Electric quantity prediction method and device and computer equipment

Abstract

The invention relates to an electric quantity prediction method and device and computer equipment, and relates to the field of energy. The method comprises the following steps: acquiring historical demand parameters and historical supply parameters of a target time period; on the basis of the historical demand parameters, predicting the remaining power demand of the target time period to obtain a first prediction result; on the basis of the first prediction result and the historical supply parameters, predicting the electric power spot price in the target time period to obtain a second prediction result; according to the first prediction result, the second prediction result, the global electric power parameters and local electric power parameters of the target electric power demand side, determining the demand electric quantity of the target electric power demand side in the current time period; the global power parameters comprise a global risk aversion level of a power supply side, a global risk aversion level of a power demand side and a global power utilization price of a power utilization side, and the local power parameters comprise a local risk aversion level, a local power utilization price and a demand power prediction proportion. The method can improve the prediction accuracy of the demand electric quantity of the power demand side.

Description

Technical Field

[0001] The present application relates to the technical field of energy, and particularly to a power consumption prediction method, apparatus, and computer device. Background Art

[0002] In a typical competitive power current hybrid time series market, as a power retailer on the power demand side, there is usually the following decision-making process: First, sign an annual medium- and long-term contract with a power plant as the power supply side according to expectations during the wholesale period, locking in part of the power consumption to be transmitted and distributed to the user side one year in advance; Subsequently, continuously adjust its own medium- and long-term power position, i.e., the power consumption demand, in the monthly, weekly, and multi-day medium- and long-term contract markets according to the progress of time and expected changes; Finally, make the final buying and selling decisions through the spot market (day-ahead trading or real-time trading) to ensure meeting the power demand on the user side in the retail market.

[0003] In an actual scenario, in order to accurately meet the delivery requirements of retail contracts, the power demand side needs to continuously update the prediction of the demand with low elasticity and strong uncertainty on the power consumption side during the process of participating in the above-mentioned competitive power current hybrid time series market, and actively manage its own medium- and long-term power position, i.e., the power consumption demand, in a timely manner in response to sudden supply-side situations such as rising primary energy prices, extreme weather, and equipment failures.

[0004] Considering that different from the traditional commodity futures and spot markets, electricity has the instantaneity characteristic of almost impossible inventory reserve (or high reserve cost), and the power demand side has only one participation opportunity in each power commodity market over time. Therefore, how the power demand side participates in the complex power current hybrid time series market and more accurately predicts the power consumption demand not only relates to whether the power demand side can avoid potential risks, but also, since the power consumption demand of the power demand side has an important impact on the electricity price in power trading, it will also have a huge impact on power trading in the power current hybrid time series market. Summary of the Invention

[0005] Based on this, in view of the above technical problems, it is necessary to provide a power consumption prediction method, apparatus, and computer device that can improve the prediction accuracy of the power consumption demand of the power demand side.

[0006] In a first aspect, the present application provides a power consumption prediction method, including:

[0007] Obtain historical demand parameters and historical supply parameters for a target period; wherein, the target period is the next period of the current period;

[0008] Predict the remaining power demand during the target period based on historical demand parameters to obtain a first prediction result, and predict the spot power price during the target period based on the first prediction result and historical supply parameters to obtain a second prediction result;

[0009] Determine the power consumption of the target power demander during the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demander;

[0010] Among them, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power price on the power consumption side. The local power parameters of the target power demander include the local risk aversion level of the target power demander, the local power price, and the demand power prediction ratio.

[0011] In one embodiment, determining the power consumption of the target power demander during the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demander includes: obtaining a power consumption prediction function; wherein, the power consumption prediction function is used to describe the influence of the remaining power demand, the spot power price, and the power parameters on the power consumption of the power demand side;

[0012] Solve the power consumption prediction function according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demander to obtain the power consumption of the target power demander during the current period.

[0013] In one embodiment, the power consumption prediction function is constructed in the following manner:

[0014] Construct a revenue function for the power supply side and a revenue function for the power demand side with the remaining power demand and the spot power price as variables;

[0015] Based on the revenue function of the power supply side, determine the optimal contract quantity function of the power supply side, and based on the revenue function of the power demand side, determine the optimal contract quantity function of the power demand side; wherein, the variables of the optimal contract quantity function of the power supply side and the optimal contract quantity function of the power demand side both include the power equilibrium price;

[0016] Determine the price function of the power equilibrium price when the function values of the optimal contract quantity function of the power supply side and the optimal contract quantity function of the power demand side are equal;

[0017] Use the price function and the local power parameter variables to assign values to the variables of the optimal contract quantity function of the power demand side to obtain the power consumption prediction function.

[0018] In one embodiment, based on the revenue function of the power supply side, determining the optimal signing quantity function of the power supply side, and based on the revenue function of the power demand side, determining the optimal signing quantity function of the power demand side, includes:

[0019] Taking the maximum function value of the revenue function of the power supply side as the objective, optimizing and solving the revenue function of the power supply side to obtain the optimal signing quantity function of the power supply side;

[0020] Taking the maximum function value of the revenue function of the power demand side as the objective, optimizing and solving the revenue function of the power demand side to obtain the optimal signing quantity function of the power demand side.

[0021] In one embodiment, taking the maximum function value of the revenue function of the power supply side as the objective, optimizing and solving the revenue function of the power supply side to obtain the optimal signing quantity function of the power supply side, includes:

[0022] Taking the power supply side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power supply side, constructing a first optimization function of the revenue function of the power supply side; wherein, the variables of the first optimization function include the signing quantity of the power supply side;

[0023] Taking the maximum function value of the first optimization function as the objective, solving the first optimization function to obtain the optimal signing quantity function of the power supply side.

[0024] In one embodiment, taking the maximum function value of the revenue function of the power demand side as the objective, optimizing and solving the revenue function of the power demand side to obtain the optimal signing quantity function of the power demand side, includes:

[0025] Taking the power demand side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power demand side, constructing a second optimization function of the revenue function of the power demand side; wherein, the variables of the second optimization function include the signing quantity of the power demand side;

[0026] Taking the maximum function value of the second optimization function as the objective, solving the second optimization function to obtain the optimal signing quantity function of the power demand side.

[0027] In one embodiment, obtaining the historical demand parameters and historical supply parameters of the target period, includes:

[0028] According to the demand window period and supply window period of the historical same period corresponding to the target period, determining the first window period and the second window period of the target period;

[0029] Obtaining the historical demand parameters of the first window period and the historical supply parameters of the second window period;

[0030] Among them, the power supply prediction result of the historical same period based on the supply window period prediction satisfies the supply prediction condition with the actual power supply value of the historical same period, and the predicted result of the remaining power demand of the historical same period based on the demand window period prediction satisfies the demand prediction condition with the actual remaining power demand value of the historical same period.

[0031] In one embodiment, based on historical demand parameters, the remaining power demand of the target period is predicted to obtain a first prediction result, and based on the first prediction result and historical supply parameters, the spot power price of the target period is predicted to obtain a second prediction result, including:

[0032] The historical demand parameters are input into a demand prediction model to obtain a prediction result of the remaining power demand of the target period as the first prediction result;

[0033] According to the historical supply parameters, the power supply prediction result of the target period is determined;

[0034] Based on the first prediction result and the power supply prediction result of the target period, the spot power price of the target period is predicted to obtain a second prediction result.

[0035] In a second aspect, the present application also provides a power quantity prediction device, including:

[0036] A parameter acquisition module for acquiring historical demand parameters and historical supply parameters of the target period; wherein the target period is the next period of the current period;

[0037] A result prediction module for predicting the remaining power demand of the target period based on historical demand parameters to obtain a first prediction result, and predicting the spot power price of the target period based on the first prediction result and historical supply parameters to obtain a second prediction result;

[0038] A power quantity determination module for determining the power demand quantity of the target power demand side in the current period according to the first prediction result, the second prediction result, the global power parameters and the local power parameters of the target power demand side;

[0039] Among them, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power price on the power consumption side, and the local power parameters of the target power demand side include the local risk aversion level, the local power price and the power demand quantity prediction ratio of the target power demand side.

[0040] In a third aspect, the present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0041] Obtain the historical demand parameters and historical supply parameters for the target period; wherein, the target period is the next period of the current period;

[0042] Based on the historical demand parameters, predict the remaining power demand for the target period to obtain a first prediction result, and based on the first prediction result and the historical supply parameters, predict the spot power price for the target period to obtain a second prediction result;

[0043] Determine the power demand quantity of the target power demand side in the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side;

[0044] Wherein, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power consumption price on the power consumption side, and the local power parameters of the target power demand side include the local risk aversion level, the local power consumption price, and the demand quantity prediction ratio of the target power demand side.

[0045] Fourthly, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0046] Obtain the historical demand parameters and historical supply parameters for the target period; wherein, the target period is the next period of the current period;

[0047] Based on the historical demand parameters, predict the remaining power demand for the target period to obtain a first prediction result, and based on the first prediction result and the historical supply parameters, predict the spot power price for the target period to obtain a second prediction result;

[0048] Determine the power demand quantity of the target power demand side in the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side;

[0049] Wherein, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power consumption price on the power consumption side, and the local power parameters of the target power demand side include the local risk aversion level, the local power consumption price, and the demand quantity prediction ratio of the target power demand side.

[0050] Fifthly, the present application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0051] Obtain the historical demand parameters and historical supply parameters for the target period; wherein, the target period is the next period of the current period;

[0052] Predict the electricity surplus demand for the target period based on historical demand parameters to obtain a first prediction result, and predict the electricity spot price for the target period based on the first prediction result and historical supply parameters to obtain a second prediction result;

[0053] Determine the electricity demand of the target electricity demander in the current period according to the first prediction result, the second prediction result, the global electricity parameters, and the local electricity parameters of the target electricity demander;

[0054] Among them, the global electricity parameters include the global risk aversion level on the electricity supply side, the global risk aversion level on the electricity demand side, and the global electricity price on the electricity consumption side. The local electricity parameters of the target electricity demander include the local risk aversion level of the target electricity demander, the local electricity price, and the demand electricity prediction ratio.

[0055] In the above electricity demand prediction method, device, and computer equipment, the electricity surplus demand for the target period is predicted through the historical demand parameters of the target period, and the electricity spot price for the target period is predicted through the electricity surplus demand obtained from the above prediction and the historical supply parameters of the target period. Then, combining the global electricity parameters and the local electricity parameters of the target electricity demander, according to the electricity surplus demand and electricity spot price for the target period obtained from the above prediction, determine the electricity demand of the above target electricity demander in the current period, where the target period is the next period of the current period. The above solution starts from the historical parameters of both the electricity supply and demand sides of the target period, and through the global electricity parameters of the three parties of the electricity supply side, the electricity demand side, and the electricity consumption side, as well as the local electricity parameters of the target electricity demander, predicts the electricity demand of the target electricity demander within the current period. In this way, the demand parameters, supply parameters, the global situation of the electricity market, and the special situation of the target electricity demander are all taken into account when predicting the electricity demand of the target electricity demander, which can achieve the consideration of the special needs of the target electricity demander on the basis of a global consideration of the electricity market, and realize the relationship fitting between the demand parameters, supply parameters, global electricity parameters, local electricity parameters of the target electricity demander, and the electricity demand of the target electricity demander, so that this solution can obtain better speculation ability, improve the accuracy of the finally predicted electricity demand, and further improve the risk aversion ability of the electricity demander, reduce the negative impact on the electricity price in the electricity transaction, and thus reduce the negative impact on the electricity transaction in the current electricity hybrid time series market. Description of the Drawings

[0056] To more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the following will briefly introduce the drawings required for use in the description of the embodiments of the present application or the related art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0057] Figure 1 It is a schematic flowchart of a power consumption prediction method provided in an embodiment.

[0058] Figure 2 It is a schematic flowchart of a process for determining the required power consumption provided in an embodiment.

