Electricity sales amount forecasting method and system
By obtaining and analyzing electricity consumption, payment and weather historical data, and entering a prediction model to determine the amount of electricity sales, the problem of inaccurate prediction of electricity sales in the prior art is solved, and higher prediction accuracy and adaptability are achieved.
Patent Information
- Application Number
- CN202111187180.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-12
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2041-10-12
AI Technical Summary
The existing technology is difficult to accurately predict the amount of electricity sales. Due to various factors such as user electricity consumption, payment habits and unit price of electricity, the prediction results are not accurate enough.
By obtaining the electricity consumption historical information at multiple historical time points, including electricity consumption characteristics, payment characteristics and weather characteristics, input the prediction model to determine the amount of electricity sales at the time points to be predicted. The prediction model includes a feature construction layer and a prediction layer, and feature construction The layer obtains power sales characteristics, and the prediction layer predicts sales amount based on these characteristics, and adjusts the prediction results through adjustment coefficients to reduce weather impact.
It improves the accuracy of forecasting of electricity sales amounts, takes into account the electricity consumption and payment habits of electricity users, reduces the interference of weather information on forecasts, and is suitable for forecasting of different time spans.
Smart Images

Figure CN113902195B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing, and in particular to a method and system for predicting the amount of electricity sales. Background Art
[0002] The grid electricity sales platform can collect electricity consumption and sales data of electricity users, and predict the electricity sales amount based on the historical electricity consumption and sales data of electricity users, so as to deploy the power supply plan, economic planning and user experience of the grid electricity sales platform. However, there are many factors that affect the electricity sales amount, such as the electricity consumption of the user unit, the payment habits of the electricity unit and the unit price of electricity.
[0003] Therefore, it is hoped to provide a method for predicting the amount of electricity sales, which can improve the accuracy of the prediction of the amount of electricity sales. Summary of the invention
[0004] One aspect of an embodiment of the present specification provides a method for predicting the amount of electricity sales, the method comprising: obtaining electricity consumption history information corresponding to multiple historical time points, the electricity consumption history information including electricity consumption characteristics, payment characteristics and historical weather characteristics; inputting the electricity consumption history information corresponding to the multiple historical time points into a prediction model to determine the amount of electricity sales at at least one time point to be predicted; wherein the prediction model comprises a feature construction layer and a prediction layer; the feature construction layer obtains the electricity sales characteristics corresponding to the multiple historical time points based on the electricity consumption history information corresponding to the multiple historical time points; the prediction layer determines the electricity sales amount at at least one time point to be predicted based on the electricity sales characteristics.
[0005] Another aspect of an embodiment of the present specification provides a system for predicting the amount of electricity sales, the system comprising: an acquisition module for acquiring electricity consumption history information corresponding to multiple historical time points, the electricity consumption history information including electricity consumption characteristics, payment characteristics and historical weather characteristics; a determination module for inputting the electricity consumption history information corresponding to the multiple historical time points into a prediction model to determine the amount of electricity sales at at least one time point to be predicted; wherein the prediction model comprises a feature construction layer and a prediction layer; the feature construction layer acquires the electricity sales characteristics corresponding to the multiple historical time points based on the electricity consumption history information corresponding to the multiple historical time points; the prediction layer determines the electricity sales amount at at least one time point to be predicted based on the electricity sales characteristics.
[0006] Another aspect of an embodiment of the present specification provides an electricity sales amount prediction device, comprising at least one storage medium and at least one processor, wherein the at least one storage medium is used to store computer instructions; the at least one processor is used to execute the computer instructions to implement the electricity sales amount prediction method as described in any of the above embodiments.
[0007] Another aspect of the embodiments of this specification provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed by a processor, the method for predicting the amount of electricity sales as described in any of the above embodiments is implemented. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:
[0009] Figure 1 is a schematic diagram of an application scenario of the power sales amount prediction system 100 according to some embodiments of this specification;
[0010] Figure 2 is an exemplary flow chart of a method for predicting the amount of electricity sales according to some embodiments of this specification;
[0011] Figure 3 is a schematic diagram of a prediction model according to some embodiments of this specification;
[0012] Figure 4 is an exemplary flow chart of adjusting the amount of electricity sales according to some embodiments of this specification;
[0013] Figure 5 It is a schematic diagram of adjusting the electricity sales amount according to some embodiments of this specification. DETAILED DESCRIPTION
[0014] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.
[0015] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.
[0016] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "comprises" and "includes" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.
[0017] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, the various steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.
[0018] Electricity users are individuals and / or organizations that purchase electricity. For example, residents, shops, factories, etc. The grid electricity sales platform is a platform for selling electricity and managing sales data. Electricity users can purchase electricity and query electricity consumption data through the grid electricity sales platform. The grid electricity sales platform can collect electricity consumption and sales data of electricity users, so as to pre-dispatch the power load.
[0019] The electricity sales amount is the total amount of electricity collected by the power grid sales platform from the electricity users during the statistical period. The outlet sales platform can make power supply plans and economic plans by predicting the electricity sales amount.
[0020] Figure 1 It is a schematic diagram of an application scenario of the electricity sales amount prediction system 100 shown in some embodiments of this specification.
[0021] The power sales amount prediction system 100 can be used to predict the power sales amount of the grid power sales platform at at least one time point to be predicted. Figure 1 As shown, the power sales amount prediction system 100 may include a server 110 , a network 120 , a network power sales platform 130 , a user terminal 140 and an electricity meter 150 .
[0022] The server 110 may process data and / or information from at least one component of the power sales amount prediction system 100. The server 110 may communicate with the grid power sales platform 130 to provide various functions of the service. For example, the server 110 may obtain data (such as power consumption history information) from the grid power sales platform 130, and output the data (such as the predicted power sales amount) to the grid power sales platform through the network 120. The server 110 may be used to process data and / or information from at least one component of the power amount prediction system 100 or an external data source (for example, a cloud data center). For example, the server may process power consumption history information to obtain the predicted power sales amount. In some embodiments, the server 110 may be a single server or a server group. The server group may be centralized or distributed (for example, the server 110 may be a distributed system). In some embodiments, the server 110 may be regional or remote. In some embodiments, the server 110 may be implemented on a cloud platform or provided in a virtual manner. By way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, or the like, or any combination thereof.
