Electric power spot transaction strategy determination method and device, equipment, medium and product

Abstract

The invention discloses an electric power spot transaction strategy determination method and device, equipment, a medium and a product. Relates to the technical field of data processing. The method comprises the steps of obtaining historical electric power spot transaction data of a target power grid node; according to the historical electric power spot transaction data, the first prediction model and the second prediction model, determining a predicted electric power spot transaction price of the target power grid node; according to historical power spot transaction data and a third prediction model, determining a predicted power spot transaction price relationship of the target power grid node; and determining an electric power spot transaction strategy corresponding to the target power grid node according to the predicted electric power spot transaction price and the predicted electric power spot transaction price relationship. The technical problem of low accuracy of the determined electric power spot transaction strategy caused by low prediction accuracy when the electric power spot transaction price and the price relationship are predicted based on the similar day algorithm in the related technology is solved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular, to a method, device, equipment, medium and product for determining a power spot trading strategy. Background Art

[0002] In the increasingly complex and changing environment of the electricity spot market, the ability to predict node electricity prices and price relationships is directly related to the profitability of power companies, and its accurate prediction helps to determine more accurate trading strategies. At present, the existing technology mainly predicts electricity spot trading prices and price relationships based on classic time series prediction models such as similar day algorithms and autoregressive moving average models. Most of these methods are limited to the mining of linear relationships and cannot effectively capture the nonlinear characteristics of electricity price fluctuations and the immediate impact of external factors on electricity prices, resulting in low accuracy in the prediction of electricity prices and price relationships, and low accuracy in the determined electricity spot trading strategies.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present application provide a method, device, equipment, medium and product for determining an electricity spot trading strategy, so as to at least solve the technical problem in the related art that the electricity spot trading price and price relationship are predicted based on the similar day algorithm, and the prediction accuracy is low, resulting in the low accuracy of the determined electricity spot trading strategy.

[0005] According to one aspect of an embodiment of the present application, a method for determining an electricity spot trading strategy is provided, including: obtaining historical electricity spot trading data of a target power grid node; determining a predicted electricity spot trading price of the target power grid node based on the historical electricity spot trading data, a first prediction model, and a second prediction model, wherein the first prediction model is used to determine a valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine a non-valley electricity price in the predicted electricity spot trading price; determining a predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; determining an electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

[0006] Furthermore, based on the historical electricity spot trading data, the first prediction model and the second prediction model, determining the predicted electricity spot trading price of the target power grid node includes: classifying the historical electricity spot trading data to obtain first-category electricity price data and second-category electricity price data; inputting the first-category electricity price data into the first prediction model to output the predicted valley electricity price; inputting the second-category electricity price data into the second prediction model to output the predicted non-valley electricity price; and determining the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0007] Furthermore, based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship, determining the electricity spot trading strategy corresponding to the target power grid node includes: using the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is greater than the predicted electricity spot trading price, then reducing the electricity spot trading volume.

[0008] Furthermore, the method also includes: if the predicted electricity spot transaction price relationship indicates that the actual electricity spot transaction price is less than the predicted electricity spot transaction price, then increasing the electricity spot transaction volume.

[0009] Furthermore, before obtaining the historical electricity spot trading data of the target power grid node, the method also includes: obtaining first training data, wherein the first training data includes at least first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that has an impact on the prediction of valley electricity prices; performing feature extraction on the covariate data corresponding to the first-category sample electricity price data to obtain multiple first data features; and training a gradient boosting tree algorithm based on the first-category sample electricity price data and the multiple first data features to obtain a first prediction model.

[0010] Furthermore, before obtaining the historical electricity spot trading data of the target power grid node, the method also includes: obtaining second training data, wherein the second training data at least includes second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; performing feature analysis on the covariate data corresponding to the second-category sample electricity price data based on Fourier decomposition and wavelet analysis to determine multiple second data features; and training the gradient boosting tree algorithm based on the second-category sample electricity price data and the multiple second data features to obtain a second prediction model.

[0011] Furthermore, the data characteristics at least include characteristics of the load rate of the thermal power units used to characterize the thermal power supply capacity.

[0012] Furthermore, before obtaining the historical electricity spot trading data of the target power grid node, the method also includes: obtaining third training data, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationship; constructing lag features corresponding to the covariate data corresponding to the sample price relationships, and training the gradient boosting tree algorithm based on the third training data and the lag features to obtain a third prediction model.

[0013] According to another aspect of an embodiment of the present application, a method for determining an electricity spot trading strategy is also provided, including: obtaining historical electricity spot trading data of a target power grid node uploaded by a client; determining a predicted electricity spot trading price of the target power grid node in a cloud server based on the historical electricity spot trading data, a first prediction model, and a second prediction model, wherein the first prediction model is used to determine a valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine a non-valley electricity price in the predicted electricity spot trading price; determining a predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the relationship between the predicted electricity spot trading price and the actual electricity spot trading price; determining an electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship; and feeding back the electricity spot trading strategy corresponding to the target power grid node to the client.

[0014] According to another aspect of an embodiment of the present application, a device for determining an electricity spot trading strategy is also provided, including: a first acquisition unit, used to acquire historical electricity spot trading data of a target power grid node; a first determination unit, used to determine a predicted electricity spot trading price of the target power grid node based on the historical electricity spot trading data, a first prediction model and a second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; a second determination unit, used to determine the predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; a third determination unit, used to determine the electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

[0015] Furthermore, the first determination unit includes: a classification subunit, used to classify historical electricity spot trading data to obtain first-category electricity price data and second-category electricity price data; a first processing subunit, used to input the first-category electricity price data into a first prediction model, and output a predicted valley electricity price; a second processing subunit, used to input the second-category electricity price data into a second prediction model, and output a predicted non-valley electricity price; the first determination subunit, used to determine the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0016] Furthermore, the third determination unit includes: a second determination subunit, used to use the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; a third determination subunit, used to reduce the electricity spot trading volume if the predicted electricity spot trading price relationship represents that the actual electricity spot trading price is greater than the predicted electricity spot trading price.

[0017] Furthermore, the device also includes: a fourth determination unit, which is used to increase the electricity spot trading volume if the predicted electricity spot trading price relationship represents that the actual electricity spot trading price is less than the predicted electricity spot trading price.

[0018] Furthermore, the device also includes: a second acquisition unit, used to acquire first training data before acquiring historical electricity spot trading data of the target power grid node, wherein the first training data at least includes first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that has an impact on the prediction of valley electricity prices; a first processing unit, used to perform feature extraction on the covariate data corresponding to the first-category sample electricity price data to obtain multiple first data features; and a second processing unit, used to train a gradient boosting tree algorithm based on the first-category sample electricity price data and multiple first data features to obtain a first prediction model.