[0059] Figure 3 It is a schematic flowchart of a process for constructing a power consumption prediction function provided in an embodiment.

[0060] Figure 4 It is a schematic flowchart of a process for determining the optimal signing quantity function provided in an embodiment.

[0061] Figure 5 It is a schematic flowchart of a process for obtaining historical demand parameters and historical supply parameters provided in an embodiment.

[0062] Figure 6 It is a schematic flowchart of a process for predicting the first prediction result and the second prediction result provided in an embodiment.

[0063] Figure 7 It is a schematic flowchart of a process for training a demand prediction model provided in an embodiment.

[0064] Figure 8 It is a schematic flowchart of a process for determining the demand window period provided in an embodiment.

[0065] Figure 9 It is a schematic flowchart of a process for determining the supply window period provided in an embodiment.

[0066] Figure 10 It is a schematic flowchart of a power consumption prediction method provided in another embodiment.

[0067] Figure 11 It is the value of Δπ under the grid values of λ r and p f provided in an embodiment; it is a schematic diagram.

[0068] Figure 12 It is the value of λ r and p f under the grid values provided in an embodiment; it is a schematic diagram of the standard deviation of Δπ.

[0069] Figure 13 For the λ provided in one embodiment r and p f Schematic diagram of the proportion of the electricity demand under the grid value in the total electricity delivery that the electricity demand side should ultimately satisfy;

[0070] Figure 14 Block diagram of the structure of a power consumption prediction device provided in one embodiment;

[0071] Figure 15 Internal structure diagram of a computer device in one embodiment. Detailed implementation manners

[0072] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0073] Different from the traditional commodity futures and spot markets, electricity has the immediacy characteristic that it is almost impossible to store in stock (or the storage cost is high), and the electricity demand side has only one participation opportunity in each electricity commodity market over time. Therefore, how the electricity demand side participates in the complex electricity spot mixed time series market and more accurately predicts the electricity demand not only relates to whether the electricity demand side can avoid potential risks, and because the electricity demand of the electricity demand side has an important impact on the electricity price in the electricity transaction, it will also have a huge impact on the electricity transaction in the electricity spot mixed time series market.

[0074] Based on this, in an exemplary embodiment, as Figure 1 shown, a power consumption prediction method is provided. This method is applied to a computer device, where the computer device can be a server or a terminal device. In this embodiment, the method includes the following steps:

[0075] S101, obtaining historical demand parameters and historical supply parameters of a target time period.

[0076] Wherein, the target time period is the next time period of the current time period.

[0077] In the above-mentioned competitive electricity spot mixed time series market, the electricity demand of the electricity demand side in the current time period will be affected by the remaining electricity demand on the electricity demand side in the next time period of the current time period and the electricity spot price in the electricity spot market in the next time period of the current time period. Therefore, when predicting the electricity demand of the target electricity demand side in the current time period, the remaining electricity demand and the electricity spot price in the next time period of the current time period can be predicted first. Among them, the next time period of the current time period can be referred to as the target time period.

[0078] Among them, after deducting the preferentially absorbed electricity from the electricity demand on the power consumption side, the remaining electricity demand is formed. Specifically, the electricity demand side or the power consumption side declares the expected electricity consumption. The power dispatching center deducts the power generation of three types of preferentially absorbed power sources from the total declared demand, including: Type A power sources (such as non-dispatchable power sources like wind power, photovoltaic power, and biomass), power sources dispatched by the local power dispatching center (power sources dispatched by the prefecture-level city power company), and power transmitted from the west to the east, and also deducts the electricity transmitted to the designated area, and calculates the remaining electricity demand (i.e., Type B spatial capacity).

[0079] Based on this, in this embodiment, the power supply during the target period refers to the part of the power supply that is competed by each power plant in the power market during the target period after deducting the power generation of the preferentially absorbed power sources. It can be understood as the remaining power supply part after deducting the power generation of the above-mentioned preferentially absorbed power sources from the total power demand on the power consumption side during the target period. Correspondingly, in this embodiment, the electricity demand of the target electricity demand side refers to the electricity obtained through competition in the above-mentioned remaining electricity demand.

[0080] Generally, when making data predictions, historical data is used to predict future data. Thus, in this embodiment, the historical data within a certain time range before the target period can be used to predict the power supply and power spot price during the target period. Therefore, it is necessary to first obtain the historical demand parameters and historical supply parameters of the target period.

[0081] In some alternative embodiments, the demand parameters and supply parameters of the historical corresponding period corresponding to the target period can be obtained and used as the historical demand parameters and historical supply parameters of the target period respectively. Among them, the historical corresponding period refers to the period within the historical cycle of the target period that is the same as the target period. For example, if the target period is April 2025, the historical corresponding period corresponding to the target period can be April 2024, April 2023, or April 2022, etc. Optionally, the historical corresponding period corresponding to the target period is the period within the previous historical cycle of the target period that is the same as the target period. That is, assuming the target period is April 2025, the historical corresponding period is April 2024.

[0082] In some alternative embodiments, the demand parameters and supply parameters of multiple consecutive periods within the historical time range before the target period can also be obtained and used as the historical demand parameters and historical supply parameters of the target period. Optionally, the above historical time range can be the time range that is before the above target period, closest to the above target period, and has the same cycle duration as the above target period.

[0083] Among them, complex economic factors, climate factors, market factors, etc. will affect the residual power demand, and the residual power demand will affect the spot power price. Therefore, the factors affecting the residual power demand can be selected as demand parameters. Optionally, as shown in Table 1 below, are the specific contents of the demand parameters.

[0084] Table 1 Specific Contents of Demand Parameters

[0085]

[0086] Among them, the above CPI is the abbreviation of Consumer Price Index, and M2 is the abbreviation of Broadmeasure of money supply.

[0087] Correspondingly, complex economic factors, climate factors, market factors, etc. will also affect the power supply, and further affect the spot power price. Therefore, the factors affecting the power supply can be selected as supply parameters. Optionally, as shown in Table 2 below, are the specific contents of the supply parameters.

[0088] Table 2 Specific Contents of Supply Parameters

[0089]

[0090]

[0091] Among them, UPS is the abbreviation of Unified Settlement Point.

[0092] S102. Based on the historical demand parameters, predict the residual power demand in the target period to obtain the first prediction result, and based on the first prediction result and the historical supply parameters, predict the spot power price in the target period to obtain the second prediction result.

[0093] As mentioned above, the spot power price in the target period is affected by the residual power demand in the target period. Therefore, when predicting the spot power price in the target period, the residual power demand in the target period can be predicted first.

[0094] After obtaining the above historical demand parameters of the target period, based on the above historical demand parameters, predict the residual power demand in the target period to obtain the first prediction result. For example, based on the above historical demand parameters, the prediction result of the residual power demand in the target period can be determined by various methods such as machine learning models and simulation prediction methods.

[0095] Furthermore, since the spot price of electricity in the target period is also affected by the electricity supply in the target period, and the electricity supply in the target period is affected by the historical supply parameters of the target period, after obtaining the above-mentioned first prediction result, the spot price of electricity in the target period can be predicted based on the above-mentioned first prediction result and historical supply parameters to obtain a second prediction result.

[0096] For example, the power supply of the target period can be first predicted based on the above-mentioned historical supply parameters to obtain the power supply prediction result of the target period, and then the spot price of electricity of the target period can be predicted based on the above-mentioned first prediction result and the power supply prediction result of the target period to obtain the second prediction result.

[0097] Usually, in the actual scenario of power duration, the power dispatching center can summarize the quotation curves (quotation data) provided by the power plants (power units) as power suppliers to form a power supply curve, which requires the above quotation curve to be horizontal or step-up. That is, for the power supplier, the essence of its power supply is the corresponding relationship between the power supply and the spot price of electricity. Optionally, the above power supply can be represented by a curve that represents the corresponding relationship between the power supply and the spot price of electricity. According to the curve, the spot price of electricity corresponding to different power supply can be determined, and the changing trend of the curve represents the changing trend of the spot price of electricity with the change of power supply.

[0098] Based on this, the essence of the power supply forecast result of the target period is the corresponding relationship between the power supply and the power spot price in the target period. Optionally, the power supply forecast result of the target period can also be represented by a curve used to characterize the corresponding relationship between the power supply and the power spot price in the target period. According to the curve, the power spot price corresponding to different power supply in the target period can be determined. The curve reflects the changing trend of the power spot price with the change of power supply in the target period.

[0099] When predicting the spot price of electricity in the target period, the remaining electricity demand in the target period matches the electricity supply. Therefore, the above-mentioned first prediction result and the electricity supply prediction result of the target period can be combined to solve the spot price of electricity in the target period as the second prediction result.

[0100] Optionally, the power supply prediction result of the above target time period is a curve in a two-dimensional coordinate system with the power supply quantity as the horizontal axis coordinate and the power spot price as the vertical axis coordinate. This curve is used to characterize the corresponding relationship between the power supply quantity and the power spot price within the target time period; the above first prediction result is a definite value of the remaining power demand. In the curve representing the power supply prediction result of the target time period, find the ordinate of the point with the abscissa being the above first prediction result. This ordinate is the power spot price of the target time period, that is, the above second prediction result.

[0101] S103. Determine the power consumption quantity of the target power demand side in the current time period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side.

[0102] Among them, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power consumption price on the power consumption side. The local power parameters of the target power demand side include the local risk aversion level of the target power demand side, the local power consumption price, and the power consumption quantity prediction ratio.

[0103] It can be understood that in the actual scenario of the power market, for the target power demand side, the characteristics of the target power demand side itself, such as the risk aversion level of the target power demand side itself (local risk aversion level), the price of supplying power to users (local power consumption price), and the share of its own power consumption quantity in the power supply of the power market (power consumption quantity prediction ratio), etc., can affect the power consumption quantity of the target power demand side in the current time period. Among them, considering that risk aversion is the attitude of investors towards investment risks, when the local risk aversion level of the above target power demand side is positive, the larger the value, the less the target power demand side is willing to bear the medium- and long-term premium (the price difference between the medium- and long-term power price and the power spot price). Therefore, the less the target power demand side is willing to sign the medium- and long-term power contract quantity.

[0104] Moreover, the power supply of each power supply side participating in the power market, the power consumption quantities declared by other power demand sides, and the power consumption price on the user side, etc., can also affect the power consumption quantity of the target power demand side in the current time period.

[0105] Based on this, in this embodiment, all power suppliers participating in the power market can be regarded as a whole. Thus, by combining the market conditions of the power market and the characteristics of each power supplier, the global risk aversion level on the power supply side can be determined, which can be referred to as the risk aversion level of the representative power supplier (representative power plant), the average risk aversion level of power suppliers, etc. Among them, considering that risk aversion is the attitude of investors towards investment risks, when the above-mentioned global risk aversion level on the power supply side is positive, the larger the value, the less the power supply side is willing to bear the medium- and long-term premium (the price difference between the medium- and long-term power price and the spot power price). Thus, the amount of medium- and long-term power contracts that the power supply side is willing to sign is less.

[0106] Correspondingly, all power demanders participating in the power market can be regarded as a whole. Thus, by combining the market conditions of the power market and the characteristics of each power demander, the global risk aversion level on the power demand side can be determined, which can be referred to as the risk aversion level of the representative power demander (representative power retailer), the average risk aversion level of power demanders, the risk aversion level of the representative power demander (representative power retailer) on the power demand side, etc. Among them, the above-mentioned global risk aversion level on the power demand side reflects the intensity of the willingness of the power demand side to participate in speculation using the medium- and long-term premium.

[0107] Furthermore, all electricity consumers participating in the power market, such as factories and enterprises, can be regarded as a whole. Thus, by combining the power retail prices provided by different power demanders for their electricity consumers, the global electricity price on the electricity consumption side can be determined, which can also be referred to as the average electricity price on the electricity consumption side, the average power retail price, etc.