[0023] The network 120 can connect the components of the system and / or connect the system with external parts. The network 120 allows the components of the system and the system and external parts to communicate with each other and promote the exchange of data and / or information. In some embodiments, the network 120 can be any one or more of a wired network or a wireless network. For example, the network 120 can include a cable network, an optical fiber network, a telecommunications network, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network (ZigBee), a near field communication (NFC), a bus within a device, a line within a device, a cable connection, etc. or any combination thereof. In some embodiments, the network connection between the various parts of the system can adopt one of the above methods or multiple methods. In some embodiments, the network 120 can be a point-to-point, shared, central, etc., various topological structures or a combination of multiple topological structures. In some embodiments, the network 120 can include one or more network access points. For example, network 120 may include wired or wireless network access points, such as base stations and / or network switching points 120-1, 120-2, ..., through which one or more components of system 100 may connect to network 120 to exchange data and / or information.
[0024] The network electricity sales platform 130 can be used to store and manage network electricity sales data. In some embodiments, the network electricity sales platform 130 can obtain the power consumption of the power user from the electric meter 150 through the network 120. In some embodiments, the network electricity sales platform 130 can also calculate the amount payable by the power user based on the power consumption of the power user, and send it to the client 140 through the network 120. In some embodiments, the network electricity sales platform 130 can also store power consumption history information, such as power consumption information, payment amount information, or weather information corresponding to the payment time point. The network electricity sales platform 130 may include one or more storage components, each of which may be an independent device or part of other devices. In some embodiments, the network electricity sales platform 130 may include a random access memory (RAM), a read-only memory (ROM), a large capacity memory, a removable memory, a volatile read-write memory, or the like, or any combination thereof. Exemplarily, the large capacity storage may include a magnetic disk, an optical disk, a solid-state disk, etc. In some embodiments, the network electricity sales platform 130 may be implemented on a cloud platform. As just an example, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, etc. or any combination thereof. In some embodiments, the network electricity sales platform 130 may be integrated or included in one or more other components of the system 100 (e.g., the server 110, the user end 140 or other possible components).
[0025] The user end 140 refers to one or more terminal devices or software used by the electricity user unit, and the electricity user unit may include an individual who pays the electricity bill (e.g., an individual who needs to pay the electricity bill) and an organization that collects the electricity bill (e.g., a company that needs to pay the electricity bill). In some embodiments, the electricity user unit can use the user end 140 to communicate with the server 110 and / or the online electricity sales platform 130 through the network 120. For example, the electricity user unit can use the user end 140 to communicate with the online electricity sales platform 130 to pay the amount of electricity used. In some embodiments, the user end 140 can be a mobile device 140-1, a tablet computer 140-2, a laptop computer 140-3, a desktop computer 140-4, other devices with input and / or output functions, etc., or any combination thereof. The above examples are only used to illustrate the wide range of the user end 140 device range and are not intended to limit its range.
[0026] The electric meter 150 is a device for measuring the power consumption of the power unit. In some embodiments, the electric meter 150 can send the power consumption of the power unit to the network power sales platform 130 through the network 120. In some embodiments, the electric meter 150 can also receive power-off and power-on instructions from the network power sales platform 150.
[0027] In some embodiments, the power sales amount prediction system 100 may be implemented on a server 110 , which includes an acquisition module and a determination module.
[0028] The acquisition module can be used to input the electricity consumption history information corresponding to the multiple historical time points into the prediction model to determine the electricity sales amount at at least one time point to be predicted. The prediction model includes a feature construction layer and a prediction layer. In some embodiments, the feature construction layer obtains the electricity sales features corresponding to the multiple historical time points based on the electricity consumption history information corresponding to the multiple historical time points. In some embodiments, the input of the feature construction layer also includes electricity unit features. The electricity unit features are obtained based on the electricity unit feature model. In some embodiments, the prediction layer determines the electricity sales amount at at least one time point to be predicted based on the electricity sales features to obtain the electricity consumption history information corresponding to multiple historical time points. For more descriptions of the prediction model, please refer to step 220, which will not be repeated here.
[0029] In some embodiments, the power sales amount prediction system 100 may further include a first module, a second module, and a third module.
[0030] The first module can be used to obtain at least one weather forecast information corresponding to the at least one time point to be predicted. In some embodiments, the weather forecast information may include at least one of temperature information, precipitation information, and wind information. For more details about the weather forecast information, see step 410, which will not be repeated here.
[0031] The second module can be used to determine at least one adjustment coefficient based on the at least one weather forecast information. In some embodiments, the adjustment coefficient is determined based on the temperature deviation rate, the precipitation deviation rate and the wind deviation rate. The temperature deviation rate is determined based on the temperature information and the reference temperature, the precipitation deviation rate is determined based on the precipitation information and the reference precipitation, and the wind deviation rate is determined based on the wind information and the reference wind. For more details on determining the adjustment coefficient, see step 420, which will not be repeated here.
[0032] The third module can be used to adjust the electricity sales amount at the at least one predicted time point based on the at least one adjustment coefficient. For more details on adjusting the electricity sales amount at the at least one predicted time point, see step 430, which will not be repeated here.
[0033] It should be understood that the system and its modules can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or a dedicated design hardware. Those skilled in the art will understand that the above-mentioned methods and systems can be implemented using computer executable instructions and / or included in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the system and its modules of this specification. Not only can the hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. be implemented, it can also be implemented with software such as executed by various types of processors, and it can also be implemented by a combination of the above-mentioned hardware circuits and software (for example, firmware).
[0034] It should be noted that the above description of the service performance information display system and its modules is for convenience of description only and does not limit this specification to the scope of the embodiments cited. It is understandable that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine the various modules, or form a subsystem to connect with other modules without deviating from this principle. For example, the service performance information determination module and the status information determination module can be combined into one module. Such variations are within the scope of this specification.
[0035] Figure 2 It is an exemplary flow chart of a method for predicting the amount of electricity sales according to some embodiments of this specification.
[0036] like Figure 2 As shown, the power sales amount prediction method 200 may include:
[0037] Step 210, obtaining electricity consumption history information corresponding to multiple historical time points.
[0038] Specifically, step 210 may be performed by an acquisition module.
[0039] A time point is a moment in time. A historical time point is a time point at which the amount of electricity sales is counted. The historical time point is a time point before (including) the current time point. For example, the current time point is 7:00 on February 28, 2021, and multiple historical time points may include 7:00 on February 28, 2021 and 7:00 on February 27, 2021.
