information corresponding to each meter position.

[0010] In one embodiment, the step of predicting the electricity load demand value of each meter position in a future time window after the current moment, based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained within the current time window, includes: The user electricity consumption pattern information and the second real-time power set corresponding to each meter position are input into the pre-trained load prediction model to obtain the electricity load demand value of each meter position in the future time window after the current time.

[0011] In one embodiment, before the step of predicting the electricity load demand value of each meter position in a future time window after the current moment based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained within the current time window, the method further includes: Determine the time type characteristics of the future time window; wherein, the time type characteristics include at least the date type and the daily load period type, and the date type is a weekday, rest day, or holiday; Based on the time type characteristics of the future time window, adjust the weight of each sub-information in the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position after weight adjustment and the second real-time power set, the electricity load demand value of each meter position in the future time window after the current moment is predicted.

[0012] In one embodiment, the method further includes: Monitor whether there are abnormal fluctuations in the real-time power of each of the aforementioned positions; If there is an abnormal fluctuation in the real-time power of any of the aforementioned meter positions, an abnormal fluctuation alarm message is generated and sent to the power grid master station and / or distribution substation.

[0013] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the electricity load prediction method as described above.

[0014] This application provides an electricity meter box, comprising: a box body, wherein the box body has multiple meter positions for installing electricity meters; multiple acquisition modules, each acquisition module being coupled one-to-one with the incoming line circuit of each meter position, for acquiring the real-time power of each meter position; and a main control module, which is located inside the box body and electrically connected to each acquisition module, for: receiving the real-time power acquired by each acquisition module; analyzing the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set received for each meter position within a preset time window; and predicting the electricity load demand value of each meter position in a future time window after the current moment based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set received for each meter position within the current time window.

[0015] Therefore, the technical solution provided in this application, by setting up a data acquisition module for each meter position within the meter box, and coupling each acquisition module to the incoming line circuit of each meter position, enables the individual acquisition of real-time power for each independent meter position within the meter box. This changes the traditional meter box's limitation of only providing electrical connections and being unable to collect precise electricity consumption data for a single meter position. Furthermore, the technical solution provided in this application also includes a main control module electrically connected to each acquisition module within the box. By receiving the real-time power from each acquisition module, this module can analyze the user's electricity consumption patterns for each meter position based on the first real-time power set within a preset time window. This allows the mining of electricity consumption patterns to no longer rely on the aggregated data from the main meter of the distribution area, but rather to be tailored to the actual electricity consumption behavior of each user. This makes the mined electricity consumption pattern information more closely aligned with the individual user's electricity consumption characteristics, possessing precise individual targeting. Based on this, the main control module combines the user's electricity consumption pattern information with the second real-time power set within the current time window to individually predict the electricity load demand value for each meter position in the future time window. This makes load prediction no longer a rough estimate based on overall data, but a refined prediction based on the actual electricity consumption pattern and real-time electricity status of each user. This effectively improves the accuracy of electricity load prediction for individual meter positions. In turn, the grid side can more accurately grasp the changes in electricity demand of end users based on these accurate single-user load prediction data, so as to rationally allocate power resources, achieve accurate matching between power generation and electricity demand, ultimately ensure the stability of grid operation, and to a certain extent avoid the waste of power resources caused by load prediction deviations.

[0016] In summary, the technical solution provided in this application can improve the accuracy of electricity load forecasting, thereby ensuring the stability of power grid operation and avoiding the waste of electrical energy resources to a certain extent. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the structure of the electricity meter provided in the first embodiment of this application; Figure 2 This is a schematic diagram of the acquisition module provided in the first embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electricity meter provided in the second embodiment of this application; Figure 4 This is a flowchart illustrating the electricity load prediction method provided in the third embodiment of this application.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings.