[0108] And, as mentioned above, the remaining power demand and the spot power price in the target period (the next period after the current period) can also affect the electricity demand of the target power demander.

[0109] Based on this, after obtaining the above first prediction result and second prediction result, the electricity demand of the target power demander in the current period can be determined according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demander.

[0110] That is, based on the current period, by predicting the remaining power demand and the spot power price in the target period (the next period after the current period), the electricity demand of the target power demander in the current period is predicted. For example, if it is desired to predict the electricity demand of the target power demander in April 2025, the remaining power demand and the spot power price in May 2025 can be predicted first. Thus, based on the predicted remaining power demand and spot power price in May 2025, the electricity demand of the target power demander in the current period is predicted.

[0111] In some alternative embodiments, the preset model may be trained by using, as inputs, the predicted results of the electricity surplus demand in historical time periods, the predicted results of the electricity spot price in historical time periods, the global electricity parameters corresponding to the historical time periods, and the local electricity parameters of different electricity demand parties, and using, as labels, the electricity consumption in the previous time period of the historical time periods of different electricity demand parties, to obtain an electricity consumption prediction model. After obtaining the above first prediction result and second prediction result, the first prediction result, the second prediction result, the global electricity parameters, and the local electricity parameters of the target electricity demand party may be input into the above electricity consumption prediction model, and the prediction result output by the electricity consumption prediction model may be obtained, and this prediction result is the electricity consumption demand of the target electricity demand party in the current time period.

[0112] In the above electricity consumption prediction method, the electricity surplus demand in the target time period is predicted by using the historical demand parameters in the target time period, and the electricity spot price in the target time period is predicted by using the predicted electricity surplus demand and the historical supply parameters in the target time period. Then, in combination with the global electricity parameters and the local electricity parameters of the target electricity demand party, based on the predicted electricity surplus demand and electricity spot price in the target time period, the electricity consumption demand of the target electricity demand party in the current time period is determined, where the target time period is the next time period of the current time period. In the above solution, starting from the historical parameters of both the power supply and demand sides, through the global electricity parameters of the power supply side, power demand side, and power consumption side, as well as the local electricity parameters of the target electricity demand party, the electricity consumption demand of the target electricity demand party within the current time period is predicted. In this way, the demand parameters, supply parameters, the overall situation of the power market, and the special situation of the target electricity demand party are all taken into consideration when predicting the electricity consumption demand of the target electricity demand party, which can achieve the consideration of the special needs of the target electricity demand party on the basis of a global consideration of the power market, realize the relationship fitting between the demand parameters, supply parameters, global electricity parameters, local electricity parameters of the target electricity demand party, and the electricity consumption demand of the target electricity demand party, so that this solution can obtain better inference ability, improve the accuracy of the finally predicted electricity consumption demand, and further improve the risk aversion ability of the electricity demand party, reduce the negative impact on the electricity price in the power transaction, and thus reduce the negative impact on the power transaction in the current mixed time series market of electricity.

[0113] Based on the above embodiments, in an exemplary embodiment, the determination method of the electricity consumption demand of the target electricity demand party in the current time period in S104 is further refined. Optionally, as Figure 2 shown, the following steps may be included:

[0114] S201, obtain an electricity consumption prediction function.

[0115] Among them, the electricity quantity prediction function is used to describe the impact of the remaining electricity demand, electricity spot price, and electricity parameters on the electricity demand quantity of the electricity demand side.

[0116] As mentioned above, the remaining electricity demand, electricity spot price, and electricity parameters can all affect the electricity demand quantity of the electricity demand side. Therefore, an electricity quantity prediction function can be obtained to describe the impact of the remaining electricity demand, electricity spot price, and electricity parameters on the electricity demand quantity of the electricity demand side. Moreover, this electricity quantity prediction function takes the remaining electricity demand, electricity spot price, and electricity parameters as variables and the electricity demand quantity of the electricity demand side as the function value.

[0117] Among them, in the above-mentioned electricity quantity prediction function, the electricity parameters include global electricity parameters and local electricity parameters of the electricity demand side, which is the electricity demand party for which the electricity demand quantity is to be predicted.

[0118] S202. Solve the electricity quantity prediction function according to the first prediction result, the second prediction result, the global electricity parameters, and the local electricity parameters of the target electricity demand side to obtain the electricity demand quantity of the target electricity demand side in the current period.

[0119] As mentioned above, the above-mentioned electricity quantity prediction function takes the remaining electricity demand, electricity spot price, and electricity parameters as variables and the electricity demand quantity of the electricity demand side as the function value. Then, after obtaining the above-mentioned electricity quantity prediction function, the remaining electricity demand of the target period predicted above (i.e., the first prediction result) can be used as the variable value of the variable of electricity supply, the electricity spot price of the target period predicted above (i.e., the second prediction result) can be used as the variable value of the variable of electricity spot price, and the above-mentioned global electricity parameters and the local electricity demand parameters of the target electricity demand side can be used as the parameter values of the variable electricity parameters.

[0120] In this way, the above-mentioned first prediction result, second prediction result, global electricity parameters, and local electricity parameters of the target electricity demand side are used to assign values to the variables of the above-mentioned electricity quantity prediction function respectively, and the electricity quantity prediction function after variable assignment is solved, so as to obtain the function value of the electricity quantity prediction function after variable assignment. This function value is the electricity demand quantity of the target electricity demand side in the current period.

[0121] In this embodiment, since the power consumption prediction function describes the influence of the remaining power demand, the power spot price, and the power parameters on the power consumption demand of the power demand side, the power consumption prediction function can reflect the correlation between the power supply side, the power demand side, the remaining power demand, and the power spot price and the power consumption demand of the power demand side. Therefore, by solving the power consumption prediction function according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side, the power consumption demand of the target power demand side at the current time period can be obtained, which can not only improve the accuracy of the predicted power consumption demand of the target power demand side at the current time period, but also reflect the real situation of the power market and improve the benefits of the power demand side.

[0122] Based on the above embodiment, in an exemplary embodiment, the construction method of the power consumption prediction function in S201 is further refined. Optionally, as Figure 3 shown, the following steps may be included:

[0123] S301, taking the remaining power demand and the power spot price as variables, construct the revenue function of the power supply side and the revenue function of the power demand side.

[0124] In this embodiment, it is assumed that the power market is in equilibrium and the power supply capacity of the power supply side is completely flexible. Then, the remaining power demand that the power demand side needs to satisfy is equal to the power supply provided by the power supply side, and its power supply volume in the power market completely depends on the remaining power demand of the power demand side.

[0125] Based on this, the revenue of the power demand side can be affected by the remaining power demand and the power spot price. Therefore, the revenue function of the power demand side can be constructed with the remaining power demand and the power spot price as variables.

[0126] Correspondingly, the revenue of the power supply side can be affected by the power supply and the power spot price, and the power supply is equal to the remaining power demand. Therefore, the revenue function of the power supply side can also be constructed with the remaining power demand and the power spot price as variables.

[0127] S302, based on the revenue function of the power supply side, determine the optimal signing volume function of the power supply side, and based on the revenue function of the power demand side, determine the optimal signing volume function of the power demand side.

[0128] Among them, the variables of the optimal signing volume function of the power supply side and the optimal signing volume function of the power demand side both include the power equilibrium price.

[0129] For the electricity supply side and the electricity demand side, both sides hope to obtain greater or even maximum profits, and the equilibrium electricity price and electricity contract volume in the electricity market can affect the profits of both the electricity supply side and the electricity demand side.

[0130] Based on this, after constructing the above-mentioned revenue function of the power supply side and the revenue function of the power demand side, the optimal contract quantity function of the power supply side can be determined based on the revenue function of the power supply side with the power equilibrium price as a variable, and the optimal contract quantity function of the power demand side can be determined based on the revenue function of the power demand side with the power equilibrium price as a variable. In this way, since the variables of the optimal contract quantity function of the power supply side and the optimal contract quantity function of the power demand side both include the power equilibrium price, the price function of the above-mentioned power equilibrium price can be solved based on the data relationship between the optimal contract quantity function of the power supply side and the optimal contract quantity function of the power demand side.

[0131] S303, determining a price function of an equilibrium power price when the function values of the optimal contract quantity function on the power supply side and the optimal contract quantity function on the power demand side are equal.

[0132] It can be understood that when the electricity market is in equilibrium, the optimal contract quantity on the electricity supply side is equal to the optimal contract quantity on the electricity demand side.

[0133] Based on this, by combining the above-mentioned optimal contract quantity function on the power supply side and the optimal contract quantity function on the power demand side, the price function of the equilibrium power price can be determined when the function values of the optimal contract quantity function on the power supply side and the optimal contract quantity function on the power demand side are equal.

[0134] S304, using the price function and the local power parameter variables, assigning variables to the optimal contract quantity function on the power demand side to obtain a power quantity prediction function.

[0135] It can be understood that, in order to meet the revenue requirements of the power demand side, the signing quantity of the power demand side needs to satisfy the above optimal signing quantity function of the power demand side. And in order to match the remaining power demand with the power supply, the variable of the power equilibrium price in the optimal signing quantity function of the power demand side needs to satisfy the above price function of the power equilibrium price. Moreover, the above-obtained optimal signing quantity function of the power demand side is obtained on the basis of regarding all power demand parties participating in the power market as a whole. Therefore, the above optimal signing quantity function of the power demand side is for the overall power demand side. For different power demand parties participating in the power market, they have local power parameters that conform to their own situations. Therefore, when using the above optimal signing quantity function of the power demand side to determine the power demand quantity of a single power demand party participating in the power market, it is necessary to first replace the power parameter variables for the overall power demand side in the above optimal signing quantity function of the power demand side with the local power parameter variables for the above single power demand party.

[0136] Based on this, the price function and local power parameter variables can be used to assign values to the variables in the optimal signing quantity function of the power demand side to obtain the power quantity prediction function. Among them, first use the local power parameter variables to assign values to the power parameter variables in the optimal signing quantity function of the power demand side, and then use the above price function of the power equilibrium price to assign values to the power equilibrium price variable in the optimal signing quantity function of the power demand side. Then the function obtained after the above variable assignment is the power quantity prediction function.

[0137] In this embodiment, a power quantity prediction function that can meet both the revenue requirements of the power demand side and the matching of the remaining power demand with the power supply can be obtained. Furthermore, when solving the above power prediction function to obtain the power demand quantity of the target power demand party at the current time period, the accuracy of the predicted power demand quantity of the target power demand party at the current time period can be improved.

[0138] In an exemplary embodiment, the determination method of the optimal signing quantity functions of the power supply side and the power demand side in S302 above is further refined. Optionally, as Figure 4 shown, the following steps may be included:

[0139] S401, aiming at maximizing the function value of the revenue function of the power supply side, optimize and solve the revenue function of the power supply side to obtain the optimal signing quantity function of the power supply side.

[0140] For the power supply side, generally, it is desired to obtain relatively large or even the largest revenue. Therefore, in order to improve the revenue of the power supply side while improving the accuracy of the predicted electricity demand of the target power demand side in the current period, the revenue function of the power supply side can be optimized and solved with the goal of maximizing the function value of the revenue function of the power supply side, so as to obtain the optimal signing quantity function of the power supply side.

[0141] In an alternative embodiment, the method for determining the optimal signing quantity function of the power supply side in S401 is further refined. Optionally, it may include the following steps:

[0142] Step 1: Taking the power supply side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power supply side, construct a first optimization function of the revenue function of the power supply side.

[0143] Among them, the variables of the first optimization function include: the signing quantity of the power supply side.

[0144] As mentioned above, in this embodiment, predicting the electricity demand of the target power demand side in the current period refers to predicting the power signing quantity of the target power demand side in the current period. Therefore, when determining the above-mentioned optimal signing quantity function of the power supply side, a first optimization function with the signing quantity of the power supply side as a variable can be constructed.

