Prepayment early warning method and system for intelligent electric meter

Abstract

The invention provides a prepaid early warning method for intelligent electric meters, which comprises the following steps of: acquiring data of each intelligent electric meter in real time and transmitting the data to a database, and transmitting the data of the database to a constructed TensorFlow at regular time; if data missing exists, automatically supplementing the historical data and the recent data through TensorFlow to form complete data, and returning the complete data to the database; the total data in the database is subjected to electricity consumption prediction after n days through TensorFlow according to historical data and recent data, and energy consumption early warning and notice are sent to a user after energy consumption balance calculation is combined. According to the technical scheme, faults in short communication time are made up by combining the data complementation and prediction capability of the AI large model, and personalized energy consumption reminding services can be carried out for different user groups.

Description

Technical Field

[0001] The present application belongs to the field of smart meter electricity consumption management, and specifically relates to a method and system for prepayment warning of smart meters. Technical Background

[0002] In the commercial energy consumption fields such as universities, hospitals, and industrial parks, the prepaid model is basically used for energy management. In these scenarios, there are many types of energy users, including students, restaurants, business groups, and other energy users with different characteristics. On the one hand, the energy consumption levels and individuals of these users are inconsistent, and it is difficult to provide personalized early warning services; on the other hand. In the current prepaid scenario, there is a problem with both wired 485 communication and wireless 4G communication services, that is, unstable communication, and communication disconnection often lasts from one week to one month, which will lead to large deductions, inaccurate early warnings, and other problems. Once the energy supply is stopped, it will lead to user complaints and some difficulties in subsequent management. Therefore, how to avoid communication problems and improve service quality is a problem that needs to be broken through at present.

[0003] Technical content

[0004] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention proposes a method and system for prepayment warning of smart meters, which compensates for short-term communication failures through the energy consumption platform combined with the data completion and prediction capabilities of the AI ​​big model, and can provide personalized energy consumption reminder services for different user groups.

[0005] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0006] In a first aspect, the present invention provides a method for prepayment warning of a smart meter, comprising:

[0007] Collect data from each smart meter in real time and transmit it to the database, and the database data is regularly transmitted to the constructed TensorFlow;

[0008] If there is any missing data, the historical data and recent data are automatically supplemented by the TensorFlow to form complete data and then returned to the database;

[0009] The full amount of data in the database is periodically used to further predict the electricity consumption n days later based on historical data and recent data through TensorFlow, and energy consumption warnings and notifications are sent to users after combining the energy balance calculation.

[0010] Optionally, if there is data missing, the data is automatically supplemented by TensorFlow based on historical data and recent data to form complete data, including:

[0011] Acquire the user's recent data and historical data, wherein the recent data includes: monthly energy consumption data for at least the previous three months, and the historical data includes: monthly energy consumption data for at least three months in the same period of the previous year;

[0012] The energy consumption trend of the recent data and the energy consumption trend of the historical data are formed. If the two energy consumption trends are inconsistent, the average value of the trend growth is taken to predict the energy consumption data of the missing months; if the two energy consumption trends are consistent, the energy consumption data of the missing months is predicted based on the year-on-year growth;

[0013] The difference between the predicted energy consumption data for the missing month and the existing data for the missing month is used as the filling data.

[0014] Optionally, if there is missing data, the data is automatically filled in by TensorFlow based on historical data and recent data to form complete data, and it also includes: verifying the predicted energy consumption data of the missing month in combination with the energy consumption data of the latest month. If the predicted energy consumption data of the missing month is greater than the energy consumption data of the latest month, taking the average of the difference between the energy consumption data of the latest month and the energy consumption data of the month before the missing month as the predicted energy consumption data and then filling in the gaps; if the predicted energy consumption data of the missing month is less than the energy consumption data of the latest month, directly filling in the missing data.

[0015] Optionally, the forming of complete data after automatic completion by the TensorFlow includes: using the Dataset object in the tf.data module to build a data pipeline; using the tf.data.Dataset.from_generator() method to create a dataset, and completing the input data by calling the function in the Dataset.padded_batch() method.

[0016] Optionally, the database data is periodically transmitted to the constructed TensorFlow, including: the data is parsed by the energy platform and then transmitted to the TensorFlow, and the system forms a data set based on the breakpoint data.