[0019] Furthermore, the device also includes: a third acquisition unit, used to acquire second training data before acquiring the historical electricity spot trading data of the target power grid node, wherein the second training data at least includes second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; a third processing unit, used to perform feature analysis on the covariate data corresponding to the second-category sample electricity price data based on Fourier decomposition and wavelet analysis to determine multiple second data features; a fourth processing unit, used to train the gradient boosting tree algorithm based on the second-category sample electricity price data and multiple second data features to obtain a second prediction model.

[0020] Furthermore, the data characteristics at least include characteristics of the load rate of the thermal power units used to characterize the thermal power supply capacity.

[0021] Furthermore, the device also includes: a fourth acquisition unit, used to acquire third training data before acquiring the historical electricity spot trading data of the target power grid node, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationship; a fifth processing unit, used to construct lag features corresponding to the covariate data corresponding to the sample price relationships, and train the gradient boosting tree algorithm based on the third training data and the lag features to obtain a third prediction model.

[0022] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein when the program is running, any one of the above methods for determining a spot electricity trading strategy is executed.

[0023] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the storage medium stores a program, wherein when the program is running, the device where the storage medium is located is controlled to execute any one of the above-mentioned methods for determining a spot electricity trading strategy.

[0024] According to another aspect of an embodiment of the present invention, there is also provided a computer program product, including a computer program or instructions, which implements any of the above methods for determining a spot electricity trading strategy when executed by a processor.

[0025] In an embodiment of the present application, historical electricity spot trading data of a target power grid node is obtained; a predicted electricity spot trading price of the target power grid node is determined based on the historical electricity spot trading data, a first prediction model, and a second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; a predicted electricity spot trading price relationship of the target power grid node is determined based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the relationship between the predicted electricity spot trading price and the actual electricity spot trading price; a method of determining the electricity spot trading strategy corresponding to the target power grid node is used based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship, and the electricity spot trading price and its relationship with the actual electricity spot trading price are effectively predicted by combining historical data with the prediction model. The relationship between the peak and the international prices provides a basis for formulating better trading strategies for power grid nodes. Aiming at the problem that the data distribution of extremely low electricity prices (i.e., valley electricity prices) and conventional electricity prices (i.e., non-valley electricity prices) in the power market is greatly different, an extreme electricity price prediction model (i.e., the first prediction model) is constructed to achieve accurate prediction of extremely low electricity prices, and a non-extreme electricity price prediction model (i.e., the second prediction model) is constructed to achieve accurate prediction of conventional electricity prices, thereby improving the accuracy and reliability of node electricity price predictions. The transaction volume can also be dynamically adjusted according to the prediction results, which helps power companies reduce transaction risks and achieves the purpose of formulating better trading strategies, thereby achieving the technical effect of improving the prediction accuracy of node electricity prices and improving the prediction accuracy of price relationships, thereby solving the technical problem in the related technology that the prediction of electricity spot transaction prices and price relationships based on similar day algorithms is low in prediction accuracy, resulting in low accuracy of the determined electricity spot trading strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0027] Figure 1 is a schematic diagram of a computer terminal provided according to Embodiment 1 of the present application;

[0028] Figure 2 It is a flow chart of a method for determining a power spot trading strategy provided in Embodiment 1 of the present application;

[0029] Figure 3 It is a schematic diagram of a process of optional day-ahead electricity price prediction provided according to the first embodiment of the present application;

[0030] Figure 4 It is a schematic diagram of a process of optional price relationship prediction provided according to the first embodiment of the present application;

[0031] Figure 5 It is a flow chart of a method for determining a power spot trading strategy provided in Embodiment 2 of the present application;

[0032] Figure 6 It is a schematic diagram of a device for determining a power spot trading strategy provided in Embodiment 3 of the present application;

[0033] Figure 7 It is a structural block diagram of an electronic device provided according to Embodiment 4 of the present application. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards in the relevant regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0037] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following explanations:

[0038] Electricity market: a place or system where buyers and sellers (power generation companies, power sales companies, large users, etc.) conduct electricity commodity transactions. It can be divided into wholesale market and retail market. The wholesale market mainly involves the electricity energy transactions between power generation companies and power sales companies, while the retail market involves the sales of electricity from power sales companies to end users.

[0039] Node price: A pricing mechanism in the wholesale electricity market. Each node in the power grid (usually a substation or a load concentration point) has a specific electricity price, called a node price or marginal price. The fluctuation of node prices reflects the influence of multiple factors such as power supply and demand, transmission network constraints, fuel prices, and changes in renewable energy output. It is a dynamic price. Among them, there are two types of node prices depending on the clearing market:

[0040] a. Day-Ahead Price: The day-ahead market is a component of the electricity spot market. Participants conduct electricity trading one day in advance based on the forecast of electricity demand in the next 24 hours or 48 hours to determine the electricity price for each hour (or 15 minutes) of the next day.

[0041] b. Real-time Price: The real-time market is a market where transactions are conducted at the actual time of electricity production and consumption, and its electricity price is dynamically adjusted according to the real-time electricity demand. The real-time electricity price directly reflects the actual operating status of the current power system, including factors such as unexpected load changes, sudden failures of generators, or uncertainty in the output of renewable energy, which can cause the real-time electricity price to deviate from the day-ahead forecast.

[0042] Price Relationship: The difference between the real-time electricity price and the day-ahead electricity price. When the real-time electricity price is higher than the day-ahead electricity price, the price relationship is positive; when the real-time electricity price is lower than the day-ahead electricity price, the price relationship is negative.

[0043] Example 1

[0044] According to an embodiment of the present application, a method for determining an electricity spot trading strategy is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0045] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 The hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for determining a power spot trading strategy is shown. Figure 1As shown, the computer terminal (or mobile device) 10 may include a processor set 102 (the processor set 102 may include but is not limited to a processing device such as a microprocessor (Microcontroller Unit, referred to as MCU) or a programmable logic device (Field Programmable Gate Array, referred to as FPGA), and the processor set 102 may include a processor set, Figure 1 102a, 102b, ..., 102n are used to illustrate), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a Universal Serial Bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. Those skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0046] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuits may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device).