[0145] Among them, considering that there is a correlation between the power supply side revenue and the signing quantity of the power supply side, therefore, taking the power supply side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power supply side, a first optimization function of the revenue function of the power supply side can be constructed.

[0146] That is, construct an expected value function and variance value function corresponding to the revenue function of the power supply side with the power supply side revenue as a variable. Thus, based on the above-mentioned expected value function and variance value function, construct a first optimization function of the revenue function of the power supply side.

[0147] Step 2: Taking the maximum value of the function value of the first optimization function as the goal, solve the first optimization function to obtain the optimal signing quantity function of the power supply side.

[0148] After obtaining the first optimization function of the above-mentioned revenue function of the power supply side, considering that the variables of the above-mentioned first optimization function include the signing quantity of the power supply side, therefore, in order to obtain the optimal signing quantity function of the power supply side when the function value of the revenue function of the power supply side is the largest, taking the maximum value of the function value of the first optimization function as the goal, solve the first optimization function to obtain the optimal signing quantity function of the power supply side.

[0149] S402. Optimize and solve the revenue function of the electricity demand side with the goal of maximizing the function value of the revenue function of the electricity demand side, and obtain the optimal signing quantity function of the electricity demand side.

[0150] Corresponding to the above electricity supply side, for the electricity demand side, it also hopes to obtain greater or even the greatest revenue. Because, in order to improve the accuracy of the predicted demand electricity quantity of the target electricity demand side at the current time period and increase the revenue of the electricity demand side, the revenue function of the electricity demand side can be optimized and solved with the goal of maximizing the function value of the revenue function of the electricity demand side, and the optimal signing quantity function of the electricity demand side can be obtained.

[0151] In an optional embodiment, further refine the determination method of the optimal signing quantity function of the electricity demand side in S402. Optionally, it may include the following steps:

[0152] Step 1. Taking the revenue of the electricity demand side as a variable, based on the expected value function and variance value function corresponding to the revenue function of the electricity demand side, construct a second optimization function of the revenue function of the electricity demand side.

[0153] Among them, the variables of the second optimization function include: the signing quantity of the electricity demand side.

[0154] Corresponding to the above electricity supply side, when determining the optimal signing quantity function of the electricity demand side, a second optimization function with the revenue of the electricity demand side as a variable can be constructed.

[0155] Among them, considering that the revenue of the electricity demand side is correlated with the signing quantity of the electricity demand side, therefore, a second optimization function of the revenue function of the electricity demand side can be constructed with the revenue of the electricity demand side as a variable, based on the expected value function and variance value function corresponding to the revenue function of the electricity demand side.

[0156] That is, construct the expected value function and variance value function corresponding to the revenue function of the electricity demand side with the revenue of the electricity demand side as a variable. Thus, based on the above expected value function and variance value function, construct a second optimization function of the revenue function of the electricity demand side.

[0157] Step 2. Taking the maximum function value of the second optimization function as the goal, solve the second optimization function to obtain the optimal signing quantity function of the electricity demand side.

[0158] After obtaining the second optimization function of the revenue function of the electricity demand side, considering that the variables of the second optimization function include the signing quantity of the electricity demand side, therefore, in order to obtain the optimal signing quantity function of the electricity demand side when the function value of the revenue function of the electricity demand side is the largest, the second optimization function can be solved with the goal of maximizing the function value of the second optimization function to obtain the optimal signing quantity function of the electricity demand side.

[0159] In this embodiment, it is possible to improve the benefits of the power supply side and the power demand side while improving the accuracy of the predicted electricity demand of the target power demand side in the current period.

[0160] Based on the above embodiment, in an exemplary embodiment, the acquisition methods of the historical demand parameters and the historical supply parameters in S101 are further refined. Optionally, as Figure 5 shown, the following steps may be included:

[0161] S501, determine the first window period and the second window period of the target period according to the demand window period and the supply window period of the historical corresponding period corresponding to the target period.

[0162] Among them, the power supply prediction result of the historical corresponding period predicted based on the supply window period satisfies the supply prediction condition with the actual value of the power supply in the historical corresponding period, and the power surplus demand prediction result of the historical corresponding period predicted based on the demand window period satisfies the demand prediction condition with the actual value of the power surplus demand in the historical corresponding period.

[0163] Generally, when making data predictions, historical data is used to predict future data. Therefore, in this embodiment, it is necessary to determine the time range before the target period and used to predict the power surplus demand and the power spot price of the target period.

[0164] To determine this time range, first, the historical corresponding period corresponding to the target period can be determined. Among them, the so-called historical corresponding period refers to the period in the historical cycle of the target period that is the same as the target period. For example, if the target period is April 2025, the historical corresponding period corresponding to the target period can be April 2024, April 2023, or April 2022, etc.

[0165] In some alternative embodiments, the historical corresponding period corresponding to the target period in this application is the period in the previous historical cycle of the target period that is the same as the target period. That is, assuming the target period is April 2025, the historical corresponding period is April 2024.

[0166] After that, the supply window period and the demand window period of the above historical same period can be determined. Among them, the above supply window period and demand window period are respectively time ranges composed of multiple consecutive time periods before the above historical same period. For example, assuming that the target time period is April 2025 and the historical same period is April 2024, the above supply window period and demand window period are respectively multiple consecutive months between April 2023 and March 2024. For example, the supply window period is from December 2023 to March 2024, and the demand window period is from January 2024 to March 2024.

[0167] It should be noted that in this embodiment, when predicting the electricity surplus demand and electricity spot price of the target time period, due to the lack of specific quotation data of the power generation units as the power supply side and the difficulty in predicting the future quotation data of the power generation units, it is assumed that the power supply remains unchanged in each time period. Thus, the above supply window period can be determined by using methods such as the grid search method with the help of the high-frequency equilibrium data of the day-ahead market. And due to the high rigidity of electricity demand, the electricity surplus demand can be directly regarded as the result of subtracting the preferentially absorbed electricity from the total electricity demand. Thus, the "error accumulation" problem caused by the possible introduction of additional errors in the separate estimation of the total electricity demand and the preferentially absorbed electricity can be effectively avoided. And similarly, the above demand window period can be determined by using methods such as the grid search method based on the high-frequency equilibrium data of the day-ahead market.

[0168] For example, it is possible to search within the time range that is closest to the above historical same period and has the same cycle duration as the above historical same period before the above historical same period to determine the supply window period and the demand window period of the above historical same period.

[0169] It can be understood that the actual electricity supply value and the actual electricity surplus demand value of the above historical same period are known, and the actual supply parameter values and actual demand parameter values within the time range to which the above supply window period and demand window period belong are also known. Furthermore, the electricity supply of the above historical same period can be predicted based on the actual supply parameter values of different window periods determined from the above time range, and the electricity surplus demand of the above historical same period can be predicted based on the actual demand parameter values of different window periods determined from the above time range. Since the predicted electricity supply prediction results and electricity surplus demand prediction results for different window periods can be different. Therefore, the supply window period can be determined according to the difference between the actual electricity supply value of the above historical same period and the predicted electricity supply prediction result of the above historical same period, and the demand window period can be determined according to the difference between the actual electricity surplus demand value of the above historical same period and the predicted electricity surplus demand prediction result of the above historical same period.

[0170] Optionally, the power supply forecast result of the historical same-period segment predicted based on the actual value of the supply parameter during the supply window period satisfies the supply forecast condition with the actual power supply value of the historical same-period segment. For example, the difference between the power supply forecast result and the actual power supply value of the above historical same-period segment is minimized, etc. The power surplus demand forecast result of the historical same-period segment predicted based on the actual value of the demand parameter during the demand window period satisfies the demand forecast condition with the actual power surplus demand value of the historical same-period segment. For example, the difference between the power surplus demand forecast result and the actual power surplus demand value of the above historical same-period segment is minimized, etc.

[0171] Since the target period and the corresponding historical same-period segment are the same-period segments within different cycles, the factors affecting power supply and power surplus demand, such as natural conditions, economic variables, and physical constraints, are relatively close between the two. Therefore, in the case where the supply window period and the demand window period of the above historical same-period segment are determined, the first window period and the second window period of the target period can be determined according to the above demand window period and supply window period.

[0172] Among them, the above first window period and second window period are respectively time ranges composed of multiple consecutive segments located within the target period. And the first window period can be the same-period segment of the above demand window period. Correspondingly, the second window period can be the same-period segment of the above supply window period.

[0173] The first window period is used to predict the power surplus demand of the target period, while the second window period is used to predict the power supply of the target period.

[0174] Optionally, the first window period and the second window period of the target period can be determined within the time range that is closest to the target period and has the same cycle duration as the target period before the target period. For example, assume that the target period is April 2025, the historical same-period segment is April 2024, the demand window period is from December 2023 to March 2024 within the time range from April 2023 to March 2024, and the supply window period is from January 2024 to March 2024 within the time range from April 2023 to March 2024. Then the first window period is from December 2024 to March 2025 within the time range from April 2024 to March 2025, and the second window period is from January 2025 to March 2025 within the time range from April 2024 to March 2025.

[0175] S502, obtain the historical demand parameter of the first window period and the historical supply parameter of the second window period.

[0176] After determining the above-mentioned first window period and second window period, the historical demand parameters of the first window period and the historical supply parameters of the second window period can be further obtained. Since the historical demand parameters of the first window period and the historical supply parameters of the second window period are known, the above-mentioned historical demand parameters and historical supply parameters can be obtained by means of network search, resource query, etc.

[0177] In this embodiment, since the power supply prediction result of the historical same period predicted based on the supply window period meets the supply prediction condition with the actual power supply value of the historical same period, and the predicted result of the remaining power demand of the historical same period predicted based on the demand window period meets the demand prediction condition with the actual remaining power demand value of the historical same period, therefore, the first window period is determined according to the demand window period, and the second window period is determined according to the supply window period. Thus, the prediction result of the remaining power demand of the target period predicted based on the demand parameters of the first window period can have high accuracy, and the prediction result of the power supply of the target period predicted based on the supply parameters of the second window period can have high accuracy, and finally the predicted result of the spot price of electricity of the target period predicted based on the prediction result of the remaining power demand of the target period and the prediction result of the power supply has high accuracy. That is, in this embodiment, the accuracy of the first prediction result and the second prediction result can be improved, and finally the accuracy of the predicted power demand of the target power demand side at the current time can be improved.

[0178] On the basis of the above embodiment, in an exemplary embodiment, the prediction methods of the first prediction result and the second prediction result in S102 are further refined. Optionally, as Figure 6 shown, the following steps can be included:

[0179] S601, input the historical demand parameters into the demand prediction model to obtain the prediction result of the remaining power demand of the target period as the first prediction result.

[0180] The demand prediction model is trained with the demand parameters as the input and the remaining power demand as the label. Therefore, after obtaining the above-mentioned historical demand parameters, the above-mentioned historical demand parameters can be input into the demand prediction model to obtain the prediction result output by the demand prediction model as the first prediction result. Then this first prediction result is the prediction result of the remaining power demand of the target period.

[0181] S602, determine the power supply prediction result of the target period according to the historical supply parameters.

[0182] After obtaining the above historical supply parameters, the power supply forecast result of the target period can be determined according to the above historical supply parameters. The essence of the power supply forecast result of the target period is the corresponding relationship between the power supply and the power spot price in the target period.

[0183] Optionally, since the power supply forecast results for the above-mentioned target period can be represented by a curve characterizing the corresponding relationship between the power supply and the electricity spot price during the target period, the power supply forecast results for the above-mentioned target period can also be called the power supply curve for the target period, which reflects the changing trend of the electricity spot price with the change of power supply during the target period.

[0184] S603, based on the first prediction result and the power supply prediction result of the target period, predict the power spot price of the target period to obtain a second prediction result.