[0021] Explanation of icon numbers: 10. Housing; 20. Acquisition module; 30. Main control module; 40. Communication module; 21. Current sampling unit; 22. Metering unit. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] This application presents a meter box according to a first embodiment; please refer to [link / reference]. Figure 1 The meter box may include: Box 10, which has multiple meter positions for installing electricity meters; Multiple acquisition modules 20 are coupled one-to-one with the incoming line circuit of each meter position to obtain the real-time power of each meter position. The main control module 30 is located inside the housing 10 and is electrically connected to each acquisition module 20. It is used for: Receive the real-time power acquired by each acquisition module 20; Based on the first real-time power set received for each meter position within the preset time window, analyze the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter and the second real-time power set received for each meter within the current time window, the electricity load demand value of each meter in the future time window after the current moment is predicted.

[0025] It should be noted that a meter position refers to the physical installation location within the meter box for installing a single energy meter, with each meter position corresponding to one end user. The main control module 30 can be an MCU (Microcontroller Unit) or other control chips with control functions; this embodiment does not impose specific limitations on this. The incoming line circuit refers to the electrical circuit through which electrical energy is input to each meter position; it is the transmission path for the end user's power consumption. Real-time power refers to the active power value on the meter position's incoming line circuit, acquired in real-time by the acquisition module 20 at a certain sampling frequency. The preset time window is the historical data time range used to analyze user electricity consumption patterns. It can be a default value, such as the past 30 days, or it can be flexibly set by the user according to actual conditions; this embodiment does not impose specific limitations on this. The first real-time power set is the set of all real-time power data for a meter position acquired within the preset time window. User electricity consumption pattern information is used to characterize the electricity consumption habits of the end user corresponding to the meter position, which may include, but is not limited to, daily peak load periods, weekly load periodic fluctuation characteristics, and seasonal electricity consumption trends; this embodiment does not impose specific limitations on this. The current time window refers to the time period during which the main control module 30 traces back and collects the latest real-time power data when performing load forecasting operations. It can have a default value, such as the past 4 hours, or it can be flexibly set by the user according to actual conditions; this embodiment does not impose specific limitations on this. The second real-time power set is the collection of all real-time power data from a single meter position collected within the current time window. The future time window refers to the future time period to be predicted, such as the next hour, the next 24 hours, or the next week. The electricity load demand value is the predicted expected electricity consumption of the user within the specified future time window.

[0026] In one feasible implementation, please refer to Figure 2 The acquisition module 20 may include: The current sampling unit 21 is coupled to the incoming circuit of a meter position in the housing 10 and is used to collect the load current on the incoming circuit of the meter position. Metering unit 22, the input terminal of metering unit 22 is electrically connected to current sampling unit 21, and the output terminal of metering unit 22 is electrically connected to main control module 30. It is used to calculate the real-time power of meter position according to load current and transmit the calculated real-time power to main control module 30.

[0027] It should be noted that the current sampling unit 21 can be a current sensor, a shunt, or other device with current acquisition function, etc., and this embodiment does not specifically limit it. The load current is the actual current value flowing through the meter's incoming line circuit, which reflects the current power consumption of the relevant user. The metering unit 22 may include a metering chip and its necessary peripheral circuitry, used to receive the signal output by the current sampling unit 21, and obtain the real-time power value through analog-to-digital conversion and power calculation.

[0028] This embodiment sets the acquisition module 20 to include a current sampling unit 21 and a metering unit 22, so that the acquisition module 20 and the meter position incoming line circuit form an integrated structure, realizing the deep integration of the meter box body and the monitoring function, eliminating the need for external monitoring devices and simplifying the installation.