[0017] Optionally, the full amount of data in the database is periodically used to further predict the electricity consumption n days later based on historical data and recent data through the TensorFlow, including:

[0018] Obtain the user's recent data and historical data, the recent data is the energy consumption data of at least the previous three months calculated from n days later, and the historical data is the energy consumption data of at least three months in the same period last year;

[0019] The energy consumption trend of the recent data and the energy consumption trend of the historical data are formed, and the average trend value is calculated to obtain the predicted energy consumption data for this month including n days.

[0020] Optionally, the timing further predicts the electricity consumption n days later using the full data in the database according to historical data and recent data through the TensorFlow, and also includes: updating user deduction data after parsing the full data through the energy consumption platform.

[0021] Optionally, the energy usage warning and notification are sent to the user after the energy usage balance is calculated, including: if the energy usage balance is insufficient, calling the deduction service, and automatically splitting the deduction according to the replenishment nodes; if the energy usage balance is close to being insufficient, issuing a collection warning to the management personnel and a payment warning to the user after calculating the reminder cycle.

[0022] Optionally, if there is missing data, the data is automatically supplemented by the TensorFlow based on historical data and recent data to form complete data, and the system also detects whether the data of the previous day is missing through a daily scheduled task; when missing data is detected, the TensorFlow is automatically called.

[0023] In a second aspect, the present invention further provides a prepayment warning system for smart meters, comprising: a user group, an AI model, a communication layer, a device layer, and data storage; wherein the device layer comprises a plurality of smart meters, the communication layer comprises a gateway 485 and 4G, and the AI ​​model is a constructed TensorFlow;

[0024] The communication layer is used to collect data from each smart meter in real time and transmit it to the MySQL database for data storage, and the database data is regularly transmitted to the TensorFlow;

[0025] The AI ​​model is used to automatically fill in the missing data based on historical data and recent data to form complete data and then return it to the database; the AI ​​model is also used to further predict the electricity consumption n days later based on the historical data and recent data at a preset time point, and the platform sends energy consumption warnings and notifications to users after calculating the energy balance.

[0026] The technical solution of the present invention utilizes a large AI model and combines it with a unique algorithm logic to complete and predict user energy consumption data. It can not only reduce the failures caused by communications and visually display user energy consumption data, but also provide personalized energy consumption services and management for multiple devices and multiple user groups of smart meters.

[0027] Instruction Manual

[0028] Figure 1 A technical interaction sequence diagram of a method for prepayment warning of a smart meter provided in an embodiment of the present invention.

[0029] Figure 2 A flowchart of a method for prepayment warning of a smart meter provided in an embodiment of the present invention.

[0030] Figure 3 A schematic diagram of a system architecture for a smart meter prepayment warning system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be interpreted as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.

[0032] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0033] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optionally" means "optional embodiments". Relevant definitions of other terms will be given in the following description. It should be noted that the concepts of "first", "second", etc. mentioned in the present invention are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0034] It should be noted that the modifications of "one" and "plurality" mentioned in the present invention are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0035] refer to Figure 1 and Figure 2 The embodiment of the present invention provides a method for prepayment warning of a smart meter, comprising:

[0036] S1. Collect data from each smart meter in real time and transmit it to the database. The database data is periodically transmitted to the constructed TensorFlow.

[0037] Specifically, the smart meter based on the device layer collects data through the communication layer, which includes gateway 485, 4G, etc. The data is parsed by the energy consumption platform and transmitted to TensorFlow. The system forms a data set based on the breakpoint data. In this embodiment, the energy consumption system will detect whether the data of the previous day is missing through a scheduled task at 2 a.m. every day. In particular, since the data is generally collected and reported regularly, there are 24 energy consumption records every day, so when it is missing, the corresponding time point will be reminded of the missing data.

[0038] S2. If there is any missing data, it will be automatically completed by TensorFlow based on historical data and recent data to form complete data and then returned to the database.