[0185] Usually, when predicting the spot price of electricity in the target period, the remaining electricity demand in the target period is matched with the electricity supply. Therefore, after obtaining the above-mentioned first prediction result and the electricity supply prediction result of the target period, the above-mentioned first prediction result and the electricity supply prediction result of the target period can be combined to solve the prediction result of the spot price of electricity in the target period as the second prediction result.

[0186] Among them, the essence of the power supply forecast result of the above-mentioned target time period is the corresponding relationship between the power supply and the power spot price during the target time period, and the above-mentioned first forecast result is a certain electricity value. Therefore, in the power supply forecast result of the above-mentioned target time period, when the power supply is the above-mentioned first forecast result, the power spot price corresponding to the power supply can be determined. Then, the above-mentioned determined power spot price is the forecast result of the power spot equilibrium price of the target time period, that is, the second forecast result.

[0187] In this embodiment, the historical demand parameters and the demand prediction model of the target period are used to predict the remaining power demand of the target period, and the historical supply parameters of the target period are used to predict the power supply prediction result of the target period. Furthermore, based on the remaining power demand and the power supply prediction result of the target period obtained from the above prediction, the prediction result of the spot price of electricity in the target period is determined. Then, by separately predicting the power supply and demand sides, the power supply and the remaining power demand of the target period can be estimated, so that the spot price of electricity when the power supply and the remaining power demand are balanced can be derived as the prediction result of the spot price of electricity in the target period. In this way, both the demand parameters and the supply parameters are taken into account when predicting the spot price of electricity, realizing the relationship fitting between the demand parameters, the supply parameters and the prediction result, so that this embodiment can obtain a larger explanation ratio and better extrapolation prediction ability, and improve the accuracy of the spot price of electricity prediction.

[0188] Based on the above embodiment, in an exemplary embodiment, as Figure 7 shown, the training method of the demand prediction model in S601 above is further refined. Optionally, the following steps may be included:

[0189] S701, obtain the demand parameters of each period within the historical time range before the historical same-period period, and the actual value of the remaining power demand in the historical same-period period.

[0190] In this embodiment, since the demand parameters and the actual values of the remaining power demand of each period before the target period are known, the historical time range before the historical same-period period can be determined first. Then, according to the period division method of the target period, the above historical time range can be divided into each period. After that, the demand parameters of each period within the above historical time range and the actual value of the remaining power demand in the historical same-period period can be obtained.

[0191] Optionally, the above historical time range may be the time range that is before the above historical same-period period, closest to the above historical same-period period, and has the same cycle duration as the above historical same-period period.

[0192] For example, assuming that the target period is April 2025 and the historical same-period period is April 2024, the above historical time range is from April 2023 to March 2024, and each period within the above historical time range is a natural month.

[0193] Optionally, if the target period is one month, then each period within the historical time range before the historical same-period period is: each month within the historical time range before the historical same-period period.

[0194] S702. Determine multiple first candidate window periods within the historical time range.

[0195] Among them, each first candidate window period includes multiple consecutive time periods, and the time periods included in different first candidate window periods are not exactly the same.

[0196] As mentioned above, if the above historical time range is divided into each time period, then multiple consecutive time periods can be divided into one window period as a candidate window period for the demand window period. Thus, multiple first candidate window periods are obtained, and the time periods included in different first candidate window periods are not exactly the same. In this way, multiple first candidate window periods can be determined within the historical time range.

[0197] For example, if the above historical time range is divided into N time periods from the earliest to the latest in order, namely periods 1 - N, then periods 1 - 2, 1 - 3... 1 - N, 2 - 3, 2 - 4... 2 - N, 3 - 4... (N - 1) - N can be respectively determined as a first candidate window period, obtaining first candidate window periods. Among them, N is a positive integer not less than 2.

[0198] Of course, in some cases, at least two window periods can also be selected from the window periods that can be formed within the above historical time range as the first candidate window periods.

[0199] S703. Use the demand parameters of each first candidate window period as the input and the actual value of the remaining power demand in the historical same - period time period as the label to train a preset model to obtain a demand prediction model.

[0200] The demand parameters of each first candidate window period are the demand parameters of each time period included in this first candidate window period. After determining multiple first candidate window periods, the preset model can be trained using the demand parameters of each first candidate window period as the input and the actual value of the remaining power demand in the historical same - period time period as the label.

[0201] Among them, during the training process of the above - mentioned preset model, the preset model will learn the relationship between the demand parameters of each first candidate window period and the actual value of the remaining power demand in the historical same - period time period, establish a prediction method for predicting the remaining power demand in the historical same - period time period based on the demand parameters of each first candidate window period, and through parameter adjustment, make the predicted result of the remaining power demand in the historical same - period time period gradually approach the actual value of the remaining power demand in the historical same - period time period until the loss value between the two satisfies the preset loss - value condition. For example, the loss value between the two is less than a preset threshold, and a demand prediction model is obtained.

[0202] Optionally, an extreme gradient boosting model can be selected as the above-mentioned preset model to train the above-mentioned demand prediction model. Among them, the extreme gradient boosting model performs excellently in dealing with non-linear relationships, the efficiency when the data scale is large, and preventing overfitting, and is suitable for the prediction requirements of this embodiment.

[0203] Optionally, the above-mentioned preset model for training the demand prediction model can also be other machine learning methods such as the LASSO (Least Absolute Shrinkage and Selection Operator) model, the Random Forest (RF) model, etc.

[0204] In this way, in this embodiment, the demand prediction model can predict the prediction result of the electricity surplus demand in the target period based on the demand parameters in the first window period according to the above-established prediction method for predicting the electricity surplus demand, and improve the accuracy of the target prediction result of the electricity spot price in the target period predicted by using the prediction result of the electricity surplus demand in the target period.

[0205] Based on the above embodiments, in an exemplary embodiment, the determination method of the demand window period in S501 is refined. Optionally, as Figure 8 shown, it may include the following steps:

[0206] S801. For each candidate window period, input the demand parameters of the first candidate window period into the demand prediction model to obtain the demand prediction result of the first candidate window period.

[0207] As described above, multiple first candidate window periods can be determined from the historical time range before the historical same period, and the demand parameters of each first candidate window period can be obtained.

[0208] In this embodiment, after training the above-mentioned demand prediction model, for each candidate window period, the demand parameters of the first candidate window period can be input into the above-mentioned demand prediction model to obtain the prediction result output by the above-mentioned demand prediction model. Then, this prediction result is the prediction result of the electricity surplus demand in the historical same period predicted based on the demand parameters of this first candidate window period. Furthermore, this prediction result can be used as the demand prediction result of this first candidate window period.

[0209] S802. Respectively determine the root mean square error between the demand prediction result of each first candidate window period and the actual value of the electricity surplus demand in the historical same period.

[0210] Since the predicted result of the electricity residual demand in the historical same period based on the predicted demand window period meets the demand prediction condition with the actual value of the electricity residual demand in the historical same period, the root mean squared error can be used as the judgment basis for whether the demand prediction condition is met. For each first candidate window period, the root mean squared error (RMSE, Root Mean Squared Error) between the predicted result of the demand in this first candidate window period and the actual value of the electricity residual demand in the historical same period can be determined, and the calculated root mean squared error is used as the root mean squared error corresponding to this first candidate window period.

[0211] S803. Determine the first candidate window period corresponding to the minimum root mean squared error as the demand window period of the historical same period.

[0212] If the above demand prediction condition is preset as the minimum root mean squared error, the first candidate window period corresponding to the minimum root mean squared error can be determined as the demand window period of the historical same period.

[0213] In this embodiment, since the root mean squared error between the predicted result of the electricity residual demand in the historical same period determined by the determined demand window period and the actual value of the electricity residual demand in the historical same period is the smallest, among the above multiple first candidate window periods, the predicted result of the electricity residual demand in the historical same period determined based on the demand parameters of the demand window period is closest to the actual value of the electricity residual demand in the historical same period, that is, the accuracy of the predicted result of the electricity residual demand in the historical same period determined based on the demand parameters of the demand window period is the highest. Therefore, the first candidate window period corresponding to the minimum root mean squared error is determined as the demand window period of the historical same period, and the above first window period is determined according to the demand window period. Thus, the predicted result of the electricity residual demand in the target period predicted based on the demand parameters of the above first window period can have higher accuracy, and thus the accuracy of the target predicted result of the electricity spot price in the target period predicted using the predicted result of the electricity residual demand in the target period can be improved.

[0214] Based on the above embodiment, in an exemplary embodiment, the determination method of the supply window period in S501 above is refined. Optionally, as Figure 9 shown, the following steps may be included:

[0215] S901. Obtain the supply parameters in the historical same period, the supply parameters in each period within the historical time range before the historical same period, and the actual electricity supply value in the historical same period.

[0216] In this embodiment, since the supply parameters and the actual power supply values for each period before the target period are known, the historical time range before the historical same-period can be determined first. Then, according to the period division method of the target period, the above historical time range can be divided into each period. After that, the supply parameters of the above historical same-period, the supply parameters of each period within the above historical time range, and the actual power supply value of the historical same-period can be obtained.

[0217] S902. Determine a plurality of second candidate window periods from the historical time range.

[0218] Wherein, each second candidate window period includes a plurality of consecutive periods, and the periods included in different second candidate window periods are not completely the same.

[0219] As described above, since the above historical time range is divided into each period, a plurality of consecutive periods can be divided into one window period as a candidate window period of the supply window period, so that a plurality of second candidate window periods are obtained, and the periods included in different second candidate window periods are not completely the same. In this way, a plurality of second candidate window periods can be determined from the historical time range.

[0220] For example, if the above historical time range is divided into a total of N periods from 1 to N in the order of time from early to late, then the periods 1-2, 1-3... 1-N, 2-3, 2-4... 2-N, 3-4... (N-1)-N can be respectively determined as one second candidate window period, obtaining second candidate window periods. Wherein, N is a positive integer not less than 2.

[0221] Of course, in some cases, at least two window periods can also be selected from the window periods that can be formed within the above historical time range as the second candidate window periods.

[0222] S903. According to the supply prediction formula, the supply parameters of each second candidate window period, and the supply parameters of the historical same-period, determine the supply prediction result of each second candidate window period.

[0223] The supply parameters of each second candidate window period are the supply parameters of each period included in this second candidate window period. After determining a plurality of second candidate window periods, for each second candidate window period, the power supply prediction result of the historical same-period can be determined according to the preset supply prediction formula, the supply parameters of this second candidate window period, and the supply parameters of the historical same-period as the supply prediction result of this second candidate window period.

[0224] In an exemplary embodiment, the method for determining the supply prediction result of each second candidate window period in S903 above is refined. Optionally, it may include the following steps:

[0225] Step 1, for each second candidate window period, solve the parameters of the supply prediction formula according to the supply parameters of the second candidate window period to obtain the prediction formula of the second candidate window period.

[0226] In this embodiment, the pre-set supply prediction formula includes unknown parameters and variables. Therefore, for each second candidate window period, the parameters of the supply prediction formula can be solved according to the supply parameters of the second candidate window period to obtain the parameter values of the unknown parameters in the supply prediction formula. Then, by assigning the above parameter values to the supply prediction formula, the prediction formula of the above variables corresponding to the second candidate window period can be obtained as the prediction formula of the second candidate window period.

[0227] Step 2, assign values to the variables of the prediction formula of the second candidate window period by using the supply parameters of the historical same period to obtain the supply prediction result of the second candidate window period.

[0228] After obtaining the prediction formula of each second candidate window period, for each second candidate window period, the supply parameters of the above historical same period can be used to assign values to the variables of the prediction formula of the second candidate window period, that is, the supply parameters of the above historical same period are used as variables and substituted into the prediction formula of the second candidate window period to calculate the calculation result of the prediction formula of the second candidate window period. Then, the calculation result is the power supply prediction result of the historical same period predicted according to the prediction formula of the second candidate window period, as the supply prediction result of the second candidate window period.