[0047] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the method for determining the power spot trading strategy in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the method for determining the power spot trading strategy described above is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0048] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0049] The display may be a touch screen type liquid crystal display, which may enable a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0050] In the increasingly complex and changing environment of the electricity spot market, the ability to predict node electricity prices and price relationships is directly related to the profitability of power companies, and its accurate prediction helps to determine more accurate trading strategies. At present, the existing technology mainly predicts electricity spot trading prices and price relationships based on classic time series prediction models such as similar day algorithms and autoregressive moving average models. Most of these methods are limited to the mining of linear relationships and cannot effectively capture the nonlinear characteristics of electricity price fluctuations and the immediate impact of external factors on electricity prices, resulting in low accuracy in the prediction of electricity prices and price relationships, and low accuracy in the determined electricity spot trading strategies.

[0051] In the above technical background, the present application provides Figure 2 The method for determining the electricity spot trading strategy is shown. Figure 2 This is a flow chart of a method for determining a power spot trading strategy according to the first embodiment of the present application. The method includes:

[0052] Step S201, obtaining historical electricity spot transaction data of the target power grid node.

[0053] Optionally, the target grid node can be any node in the grid (such as a substation or a load concentration point), and the historical power spot transaction data can be data obtained by preprocessing the collected original spot data. The original spot data includes transaction time series data and covariate data (also called market boundary data, as shown in Table 1). The data in Table 1 are all called market boundary data, and each transaction time series data has these market boundary data.

[0054] It should be noted that the factors affecting transaction prices (i.e. covariate data) and the predicted transaction prices are in line with objective natural laws, and the factors affecting transaction prices and the actual transaction prices are also in line with objective natural laws and are not affected by economic laws or other policies.

[0055] Table 1

[0056]

[0057] Optionally, the collected raw spot data is preprocessed to clean the data based on threshold screening and linear filling of missing values. Among them, threshold screening is to check the quality of the raw data by setting a reasonable numerical range threshold (for example, a reasonable range of variation between the real-time electricity price and the day-ahead electricity price), remove abnormal values ​​or extreme values, and ensure the accuracy and reliability of the data. In view of the problem of missing values ​​that may appear in the data set, considering the continuity of the time series, linear interpolation methods can be used in specific operations to fill in the missing values ​​to maintain the coherence and trend of the data series. In some cases, if the amount of missing data is large, a time series prediction model can be used to estimate the missing part before interpolation.

[0058] Step S202, determine the predicted electricity spot trading price of the target power grid node based on historical electricity spot trading data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price.

[0059] Optionally, the first prediction model is an extreme electricity price prediction model, wherein the extreme electricity price may have different values ​​in different electricity markets. For example, the extreme electricity price is zero in the electricity market of Province A and is negative in the electricity market of Province B. Taking the extreme electricity price as zero as an example, the valley electricity price is zero electricity price, and the first prediction model can also be called a zero electricity price prediction model. The second prediction model is a prediction model for predicting conventional electricity prices other than zero electricity price, that is, a non-extreme electricity price prediction model, and the non-valley electricity price is the conventional electricity price other than zero electricity price. The predicted electricity spot transaction price is the day-ahead electricity price, including zero electricity price and conventional electricity price.

[0060] For example, a power generation company needs to predict the day-ahead electricity price of a certain node in the power grid and provide it to the electricity seller. After obtaining the historical electricity spot transaction data through data preprocessing, the day-ahead electricity price of the node can be obtained by prediction based on the historical electricity spot transaction data, the zero electricity price prediction model, and the non-extreme electricity price prediction model.

[0061] Step S203, determining the predicted electricity spot transaction price relationship of the target power grid node based on the historical electricity spot transaction data and the third prediction model, wherein the predicted electricity spot transaction price relationship represents the size relationship between the predicted electricity spot transaction price and the actual electricity spot transaction price.

[0062] Optionally, the third prediction model is a price relationship prediction model, which is based on the historical electricity spot transaction data and the price relationship prediction model to predict, and the predicted electricity spot transaction price relationship of the power grid node, that is, a positive or negative value, can be obtained. By inputting the historical electricity spot transaction data into the price relationship prediction model, the prediction is made based on the model's ability to capture the nonlinear characteristics of electricity price fluctuations and the immediate impact of external factors on electricity prices, and the predicted electricity spot transaction price relationship is output. If the predicted electricity spot transaction price relationship is a positive value, it means that the actual electricity spot transaction price (that is, the real-time electricity price) is greater than the predicted electricity spot transaction price (that is, the day-ahead electricity price); if the predicted electricity spot transaction price relationship is a negative value, it means that the real-time electricity price is less than the day-ahead electricity price.

[0063] Step S204: determining the electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

[0064] Optionally, based on the predicted day-ahead electricity price and price relationship, the electricity spot trading strategy corresponding to the grid node can be determined. For example, when the price relationship is positive, that is, the real-time electricity price is greater than the day-ahead electricity price, the electricity sales in the day-ahead market can be reduced according to the predicted day-ahead electricity price; when the price relationship is negative, that is, the real-time electricity price is less than the day-ahead electricity price, the electricity sales in the day-ahead market can be increased according to the predicted day-ahead electricity price.

[0065] In this scheme, by combining historical data with the prediction model, the spot electricity transaction price and its relationship with the actual price are effectively predicted, which provides a basis for the grid nodes to formulate better trading strategies. In view of the large difference in data distribution between extremely low electricity prices (i.e., valley electricity prices) and conventional electricity prices (i.e., non-valley electricity prices) in the electricity market, an extreme electricity price prediction model (i.e., the first prediction model) is constructed to achieve accurate prediction of extremely low electricity prices, and a non-extreme electricity price prediction model (i.e., the second prediction model) is constructed to achieve accurate prediction of conventional electricity prices, thereby improving the accuracy and reliability of node electricity price predictions. It can also dynamically adjust the transaction volume according to the prediction results, which helps power companies reduce transaction risks and achieves the purpose of formulating better trading strategies, thereby achieving the technical effect of improving the prediction accuracy of node electricity prices and improving the prediction accuracy of price relationships, and further solving the technical problem in the related technology that the prediction of electricity spot transaction prices and price relationships based on similar day algorithms has low prediction accuracy, resulting in low accuracy of the determined electricity spot trading strategy.