[0040] In some embodiments, each historical time point may be separated by the same statistical period. For example, if one day is a statistical period, then each interval of one day is a historical time point. For another example, if one month is a statistical period, then each interval of one natural month is a historical time point.
[0041] It can be understood that each historical time point may correspond to a statistical period between the historical time point and the previous historical time point.
[0042] In some embodiments, the time interval of each statistical cycle is the same. For example, taking 1 day as a statistical cycle, the historical time point 7:00 on February 27, 2021 can correspond to the statistical cycle between 7:00 on February 26, 2021 and 7:00 on February 27, 2021, and the historical time point 7:00 on February 28, 2021 can correspond to the statistical cycle between 7:00 on February 27, 2021 and 7:00 on February 28, 2021, and the time interval of each statistical cycle is 24 hours.
[0043] In some embodiments, the time interval of each statistical cycle may also be different. For example, taking one natural month as one statistical cycle, the historical time point 7:00 on February 27, 2021 may correspond to the statistical cycle between 7:00 on January 27, 2021 and 7:00 on February 27, 2021, and the historical time point 7:00 on March 27, 2021 may correspond to the statistical cycle between 7:00 on February 27, 2021 and 7:00 on March 27, 2021, and the time intervals of the corresponding statistical cycles are 31 days and 28 days respectively.
[0044] In some embodiments, the time interval of each statistical period may be a preset default parameter, or may be determined based on a parameter input by a user.
[0045] In some embodiments, the time interval of each statistical cycle can also be automatically determined based on the time interval between the time point to be predicted and the current time point. For a detailed description of the time point to be predicted, please refer to the detailed description of step 220, which will not be repeated here.
[0046] For example, if the time interval between the to-be-predicted time point and the current time point is less than 30 days, one day is taken as a statistical period; if the time interval between the to-be-predicted time point and the current time point is greater than 30 days and less than three months, one natural month is taken as a statistical period; if the time interval between the to-be-predicted time point and the current time point is greater than three months, one quarter is taken as a statistical period.
[0047] For example, if the current time point is 7:00 on February 28, 2021, and the time point to be predicted is 7:00 on March 1, 2021, then one day is one statistical period. For another example, if the current time point is 7:00 on February 28, 2021, and the time point to be predicted is 7:00 on March 29, 2021, then one month is one statistical period.
[0048] Some embodiments of the present specification determine the time intervals of multiple statistical periods corresponding to multiple historical time points based on the time interval between the time point to be predicted and the current time point, and can obtain statistical data of time density that matches the time point to be predicted, thereby ensuring that the accuracy of predicting the electricity sales amount can be improved in different time spans.
[0049] In some embodiments, the multiple historical time points may be determined based on the number of historical time points N and the current time point. Specifically, the multiple historical time points may include the current time point and N-1 historical time points before the current time point.
[0050] Continuing with the above example, the number of historical time points is 5, and the current time point is 7:00 on February 28, 2021. The multiple historical time points may include 7:00 on February 28, 2021, 7:00 on February 27, 2021, 7:00 on February 26, 2021, 7:00 on February 25, 2021, and 7:00 on February 24, 2021, which correspond to the statistical period from 7:00 on February 27, 2021 to 7:00 on February 28, 2021, the statistical period from 7:00 on February 26, 2021 to 7:00 on February 27, 2021, ... and the statistical period from 7:00 on February 23, 2021 to 7:00 on February 24, 2021.
[0051] In some embodiments, the number of historical time points may be a preset default parameter, or may be determined based on parameters input by a user.
[0052] The power consumption history information corresponding to the multiple historical time points is the power consumption history information within the multiple statistical periods corresponding to the multiple historical time points.
[0053] The electricity usage history information can reflect the electricity usage information and payment information of all electricity users within a statistical period. In some embodiments, the electricity usage history information may include electricity usage characteristics, payment characteristics, and historical weather characteristics.
[0054] The electricity consumption characteristics are the characteristics of the electricity consumption statistical information of all electricity users in the statistical period corresponding to the historical time point.
[0055] The electricity user can pay the electricity bill through the online electricity sales platform. In some embodiments, the types of electricity users may include residential electricity users, industrial electricity users, commercial electricity users, and administrative electricity users.
[0056] In some embodiments, the electricity usage characteristics can be determined based on the total electricity usage and total electricity usage amount of different types of electricity users. Specifically, the acquisition module can obtain the total electricity usage and total electricity usage amount of different types of electricity users from the online electricity sales platform through the network, and obtain the electricity usage characteristics by statistics.
[0057] For example, during the statistical period from 7:00 a.m. on February 27, 2021 to 7:00 a.m. on February 28, 2021, which corresponds to the historical time point of 7:00 a.m. on February 28, 2021, the total electricity consumption of residential electricity users, industrial electricity users, commercial electricity users, and administrative electricity users was 300,000 kWh, 400,000 kWh, 200,000 kWh, and 100,000 kWh, respectively, and the total electricity consumption amount was 200,000 yuan, 800,000 yuan, 400,000 yuan, and 200,000 yuan, respectively. The electricity consumption characteristics corresponding to the historical time point of 7:00 a.m. on February 28, 2021 can be Similarly, the acquisition module can further acquire the power consumption characteristics X2, X3, X4 and X5 corresponding to the historical time points of 7:00 am on February 27, 2021, 7:00 am on February 26, 2021, 7:00 am on February 25, 2021 and 7:00 am on February 24, 2021.
[0058] The payment feature is a feature that characterizes the payment statistical information of all electricity users in the statistical period corresponding to the historical time point. In some embodiments, the payment features corresponding to multiple historical times can be determined based on the payment delay statistical information, payment method statistical information and payment frequency statistical information of all electricity users in the statistical period corresponding to multiple historical time points.