[0229] S904, respectively determine the root mean square error between the supply prediction result of each second candidate window period and the actual value of the power supply in the historical same period.

[0230] Since the power supply prediction result of the historical same period predicted based on the supply window period satisfies the supply prediction condition with the actual value of the power supply in the historical same period, the root mean square error can be used as the judgment basis for whether the supply prediction condition is satisfied. Therefore, for each second candidate window period, the root mean square error between the power supply of the second candidate window period and the actual value of the power supply in the historical same period can be determined, and the calculated above root mean square error is used as the root mean square error corresponding to the second candidate window period.

[0231] S905, determine the second candidate window period corresponding to the smallest root mean square error as the supply window period of the historical same period.

[0232] If the above demand prediction conditions are preset to minimize the root mean square error, the first candidate window period corresponding to the minimum root mean square error can be determined as the demand window period for the historical same period.

[0233] In this embodiment, since the root mean square error between the predicted result of the power supply for the historical same period determined by the determined supply window period and the actual value of the power supply for the historical same period is the smallest, among the above-mentioned multiple first and second candidate window periods, the predicted result of the power supply for the historical same period determined based on the supply parameters of the supply window period is closest to the actual value of the power supply for the historical same period, that is, the accuracy of the predicted result of the power supply for the historical same period determined based on the demand parameters of the supply window period is the highest. Therefore, the second candidate window period corresponding to the minimum root mean square error is determined as the supply window period for the historical same period, and the above-mentioned second window period is determined according to the supply window period. Thus, the predicted result of the power supply for the target period predicted based on the supply parameters of the above-mentioned second window period can have high accuracy, thereby improving the accuracy of the target prediction result of the spot price of electricity for the target period predicted using the predicted result of the power supply for the target period.

[0234] In addition, the above-mentioned multiple first candidate window periods and the above-mentioned multiple second window periods may include the same multiple window periods, or may include multiple window periods that are not completely the same. For example, the above-mentioned multiple first candidate window periods and the above-mentioned multiple second window periods include completely different multiple window periods, or among the above-mentioned multiple first candidate window periods and the above-mentioned multiple second window periods, there are some identical window periods and some different window periods.

[0235] Based on this, the determined demand window period and supply window period may be the same or different.

[0236] On the basis of the above embodiment, in an exemplary embodiment, the determination method of the predicted result of the power supply for the target period in S602 may be to solve the parameters of the supply prediction formula according to the historical supply parameters to obtain the predicted result of the power supply for the target period.

[0237] Similar to the determination process of the predicted result of the power supply for each of the above-mentioned second candidate window periods, in this implementation, the preset supply prediction formula includes unknown parameters and variables. Then, according to the historical supply parameters of the second window period for the target period, the parameters of the supply prediction formula can be solved to obtain the parameter values of the unknown parameters in the supply prediction formula. Then, using the above parameter values to assign parameters to the above supply prediction formula, the supply prediction formula for the above variables corresponding to the target period can be obtained as the predicted result of the power supply for the target period.

[0238] Thus, in this embodiment, by using the above supply prediction method and combining the historical supply parameters of the second window period of the target period, the power supply of the target period can be predicted. Considering that the second window period of the target period is determined based on the supply window period of the corresponding historical same period of the target period, and the power supply prediction result of the above historical same period predicted based on the above supply window period meets the supply prediction condition with the actual value of the power supply of the above historical same period, therefore, by means of the above historical supply parameters, the accuracy of the power supply prediction result of the target period obtained can be improved. Furthermore, the accuracy of the target prediction result of the power spot price of the target period predicted by using the power supply prediction result of the target period can be improved.

[0239] Based on the above embodiments, in an exemplary embodiment, optionally, as Figure 10 shown, the electricity quantity prediction method may include the following steps:

[0240] S1001. Construct a revenue function on the power supply side and a revenue function on the power demand side with the power residual demand and the power spot price as variables.

[0241] S1002. With the revenue on the power supply side as a variable, based on the expected value function and the variance value function corresponding to the revenue function on the power supply side, construct a first optimization function of the revenue function on the power supply side.

[0242] S1003. With the goal of maximizing the function value of the first optimization function, solve the first optimization function to obtain the optimal signing quantity function on the power supply side.

[0243] S1004. With the revenue on the power demand side as a variable, based on the expected value function and the variance value function corresponding to the revenue function on the power demand side, construct a second optimization function of the revenue function on the power demand side.

[0244] S1005. With the goal of maximizing the function value of the second optimization function, solve the second optimization function to obtain the optimal signing quantity function on the power demand side.

[0245] S1006. Determine the price function of the power equilibrium price when the function values of the optimal signing quantity function on the power supply side and the optimal signing quantity function on the power demand side are equal.

[0246] S1007. Use the price function to assign values to the variables of the optimal signing quantity function on the power demand side to obtain the electricity quantity prediction function.

[0247] S1008. According to the demand window period and the supply window period of the corresponding historical same period of the target period, determine the first window period and the second window period of the target period.

[0248] S1009, Obtain the historical demand parameters for the first window period and the historical supply parameters for the second window period.

[0249] S1010, Input the historical demand parameters into the demand forecasting model to obtain the forecasting result of the remaining electricity demand for the target period, which is used as the first forecasting result.

[0250] S1011, Determine the electricity supply forecasting result for the target period based on the historical supply parameters.

[0251] S1012, Forecast the spot electricity price for the target period based on the first forecasting result and the electricity supply forecasting result for the target period to obtain the second forecasting result.

[0252] S1013, Solve the electricity quantity forecasting function based on the first forecasting result, the second forecasting result, the global electricity parameters, and the local electricity parameters of the target electricity demand side to obtain the electricity demand quantity of the target electricity demand side for the target period.

[0253] It should be noted that the specific implementation manners of the above S1001 - S1013 are the same as those in the above embodiments and will not be elaborated herein.

[0254] Based on the above embodiments, in an exemplary embodiment, assuming that when the electricity market reaches equilibrium, represents the remaining electricity demand on the electricity demand side, and q represents the electricity supply on the electricity supply side. Then, when the electricity market reaches equilibrium, Therefore, the spot electricity price p at the equilibrium of the electricity market can be obtained. s , and since then p s can be expressed as Additionally, p u represents the global electricity price on the electricity consumption side.

[0255] It can be understood that in the actual application scenario, taking monthly forecasting as an example, according to the time sequence, the electricity demand side can first determine its signing strategy in the monthly medium - and long - term contract market (i.e., determine the electricity demand quantity) based on the forecasting situation of the electricity market one month later. Among them, the monthly medium - and long - term contract market is an over - the - counter market, and the equilibrium price of the medium - and long - term contract is determined by the bids of the electricity demand side and the electricity supply side, that is, bilateral negotiation transactions. When the electricity demand side and the electricity supply side both make optimal decisions, an electricity transaction price (i.e., the electricity equilibrium price) can be formed in the electricity market, and the electricity market reaches equilibrium.

[0256] Since In this embodiment, it is assumed that the power supply capacity of the power supply side is completely flexible, and its power supply quantity to the power market completely depends on the remaining electricity demand of the power demand side.

[0257] 1) Construct a revenue function for the power supply side with the residual power demand and the spot power price as variables. Among them, the above-mentioned revenue function for the power supply side is shown in the following formula (1).

[0258]

[0259] Among them, π p represents the revenue of the power supply side, f p represents the contracted quantity of the power supply side, p f represents the power equilibrium price, p s represents the spot power price at the power market equilibrium, represents the residual power demand of the power demand side, is the power generation cost of the power supply side.

[0260] As shown in the above formula (1), f p p f represents the immediate cash flow of the power supply side for signing medium- and long-term power, represents the cash flow of the power supply side for supplying the remaining part of the power in the power spot market, and

[0261] Furthermore, with the revenue of the power supply side as a variable, based on the expected value function and the variance value function corresponding to the revenue function of the power supply side, construct the first optimization function of the revenue function of the power supply side. Among them, the above-mentioned first optimization function is shown in the following formula (2).

[0262]

[0263] Among them, π p represents the revenue of the power supply side, f p represents the contracted quantity of the power supply side, R represents the set of real numbers, E(π p ) represents the expected value function corresponding to the revenue function of the power supply side, Var(π p ) represents the variance value function corresponding to the revenue function of the power supply side, λ p represents the global risk aversion level of the power supply side.

[0264] And,

[0265]

[0266] Among them, π p represents the revenue of the power supply side, p s represents the spot power price at the power market equilibrium, represents the residual power demand of the power demand side, fp Denote the quantity contracted by the power supply side as \(p\). f Denote the power equilibrium price as is the power generation cost of the power supply side. \(E\) represents the expected value calculation, \(Var\) represents the variance value calculation, and \(cov\) represents the covariance value calculation.

[0267] Obviously, after substituting the specific contents of the above \(E(\pi p )\) and \(Var(\pi p )\) into the above first optimization function, the variables of the obtained first optimization function include: the quantity contracted by the power supply side \(f p \).

[0268] Furthermore, aiming at maximizing the function value of the above first optimization function, solve the above first optimization function to obtain the optimal quantity contracting function of the power supply side.

[0269] Among them, after substituting the specific contents of the above \(E(\pi p )\) and \(Var(\pi p )\) into the above first optimization function, it can be determined that \(V(\pi p )\) is a concave function with respect to \(f p \). Then, when maximizing \(V(\pi p )\), it can be set that Then the optimal quantity contracting function of the power supply side can be solved. Among them, the optimal quantity contracting function of the power supply side is shown in the following formula (3).

[0270]

[0271] Among them, Denote the optimal quantity contracted by the power supply side as \(p f Denote the power equilibrium price as \(p s Denote the power spot price at the power market equilibrium as Denote the remaining power demand on the power demand side as is the power generation cost of the power supply side, and \(\lambda p Denote the global risk aversion level of the power supply side. \(E\) represents the expected value calculation, \(Var\) represents the variance value calculation, and \(cov\) represents the covariance value calculation.

[0272] In the above optimal quantity contracting function of the power supply side, the first item on the right side of the equal sign reflects the speculative motivation of the power supply side in the medium and long-term market: \(p f - E(p s ) represents the difference between the actual medium and long-term power price (the power equilibrium price obtained in subsequent solutions) and the predicted power spot price. If the medium and long-term price to be contracted currently is higher than the predicted power spot price, the power supply side will increase the electricity quantity sold through medium and long-term contracts. \(\lambda prepresents the global risk aversion level on the power supply side. When the value of λ p is positive, the larger λ p is, the less the power supply side is willing to bear the medium- and long-term premium (the price difference between the medium- and long-term power price and the power spot price), and the less the quantity of medium- and long-term contracts signed by the power supply side.

[0273] Correspondingly, in the above optimal signing quantity function of the power supply side, the second term on the right side of the equal sign reflects the hedging motivation: under the assumption of not participating in the medium- and long-term power market, the stronger the correlation between the expected power sales profit of the power supply side and the power spot price, the stronger the willingness of the power supply side to take a long position in signing medium- and long-term power contracts to avoid the uncertainty risk of the power spot market, that is, the less medium- and long-term power contracts signed by the power supply side. Then this part of the position (i.e., the medium- and long-term power contracts signed by the power supply side) is the same as the position taken for speculative motives (the medium- and long-term power contracts signed by the power supply side for speculative motives), and both are adjusted by the predicted fluctuation range of the power spot price Var(p s ). That is to say, it can be understood that the larger the predicted fluctuation range of the power spot price, the lower the power supply side's grasp of the realized power spot price in actual power spot transactions, and the fewer monthly medium- and long-term contracts signed by the power supply side.