[0066] In order to accurately predict the day-ahead electricity price, in the method for determining the electricity spot trading strategy provided in Example 1 of the present application, the predicted electricity spot trading price of the target power grid node is determined based on historical electricity spot trading data, a first prediction model and a second prediction model, including: classifying the historical electricity spot trading data to obtain first-category electricity price data and second-category electricity price data; inputting the first-category electricity price data into the first prediction model to output a predicted valley electricity price; inputting the second-category electricity price data into the second prediction model to output a predicted non-valley electricity price; and determining the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0067] Optionally, according to whether the price is an extreme low limit price (i.e., an extreme electricity price), the historical electricity spot transaction data is classified to obtain zero electricity price data (i.e., the first type of electricity price data) and conventional non-zero electricity price data (i.e., the second type of electricity price data), and then the zero electricity price data is input into the zero electricity price prediction model, which predicts based on the characteristics of the input data learned during the training process and the potential laws and patterns of the occurrence of zero electricity prices, and outputs the predicted zero electricity price and the time of occurrence of zero electricity prices, and the conventional non-zero electricity price data is input into the non-extreme electricity price prediction model, which predicts based on the characteristics of the input data learned during the training process and the potential laws and patterns of the occurrence of non-zero electricity prices, and outputs the predicted conventional electricity price, and then the day-ahead electricity price can be determined based on the predicted zero electricity price and the predicted conventional electricity price. For example, the zero electricity price prediction model predicts that the electricity price at 4 o'clock the next day is zero electricity price, and the non-extreme electricity price prediction model predicts the conventional electricity price of each hour except 4 o'clock the next day, and the combination of the two can obtain the complete day-ahead electricity price of the next day.

[0068] It should be noted that the use of a zero electricity price prediction model to achieve accurate predictions of extremely low electricity prices, and the use of a non-extreme electricity price prediction model to achieve accurate predictions of conventional electricity prices, effectively solves the problem of large differences in data distribution between extremely low electricity prices and conventional electricity prices in the power market, and improves the accuracy and reliability of node electricity price predictions.

[0069] In order to determine a better trading strategy, in the method for determining the electricity spot trading strategy provided in Example 1 of the present application, the electricity spot trading strategy corresponding to the target power grid node is determined based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship, including: using the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is greater than the predicted electricity spot trading price, then reducing the electricity spot trading volume.

[0070] In the method for determining the electricity spot trading strategy provided in the first embodiment of the present application, if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is less than the predicted electricity spot trading price, the electricity spot trading volume is increased.

[0071] Optionally, the predicted electricity spot transaction price is used as the transaction price in the electricity spot transaction strategy. If the predicted electricity spot transaction price relationship indicates that the real-time electricity price is greater than the day-ahead electricity price, that is, the price relationship is a positive value, the electricity spot transaction volume is reduced. For example, the electricity sales volume in the day-ahead market can be reduced according to the predicted day-ahead electricity price. If the predicted electricity spot transaction price relationship indicates that the real-time electricity price is less than the day-ahead electricity price, that is, the price relationship is a negative value, the electricity spot transaction volume is increased. For example, the electricity sales volume in the day-ahead market can be increased according to the predicted day-ahead electricity price.

[0072] It should be noted that for power generation companies, a larger positive spread means that they can make more profits by participating in the real-time market because they can sell excess electricity at a higher price. Therefore, based on the predicted price relationship, it can provide data reference for power generation companies to determine trading strategies. For example, traders can combine the predicted price relationship with other relevant factors to comprehensively determine trading strategies. By dynamically adjusting the trading volume based on the forecast results, it helps power companies reduce trading risks.

[0073] In order to improve the prediction accuracy of node electricity prices, in the method for determining the electricity spot trading strategy provided in Example 1 of the present application, before obtaining the historical electricity spot trading data of the target power grid node, first training data is obtained, wherein the first training data at least includes first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that has an impact on the prediction of valley electricity prices; feature extraction is performed on the covariate data corresponding to the first-category sample electricity price data to obtain multiple first data features; and a gradient boosting tree algorithm is trained based on the first-category sample electricity price data and the multiple first data features to obtain a first prediction model.

[0074] In order to improve the prediction accuracy of node electricity prices, in the method for determining the electricity spot trading strategy provided in Example 1 of the present application, before obtaining the historical electricity spot trading data of the target power grid node, second training data is obtained, wherein the second training data at least includes second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; based on Fourier decomposition and wavelet analysis, feature analysis is performed on the covariate data corresponding to the second-category sample electricity price data to determine multiple second data features; and a gradient boosting tree algorithm is trained based on the second-category sample electricity price data and the multiple second data features to obtain a second prediction model.

[0075] In the method for determining the electricity spot trading strategy provided in the first embodiment of the present application, the data characteristics at least include the load rate characteristics of the thermal power units used to characterize the thermal power supply capacity.

[0076] Optionally, first obtain the original sample data (including transaction time series data and covariate data), clean the data based on threshold screening, missing value linear filling and other preprocessing, and then perform training sample classification, that is, classify the preprocessed data according to whether the price is an extreme low limit price, and obtain zero electricity price sample data (i.e., the first type of sample electricity price data) and its corresponding covariate data, as well as conventional non-zero electricity price sample data (i.e., the second type of sample electricity price data) and its corresponding covariate data. Then, build a zero electricity price prediction model and a non-extreme electricity price prediction model based on the corresponding sample data.

[0077] Optionally, a zero electricity price prediction model is constructed. Feature extraction is performed on the covariate data corresponding to the zero electricity price sample data to obtain multiple data features, such as approximate load rate, bidding space, new energy output prediction, etc., and a classification model is constructed based on the gradient boosting decision tree algorithm (GBDT). For example, a classification model is constructed based on CatBoost (a branch in GBDT) of GBDT, and a sample weighted training strategy is used for model training to solve the sample imbalance problem of a small number of zero samples, and a zero electricity price prediction model is obtained, that is, the zero electricity price prediction model is a classification model, and the model output is a probability value between 0-1. Define the classification threshold hyperparameter beta (value 0-1, for example, the value is 0.8). When the model output value is greater than beta, it is considered that the current electricity price is zero electricity price, that is, the prediction result is zero electricity price.

[0078] Optionally, a non-extreme electricity price prediction model is constructed. Based on Fourier decomposition and wavelet analysis, feature analysis is performed on the covariate data corresponding to the conventional non-zero electricity price sample data, multiple data features are determined, power market signals and price information are mined, and high-limit price samples (i.e., larger price values) are removed. For example, Fourier decomposition-related features: 1st, 3rd, 5th, and 7th harmonic values ​​of the unified load forecast value; wavelet analysis-related features: 5-layer decomposition values ​​of the discrete wavelet transform function of the unified load forecast value. Then, a regression model is constructed based on CatBoost and model training is performed to obtain a non-extreme electricity price prediction model.

[0079] Optionally, the load factor characteristics of thermal power units are constructed based on the covariate data to characterize the local thermal power supply capacity. The formula is as follows:

[0080]

[0081] It should be noted that in order to address the problem of large distribution differences between extremely low electricity price data and conventional electricity price data in the electricity market, this solution provides a zero electricity price prediction model to achieve accurate prediction of extremely low electricity prices, and can effectively distinguish and predict extreme situations in the electricity market; combined with signal processing technologies such as Fourier decomposition and wavelet analysis, it optimizes the analysis and extraction of market signals, deeply explores the complex information hidden behind price fluctuations, and further explores the information hidden behind the high correlation between electricity prices and bidding space by constructing the load rate characteristics of thermal power units, thereby improving the ability of the prediction model to capture dynamic changes in the market, enhancing the accuracy of price predictions and the overall effectiveness of the model.