[0029] As can be seen from the above, the technical solution provided in this embodiment achieves the individual acquisition of real-time power for each independent meter position within the meter box by setting up a data acquisition module 20 for each meter position in the meter box, with each module 20 coupled to the incoming line circuit of each meter position. This changes the traditional meter box, which only provides electrical connections and cannot collect precise electricity consumption data for a single meter position. Furthermore, the technical solution provided in this embodiment also includes a main control module 30 electrically connected to each data acquisition module 20 within the box body 10. By receiving the real-time power from each data acquisition module 20, the main control module 30 can analyze the user's electricity consumption pattern information corresponding to each meter position based on the first real-time power set of each meter position within a preset time window. This makes the mining of electricity consumption patterns no longer dependent on the aggregated data of the main meter of the transformer area, but rather tailored to the actual electricity consumption behavior of each user. This makes the mined electricity consumption pattern information more closely match the electricity consumption characteristics of individual users, possessing precise individual targeting. Based on this, the main control module 30 combines the user's electricity consumption pattern information with the second real-time power set within the current time window to individually predict the electricity load demand value for each meter position in the future time window. This makes the load prediction no longer a rough estimate based on overall data, but a refined prediction based on the actual electricity consumption pattern and real-time electricity status of each user. This effectively improves the accuracy of the electricity load prediction for a single meter position. In turn, the power grid can more accurately grasp the changes in the electricity demand of end users based on these accurate single-user load prediction data, so as to rationally allocate power resources, achieve accurate matching between power generation and electricity demand, ultimately ensure the stability of power grid operation, and to a certain extent avoid the waste of power resources caused by load prediction deviations.

[0030] Therefore, the technical solution provided in this embodiment can improve the accuracy of electricity load prediction, so as to ensure the stability of power grid operation and avoid the waste of electrical energy resources to a certain extent.

[0031] Based on the first embodiment described above, a second embodiment of the meter box of this application is proposed. For the second embodiment, please refer to... Figure 3 The meter box may also include: Communication module 40 is electrically connected to main control module 30 and is used for data exchange with power grid master station and / or distribution substation; The main control module 30 is also used to send the predicted power load demand values ​​of each meter position and the received real-time power of each meter position to the power grid master station and / or distribution substation via the communication module 40.

[0032] It should be noted that the communication module 40 can use wired communication methods (such as RS485, Ethernet, power line carrier) or wireless communication methods (such as 4G / 5G, Wi-Fi) to achieve communication; this embodiment does not impose specific limitations on this. The power grid master station refers to the master station layer of the distribution automation system, usually located in the dispatch center or operation and maintenance center, responsible for large-scale power grid monitoring, dispatching, and management. A distribution transformer area refers to a distribution transformer and its power supply range, usually equipped with an area edge gateway or area terminal, responsible for data aggregation and edge computing from multiple meter boxes within that area.

[0033] In this embodiment, a communication module 40 is installed in the meter box to achieve data exchange with the upper-level system, transforming the meter box from a simple passive metering device into an active sensing node with data reporting capabilities. Furthermore, the main control module 30 can transmit the prediction results and real-time data to the power grid master station and / or distribution substation via the communication module 40, enabling the upper-level dispatching system to obtain refined prediction data at the micro-user level. This provides accurate basic data support for short-term and ultra-short-term load forecasting, thereby improving the stability of power grid operation, reducing energy waste, and achieving refined management of the distribution network.

[0034] The third embodiment of this application provides a method for predicting electricity load. Please refer to [link / reference]. Figure 4 The electricity load forecasting method may include steps S10 to S30: Step S10: Obtain the real-time power of each meter position in the meter box; Step S20: Analyze the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set of each meter position obtained within the preset time window; In one feasible implementation, step S20 may include steps S21-S22: Step S21: For any meter position, based on each real-time power in the first real-time power set of the meter position and the acquisition time corresponding to each real-time power, generate the power curve of the meter position within a preset time window. Step S22: Analyze the power curve to extract the user electricity consumption pattern information corresponding to the meter position; wherein, the user electricity consumption pattern information includes at least the daily load peak period, the weekly load periodic fluctuation characteristics, and the seasonal electricity consumption trend.

[0035] It should be noted that the power curve is a continuous curve plotted with time on the horizontal axis and power value on the vertical axis. It is used to intuitively reflect the change of power consumption of a user within a preset time window. When generating the power curve of a meter within the preset time window based on the real-time power of each meter position in the first real-time power set and the acquisition time of each real-time power, only one power curve can be generated, or multiple power curves can be generated (such as generating daily power curves, weekly power curves, monthly power curves, etc. within the preset time window at the same time). This embodiment does not make specific limitations on this. The daily load peak period is the continuous period of time when the user's daily electricity consumption is the highest or the highest power value; the weekly load periodic fluctuation characteristics are the regular changes in the user's electricity consumption behavior on the weekly dimension, such as the lower electricity consumption on weekdays than on weekends, and the peak electricity consumption on Friday nights; the seasonal electricity consumption trend is the trend of the user's electricity consumption behavior with the seasons, such as the significant increase in electricity consumption due to air conditioning load in summer and the significant increase in electricity consumption due to heating load in winter.