[0039] When missing data is detected, the system will automatically call the TensorFlow model. The idea of ​​filling missing data is: first obtain historical data and recent data, then make a prediction, and verify the prediction results with the current latest energy consumption data, and finally fill in the missing data. The details are as follows:

[0040] S21. Obtain the user's recent data and historical data. In this embodiment, the recent data is the monthly energy consumption data of at least the previous three months, and the historical data is the monthly energy consumption data of at least three months in the same period of the previous year. Of course, energy consumption data of more consecutive months can also be selected according to specific circumstances.

[0041] S22. Form the energy consumption trend of recent data and the energy consumption trend of historical data. If the two energy consumption trends are inconsistent, the average value of the trend growth is taken to predict the energy consumption data of the missing months; if the two energy consumption trends are consistent, the energy consumption data of the missing months is predicted based on the year-on-year growth.

[0042] S23. The difference between the predicted energy consumption data for the missing month and the existing data for the missing month is used as the filling data.

[0043] S24. Verify the predicted energy consumption data of the missing months in combination with the energy consumption data of the latest month. If the predicted energy consumption data of the missing months is greater than the energy consumption data of the latest month, take the average of the difference between the energy consumption data of the latest month and the energy consumption data of the month before the missing month as the predicted energy consumption data and then fill in the gaps; if the predicted energy consumption data of the missing months is less than the energy consumption data of the latest month, fill in the missing data directly.

[0044] S25. The completed data is returned to the database to form complete full data.

[0045] For example: If it is detected that there are M days of missing data in April this year, the energy consumption data for the same three months of the previous year (January, February, and March of the previous year) were 100, 200, and 300 respectively, and the energy consumption data for the first three months of this year (January, February, and March of this year) were 200, 300, and 400 respectively, and the energy consumption data for the latest month, the 5th month, was 600. Then, the difference is 100 compared with the same trend. According to the trend, 100 are added, the data for the 4th month of the previous year is 400, and the predicted energy consumption data for the missing month this year is 500. Further, if the predicted energy consumption data is less than the latest data of the 5th month, which is 600, the predicted energy consumption data of 500 is directly subtracted from the data collected and reported in April this year, and the data obtained is the data that needs to be filled in for M days. If the missing M days are 30 days of a whole month, it means that the data collected and reported in April this year is 0, then 500 is directly used as the missing data to fill.

[0046] If, in the above example, the historical data and recent data remain unchanged, and the energy consumption data for May this year is 480, then since the energy consumption data 500 predicted during verification is greater than the latest data of the 5th month 480, it is necessary to perform a difference average calculation, that is, (May energy consumption data 480-March energy consumption data 400) / 2=the energy consumption data 440 predicted for April after verification. At this time, it is necessary to subtract the data collected and reported in April from 440 to obtain the data that needs to be supplemented for M days.

[0047] If it is detected that there are M days of missing data in April this year, the energy consumption data for the same three months of the previous year (January, February, and March of the previous year) are 100, 200, and 400 respectively, and the energy consumption data for the first three months of this year (January, February, and March of this year) are 200, 300, and 400 respectively, and the energy consumption data for the latest month, the fifth month, is 600. Since the trends in the same period this year and last year are different, the average value of the growth is [(400-200)+(400-300)] / 2=150, and the predicted energy consumption data for April is 400+150=550, and after verification, the predicted energy consumption data of 550 is less than the energy consumption data of 600 in May, then the value 550 can be directly used as the basis for completing the data.

[0048] In particular, the "one month" mentioned in the technical solution of the present invention is calculated based on the 30 days before the daily scheduled task. For example, when the scheduled task on April 15, 2024 is processed, this month is from March 16, 2024 to April 15, 2024, and the dimension of the previous month is from February 15, 2024 to March 16, 2024. Similarly, the dimensions of the first two months and the first three months are calculated similarly.

[0049] S3. Regularly use TensorFlow to further predict the electricity consumption in n days based on the historical data and recent data in the database, and send energy consumption warnings and notifications to users after combining the energy balance calculation.

[0050] Specifically, the energy consumption platform analyzes the full amount of data, updates the user deduction data, and then predicts the energy consumption in the next 7 days through TensorFlow. The prediction method is the same as the prediction and completion method of the missing data mentioned above. That is, the recent data is the energy consumption data of the previous three months, and the historical data is the energy consumption data of the same period last year within three months; the energy consumption data in 7 days is predicted based on the trend difference between the recent data and the historical data.