[0082] In an optional embodiment, it is possible to use Figure 3 The schematic diagram shown in Figure 1 realizes the prediction of day-ahead electricity prices. Figure 3 is a schematic diagram of an optional process of predicting the day-ahead electricity price provided in Example 1 of the present application, such as Figure 3 As shown, after obtaining the spot data, data preprocessing is performed, and the samples are classified into zero electricity price samples and other samples (i.e., non-zero electricity price samples) through sample screening, and the zero electricity price samples are divided into training samples and prediction samples, and the training samples are used to train the zero electricity price classification model to obtain a trained zero electricity price classification model, and the prediction samples and the trained zero electricity price classification model are used for prediction to obtain the predicted zero electricity price and the occurrence time of the zero electricity price; the other samples are divided into training samples and prediction samples, and the top grid electricity price (i.e., the larger value of the price) is removed, and the electricity price regression model is trained using the training samples to obtain a non-extreme electricity price prediction model, and the prediction samples and the non-extreme electricity price prediction model are used for prediction to obtain the predicted conventional electricity price other than the zero electricity price, and then the predicted electricity price is post-processed, for example, the zero electricity price prediction model predicts that the electricity price at 4 o'clock the next day is zero electricity price, and the non-extreme electricity price prediction model predicts the conventional electricity price of each hour except 4 o'clock the next day, and the combination of the two can obtain the complete day-ahead electricity price of the next day. Among them, the specific implementation details refer to the aforementioned embodiment, which will not be repeated here.

[0083] Through the above scheme, the limitations of existing forecasting methods that ignore extreme electricity price forecasts and deep market law mining are solved, the accuracy and practicality of price forecasts at electricity spot trading nodes are improved, and advanced technical support is provided for the formulation of electricity market trading strategies.

[0084] In order to improve the prediction accuracy of price relationships, in the method for determining the electricity spot trading strategy provided in Example 1 of the present application, before obtaining the historical electricity spot trading data of the target power grid node, third training data is obtained, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of price relationships; lag features corresponding to the covariate data corresponding to the sample price relationships are constructed, and the gradient boosting tree algorithm is trained based on the third training data and the lag features to obtain a third prediction model.

[0085] Optionally, first obtain the original sample data (including transaction time series data and covariate data), clean the data based on threshold screening, missing value linear filling and other preprocessing, and then perform feature engineering: construct the lag feature corresponding to the covariate data, for example, lag feature: set the current time T, take the T-96 time of the unified load forecast value as the new feature. Then construct a classification model based on CatBoost, apply the automatic sample balancing strategy to train the model, obtain the price relationship prediction model, and solve the sample imbalance problem. Among them, the automatic sample balancing strategy is: for the imbalance of positive and negative price difference ratios that may occur in the power spot market, use strategies such as oversampling (copying minority class samples), undersampling (deleting majority class samples) or synthetic samples to balance the number of samples in each category to ensure that the model learns the features of all categories and avoid prediction bias.

[0086] It should be noted that the construction of lag features is crucial to capturing the time dependency and sequence patterns in the data. The functions of lag features are as follows:

[0087] (1) Capturing the dynamics of time series: Both electricity prices and market boundary data show a certain degree of continuity or periodicity, and past data points can provide valuable information about future data points;

[0088] (2) Modeling sequence dependency: By introducing lag features, the impact of the context of the variable sequence on the current point in time can be directly expressed in the model, characterizing the time-dependent characteristics of the sequence;

[0089] (3) Capturing periodicity: For data with obvious periodicity, such as seasonal changes, these periodic patterns can be effectively captured by selecting features with appropriate lag lengths.

[0090] Since the law of the price difference itself has a certain correlation with the contextual information of its adjacent time intervals, it is judged appropriate to introduce historical data points as features. Therefore, based on the lagged feature engineering technology, the time series characteristics of the price difference data and market boundary information are captured, and the historical price difference data and market boundary information are effectively integrated as the key input features of the prediction model. The historical contextual information of the time series is fully utilized, which enhances the model's understanding and extraction capabilities of the time series dependence of the price difference, thereby improving the accuracy and practicality of the prediction.

[0091] In an optional embodiment, it is possible to use Figure 4 The schematic diagram shown enables the prediction of price relationships. Figure 4 is a schematic diagram of a process for predicting optional price relationships according to the first embodiment of the present application, such as Figure 4 As shown, after collecting spot data, data preprocessing is performed, and then feature engineering is performed to construct lag features, and training samples and prediction samples are divided, and then the price difference model is trained, and prediction is performed through the trained model. The specific implementation details refer to the above embodiment and are not repeated here.

[0092] In an embodiment of the present application, historical electricity spot trading data of a target power grid node is obtained; a predicted electricity spot trading price of the target power grid node is determined based on the historical electricity spot trading data, a first prediction model, and a second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; a predicted electricity spot trading price relationship of the target power grid node is determined based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the relationship between the predicted electricity spot trading price and the actual electricity spot trading price; a method of determining the electricity spot trading strategy corresponding to the target power grid node is used based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship, and the electricity spot trading price and its relationship with the actual electricity spot trading price are effectively predicted by combining historical data with the prediction model. The relationship between the peak and the international prices provides a basis for formulating better trading strategies for power grid nodes. Aiming at the problem that the data distribution of extremely low electricity prices (i.e., valley electricity prices) and conventional electricity prices (i.e., non-valley electricity prices) in the power market is greatly different, an extreme electricity price prediction model (i.e., the first prediction model) is constructed to achieve accurate prediction of extremely low electricity prices, and a non-extreme electricity price prediction model (i.e., the second prediction model) is constructed to achieve accurate prediction of conventional electricity prices, thereby improving the accuracy and reliability of node electricity price predictions. The transaction volume can also be dynamically adjusted according to the prediction results, which helps power companies reduce transaction risks and achieves the purpose of formulating better trading strategies, thereby achieving the technical effect of improving the prediction accuracy of node electricity prices and improving the prediction accuracy of price relationships, thereby solving the technical problem in the related technology that the prediction of electricity spot transaction prices and price relationships based on similar day algorithms is low in prediction accuracy, resulting in low accuracy of the determined electricity spot trading strategies.