[0036] When analyzing power curves to extract user electricity consumption patterns corresponding to meter locations, in one feasible implementation, peak and valley detection can be performed on the power curves to identify the daily highest and lowest power points and their corresponding times, and the concentrated distribution intervals of peak occurrence times over multiple days can be statistically analyzed to determine the daily peak load period. Fourier transform can be performed on the power curves to extract the amplitude and phase of the weekly time and seasonal components to analyze the periodic fluctuation characteristics of weekly load and seasonal electricity consumption trends. In another feasible implementation, multiple typical user electricity consumption templates can be pre-constructed, such as those with fixed work schedules, fluctuating work schedules, and seasonal electricity consumption patterns. The generated power curves can then be compared with each typical template using similarity calculations (e.g., using dynamic time warping algorithms or cosine similarity), and the user electricity consumption template with the highest similarity can be selected as the target template. The electricity consumption pattern information in this target template can then be used as the user electricity consumption pattern information corresponding to the meter location. This embodiment does not specifically limit the implementation of step S22.

[0037] In this embodiment, discrete real-time power data is integrated into a continuous power curve, achieving a structured organization of electricity consumption data from points to lines. Then, by analyzing and extracting features from the power curve, the abstract curve shape is transformed into specific user electricity consumption pattern information, including daily peak load periods, weekly load periodic fluctuations, and seasonal electricity consumption trends. This layer-by-layer abstraction process, from data to curves to patterns, improves the accuracy and interpretability of electricity consumption behavior identification and compresses complex time-series data into concise feature vectors, reducing the computational complexity of subsequent prediction models. Therefore, this embodiment achieves refined perception and quantitative description of each user's electricity consumption habits, providing a reliable basis for subsequent load forecasting and effectively solving the problem of existing technologies failing to capture micro-level changes in user electricity consumption habits.

[0038] This embodiment does not specifically limit the implementation of step S20. For example, in other feasible implementations, the real-time power in the first real-time power set can be organized according to time series to form a daily load curve for each meter position; the time series features of each daily load curve can be extracted using the sliding window method, including peak occurrence time, rising slope, falling slope, peak-valley difference, etc.; clustering algorithm can be used to perform cluster analysis on the extracted features, grouping users with similar electricity consumption patterns into one category, and generating a typical electricity consumption template for each category of users; for each meter position, the electricity consumption pattern information in the typical template of its category is used as the user electricity consumption pattern information corresponding to that meter position.

[0039] Furthermore, in one feasible implementation, the electricity load prediction method may further include steps S201-S202 prior to step S20: Step S201: Perform data cleaning processing on the first real-time power set of each table position; Step S202: Based on the first real-time power set of each meter position after data cleaning and processing, analyze the user electricity consumption pattern information corresponding to each meter position.

[0040] It should be noted that data cleaning is the process of preprocessing raw real-time power data to identify and process outliers, missing values, noise points, etc., in order to improve data quality and the accuracy of subsequent analysis.

[0041] When performing data cleaning on the first real-time power set for each table position, a reasonable power range can be set. By traversing each data point in the first real-time power set, data points exceeding the power range are marked as outliers and removed. For data gaps resulting from removal, interpolation algorithms (such as linear interpolation or spline interpolation) are used to fill them in. Alternatively, the mean and standard deviation of the first real-time power set can be calculated, and data points deviating from the mean and exceeding a preset multiple of the standard deviation (such as 3 times the standard deviation) are identified as outliers and removed. For gaps resulting from removal, moving average or forward imputation methods are used to fill them in. This embodiment does not specifically limit the implementation method of step S201.