[0051] For example, a scheduled task on December 15th predicts the energy consumption data for December 22nd, seven days later. Then, the system can default to November 22nd to December 22nd for this monthly period, and the monthly energy consumption data for the previous three months are obtained as "October 23rd to November 22nd", "September 23rd to October 23rd", and "August 24th to September 23rd". Similarly, the same three-month dimension of the previous year is also the same. Assuming that this year's monthly energy consumption data of "10.23-11.22", "9.23-10.23" and "8.24-9.23" are 100, 200 and 400 respectively, and last year's monthly energy consumption data of "10.23-11.22", "9.23-10.23" and "8.24-9.23" are 100, 200 and 500 respectively, then the average growth value is [(400-200)+(500-200)] / 2=250, then the energy consumption data for this month's 11.22-12.22 is predicted to be 650, that is, the energy consumption data 7 days later will be 650. Combined with user deduction data, intelligent control provides personalized services and notifications within a period of nearly 7 days after energy consumption based on the energy consumption levels of different user groups: if the energy balance is insufficient, the deduction service is called to automatically split the deduction according to the replenishment nodes; if the energy balance is close to being insufficient, the reminder cycle is calculated and the management personnel are warned to pay, and the users are warned to pay.

[0052] Since this technical solution is mainly aimed at commercial electricity users, the monthly energy consumption data varies significantly. Therefore, as long as the prediction is made based on the monthly energy consumption trend, the accuracy of the prediction can be guaranteed to the greatest extent.

[0053] In addition, an embodiment of the present invention further provides a prepayment warning system for smart meters, including: a user group, an AI model, a communication layer, a device layer, and data storage; wherein the device layer includes a plurality of smart meters, the communication layer includes a gateway 485 and 4G, and the AI ​​model is a constructed TensorFlow;

[0054] The communication layer is used to collect data from each smart meter in real time and transmit it to the MySQL database for data storage. The database data is regularly transmitted to TensorFlow;

[0055] The AI ​​model is used to automatically fill in the missing data based on historical and recent data to form complete data and then return it to the database; the AI ​​model is also used to further predict the electricity consumption n days later based on the historical and recent data at a preset time point. The platform calculates the energy balance and sends energy warnings and notifications to users.

[0056] The technical effect that can be achieved by this embodiment is basically the same as the technical effect that can be achieved by the aforementioned method for prepayment warning of smart meters, and will not be repeated here.

[0057] In particular, the specific execution operations to achieve the purpose of this embodiment can be referred to as follows:

[0058] 1. There are many energy consumption platforms on the market now, you can choose one.

[0059] 2. Install TensorFlow CPU version: You can download and install TensorFlow CPU version from the TensorFlow official website.

[0060] 3. Use the Docker image officially provided by TensorFlow.

[0061] 4. Create a virtual environment in Anaconda and install TensorFlow: Use Anaconda to create a virtual environment and install TensorFlow in it.

[0062] 5. Use TensorFlow Lite: You can perform reasoning without the need for high-performance hardware (such as GPU or TPU).

[0063] 6. After installing TensorFlow, use TensorFlow to debug and reason about the model through the Python API or command line tools.

[0064] 7. The energy consumption platform reports the breakpoint data set;

[0065] 8. Use the Dataset object in the tf.data module to build a data pipeline. Use the tf.data.Dataset.from_generator() method to create a dataset and call the function in the Dataset.padded_batch() method to padded the input data.

[0066] 9. The energy consumption platform completes and updates the deduction data, and reports the full data to the big model;

[0067] 10. Use the predict() method to make model predictions. The data is returned to the energy consumption platform;

[0068] 11. The energy consumption platform calculates the cycle and determines whether to issue a reminder.

[0069] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0070] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0072] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0073] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc. In the present application, the unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment of the present invention. In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0074] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A method for prepayment warning of smart meters, characterized in that include: Collect data from each smart meter in real time and transmit it to the database, and the database data is regularly transmitted to the constructed TensorFlow; If there is any missing data, the historical data and recent data are automatically supplemented by the TensorFlow to form complete data and then returned to the database; The full amount of data in the database is periodically used to further predict the electricity consumption n days later through the TensorFlow according to historical data and recent data, and energy consumption warnings and notifications are sent to users after combining the energy balance calculation.