[0093] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0095] Example 2

[0096] According to an embodiment of the present application, a method for determining a power spot trading strategy is also provided, such as Figure 5 As shown, the method includes:

[0097] Step S501, obtaining historical power spot transaction data of the target power grid node uploaded by the client;

[0098] Step S502, in the cloud server, based on the historical electricity spot trading data, the first prediction model and the second prediction model, determine the predicted electricity spot trading price of the target power grid node, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; based on the historical electricity spot trading data and the third prediction model, determine the predicted electricity spot trading price relationship of the target power grid node, wherein the predicted electricity spot trading price relationship represents the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship, determine the electricity spot trading strategy corresponding to the target power grid node;

[0099] Step S503: Feedback the power spot trading strategy corresponding to the target power grid node to the client.

[0100] Through the above scheme, by combining historical data with the prediction model, the spot electricity transaction price and its relationship with the actual price are effectively predicted, which provides a basis for the grid nodes to formulate better trading strategies. In view of the large difference in data distribution between extremely low electricity prices (i.e., valley electricity prices) and conventional electricity prices (i.e., non-valley electricity prices) in the electricity market, an extreme electricity price prediction model (i.e., the first prediction model) is constructed to achieve accurate prediction of extremely low electricity prices, and a non-extreme electricity price prediction model (i.e., the second prediction model) is constructed to achieve accurate prediction of conventional electricity prices, thereby improving the accuracy and reliability of node electricity price predictions. The transaction volume can also be dynamically adjusted according to the prediction results, which helps power companies reduce transaction risks and achieves the purpose of formulating better trading strategies, thereby achieving the technical effect of improving the prediction accuracy of node electricity prices and improving the prediction accuracy of price relationships, thereby solving the technical problem in the related technology of predicting electricity spot transaction prices and price relationships based on similar day algorithms, resulting in low prediction accuracy and low accuracy of the determined electricity spot trading strategy.

[0101] In the cloud server, the specific method for determining the electricity spot trading strategy is the same as the method in Example 1, and will not be repeated here.

[0102] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0103] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0104] Example 3

[0105] According to an embodiment of the present application, a device for determining a power spot trading strategy for implementing the above-mentioned method for determining a power spot trading strategy is also provided, such as Figure 6As shown, the device includes: a first acquisition unit 601, a first determination unit 602, a second determination unit 603, and a third determination unit 604.

[0106] The first acquisition unit 601 is used to acquire historical power spot transaction data of a target power grid node;

[0107] The first determining unit 602 is used to determine the predicted power spot transaction price of the target power grid node based on the historical power spot transaction data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley power price in the predicted power spot transaction price, and the second prediction model is used to determine the non-valley power price in the predicted power spot transaction price;

[0108] The second determining unit 603 is used to determine the predicted power spot transaction price relationship of the target power grid node based on the historical power spot transaction data and the third prediction model, wherein the predicted power spot transaction price relationship represents the size relationship between the predicted power spot transaction price and the actual power spot transaction price;

[0109] The third determining unit 604 is used to determine the power spot trading strategy corresponding to the target power grid node according to the predicted power spot trading price and the predicted power spot trading price relationship.

[0110] In the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the historical electricity spot trading data of the target power grid node is acquired by the first acquisition unit 601; the first determination unit 602 determines the predicted electricity spot trading price of the target power grid node based on the historical electricity spot trading data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; the second determination unit 603 determines the predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship characterizes the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; the third determination unit 604 determines the electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship. In this scheme, by combining historical data with the prediction model, the spot electricity transaction price and its relationship with the actual price are effectively predicted, which provides a basis for the grid nodes to formulate better trading strategies. In view of the large difference in data distribution between extremely low electricity prices (i.e., valley electricity prices) and conventional electricity prices (i.e., non-valley electricity prices) in the electricity market, an extreme electricity price prediction model (i.e., the first prediction model) is constructed to achieve accurate prediction of extremely low electricity prices, and a non-extreme electricity price prediction model (i.e., the second prediction model) is constructed to achieve accurate prediction of conventional electricity prices, thereby improving the accuracy and reliability of node electricity price predictions. It can also dynamically adjust the transaction volume according to the prediction results, which helps power companies reduce transaction risks and achieves the purpose of formulating better trading strategies, thereby achieving the technical effect of improving the prediction accuracy of node electricity prices and improving the prediction accuracy of price relationships, and further solving the technical problem in the related technology that the prediction of electricity spot transaction prices and price relationships based on similar day algorithms has low prediction accuracy, resulting in low accuracy of the determined electricity spot trading strategy.

[0111] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the first determination unit 602 includes: a classification subunit, used to classify historical electricity spot trading data to obtain first-category electricity price data and second-category electricity price data; a first processing subunit, used to input the first-category electricity price data into a first prediction model, and output a predicted valley electricity price; a second processing subunit, used to input the second-category electricity price data into a second prediction model, and output a predicted non-valley electricity price; the first determination subunit, used to determine the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0112] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the third determination unit 604 includes: a second determination subunit, used to use the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; a third determination subunit, used to reduce the electricity spot trading volume if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is greater than the predicted electricity spot trading price.

[0113] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the device also includes: a fourth determination unit, used to increase the electricity spot trading volume if the predicted electricity spot trading price relationship represents that the actual electricity spot trading price is less than the predicted electricity spot trading price.

[0114] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the device also includes: a second acquisition unit, used to acquire first training data before acquiring the historical electricity spot trading data of the target power grid node, wherein the first training data includes at least first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that affects the prediction of valley electricity prices; a first processing unit, used to perform feature extraction on the covariate data corresponding to the first-category sample electricity price data to obtain multiple first data features; and a second processing unit, used to train the gradient boosting tree algorithm based on the first-category sample electricity price data and the multiple first data features to obtain a first prediction model.

[0115] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the device also includes: a third acquisition unit, used to acquire second training data before acquiring the historical electricity spot trading data of the target power grid node, wherein the second training data includes at least second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; a third processing unit, used to perform feature analysis on the covariate data corresponding to the second-category sample electricity price data based on Fourier decomposition and wavelet analysis, and determine multiple second data features; a fourth processing unit, used to train the gradient boosting tree algorithm based on the second-category sample electricity price data and multiple second data features to obtain a second prediction model.

[0116] Optionally, the data characteristics include at least a load rate characteristic of a thermal power unit for characterizing the thermal power supply capacity.

[0117] Optionally, in the device for determining the electricity spot trading strategy provided in Example 3 of the present application, the device also includes: a fourth acquisition unit, used to acquire third training data before acquiring the historical electricity spot trading data of the target power grid node, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationship; a fifth processing unit, used to construct lag features corresponding to the covariate data corresponding to the sample price relationships, and train the gradient boosting tree algorithm based on the third training data and the lag features to obtain a third prediction model.