[0274] 2) Construct the revenue function of the power demand side with the power residual demand and the power spot price as variables. Among them, the above revenue function of the power demand side is shown in the following formula (4).

[0275]

[0276] Among them, π r represents the revenue of the power demand side, p f represents the power equilibrium price, f r represents the signing quantity of the power demand side, p s represents the power spot price at the power market equilibrium, represents the power residual demand of the power demand side, p u represents the global electricity price on the electricity consumption side.

[0277] As shown in the above formula (4), -p f f r is the immediate cash flow for the power demand side to sign medium- and long-term power, is the cash flow for the power supply side to buy and sell electricity in the power spot market to exactly meet the remaining part of the power demand.

[0278] Furthermore, taking the revenue of the electricity demand side as a variable, based on the expected value function and variance value function corresponding to the revenue function of the electricity demand side, a second optimization function of the revenue function of the electricity demand side is constructed. Among them, the above-mentioned second optimization function is shown in the following formula (5).

[0279]

[0280] Among them, π r represents the revenue of the electricity demand side, f r represents the contracted quantity of the electricity demand side, R represents the set of real numbers, E(π r ) represents the expected value function corresponding to the revenue function of the electricity demand side, Var(π r ) represents the variance value function corresponding to the revenue function of the electricity demand side, and λ r represents the global risk aversion level of the electricity demand side.

[0281] Among them,

[0282]

[0283] Among them, π r represents the revenue of the electricity demand side, p s represents the electricity spot price at the electricity market equilibrium, represents the remaining electricity demand of the electricity demand side, f r represents the contracted quantity of the electricity demand side, p f electricity equilibrium price, p u represents the global electricity price on the electricity consumption side, E represents the calculation of the expected value, Var represents the calculation of the variance value, and cov represents the calculation of the covariance value.

[0284] Obviously, after substituting the specific contents of the above E(π r ) and Var(π r ) into the above second optimization function, the variables of the second optimization function obtained include: the contracted quantity f r of the electricity demand side.

[0285] Furthermore, aiming at maximizing the function value of the above second optimization function, the above first optimization function is solved to obtain the optimal contracted quantity function of the electricity demand side.

[0286] Among them, after substituting the specific contents of the above E(π r ) and Var(π r ) into the above second optimization function, it can be determined that U(π r ) is a concave function with respect to f r , then when maximizing U(π r ), it can be set Then the optimal contracting quantity function of the electricity demand side can be solved. Among them, the optimal contracting quantity function of the electricity demand side is shown in the following formula (6).

[0287]

[0288] Among them, represents the optimal contracting quantity of the electricity demand side, p s represents the electricity spot price at the equilibrium of the electricity market, p f represents the electricity equilibrium price, p u represents the global electricity price on the electricity consumption side, represents the remaining electricity demand of the electricity demand side, λ r represents the global risk aversion level of the electricity demand side. E represents the calculation of the expected value, Var represents the calculation of the variance value, and cov represents the calculation of the covariance value.

[0289] In the above optimal contracting quantity function of the electricity demand side, the first term on the right side of the equal sign reflects the speculative motive of the electricity demand side in the medium and long-term market: E(p s ) - p f represents the difference between the predicted electricity spot price and the actual medium and long-term electricity price (the electricity equilibrium price obtained by subsequent solution). If the predicted electricity spot price is higher than the actual medium and long-term electricity price, the electricity demand side signs more long positions in medium and long-term electricity contracts, that is, the electricity demand side buys more medium and long-term electricity contracts. Conversely, if the predicted electricity spot price is lower than the actual medium and long-term electricity price, the electricity demand side signs more short positions in medium and long-term electricity contracts, that is, the electricity demand side buys and sells more medium and long-term electricity contracts. λ r represents the global risk aversion level of the electricity demand side, that is, the risk aversion level of the electricity demand side. λ r reflects the intensity of the willingness of the electricity demand side to participate in speculation by using the medium and long-term premium.

[0290] Correspondingly, in the above optimal contracting quantity function of the electricity demand side, the second term on the right side of the equal sign reflects the hedging motive: on the premise of not participating in the medium and long-term electricity market, if the correlation between the expected electricity sales profit of the electricity demand side and the electricity spot price is negative, the higher the above negative correlation degree, the stronger the willingness of the electricity demand side to sign long positions in medium and long-term electricity contracts. Similarly, both the medium and long-term electricity contracts signed by the electricity demand side and the medium and long-term electricity contracts signed by the electricity demand side for speculative motives may be reduced due to the large fluctuation Var(p s ) of the predicted electricity spot price and the lack of control of the electricity demand side over the realized electricity spot price in the actual electricity spot transaction.

[0291] 3) After obtaining the optimal contract quantity functions for the power supply side and the power demand side as described above, the price function of the power equilibrium price can be determined when the function values of the optimal contract quantity function for the power supply side and the optimal contract quantity function for the power demand side are equal.

[0292] When the power market reaches equilibrium, the optimal contract quantity on the power demand side as described above is equal to the optimal contract quantity representing the power supply side, that is Then substitute into the above equations (3) and (6), and the price function of the power equilibrium price can be solved. Among them, the price function of the above power equilibrium price is shown in equation (7) below.

[0293]

[0294] Among them, represents the power equilibrium price at the power market equilibrium, p s represents the power spot price at the power market equilibrium, λ p represents the global risk aversion level of the power supply side, λ r represents the global risk aversion level of the power demand side, p u represents the global electricity price on the electricity consumption side, represents the remaining power demand on the power demand side, is the power generation cost of the power supply side, E represents the expected value calculation, and cov represents the covariance value calculation.

[0295] As shown in the above equation (7), the power equilibrium price at the power market equilibrium is based on the general expectation of the power market for the power spot price, and at the same time, with the harmonic mean of the global risk aversion levels of the power supply side and the power demand side as the weight, it is adjusted by combining the correlation between the power spot price and the overall revenue of all power market participants (i.e., the power supply side, the power demand side, and the electricity consumption side).

[0296] And, substitute into the above equations (3) and (6), and the contract quantity function of the medium- and long-term power contract quantity can also be solved. Among them, the contract quantity function of the above medium- and long-term power contract quantity is shown in equation (8) below.

[0297]

[0298] Among them, represents the medium- and long-term power contract quantity at the power market equilibrium, p s represents the power spot price at the power market equilibrium, λ p represents the global risk aversion level of the power supply side, λ r represents the global risk aversion level of the power demand side, is the generation cost on the power supply side, p u represents the global electricity price on the electricity consumption side, Var represents the variance value calculation, and cov represents the covariance value calculation.

[0299] As shown in the above formula (8), reflects the correlation between the power supply side and the electricity spot price, reflects the correlation between the power demand side and the electricity spot price, reflects the correlation between the electricity consumption side and the electricity spot price, and moreover, the above multiple correlations are all adjusted by the volatility of the predicted electricity spot price.

[0300] 4) After obtaining the price function of the above power equilibrium price, using the above price function and local power parameter variables, variable assignment is performed on the optimal contract quantity function of the above power demand side to obtain the electricity quantity prediction function.

[0301] According to the local power parameters of the power demand side, local power parameter variables are set. Among them, λ r,i represents the local risk aversion level of the i-th power demand side, p u,i represents the local electricity price of the i-th power demand side, τ represents the predicted proportion of the electricity demand of the i-th power demand side, that is, the proportion of the demand electricity quantity to be contracted by the i-th power demand side in the remaining power demand on the power demand side in, and moreover, 0 < τ < 1, and for different power demand sides, τ can be a constant.

[0302] In this way, using the price function of the power equilibrium price represented by the above formula (7), and the above local power parameter variables, variable assignment is performed on the optimal contract quantity function of the power demand side shown in the above formula (6), and the electricity quantity prediction function can be obtained. Among them, the above electricity quantity prediction function is as shown in the following formula (9).

[0303]

[0304] Among them, represents the electricity demand of the i-th power demand side, λ r,i represents the local risk aversion level of the i-th power demand side, λ p represents the global risk aversion level of the power supply side, λ r represents the global risk aversion level of the power demand side, p s represents the electricity spot price at the power market equilibrium, p u represents the global electricity price on the electricity consumption side, the remaining power demand on the power demand side p u,iDenote the local electricity price of the \(i\) -th electricity demand side, \(\tau\) represents the predicted proportion of the electricity demand volume of the \(i\) -th electricity demand side, Var represents the calculation of the variance value, and cov represents the calculation of the covariance value.

[0305] In summary, in this embodiment, after predicting the remaining electricity demand in the target period to obtain the first prediction result, and predicting the spot electricity price in the target period to obtain the second prediction result, the above - mentioned first prediction result can be used as the variable value in the above formula (9), the above - mentioned second prediction result can be used as \(p\) in the above formula (9) s variable value, the local risk - aversion level of the target electricity demand side can be used as \(\lambda\) in the above formula (9) r,i variable value, the local electricity price of the target electricity demand side can be used as \(p\) in the above formula (9) u,i variable value, the predicted proportion of the electricity demand volume of the target electricity demand side can be used as the variable value of \(\tau\) in the above formula (9), the global risk - aversion level on the power supply side can be used as \(\lambda\) in the above formula (9) p variable value, the global risk - aversion level on the power demand side can be used as \(\lambda\) in the above formula (9) r variable value, and the global electricity price on the power consumption side can be used as \(p\) in the above formula (9) u variable value, assign variable values to the above formula (9) and solve Then the finally obtained solution result is the electricity demand volume of the target electricity demand side in the current period.

[0306] Based on the formulas (1)-(9) provided in the above - mentioned embodiment, optionally, considering that electricity demand sides with different positions can have different local risk - aversion levels, local electricity prices, and predicted proportions of electricity demand volume, it can be simply assumed that \(\tau = 1\). Then the local risk - aversion level of the electricity demand side is the same as the global risk - aversion level on the power demand side, and the local electricity price of the electricity demand side is the same as the global electricity price on the user side. Thus, Define the relative profitability \(\Delta\pi\) as the revenue increase amplitude of the electricity demand side predicting the electricity demand volume according to this embodiment compared with the situation where the electricity demand side does not sign any medium - and - long - term contracts and all purchases electricity demand volume on the electricity spot market as needed. Then substitute the optimal signing quantity function on the power demand side shown in the above formula (6) into the relative profitability \(\Delta\pi\), and the expressions of the value of \(\Delta\pi\) and the variance value of \(\Delta\pi\) (\(Var(\Delta\pi)\)) can be obtained. The above \(\Delta\pi\) and \(Var(\Delta\pi)\) are shown in the following formulas (10) and (11) respectively.

[0307]

[0308] Next, based on the formulas (1)-(11) provided in the above embodiments, the empirical analysis and numerical simulation of this embodiment are carried out using the electricity market data of a certain region as a representative sample. Based on the monthly electricity residual demand and electricity spot price prediction methods, combined with the real historical transaction data of the electricity market in this region, key parameters are fitted, and the improvement effect of this embodiment on the profitability of electricity demand sides with a general risk aversion level is tested.

[0309] Among them, by substituting conditions such as the medium- and long-term electricity price actually formed in the electricity market in this region, the medium- and long-term electricity contract volume, the cost function on the electricity supply side, the electricity price on the electricity consumption side, the expected value and variance value of the electricity spot price, and the covariance value corresponding to the electricity spot price into the above formulas, the global risk aversion level (average risk aversion level) λ of the electricity supply side can be inversely deduced. p =1.51*10 -4 , and the global risk aversion level (average risk aversion level) λ of the electricity demand side r =2.41*10 -4 . Furthermore, by substituting the electricity demand side with τ = 1, and Δπ can be calculated. Compared with the benchmark scenario where the electricity demand side does not sign any medium- and long-term contracts and all purchases electricity demand through the electricity spot market as needed, the solution of this embodiment significantly improves the overall revenue ability of the electricity demand side in the electricity market. The overall revenue increase compared with the above benchmark scenario is as high as 93.83%. The strategy performances of the above benchmark scenario and the solution of this embodiment are shown in Table 3 below.