2. A method for prepayment warning of smart meters as claimed in claim 1, characterized in that If there is data missing, the historical data and recent data are automatically supplemented by TensorFlow to form complete data, including: Acquire the user's recent data and historical data, wherein the recent data includes: monthly energy consumption data for at least the previous three months, and the historical data includes: monthly energy consumption data for at least three months in the same period of the previous year; The energy consumption trend of the recent data and the energy consumption trend of the historical data are formed. If the two energy consumption trends are inconsistent, the average value of the trend growth is taken to predict the energy consumption data of the missing months; if the two energy consumption trends are consistent, the energy consumption data of the missing months is predicted based on the year-on-year growth; The difference between the predicted energy consumption data for the missing month and the existing data for the missing month is used as the filling data.

3. A method for prepayment warning of smart meter as claimed in claim 2, characterized in that If there is missing data, the data is automatically filled in by TensorFlow based on historical data and recent data to form complete data, which also includes: verifying the predicted energy consumption data of the missing month in combination with the energy consumption data of the latest month. If the predicted energy consumption data of the missing month is greater than the energy consumption data of the latest month, taking the average of the difference between the energy consumption data of the latest month and the energy consumption data of the month before the missing month as the predicted energy consumption data and then filling in the gaps; if the predicted energy consumption data of the missing month is less than the energy consumption data of the latest month, directly filling in the missing data.

4. A method for prepayment warning of smart meter as claimed in claim 1, 2 or 3, characterized in that The automatic completion by TensorFlow forms complete data, including: using the Dataset object in the tf.data module to build a data pipeline; using the tf.data.Dataset.from_generator() method to create a data set, and completing the input data by calling the function in the Dataset.padded_batch() method.

5. A method for prepayment warning of smart meter as claimed in claim 1, characterized in that The database data is transmitted to the constructed TensorFlow at a regular interval, including: the data is parsed by the energy consumption platform and then transmitted to the TensorFlow, and the system forms a data set based on the breakpoint data.

6. A method for prepayment warning of smart electricity meters as claimed in claim 1, characterized in that The timing further predicts the power consumption n days later using the full amount of data in the database according to historical data and recent data through the TensorFlow, including: Obtain the user's recent data and historical data, the recent data is the energy consumption data of at least the previous three months calculated from n days later, and the historical data is the energy consumption data of at least three months in the same period last year; The energy consumption trend of the recent data and the energy consumption trend of the historical data are formed, and the average trend value is calculated to obtain the predicted energy consumption data for this month including n days.

7. A method for prepayment warning of smart meter as claimed in claim 6, characterized in that The method further predicts the electricity consumption n days later by timing the full data in the database according to historical data and recent data through the TensorFlow, and also includes: updating the user deduction data after parsing the full data through the energy consumption platform.

8. A method for prepayment warning of smart meter as claimed in claim 1 or 7, characterized in that The energy usage warning and notification are sent to the user after the energy usage balance is calculated, including: if the energy usage balance is insufficient, the deduction service is called to automatically split the deduction according to the replenishment node; if the energy usage balance is close to being insufficient, the management personnel are issued a payment reminder warning after calculating the reminder cycle, and the user is issued a payment warning.

9. A method for prepayment warning of smart meter as claimed in claim 1, characterized in that If there is missing data, the data will be automatically supplemented by the TensorFlow based on historical data and recent data to form complete data. The system will detect whether the data of the previous day is missing through a daily scheduled task; when missing data is detected, the TensorFlow will be automatically called.

10. A prepayment warning system for smart meters as claimed in claim 1, characterized in that include: User groups, AI models, communication layers, device layers, and data storage; wherein the device layer includes multiple smart meters, the communication layer includes gateway 485 and 4G, and the AI ​​model is a constructed TensorFlow; The communication layer is used to collect data from each smart meter in real time and transmit it to the MySQL database for data storage, and the database data is regularly transmitted to the TensorFlow; The AI ​​model is used to automatically fill in the missing data based on historical data and recent data to form complete data and then return it to the database; the AI ​​model is also used to further predict the electricity consumption n days later based on the historical data and recent data at a preset time point, and the platform sends energy consumption warnings and notifications to users after calculating the energy balance.

Images