[0059] The payment delay statistical information can characterize the degree of payment delay of all electricity users within the statistical period corresponding to the historical time point. In some embodiments, the payment delay statistical information corresponding to each historical time point can be obtained by counting the delayed payment time of all electricity users within the corresponding statistical period. Specifically, the acquisition module can first obtain the time difference between the notification payment time and the corresponding actual payment time of each electricity user within each statistical period from the grid electricity sales platform through the network. For example, in the statistical period between 7:00 on February 27, 2021 and 7:00 on February 28, 2021 corresponding to the historical time point 7:00 on February 28, 2021, the notification payment time of electricity user A is 8:00 on February 27, 2021, and the corresponding actual payment time is 18:00 on February 27, 2021. Then, the payment delay time of electricity user A in the statistical period corresponding to the historical time point 7:00 on February 28, 2021 is 10 hours. Similarly, the acquisition module can obtain the payment delay time of power unit B, power unit C, ... in the statistical period corresponding to the historical time point of 7:00 on February 28, 2021, which is 8 hours, 3 hours, ... respectively. Among them, if the power unit does not receive a payment notice during the statistical period, it will not be included in the statistics. Further, the acquisition module can divide the payment delay time of the power unit into multiple time periods, and count the number of power units corresponding to each time period. For example, in the statistical period corresponding to the historical time point of 7:00 on February 28, 2021, the acquisition module counts that the number of power units with a payment delay time of 12 hours is 100, the number of power units with a payment delay time of 12-24 hours is 200, the number of power units with a payment delay time of 24-48 hours is 150, and the number of power units with a payment delay time of more than 48 hours is 50. Then the vector a1 = (100, 200, 150, 50) can be used to represent the payment delay statistical information corresponding to the historical time point.
[0060] The payment method statistical information can represent the payment methods of all electricity users in the statistical period corresponding to the historical time point. In some embodiments, based on the relationship between the amount payable and the actual amount paid, the payment methods may include pre-deposit, arrears and full payment, corresponding to the amount payable being greater than the actual amount paid, the amount payable being less than the actual amount paid and the amount payable being equal to the actual amount paid, respectively. The acquisition module can obtain the payment methods of all electricity users in the statistical period corresponding to the historical time point from the outlet sales platform through the network, and count the number of electricity users corresponding to each payment method in multiple statistical periods corresponding to multiple historical time points. For example, in the statistical period corresponding to the historical time point of 7:00 on February 28, 2021, the number of pre-deposited electricity users is 200, the number of arrears electricity users is 100, and the number of full payment electricity users is 200. Then the vector b1=(200, 100, 200) can be used to represent the payment method statistical information corresponding to the historical time point.
[0061] The payment frequency statistical information can characterize the payment frequency of all electricity users within the statistical period corresponding to the historical time point. Specifically, the acquisition module can obtain the payment frequency of all electricity users within the statistical period corresponding to the historical time point from the grid electricity sales platform, and then divide the number of payments by the electricity users within the statistical period into multiple ranges, and count the number of electricity users corresponding to each range. For example, in the statistical period corresponding to the historical time point of 7:00 on February 28, 2021, the acquisition module counted that the number of electricity users who paid 0 times was 120, the number of electricity users who paid 1 time was 280, and the number of electricity users who paid more than 1 time was 200. The payment frequency statistical information corresponding to the historical time point can be represented by the vector c1=(120, 280, 200).
[0062] In some embodiments, the acquisition module can obtain the payment features corresponding to each historical time point by processing the payment delay statistics, payment method statistics, and payment frequency statistics corresponding to each historical time point. The processing methods may include but are not limited to splicing, dimensionality reduction, and linear transformation.
[0063] For example, the acquisition module can directly splice the payment delay statistical information a1, payment method statistical information b1 and payment frequency statistical information c1 corresponding to the historical time point 7:00 on February 28, 2021, to obtain the payment feature Y1=(100, 200, 150, 50, 200, 100, 200, 120, 280, 200) corresponding to the historical time point. Similarly, the acquisition module can further obtain the payment features Y2, Y3, Y4 and Y5 corresponding to the historical time points 7:00 on February 27, 2021, 7:00 on February 26, 2021, 7:00 on February 25, 2021 and 7:00 on February 24, 2021.
[0064] The historical weather feature is a feature that characterizes the historical weather information corresponding to the historical time point. In some embodiments, the historical weather information may include but is not limited to at least one of historical temperature information, historical wind information, and historical precipitation information.
[0065] Historical temperature information is information that characterizes the temperature in the statistical period corresponding to the historical time point. For example, the highest temperature within 24 hours is 36°C. Another example is the average temperature within 24 hours is 28°C.
[0066] Historical wind force information is information that characterizes the wind force in the statistical period corresponding to the historical time point. For example, the average wind force level within 24 hours is level 2. Another example is the average wind speed within 24 hours is 11 km / h.
[0067] Historical precipitation information is information that characterizes the average precipitation in the statistical period corresponding to the historical time point. For example, the precipitation in 8 hours is 50 mm. Another example is the precipitation in 24 hours is 70 mm.
[0068] In some embodiments, the acquisition module may acquire historical weather information within a statistical period corresponding to a historical time point from a point-of-sale platform, a meteorological statistics platform, or other storage device.
[0069] Furthermore, the determination module may determine the historical weather characteristics based on at least one historical weather information. For example, in a hydroelectric power generation area, the determination module may determine the historical weather characteristics corresponding to the historical time point based on the historical temperature information and the historical precipitation information. For another example, in a hydro-wind hybrid power generation area, the determination module may determine the historical weather characteristics corresponding to the historical time point based on the historical temperature information, the historical wind information, and the historical precipitation information.
[0070] In some embodiments, the determination module may also determine historical weather characteristics based on weather information such as historical light intensity, historical geothermal temperature, historical tidal water level, etc., which will not be elaborated here.
[0071] Exemplarily, the average temperature of the statistical period corresponding to the historical time point 7:00 on February 27, 2021 is 30°C, the wind force is level 3, and the 24-hour precipitation is 50 mm, then the historical weather feature corresponding to the historical time point can be Z1=(30, 3, 50). Similarly, the acquisition module can further acquire the historical weather features Z2, Z3, Z4 and Z5 corresponding to the historical time points 7:00 on February 27, 2021, 7:00 on February 26, 2021, 7:00 on February 25, 2021 and 7:00 on February 24, 2021.
[0072] Step 220: inputting the electricity consumption history information corresponding to the multiple historical time points into a prediction model to determine the electricity sales amount at least one time point to be predicted.
[0073] Specifically, step 220 may be performed by a determination module.
[0074] Figure 3 is a schematic diagram of a prediction model according to some embodiments of this specification. Figure 3 As shown, the prediction model includes a feature construction layer 212 and a prediction layer 221 .
[0075] The feature construction layer 212 may obtain the power sales features corresponding to the multiple historical time points based on the power consumption history information corresponding to the multiple historical time points. The power sales features are vectors that characterize the temporal regularity of power sales.