[0042] In this embodiment, by adding a data cleaning step before step S20, outliers, noise points, and missing data in the first real-time power set are identified, removed, and supplemented, which effectively improves the quality of the basic data. This allows the extraction of user electricity consumption pattern information to be based on high-quality data, thereby improving the accuracy and reliability of electricity consumption pattern analysis. This provides a more reliable basis for subsequent load forecasting and further improves the accuracy of the meter box in predicting electricity load.

[0043] Step S30: Based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained in the current time window, predict the electricity load demand value of each meter position in the future time window after the current moment.

[0044] In one feasible implementation, step S30 may include: inputting the user electricity consumption pattern information and the second real-time power set corresponding to each meter position into a pre-trained load prediction model to obtain the electricity load demand value of each meter position in the future time window after the current time.

[0045] It should be noted that the load forecasting model is a mathematical model pre-trained using historical data, used to learn the mapping relationship between user electricity consumption patterns, real-time power data, and future electricity load. The load forecasting model can employ deep learning models (such as long short-term memory neural networks), statistical models (such as autoregressive moving averages), or machine learning models, etc., and this embodiment does not impose specific limitations on this approach.

[0046] Understandably, user electricity consumption patterns reflect users' long-established electricity usage habits (such as daily peak load periods, weekly load cyclical fluctuations, and seasonal electricity trends), providing stable prior knowledge for forecasting. The second real-time power set reflects users' recent electricity consumption status and trend changes, providing dynamic real-time information for forecasting. Therefore, this implementation method deeply integrates these two aspects using a load forecasting model, ensuring that the forecast results not only conform to users' long-term electricity consumption patterns but also respond promptly to recent changes in electricity consumption. Thus, this implementation method achieves refined and personalized forecasting of each user's electricity load. The forecast results can directly serve as an important basis for grid dispatching, providing accurate micro-level user forecast data to the upper-level system, effectively supporting short-term and ultra-short-term load forecasting, improving the stability of grid operation, and to a certain extent avoiding the waste of electrical resources.

[0047] This embodiment does not specifically limit the implementation of step S30. For example, in other feasible implementations, at least two different types of prediction models can run in parallel in the main control module, such as simultaneously deploying a deep learning model based on a long short-term memory neural network and a statistical model based on an autoregressive moving average. Each model generates its own prediction results based on user electricity consumption patterns and the second real-time power set. The main control module combines the prediction results of each model through weighted fusion to generate the final electricity load demand value. The fusion weights can be dynamically adjusted according to the performance of each model in historical predictions, with models that perform better being assigned higher weights.

[0048] In one feasible implementation, for newly connected users with insufficient historical data (less than a preset number of days), the main control module cannot acquire enough data to analyze their electricity consumption patterns. In this case, the main control module can employ a transfer learning strategy to identify target users with similar electricity consumption patterns from other users connected to the meter box. Then, using the target user's electricity consumption pattern information and the second real-time power set corresponding to the newly connected user, the module predicts the newly connected user's electricity load demand within a future time window after the current moment. Thus, this implementation effectively solves the cold start problem for new users, enabling relatively accurate load forecasts even when data accumulation is insufficient.

[0049] Furthermore, in one feasible implementation, the electrical load prediction method may further include steps S301 to S303 prior to step S30: Step S301: Determine the time type characteristics of the future time window; wherein, the time type characteristics include at least the date type and the daily load period type, and the date type is weekday, rest day or holiday; It should be noted that time type characteristics are descriptive features of the time attributes to which a future time window belongs. Weekdays refer to Monday through Friday (excluding statutory holidays). User electricity consumption on weekdays typically exhibits two peak periods: morning and evening. Rest days refer to Saturdays and Sundays. User electricity consumption on rest days usually differs significantly from weekdays, such as higher daytime consumption and a potentially later evening peak. Holidays refer to statutory holidays (such as Spring Festival and National Day). User electricity consumption on holidays may change significantly, such as decreased consumption due to travel and increased consumption due to staying at home. Daily load period types divide the day into different load characteristic periods, such as morning peak, midday off-peak, evening peak, and nighttime off-peak. Electricity consumption patterns differ significantly between these different periods.