[0118] It should be noted that the first acquisition unit 601, the first determination unit 602, the second determination unit 603, and the third determination unit 604 correspond to steps S201 to S204 in Example 1, and the examples and application scenarios implemented by the above units and the corresponding steps are the same, but are not limited to the contents disclosed in the above Example 1. It should be noted that the above modules as part of the device can be run in the computer terminal 10 provided in Example 1.

[0119] It should be noted that the preferred implementation scheme involved in the above embodiments of the present application is the same as the scheme provided in Example 1, as well as the application scenario and implementation process, but is not limited to the scheme provided in Example 1.

[0120] Example 4

[0121] The embodiment of the present application may provide an electronic device, which may be any electronic device in a group of electronic devices. Optionally, in this embodiment, the electronic device may also be replaced by a terminal device such as a mobile terminal.

[0122] Optionally, in this embodiment, the electronic device may be located in at least one network device among a plurality of network devices of a computer network.

[0123] In this embodiment, the above-mentioned electronic device can execute the program code of the following steps in the method for determining the electricity spot trading strategy: obtaining historical electricity spot trading data of the target power grid node; determining the predicted electricity spot trading price of the target power grid node based on the historical electricity spot trading data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; determining the predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship represents the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; determining the electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

[0124] The above-mentioned electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: classify the historical electricity spot trading data to obtain the first category of electricity price data and the second category of electricity price data; input the first category of electricity price data into the first prediction model, and output the predicted valley electricity price; input the second category of electricity price data into the second prediction model, and output the predicted non-valley electricity price; determine the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0125] The above-mentioned electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: using the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is greater than the predicted electricity spot trading price, then reducing the electricity spot trading volume.

[0126] The electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is less than the predicted electricity spot trading price, then the electricity spot trading volume is increased.

[0127] The above-mentioned electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: before obtaining the historical electricity spot trading data of the target power grid node, obtaining first training data, wherein the first training data at least includes first-class sample electricity price data and covariate data corresponding to the first-class sample electricity price data, and the covariate data corresponding to the first-class sample electricity price data is data that has an impact on the prediction of valley electricity prices; performing feature extraction on the covariate data corresponding to the first-class sample electricity price data to obtain multiple first data features; training the gradient boosting tree algorithm based on the first-class sample electricity price data and the multiple first data features to obtain a first prediction model.

[0128] The above-mentioned electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: before obtaining the historical electricity spot trading data of the target power grid node, obtaining second training data, wherein the second training data at least includes second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; based on Fourier decomposition and wavelet analysis, feature analysis is performed on the covariate data corresponding to the second-category sample electricity price data to determine multiple second data features; based on the second-category sample electricity price data and multiple second data features, the gradient boosting tree algorithm is trained to obtain a second prediction model.

[0129] The electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: the data characteristics at least include the load rate characteristics of the thermal power unit used to characterize the thermal power supply capacity.

[0130] The above-mentioned electronic device can also execute the program code of the following steps in the method for determining the electricity spot trading strategy: before obtaining the historical electricity spot trading data of the target power grid node, obtaining third training data, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationship; constructing lag features corresponding to the covariate data corresponding to the sample price relationships, and training the gradient boosting tree algorithm based on the third training data and the lag features to obtain a third prediction model.

[0131] Optionally, Figure 7 is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 7 As shown, the electronic device 70 may include: one or more ( Figure 7 Only one is shown in the figure) processor 702, memory 704. The electronic device 70 may also include a storage controller, through which the memory 704 is controlled and managed; the electronic device 70 may also include a peripheral interface, through which the radio frequency module, audio module and display screen are connected.

[0132] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the method and device for determining the power spot trading strategy in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, the above-mentioned method for determining the power spot trading strategy is realized. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories can be connected to the terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0133] The processor can call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the historical electricity spot trading data of the target power grid node; determine the predicted electricity spot trading price of the target power grid node based on the historical electricity spot trading data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot trading price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot trading price; determine the predicted electricity spot trading price relationship of the target power grid node based on the historical electricity spot trading data and the third prediction model, wherein the predicted electricity spot trading price relationship represents the size relationship between the predicted electricity spot trading price and the actual electricity spot trading price; determine the electricity spot trading strategy corresponding to the target power grid node based on the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

[0134] Optionally, the processor may also execute the program code of the following steps: classify historical electricity spot trading data to obtain first-category electricity price data and second-category electricity price data; input the first-category electricity price data into a first prediction model to output a predicted valley electricity price; input the second-category electricity price data into a second prediction model to output a predicted non-valley electricity price; and determine the predicted electricity spot trading price of the target power grid node based on the valley electricity price and the non-valley electricity price.

[0135] Optionally, the processor may also execute the program code of the following steps: using the predicted electricity spot trading price as the trading price in the electricity spot trading strategy; if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is greater than the predicted electricity spot trading price, reducing the electricity spot trading volume.

[0136] Optionally, the processor may also execute the program code of the following steps: if the predicted electricity spot trading price relationship indicates that the actual electricity spot trading price is less than the predicted electricity spot trading price, then increasing the electricity spot trading volume.

[0137] Optionally, the processor may also execute the program code of the following steps: before obtaining the historical electricity spot trading data of the target power grid node, obtaining first training data, wherein the first training data includes at least first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that has an impact on the prediction of valley electricity prices; performing feature extraction on the covariate data corresponding to the first-category sample electricity price data to obtain multiple first data features; and training the gradient boosting tree algorithm based on the first-category sample electricity price data and the multiple first data features to obtain a first prediction model.

[0138] Optionally, the processor may also execute the program code of the following steps: before obtaining the historical electricity spot trading data of the target power grid node, obtaining second training data, wherein the second training data includes at least second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of non-valley electricity prices; performing feature analysis on the covariate data corresponding to the second-category sample electricity price data based on Fourier decomposition and wavelet analysis to determine multiple second data features; and training the gradient boosting tree algorithm based on the second-category sample electricity price data and the multiple second data features to obtain a second prediction model.

[0139] Optionally, the processor may also execute program code of the following steps: the data characteristics at least include load rate characteristics of the thermal power unit for characterizing the thermal power supply capacity.

[0140] Optionally, the processor may also execute the program code of the following steps: before obtaining the historical electricity spot trading data of the target power grid node, obtaining third training data, wherein the third training data includes at least sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationship; constructing lag features corresponding to the covariate data corresponding to the sample price relationships, and training the gradient boosting tree algorithm based on the third training data and the lag features to obtain a third prediction model.