[0076] like Figure 3As shown, the power consumption history information 211 input by the feature construction layer 212 includes power consumption characteristics 211-1, payment characteristics 211-2 and historical weather characteristics 211-3 at multiple historical time points. For example, the input of the feature construction layer may include the power consumption characteristics corresponding to the above five historical time points: X1, X2, ...X5; payment characteristics Y1, Y2, ...Y5; and historical weather characteristics: Z1, Z2, ...Z5, and the output may include power sales characteristics 213.
[0077] In some embodiments, the input to the feature construction layer 212 also includes power unit features 211 - 4 .
[0078] The power consumption unit characteristics are characteristics that characterize the power consumption unit statistical information. In some embodiments, the power consumption unit statistical information can be obtained based on unit type statistical information, unit region statistical information, and unit power generation type statistical information.
[0079] As mentioned above, the types of power units may include residential power units, industrial power units, commercial power units, and administrative power units. Unit type statistics may include the number of power units of each unit type. In some embodiments, unit type statistics may be represented by a vector. For example, if the numbers of residential power units, industrial power units, commercial power units, and administrative power units are 200, 100, 100, and 100, respectively, then the unit type statistics may be represented by a vector (200, 100, 100, 100).
[0080] The unit region is the region where the electricity-consuming unit is located. The unit region can be divided into hot areas with high electricity consumption in summer, cold areas with high electricity consumption in winter, and normal temperature areas with relatively average electricity consumption based on the electricity consumption habits of the electricity-consuming units. The unit region statistical information may include the number of electricity-consuming units in each unit region. In some embodiments, the unit region statistical information may be represented by a vector. For example, if the number of electricity-consuming units in hot areas, cold areas, and normal temperature areas is 200, 200, and 100, respectively, the unit region statistical information may be represented by a vector (200, 200, 100).
[0081] The unit generation type is the generation method of the region where the power user is located. The generation method may include wind power generation, hydroelectric power generation, solar power generation, tidal power generation, etc. The unit generation type statistical information may include the number of power users corresponding to each generation method. In some embodiments, the unit generation type statistical information may be represented by a vector. For example, if the number of power users in the regions of wind power generation, hydroelectric power generation, and solar power generation are 100, 300, and 100, respectively, the unit generation type statistical information may be represented by a vector (100, 300, 100).
[0082] In some embodiments, the power unit characteristics may be generated based on a power unit characteristics model.
[0083] Specifically, the input of the power unit characteristic model may include unit type statistical information, unit regional statistical information and unit power generation type statistical information, and the output is the power unit characteristic. For example, the input unit type statistical information: vector (200, 100, 100, 100), unit regional statistical information: (200, 200, 100) and unit power generation type statistical information: vector (100, 300, 100) are used, and the output is the power unit characteristic Z.
[0084] In some embodiments, the electricity consumption unit feature model may include but is not limited to a convolutional neural network (CNN) model and a deep neural network (DNN) model.
[0085] In some embodiments, the power consumption unit characteristic model can be jointly trained and acquired with the prediction model. A detailed description of the power consumption unit characteristic model training can be found below and will not be repeated here.
[0086] In some embodiments, the input of the feature construction layer 212 also includes a weight coefficient vector 211 - 5 . The weight coefficient vector represents the degree of influence of weather characteristics on the amount of electricity sales.
[0087] In some embodiments, the weight coefficient vector can be determined based on the weight of the temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate. The weight of the temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate can respectively represent the degree of influence of the temperature on the electricity sales amount, the degree of influence of the precipitation on the electricity sales amount, and the degree of influence of the wind on the electricity sales amount.
[0088] For a detailed description of the weights of the temperature deviation rate, precipitation deviation rate, and wind deviation rate, see Figure 4 And the related description will not be repeated here.
[0089] Exemplarily, the weight of the temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate are 0.3, 0.5, and 0.2, respectively, and the weight coefficient vector α=(0.3, 0.5, 0.2).
[0090] Specifically, the feature construction layer can construct features for the input power consumption features, payment features, historical weather features, power unit features and weight coefficient vectors, and output power sales features. In some embodiments, the feature construction layer can be a machine learning model. Preferably, the feature construction layer can be a machine learning model that can process time series data. In some embodiments, the feature construction layer can include but is not limited to a recurrent neural network (RNN) model, a long short-term memory network (LSTM) model, and a gated recurrent unit (GRU) model. Exemplarily, the feature construction layer can be based on power consumption features X1, X2, ...X5; payment features Y1, Y2, ...Y5; historical weather features: Z1, Z2, ...Z5, power unit features Z and weight coefficient vector α, and output power sales features D.
[0091] Some embodiments of this specification construct power sales features based on power consumption features, so that the power sales features can reflect the power consumption and power consumption amount patterns of power users. In addition, the time when the power user generates the power consumption amount and the time when the grid power sales platform actually collects the power consumption amount from the power user are not synchronized. Some embodiments of this specification construct power sales features based on payment features, taking into account the different payment methods and payment habits of different power users, and avoiding the impact of early payment, delayed payment, installment payment, etc. of power users on the predicted power sales amount.
[0092] Some embodiments of the present specification also take into account the impact of weather information on the amount of electricity sales (for example, electricity consumption in extremely cold and / or extremely hot temperatures is higher than that in normal temperatures; for example, the amount of electricity consumption in the flood season with high rainfall is lower than that in the dry season with low rainfall), thereby constructing electricity sales characteristics based on historical weather characteristics, reducing the interference of different historical weather information on historical electricity sales characteristics, and improving the accuracy of predicting the amount of electricity sales.
[0093] The prediction layer 221 may determine the electricity sales amount at at least one time point to be predicted.
[0094] The time point to be predicted is the time point at which the amount of electricity sales needs to be predicted. The time point to be predicted is the time point after the current time point.
[0095] In some embodiments, the time point to be predicted may be separated from the current time point by one or more complete statistical cycles. For example, the current time point is 7:00 on February 28, 2021, the statistical cycle is 1 month, and at least one time point to be predicted may include 7:00 on April 28, 2021.
[0096] In some embodiments, the time point to be predicted may also be separated from the current time point by one or more incomplete statistical cycles. For example, the current time point is 7:00 on February 28, 2021, the statistical cycle is 1 day, and at least one time point to be predicted may include 8:00 on March 3, 2021.