[0050] When determining the time type characteristics of a future time window, in one feasible implementation, the main control module can have a built-in calendar information database that stores information on weekdays, rest days, and holidays for the current year. Thus, the main control module can determine the date type (weekday, rest day, or holiday) of the future time window by querying the calendar information database, and simultaneously determine the load period type based on the hour range corresponding to the future time window. In another feasible implementation, the main control module can also determine the time type characteristics of the future time window from the power grid master station or a third-party service via a communication module. This embodiment does not specifically limit the implementation of step S301.

[0051] Step S302: Adjust the weight of each sub-information in the user electricity consumption pattern information corresponding to each meter position according to the time type characteristics of the future time window; It should be noted that the sub-information in user electricity consumption pattern information refers to the multiple sub-features contained within it, such as the power characteristics during daily peak load periods, the periodic fluctuation characteristics of weekly load, and the seasonal electricity consumption trend characteristics. Adjustment weights dynamically adjust the relative importance of each part of the user electricity consumption pattern information based on time-type characteristics, so that pattern information more relevant to the current forecast period plays a greater role in forecasting.

[0052] When adjusting the weights of each sub-information in the user electricity consumption pattern information corresponding to each meter position based on the time type characteristics of the future time window, the main control module can pre-set weight adjustment rules. For example, when the future time window is the evening peak period (such as 19:00-21:00), the weight of power data in the daily load peak period is increased; when the future time window is a rest day, the weight of weekend electricity consumption pattern in the weekly load periodic fluctuation characteristics is increased; when the future time window is a holiday, the weight of the corresponding season in the seasonal electricity consumption trend is increased.

[0053] Step S303: Based on the user electricity consumption pattern information and the second real-time power set corresponding to each meter position after weight adjustment, predict the electricity load demand value of each meter position in the future time window after the current moment.

[0054] In this embodiment, by adding time type feature identification and weight adjustment steps before forecasting, user electricity consumption pattern information can be dynamically adapted according to the specific attributes of future time windows. Specifically, by determining the time type characteristics of future time windows, the weights of each sub-information in the user electricity consumption pattern information are adjusted, strengthening electricity consumption patterns highly correlated with the forecast period and weakening those less correlated. This effectively improves the accuracy of load forecasting during specific periods (such as peak hours and holidays), providing more reliable forecast data for power grid dispatching, effectively supporting refined management of the distribution network and load control decisions, thereby further ensuring the stability of power grid operation and further avoiding the waste of electrical resources.

[0055] The electricity load forecasting method provided in this application can improve the accuracy of electricity load forecasting, thereby ensuring the stability of power grid operation and avoiding the waste of electrical energy resources to a certain extent. Compared with the prior art, the beneficial effects of the electricity load forecasting method provided in this application are the same as those of the electricity meter provided in the above embodiments, and will not be repeated here.

[0056] Based on the third embodiment described above, a fourth embodiment of the electricity load forecasting method of this application is proposed. In the fourth embodiment, the electricity load forecasting method may further include steps S01 to S02: Step S01: Monitor whether there are abnormal fluctuations in the real-time power of each meter position; It should be noted that abnormal fluctuations in real-time power are possible, meaning that the real-time power data may exhibit sudden changes that do not conform to normal power consumption patterns, such as a sudden drop in power to zero or a surge in power exceeding a preset threshold.

[0057] When monitoring the real-time power of each meter position for abnormal fluctuations, for any meter position, if the real-time power of that meter position is less than a preset lower power threshold and remains below it for a preset duration, it is determined to be an abnormal power drop; if the real-time power of that meter position is greater than a preset upper power threshold, it is determined to be an abnormal power surge. Both the lower and upper power thresholds can be default values ​​or can be flexibly set by the user according to actual conditions; this embodiment does not impose specific limitations on them.

[0058] Step S02: If there is an abnormal fluctuation in the real-time power of any meter position, generate an abnormal fluctuation alarm message and send the abnormal fluctuation alarm message to the power grid master station and / or distribution substation.