[0141] It can be understood by those skilled in the art that Figure 7 The structure shown is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, and other terminal devices. Figure 7 The structure of the electronic device is not limited. Figure 7 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 7 Different configurations shown.

[0142] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0143] Example 5

[0144] The embodiment of the present application also provides a computer-readable storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method for determining the power spot trading strategy provided in the first embodiment.

[0145] Optionally, in this embodiment, the above storage medium may be located in any electronic device in a group of electronic devices in a computer network, or in any mobile terminal in a group of mobile terminals.

[0146] Example 6

[0147] The embodiment of the present application also provides a computer program product. Optionally, in this embodiment, the computer program product may include a computer program, and when the computer program is executed by a processor, the method for determining the power spot trading strategy provided in the first embodiment is implemented.

[0148] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0149] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0150] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0151] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0152] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0153] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0154] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for determining a power spot trading strategy, characterized in that: include: Obtain historical electricity spot transaction data of target power grid nodes; Determine the predicted electricity spot transaction price of the target power grid node according to the historical electricity spot transaction data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot transaction price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot transaction price; Determine the predicted power spot transaction price relationship of the target power grid node based on the historical power spot transaction data and the third prediction model, wherein the predicted power spot transaction price relationship represents the size relationship between the predicted power spot transaction price and the actual power spot transaction price; According to the predicted electricity spot trading price and the predicted electricity spot trading price relationship, the electricity spot trading strategy corresponding to the target power grid node is determined.

2. The method according to claim 1, characterized in that Determining the predicted power spot transaction price of the target power grid node according to the historical power spot transaction data, the first prediction model and the second prediction model includes: Classifying the historical electricity spot transaction data to obtain first-category electricity price data and second-category electricity price data; Inputting the first type of electricity price data into the first prediction model, and outputting the predicted valley electricity price; Inputting the second type of electricity price data into the second prediction model, and outputting the predicted non-valley electricity price; The predicted electricity spot transaction price of the target power grid node is determined based on the off-peak electricity price and the non-off-peak electricity price.

3. The method according to claim 1, characterized in that Determining the power spot trading strategy corresponding to the target power grid node according to the predicted power spot trading price and the predicted power spot trading price relationship includes: Using the predicted electricity spot transaction price as the transaction price in the electricity spot transaction strategy; If the predicted electricity spot transaction price relationship indicates that the actual electricity spot transaction price is greater than the predicted electricity spot transaction price, the electricity spot transaction volume is reduced.

4. The method according to claim 3, characterized in that The method further comprises: If the predicted electricity spot transaction price relationship indicates that the actual electricity spot transaction price is less than the predicted electricity spot transaction price, the electricity spot transaction volume is increased.

5. The method according to claim 1, characterized in that: Before acquiring the historical power spot transaction data of the target power grid node, the method further includes: Acquire first training data, wherein the first training data at least includes first-category sample electricity price data and covariate data corresponding to the first-category sample electricity price data, and the covariate data corresponding to the first-category sample electricity price data is data that has an impact on the prediction of the valley electricity price; Extracting features of covariate data corresponding to the first type of sample electricity price data to obtain a plurality of first data features; The gradient boosting tree algorithm is trained according to the first type of sample electricity price data and the plurality of first data features to obtain the first prediction model.

6. The method according to claim 1, characterized in that Before acquiring the historical power spot transaction data of the target power grid node, the method further includes: Acquire second training data, wherein the second training data at least includes second-category sample electricity price data and covariate data corresponding to the second-category sample electricity price data, and the covariate data corresponding to the second-category sample electricity price data is data that has an impact on the prediction of the non-valley electricity price; Performing feature analysis on the covariate data corresponding to the second type of sample electricity price data based on Fourier decomposition and wavelet analysis to determine a plurality of second data features; The gradient boosting tree algorithm is trained according to the second type of sample electricity price data and the plurality of second data features to obtain the second prediction model.

7. The method according to claim 5 or 6, characterized in that: The data characteristics at least include the load rate characteristics of the thermal power units used to characterize the thermal power supply capacity.

8. The method according to claim 1, characterized in that Before acquiring the historical power spot transaction data of the target power grid node, the method further includes: Acquire third training data, wherein the third training data at least includes sample price relationships and covariate data corresponding to the sample price relationships, and the covariate data corresponding to the sample price relationships are data that have an impact on the prediction of the price relationships; Constructing the lag features corresponding to the covariate data corresponding to the sample price relationship, and training the gradient boosting tree algorithm based on the third training data and the lag features to obtain the third prediction model.

9. A method for determining a power spot trading strategy, characterized in that: include: Obtain the historical electricity spot transaction data of the target power grid node uploaded by the client; In the cloud server, the predicted electricity spot transaction price of the target power grid node is determined based on the historical electricity spot transaction data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted electricity spot transaction price, and the second prediction model is used to determine the non-valley electricity price in the predicted electricity spot transaction price; based on the historical electricity spot transaction data and the third prediction model, the predicted electricity spot transaction price relationship of the target power grid node is determined, wherein the predicted electricity spot transaction price relationship represents the size relationship between the predicted electricity spot transaction price and the actual electricity spot transaction price; based on the predicted electricity spot transaction price and the predicted electricity spot transaction price relationship, the electricity spot transaction strategy corresponding to the target power grid node is determined; The power spot trading strategy corresponding to the target power grid node is fed back to the client.

10. A device for determining a power spot trading strategy, characterized in that: include: A first acquisition unit, used to acquire historical power spot transaction data of a target power grid node; A first determination unit is used to determine the predicted power spot transaction price of the target power grid node based on the historical power spot transaction data, the first prediction model and the second prediction model, wherein the first prediction model is used to determine the valley electricity price in the predicted power spot transaction price, and the second prediction model is used to determine the non-valley electricity price in the predicted power spot transaction price; A second determining unit is used to determine the predicted power spot transaction price relationship of the target power grid node based on the historical power spot transaction data and the third prediction model, wherein the predicted power spot transaction price relationship represents the size relationship between the predicted power spot transaction price and the actual power spot transaction price; The third determining unit is used to determine the electricity spot trading strategy corresponding to the target power grid node according to the predicted electricity spot trading price and the predicted electricity spot trading price relationship.

11. An electronic device, characterized in that: include: A memory storing an executable program; A processor is used to run the program, wherein the program, when running, executes the method for determining the electricity spot trading strategy of any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the storage medium is located is controlled to execute the method for determining the electricity spot trading strategy of any one of claims 1 to 9.

13. A computer program product, characterized in that The method comprises a computer program or an instruction, which, when executed by a processor, implements a method for determining a power spot trading strategy according to any one of claims 1 to 9.

Images