[0097] like Figure 3 As shown, the input of the prediction layer 221 includes power sales features 213 corresponding to multiple historical time points and at least one time point to be predicted 222, and the output is the power sales amount 223 of at least one time point to be predicted. Specifically, for each time point to be predicted, the prediction layer can map the input power sales features and the time point to be predicted into numerical values or probabilities, and then obtain the corresponding power sales amount based on the numerical values or probabilities.
[0098] For example, the determination module can output the corresponding electricity sales amount of 2 million yuan based on the electricity sales feature D corresponding to the predicted time point of 7:00 on April 28, 2021.
[0099] In some embodiments, the prediction layer may include but is not limited to a convolutional neural network (CNN) model, an ELMo (Embedding from Language Models) model, a GPT (Generative Pre-Traxining) model, and a BERT (Bidirectional Encoder Representation from Transformers) model, etc.
[0100] In some embodiments, the power unit characteristic model can be obtained by joint training with the prediction model. The training module can train the initial power unit characteristic model and the initial prediction model based on a large number of labeled training samples. Specifically, the labeled training samples are input into the initial power unit characteristic model, and the parameters of the initial power unit characteristic model and the initial prediction model are updated through training until the trained intermediate power unit characteristic model and the intermediate prediction model meet the preset conditions, and the trained power unit characteristic model and the prediction model are obtained, wherein the preset conditions can be that the loss function is less than a threshold, converges, or the training cycle reaches a threshold.
[0101] The training samples include sample electricity consumption unit historical information and electricity consumption historical information at multiple sample historical time points, and the identification of the training samples is the actual electricity sales amount corresponding to the sample time points corresponding to the multiple sample historical time points. The sample time point is a time point after the multiple sample historical time points.
[0102] As mentioned above, weather information may affect the amount of electricity sold. In some embodiments, the electricity sales amount prediction system may also adjust the electricity sales amount based on at least one piece of weather forecast information corresponding to at least one time point to be predicted.
[0103] Figure 4 This is an exemplary flow chart of adjusting the electricity sales amount according to some embodiments of this specification.
[0104] like Figure 4 As shown, the method 400 for adjusting the electricity sales amount may include:
[0105] Step 410: Obtain at least one piece of weather forecast information corresponding to the at least one time point to be predicted.
[0106] Weather forecast information is weather information corresponding to a time point to be predicted. Similar to historical weather information, weather forecast information may include but is not limited to at least one of temperature information, precipitation information, and wind information. For detailed descriptions of temperature information, precipitation information, and wind information, please refer to historical temperature information, historical wind information, and historical precipitation information, which will not be repeated here.
[0107] In some embodiments, the first module can obtain weather forecast information corresponding to the predicted time point from a weather forecast platform or other storage device. For example, the first module can obtain the temperature information, precipitation information, and wind force information corresponding to the predicted time point of 7:00 on April 28, 2021 from the weather forecast platform, which are 28°C, 0mm of precipitation in 24 hours, and wind force level 1.
[0108] In some embodiments, weather forecast information may also include light intensity, geothermal temperature, tidal level, etc.
[0109] Step 420: Determine at least one adjustment coefficient based on the at least one piece of weather forecast information.
[0110] The adjustment coefficient is a parameter for adjusting the electricity sales amount corresponding to at least one forecast time point, and can adjust the impact of weather on the forecast electricity sales amount.
[0111] In some embodiments, the second module may determine the adjustment factor based on the air temperature deviation rate, the precipitation deviation rate, and the wind deviation rate.
[0112] Figure 5 It is a schematic diagram of adjusting the electricity sales amount according to some embodiments of this specification.
[0113] like Figure 5As shown, the second module can determine the adjustment coefficient 570 based on the weather forecast information 510 (including the temperature deviation rate 511, the precipitation deviation rate 512, and the wind deviation rate 513) and the corresponding weights of the temperature deviation rate, the precipitation deviation rate, and the wind deviation rate.
[0114] The reference temperature is a temperature reference value that does not affect the amount of electricity sold. In some embodiments, the reference temperature reference value may be a preset default parameter or may be determined based on a parameter input by a user.
[0115] The temperature deviation rate is a parameter that measures the deviation between the temperature information corresponding to the predicted time point and the reference temperature. In some embodiments, the temperature deviation rate can be determined based on the deviation ratio between the temperature information corresponding to the predicted time point and the reference temperature information. For example, the temperature deviation rate between the temperature information 28°C corresponding to the predicted time point 7:00 on April 28, 2021 and the reference temperature information 25°C is (28-25) / 25=0.12.
[0116] The reference precipitation amount is a precipitation amount reference value that does not affect the amount of electricity sold. In some embodiments, the reference precipitation amount reference value may be a preset default parameter, or may be determined based on a parameter input by a user.
[0117] The precipitation deviation rate is a parameter that measures the deviation between the precipitation information corresponding to the time point to be predicted and the reference precipitation. In some embodiments, the precipitation deviation rate can be determined based on the deviation ratio between the precipitation corresponding to the time point to be predicted and the reference precipitation information.
[0118] For example, the precipitation deviation rate between the precipitation information 0 mm and the benchmark precipitation information 50 mm corresponding to the predicted time point 7:00 on April 28, 2021 is (0-50) / 50=-1.
[0119] The reference wind force is a wind force reference value that does not affect the amount of electricity sales. In some embodiments, the reference wind force reference value may be a preset default parameter, or may be determined based on a parameter input by a user.
[0120] The wind force deviation rate is a parameter that measures the deviation between the wind force information corresponding to the time point to be predicted and the reference wind force. In some embodiments, the wind force deviation rate can be determined based on the deviation ratio between the wind force information corresponding to the time point to be predicted and the reference wind force information.
[0121] For example, the wind force deviation rate between the wind force information level 1 and the benchmark wind force information level 1 corresponding to the predicted time point 7:00 on April 28, 2021 is (1-1) / 1=0.
[0122] It can be understood that the greater the temperature deviation rate, precipitation deviation rate and wind deviation rate, the greater the impact of weather forecast information on the electricity sales amount.
[0123] As mentioned above, the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate can respectively represent the degree of influence of temperature on the electricity sales amount, the degree of influence of precipitation on the electricity sales amount and the degree of influence of wind on the electricity sales amount.
[0124] In some embodiments, the second module may determine the weight of the air temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate based on a weight calculation model.
[0125] like Figure 5 As shown, the input of the weight calculation model 560 may include the electricity unit characteristics 550 .
[0126] As mentioned above, the power unit characteristics are characteristics that characterize the statistical information of the power unit, including unit type statistical information, unit region statistical information, and unit power generation type statistical information. For a detailed description of the power unit characteristics, see step 220, which will not be repeated here.