[0059] It should be noted that the abnormal fluctuation alarm message is a data packet used to describe the abnormal fluctuation event. It may include, but is not limited to, information such as abnormal position identifier, abnormal occurrence time, abnormal type (such as power drop or power surge) and power value at the time of abnormal occurrence. This embodiment does not make specific limitations on this.

[0060] In this embodiment, by monitoring the real-time power of each meter unit for abnormal fluctuations, alarm information is generated promptly and sent to the main power grid station and / or distribution substation when an anomaly occurs, thereby achieving early warning of faults. Thus, this embodiment further expands the anomaly monitoring and alarm capabilities of the meter box beyond its load sensing and prediction functions. It can promptly detect abnormal events such as user power outages, line overloads, and data collection faults, providing maintenance personnel with a basis for rapid response, shortening fault handling time, and improving power supply reliability.

[0061] Based on the third and / or fourth embodiments described above, a fifth embodiment of the electricity load forecasting method of this application is proposed. In the fifth embodiment, the electricity load forecasting method may further include: performing similarity analysis on the electricity consumption pattern information of users at each meter location to identify user groups whose electricity consumption pattern information is highly similar (i.e., the similarity is greater than a preset similarity threshold); for meter locations belonging to the same user group, the predicted electricity load demand value of each meter location in the user group can be reversed using the group forecasting result of the user group.

[0062] Among these, the high similarity of user electricity consumption patterns—meaning that the distance between the electricity consumption patterns of multiple meter locations is small in the feature space—indicates a high degree of homogeneity in the electricity consumption behavior patterns of these users. Similarity analysis can be performed using methods such as cosine similarity, Euclidean distance, or Pearson correlation coefficient; this embodiment does not impose specific limitations on these methods. The group prediction result refers to the overall predicted value obtained by accumulating or fusing the electricity load demand values ​​of all meter locations within the user group, reflecting the overall trend of the group's electricity consumption behavior. Reverse correction utilizes the statistical characteristics of the group prediction result to adjust individual prediction values, making the individual prediction values ​​converge towards the group trend while maintaining the overall group prediction value unchanged.

[0063] A preset similarity threshold is used as the basis for determining whether the user electricity consumption pattern information of each table position is highly similar. It can be a default value or it can be flexibly set by the user according to the actual situation. This embodiment does not make specific limitations on this.

[0064] In this embodiment, by identifying user groups with highly similar electricity consumption patterns, the statistical stability of group electricity consumption behavior is used to correct any abnormal deviations that may occur in individual predictions, thereby further improving the accuracy and reliability of electricity load prediction.

[0065] This application also provides a computer-readable storage medium storing a computer program that can run on a processor. The computer program is used to execute the power load forecasting method in the above embodiments.

[0066] The computer-readable storage medium provided in this application embodiment may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0067] The aforementioned computer-readable storage medium may be included in the energy meter; or it may exist independently and not be installed in the energy meter.

[0068] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electricity meter, cause the electricity meter to: acquire the real-time power of each meter position in the meter box; analyze the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set acquired within a preset time window; and predict the electricity load demand value of each meter position in a future time window after the current moment based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set acquired within the current time window; wherein, the step of acquiring the real-time power of each meter position in the meter box includes: acquiring the load current on the incoming circuit of each meter position; and calculating the real-time power of each meter position based on the load current of each meter position.

[0069] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0070] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0071] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0072] The computer-readable storage medium provided in this application embodiment stores computer-readable program instructions for executing the above-described electricity load forecasting method, which can improve the accuracy of electricity load forecasting, thereby ensuring the stability of power grid operation and avoiding the waste of electrical energy resources to a certain extent. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application embodiment are the same as the beneficial effects of the electricity load forecasting method provided in the above embodiments, and will not be repeated here.

[0073] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the power load prediction method described above.

[0074] The computer program product provided in this application can improve the accuracy of electricity load forecasting, thereby ensuring the stability of power grid operation and avoiding the waste of electrical energy resources to a certain extent. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the electricity load forecasting method provided in the above embodiments, and will not be repeated here.