[0127] Specifically, the weight calculation model can output the weight of the temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate based on the input unit type statistical information, unit regional statistical information, and unit power generation type statistical information. For example, the weight calculation model can output the weight W1 of the temperature deviation rate, the weight W2 of the precipitation deviation rate, and the weight W3 of the wind deviation rate based on the input vectors (200, 100, 100, 100), (200, 200, 100), and (100, 300, 100), which are 0.3, 0.2, and 0.5, respectively.
[0128] In some embodiments, the weight calculation model may include but is not limited to a support vector machine model, a logistic regression model, a naive Bayes classification model, a Gaussian distribution Bayes classification model, a decision tree model, a random forest model, a KNN classification model, and a neural network model.
[0129] In some embodiments, the second module may determine the adjustment coefficient based on formula (1):
[0130] R=a*W1+b*W2+c*W3 (1)
[0131] Among them, R represents the adjustment coefficient, a, b and c represent the temperature deviation rate, the precipitation deviation rate and the wind deviation rate respectively, and W1, W2 and W3 represent the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate respectively.
[0132] Exemplarily, based on the temperature deviation rate of 0.12, precipitation deviation rate of -1 and hydraulic deviation rate of 0.5 corresponding to the predicted time point of 7:00 on April 28, 2021, as well as the weight of the temperature deviation rate of 0.3, the weight of the precipitation deviation rate of -1 and the weight of the wind deviation rate of 0.5, the adjustment coefficient R = 0.12*0.3+-1*0.2+0*0.5 = 0.16 can be obtained.
[0133] In some embodiments, the adjustment coefficient can also be obtained based on weather forecast information through an adjustment coefficient model. Specifically, the adjustment coefficient model can fuse the input weather forecast information into a vector, map the vector into a value or probability, and then obtain the corresponding adjustment coefficient based on the value or probability.
[0134] Exemplarily, the adjustment coefficient model can output an adjustment coefficient of 0.16 based on the temperature information, precipitation information and wind force information corresponding to the input prediction time point of 7:00 on April 28, 2021, which are 28°C, 24-hour precipitation 0mm and wind force level 1.
[0135] In some embodiments, the adjustment coefficient model may include but is not limited to a support vector machine model, a logistic regression model, a naive Bayes classification model, a Gaussian distribution Bayes classification model, a decision tree model, a random forest model, a KNN classification model, and a neural network model.
[0136] In some embodiments, the adjustment coefficient model can be obtained by training based on historical electricity usage information, wherein the historical electricity usage information includes historical weather characteristics.
[0137] In some embodiments, the training module may use the historical weather characteristics corresponding to the multiple sample historical time points as training samples. The identification of the training sample may be a historical adjustment coefficient. In some embodiments, the historical adjustment coefficient may be determined based on the predicted electricity sales amount and the actual historical electricity sales amount corresponding to the sample time point. The sample time point is a time point after the multiple sample historical time points.
[0138] Specifically, the training samples with the mark are input into the initial adjustment coefficient model, and the parameters of the initial adjustment coefficient model are updated through training. When the trained model meets the preset conditions, the training ends and the trained adjustment coefficient model is obtained.
[0139] In some embodiments, the adjustment coefficient model can also be jointly trained with the feature construction layer and the prediction layer. A detailed description of the joint training can be found in step 220, which will not be repeated here.
[0140] Some embodiments of the present application determine an adjustment coefficient based on predicted weather information, so that the adjusted electricity sales amount is not affected by the weather, thereby improving the prediction accuracy.
[0141] Step 430: adjusting the electricity sales amount at the at least one predicted time point based on the at least one adjustment coefficient.
[0142] The adjustment may include at least one of addition, subtraction, multiplication, division, etc. For example, the third module may adjust the electricity sales amount of 2 million yuan corresponding to the predicted time point of 7:00 on April 28, 2021 based on the adjustment coefficient 0.16, and obtain the adjusted electricity sales amount of 200×(1-0.16)=1.88 million yuan.
[0143] The embodiment of the present specification also provides a computer-readable storage medium. The storage medium stores computer instructions, and when a computer reads the computer instructions in the storage medium, the computer implements the aforementioned power sales amount prediction method.
[0144] An embodiment of the present specification also provides a service fulfillment information display device, including a storage medium and a processor, wherein the storage medium is used to store computer instructions, and the processor is used to execute at least a portion of the computer instructions to implement the aforementioned electricity sales amount prediction method.
[0145] The beneficial effects that may be brought about by the embodiments of this specification include but are not limited to: (1) constructing electricity sales features based on electricity consumption features, so that the electricity sales features can reflect the time pattern of electricity consumption and electricity consumption amount of electricity users; at the same time, constructing electricity sales features based on payment features, taking into account the different payment methods and payment habits of different electricity users, avoiding the influence of early payment, delayed payment, installment payment, etc. of electricity users on the predicted electricity sales amount, and improving the accuracy of prediction; (2) considering the influence of weather information on the electricity sales amount (for example, the electricity consumption in extremely cold and / or extremely hot temperatures is higher than that in normal temperatures; for example, the electricity consumption in the flood season with high rainfall is lower than that in the dry season with low rainfall), on the one hand, constructing electricity sales features based on historical weather features, reducing the interference of different historical weather information on historical electricity sales features; on the other hand, determining the adjustment coefficient based on the predicted weather information, so that the adjusted electricity sales amount is not affected by the weather; (3) determining the time interval of each statistical period based on the time span between the time point to be predicted and the current time point, which can avoid the inapplicability of historical data, thereby ensuring that the accuracy of the predicted electricity sales amount can be improved in different time spans.
[0146] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are suggested in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.
[0147] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.
[0148] In addition, it will be understood by those skilled in the art that various aspects of this specification may be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of this specification may be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of this specification may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.
[0149] A computer storage medium may include a propagated data signal containing computer program code, for example, in baseband or as part of a carrier wave. The propagated signal may be in a variety of forms, including electromagnetic, optical, etc., or a suitable combination. A computer storage medium may be any computer-readable medium other than a computer-readable storage medium, which can be connected to an instruction execution system, device or apparatus to communicate, propagate or transmit the program for use. The program code on the computer storage medium may be transmitted via any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the above media.
[0150] The computer program code required for the operation of each part of this specification can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., conventional procedural programming languages such as C language, Visual Basic, Fortran2003, Perl, COBOL2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages, etc. The program code can be run entirely on the user's computer, or run on the user's computer as an independent software package, or run partially on the user's computer and partially on a remote computer, or run entirely on a remote computer or processing device. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).
[0151] In addition, unless explicitly stated in the claims, the order of the processing elements and sequences described in this specification, the use of alphanumeric characters, or the use of other names are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some invention embodiments that are currently considered useful through various examples, it should be understood that such details are only for illustrative purposes, and the attached claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the essence and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing processing device or mobile device.
[0152] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.
[0153] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values is as accurate as possible within the feasible range.
[0154] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, documents, etc., cited in this specification are hereby incorporated by reference in their entirety. Except for application history documents that are inconsistent with or conflicting with the contents of this specification, documents that limit the broadest scope of the claims of this specification (currently or later attached to this specification) are also excluded. It should be noted that if the descriptions, definitions, and / or use of terms in the materials attached to this specification are inconsistent or conflicting with the contents described in this specification, the descriptions, definitions, and / or use of terms in this specification shall prevail.
[0155] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.
Claims
1. A method for predicting the amount of electricity sales, characterized in that: The method comprises: Obtaining electricity consumption unit characteristics and electricity consumption characteristics, payment characteristics and historical weather characteristics corresponding to multiple historical time points; the electricity consumption unit characteristics are characteristics that characterize statistical information of the electricity consumption unit, and the electricity consumption unit statistical information is obtained based on unit type statistical information, unit regional statistical information and unit power generation type statistical information; Input the power consumption unit characteristics and the power consumption characteristics, payment characteristics and historical weather characteristics corresponding to the multiple historical time points into the prediction model to determine the power sales amount at at least one time point to be predicted; wherein the prediction model includes a feature construction layer and a prediction layer; the feature construction layer obtains the power sales characteristics corresponding to the multiple historical time points based on the power consumption unit characteristics and the power consumption characteristics, payment characteristics and historical weather characteristics corresponding to the multiple historical time points; the prediction layer determines the power sales amount at at least one time point to be predicted based on the power sales characteristics; Acquire at least one piece of weather forecast information corresponding to the at least one time point to be predicted; the weather forecast information includes temperature information, precipitation information and wind force information; Determine the temperature deviation rate, the precipitation deviation rate and the wind deviation rate based on the at least one weather forecast information; determine the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate based on the characteristics of the electricity unit using a weight calculation model; determine at least one adjustment coefficient based on the temperature deviation rate, the precipitation deviation rate, the wind deviation rate, the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate; The electricity sales amount at the at least one to-be-predicted time point is adjusted based on the at least one adjustment coefficient.
2. The method according to claim 1, characterized in that The determining of the temperature deviation rate, the precipitation deviation rate and the wind deviation rate based on the at least one piece of weather forecast information comprises: The temperature deviation rate is determined based on the temperature information and the reference temperature, the precipitation deviation rate is determined based on the precipitation information and the reference precipitation, and the wind force deviation rate is determined based on the wind force information and the reference wind force; wherein the reference temperature is a temperature reference value that does not affect the electricity sales amount, the reference precipitation is a precipitation reference value that does not affect the electricity sales amount, and the reference wind force is a wind force reference value that does not affect the electricity sales amount.
3. The method according to claim 1, characterized in that The input of the feature construction layer also includes a weight coefficient vector, which is determined based on the weight of the air temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate.
4. The method according to claim 1, characterized in that The electricity consumption unit characteristics are obtained based on an electricity consumption unit characteristic model.
5. A power sales amount prediction system, characterized in that: The system comprises: An acquisition module, used to acquire the characteristics of the power consumption unit and the power consumption characteristics, payment characteristics and historical weather characteristics corresponding to multiple historical time points; the characteristics of the power consumption unit are characteristics that characterize the statistical information of the power consumption unit, and the statistical information of the power consumption unit is acquired based on the statistical information of the unit type, the statistical information of the unit region and the statistical information of the unit power generation type; A determination module, configured to input the power consumption unit characteristics and the power consumption characteristics, payment characteristics and historical weather characteristics corresponding to the multiple historical time points into a prediction model, and determine the power sales amount at at least one time point to be predicted; wherein the prediction model comprises a feature construction layer and a prediction layer; the feature construction layer obtains the power sales characteristics corresponding to the multiple historical time points based on the power consumption unit characteristics and the power consumption characteristics, payment characteristics and historical weather characteristics corresponding to the multiple historical time points; the prediction layer determines the power sales amount at at least one time point to be predicted based on the power sales characteristics; The first module is used to obtain at least one piece of weather forecast information corresponding to the at least one time point to be predicted; the weather forecast information includes temperature information, precipitation information and wind force information; The second module is used to determine the temperature deviation rate, the precipitation deviation rate and the wind deviation rate based on the at least one weather forecast information; determine the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate based on the characteristics of the electricity unit by using a weight calculation model; determine at least one adjustment coefficient based on the temperature deviation rate, the precipitation deviation rate, the wind deviation rate, the weight of the temperature deviation rate, the weight of the precipitation deviation rate and the weight of the wind deviation rate; The third module is used to adjust the electricity sales amount at the at least one predicted time point based on the at least one adjustment coefficient.
6. The system according to claim 5, characterized in that The temperature deviation rate is determined based on the temperature information and a reference temperature, the precipitation deviation rate is determined based on the precipitation information and a reference precipitation amount, and the wind force deviation rate is determined based on the wind force information and a reference wind force; Among them, the reference temperature is a temperature reference value that does not affect the electricity sales amount, the reference precipitation is a precipitation reference value that does not affect the electricity sales amount, and the reference wind force is a wind force reference value that does not affect the electricity sales amount.
7. The system according to claim 5, characterized in that The input of the feature construction layer also includes a weight coefficient vector, which is determined based on the weight of the air temperature deviation rate, the weight of the precipitation deviation rate, and the weight of the wind deviation rate.
8. The system according to claim 5, characterized in that The electricity consumption unit characteristics are obtained based on an electricity consumption unit characteristic model.
9. An electricity sales amount prediction device, comprising at least one storage medium and at least one processor, wherein the at least one storage medium is used to store computer instructions; and the at least one processor is used to execute the computer instructions to implement the method according to any one of claims 1 to 4.
10. A computer-readable storage medium, characterized in that: The storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 4 is implemented.