[0075] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. An electric meter box, characterized in that, include: The enclosure contains multiple mounting positions for installing electricity meters; Multiple acquisition modules are coupled one-to-one with the incoming line circuit of each meter position to acquire the real-time power of each meter position. A main control module, located inside the enclosure and electrically connected to each of the acquisition modules, is used for: Receive the real-time power acquired by each of the acquisition modules; Based on the first real-time power set received for each meter position within a preset time window, analyze the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set received for each meter position within the current time window, predict the electricity load demand value of each meter position in the future time window after the current moment; The acquisition module includes: A current sampling unit is coupled to the incoming circuit of a meter position in the enclosure and is used to collect the load current on the incoming circuit of the meter position. The metering unit has its input terminal electrically connected to the current sampling unit and its output terminal electrically connected to the main control module. It is used to calculate the real-time power of the meter position based on the load current and transmit the calculated real-time power to the main control module.

2. The meter box as described in claim 1, characterized in that, The meter box also includes: A communication module, which is electrically connected to the main control module, is used to exchange data with the power grid master station and / or distribution substation. The main control module is also used to send the predicted power load demand value of each meter position and the received real-time power of each meter position to the power grid master station and / or distribution substation through the communication module.

3. A method for predicting electricity load, characterized in that, Applied to the meter box as described in claim 1 or 2, the method includes: Obtain the real-time power of each meter position in the meter box; Based on the first real-time power set of each meter position obtained within the preset time window, analyze the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained within the current time window, predict the electricity load demand value of each meter position in the future time window after the current moment; The step of obtaining the real-time power of each meter position in the meter box includes: Obtain the load current on the incoming circuit of each of the aforementioned positions; The real-time power of each meter position is calculated based on the load current of each meter position.

4. The method as described in claim 3, characterized in that, The step of analyzing the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set obtained within a preset time window includes: For any of the aforementioned positions, a power curve for the position within the preset time window is generated based on each real-time power in the first real-time power set of the position and the acquisition time corresponding to each real-time power. The power curve is analyzed to extract the user electricity consumption pattern information corresponding to the meter position; wherein, the user electricity consumption pattern information includes at least the daily peak load period, the weekly load periodic fluctuation characteristics, and the seasonal electricity consumption trend.

5. The method as described in claim 3 or 4, characterized in that, Before the step of analyzing the user electricity consumption pattern information corresponding to each meter position based on the first real-time power set obtained within a preset time window, the method further includes: Data cleaning processing is performed on the first real-time power set of each of the aforementioned positions; Based on the first real-time power set of each meter position after data cleaning and processing, analyze the user electricity consumption pattern information corresponding to each meter position.

6. The method as described in claim 3, characterized in that, The step of predicting the electricity load demand value of each meter position in the future time window after the current moment, based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained within the current time window, includes: The user electricity consumption pattern information and the second real-time power set corresponding to each meter position are input into the pre-trained load prediction model to obtain the electricity load demand value of each meter position in the future time window after the current time.

7. The method as described in claim 3 or 6, characterized in that, Before the step of predicting the electricity load demand value of each meter position in the future time window after the current moment based on the user electricity consumption pattern information corresponding to each meter position and the second real-time power set of each meter position obtained in the current time window, the method further includes: Determine the time type characteristics of the future time window; wherein, the time type characteristics include at least the date type and the daily load period type, and the date type is a weekday, rest day, or holiday; Based on the time type characteristics of the future time window, adjust the weight of each sub-information in the user electricity consumption pattern information corresponding to each meter position; Based on the user electricity consumption pattern information corresponding to each meter position after weight adjustment and the second real-time power set, the electricity load demand value of each meter position in the future time window after the current moment is predicted.

8. The method as described in claim 3, characterized in that, The method further includes: Monitor whether there are abnormal fluctuations in the real-time power of each of the aforementioned positions; If there is an abnormal fluctuation in the real-time power of any of the aforementioned meter positions, an abnormal fluctuation alarm message is generated and sent to the power grid master station and / or distribution substation.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the power load forecasting method as described in any one of claims 3 to 8.

Images