[0310] Table 3 Strategy performances of the benchmark scenario and the solution of this embodiment

[0311]

[0312] Among them, when the global electricity price (p u ) on the electricity consumption side is fixed, the relative profitability of the electricity demand side is affected by the global risk aversion level (λ r ) of the electricity demand side and the electricity equilibrium price (p f ). According to the above formula (7), the electricity equilibrium price (p f ) is affected by the global risk aversion level (λ r ) of the electricity demand side and the global risk aversion level (λ p ) of the electricity supply side. For the above global risk aversion level (λ r ) of the electricity demand side and the electricity equilibrium price (p f ), the global risk aversion level (λ rThe median, as well as the weighted average of monthly electricity contract signings. Thus, in order to verify the robustness of the embodiments of the present application, λ r is set in the value range of [1, 5], and p f is set in the value range of [450, 550]. And the relationship between the electricity demand quantity, relative profitability level and its standard deviation of the electricity demand side predicted by the solution of this embodiment and the variation of relevant parameters is analyzed.

[0313] Specifically, as Figure 11 shown, it is a schematic diagram of the value of Δπ under the grid values of λ r and p f . Among them, when the local risk aversion level of the retailer (electricity demand side) and the true medium- and long-term electricity contract price are both in the lower range, the relative profitability of the electricity demand quantity prediction method (management strategy) of the electricity demand side in this embodiment performs best. When the true medium- and long-term electricity contract price is too high, far exceeding the predicted electricity spot price level, the electricity demand side can only ensure a positive relative profitability at a lower local risk aversion level. If the electricity demand side cannot sign a sufficient quantity of electricity demand (contract positions) in the medium- and long-term electricity market that significantly deviates from the electricity spot price, it will instead make the final revenue situation lower than the above benchmark situation.

[0314] As Figure 12 shown, it is a schematic diagram of the standard deviation of Δπ under the grid values of λ r and p f , which shows the variation of the standard deviation of the relative profitability of electricity demand with λ r and p f by using the electricity demand quantity prediction method of the electricity demand side in this embodiment. Among them, the fluctuation of the profitability of electricity demand is mainly affected by the global risk aversion level on the electricity demand side, and the bargaining power corresponding to the medium- and long-term electricity price has an insignificant impact on the profit fluctuation.

[0315] Combining Figure 11 and Figure 12 , it can be seen that when other conditions remain unchanged, the relative profitability of the electricity demand side increases with the increase of the local risk aversion level of the electricity demand side, but the revenue of the electricity demand side also increases with the increase of the local risk aversion level of the electricity demand side.

[0316] As Figure 13 shown, it is for λ r and p fSchematic diagram of the proportion of the electricity demand under grid value-taking in the total electricity delivery that the electricity demand side should ultimately satisfy. Among them, the electricity demand predicted by the electricity demand prediction method of this embodiment, the proportion in the total final electricity demand of the electricity demand side falls within the range of 30% to 90%, which is consistent with the general reality of the above-mentioned region. When λ r is low, the deviation degrees between the real electricity spot price and the predicted electricity spot price reach the positive and negative extreme values respectively corresponding to the negative and positive extreme values of the electricity demand of the electricity demand side. That is, when the real electricity spot price is higher or lower than the predicted electricity spot price, the electricity demand of the electricity demand side changes in the opposite direction with λ r . This is because the relative magnitude relationship between the real electricity spot price and the predicted electricity spot price determines the sign of the contract position (electricity demand) of the electricity demand side due to speculative motives, as shown in the above formula (6). And when λ r is high, the absolute value of the electricity demand signed by the electricity demand side due to speculative motives decreases, making the predicted electricity demand of the electricity demand side converge to the direction of the hedging motive position of the electricity demand side in both cases where λ r is low and λ r is high.

[0317] Among them, in Figures 11 - 13 , the risk aversion coefficient of the retailer refers to the local risk aversion level of the electricity demand side, and the medium- and long-term price refers to the medium- and long-term electricity price in the electricity market. And, Figure 11 "Scenario 4: 93.83%" in Figure 12 refers to the relative profitability of the solution of this embodiment, Figure 13 "Scenario 4: 3.58%" in

[0318] refers to the variance value of the relative profitability of the solution of this embodiment, "Scenario 4: 64.57%" in

[0318] should be understood that, although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the indication of the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same moment, but can be executed at different moments. The execution order of these steps or stages does not necessarily need to be sequential, but can be executed alternately or alternately with at least a part of other steps or steps or stages in other steps.

[0319] Based on the same inventive concept, an embodiment of the present application further provides an electricity consumption prediction device for implementing the electricity consumption prediction method involved above. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the electricity consumption prediction device provided below can refer to the limitations on the electricity consumption prediction method in the above text, and will not be elaborated here.

[0320] In an exemplary embodiment, as Figure 14 shown, an electricity consumption prediction device is provided, including:

[0321] A parameter acquisition module 1410, configured to acquire historical demand parameters and historical supply parameters of a target period; wherein, the target period is the next period of the current period;

[0322] A result prediction module 1420, configured to predict the remaining electricity demand of the target period based on the historical demand parameters to obtain a first prediction result, and predict the electricity spot price of the target period based on the first prediction result and the historical supply parameters to obtain a second prediction result;

[0323] An electricity consumption determination module 1430, configured to determine the electricity consumption of the target electricity demand side in the current period according to the first prediction result, the second prediction result, the global electricity parameters, and the local electricity parameters of the target electricity demand side; wherein, the global electricity parameters include the global risk aversion level on the electricity supply side, the global risk aversion level on the electricity demand side, and the global electricity price on the electricity consumption side, and the local electricity parameters of the target electricity demand side include the local risk aversion level, the local electricity price, and the electricity consumption prediction ratio of the target electricity demand side.

[0324] In an exemplary embodiment, the electricity consumption determination module 1430 is specifically configured to:

[0325] Acquire an electricity consumption prediction function; wherein, the electricity consumption prediction function is used to describe the influence of the remaining electricity demand, the electricity spot price, and the electricity parameters on the electricity consumption of the electricity demand side;

[0326] Solve the electricity consumption prediction function according to the first prediction result, the second prediction result, the global electricity parameters, and the local electricity parameters of the target electricity demand side to obtain the electricity consumption of the target electricity demand side in the current period.

[0327] In an exemplary embodiment, the electricity consumption prediction device further includes:

[0328] A first construction module, configured to construct a revenue function on the electricity supply side and a revenue function on the electricity demand side with the remaining electricity demand and the electricity spot price as variables;

[0329] A second construction module, configured to determine an optimal contract quantity function for the power supply side based on the revenue function of the power supply side, and to determine an optimal contract quantity function for the power demand side based on the revenue function of the power demand side; wherein, the variables of the optimal contract quantity function for the power supply side and the optimal contract quantity function for the power demand side both include the power equilibrium price;

[0330] A third construction module, configured to determine a price function of the power equilibrium price when the function values of the optimal contract quantity function for the power supply side and the optimal contract quantity function for the power demand side are equal;

[0331] A fourth construction module, configured to use the price function to assign values to the variables of the optimal contract quantity function for the power demand side, and obtain an electricity quantity prediction function.

[0332] In an exemplary embodiment, the second construction module includes:

[0333] A first construction unit, configured to optimize and solve the revenue function of the power supply side with the goal of maximizing the function value of the revenue function of the power supply side, and obtain the optimal contract quantity function for the power supply side;

[0334] A second construction unit, configured to optimize and solve the revenue function of the power demand side with the goal of maximizing the function value of the revenue function of the power demand side, and obtain the optimal contract quantity function for the power demand side.

[0335] In an exemplary embodiment, the first construction unit is specifically configured to:

[0336] Taking the power supply side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power supply side, construct a first optimization function of the revenue function of the power supply side; wherein, the variables of the first optimization function include: the power supply side contract quantity;

[0337] Taking the maximum function value of the first optimization function as the goal, solve the first optimization function, and obtain the optimal contract quantity function for the power supply side.

[0338] In an exemplary embodiment, the second construction unit is specifically configured to:

[0339] Taking the power demand side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function of the power demand side, construct a second optimization function of the revenue function of the power demand side; wherein, the variables of the second optimization function include: the power demand side contract quantity;

[0340] Taking the maximum function value of the second optimization function as the goal, solve the second optimization function, and obtain the optimal contract quantity function for the power demand side.

[0341] In an exemplary embodiment, the parameter acquisition module 1410 is specifically configured to:

[0342] Determine a first window period and a second window period of the target period according to the demand window period and the supply window period of the historical same-period period corresponding to the target period;

[0343] Obtain the historical demand parameters of the first window period and the historical supply parameters of the second window period;

[0344] Wherein, the power supply forecast result of the historical same-period period predicted based on the supply window period satisfies the supply forecast condition with the actual value of the power supply of the historical same-period period, and the power remaining demand forecast result of the historical same-period period predicted based on the demand window period satisfies the demand forecast condition with the actual value of the power remaining demand of the historical same-period period.

[0345] In an exemplary embodiment, the result prediction module 1420 is specifically configured to:

[0346] Input the historical demand parameters into the demand prediction model to obtain the prediction result of the power remaining demand of the target period as the first prediction result;

[0347] Determine the power supply forecast result of the target period according to the historical supply parameters;

[0348] Based on the first prediction result and the power supply forecast result of the target period, predict the spot price of electricity in the target period to obtain a second prediction result.

[0349] Each module in the above power quantity prediction device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above respective modules.

[0350] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 15As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as demand parameters and supply parameters. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. The computer program, when executed by the processor, implements a power prediction method.

[0351] Those skilled in the art can understand that Figure 15 the structure shown in the figure is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0352] In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0353] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0354] In one embodiment, a computer program product is provided, including a computer program. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0355] It should be noted that the information (including but not limited to device information, market information, demand parameters, supply parameters, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of the relevant data need to comply with the relevant regulations.

[0356] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, a database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in this application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0357] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope recorded in this application.

[0358] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A power consumption prediction method, characterized in that, The method includes: Obtaining historical demand parameters and historical supply parameters for a target period; wherein, the target period is the next period of the current period; Based on the historical demand parameters, predicting the remaining power demand for the target period to obtain a first prediction result, and based on the first prediction result and the historical supply parameters, predicting the spot power price for the target period to obtain a second prediction result; Determining the power demand quantity of the target power demand side in the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side; Wherein, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power price on the power consumption side, and the local power parameters of the target power demand side include the local risk aversion level, the local power price, and the demand quantity prediction ratio of the target power demand side.

2. The method according to claim 1, characterized in that, The determining the power demand quantity of the target power demand side in the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side includes: Obtaining a power demand quantity prediction function; wherein, the power demand quantity prediction function is used to describe the influence of the remaining power demand, the spot power price, and the power parameters on the power demand quantity on the power demand side; Solving the power demand quantity prediction function according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demand side to obtain the power demand quantity of the target power demand side in the current period.

3. The method according to claim 1, wherein The power demand quantity prediction function is constructed in the following manner: Taking the remaining power demand and the spot power price as variables, constructing a revenue function on the power supply side and a revenue function on the power demand side; Based on the revenue function on the power supply side, determining the optimal contract quantity function on the power supply side, and based on the revenue function on the power demand side, determining the optimal contract quantity function on the power demand side; wherein, the variables of the optimal contract quantity function on the power supply side and the optimal contract quantity function on the power demand side both include the power equilibrium price; Determining the price function of the power equilibrium price when the function values of the optimal contract quantity function on the power supply side and the optimal contract quantity function on the power demand side are equal; Using the price function and the local power parameter variables to assign values to the variables of the optimal contract quantity function on the power demand side to obtain the power demand quantity prediction function.

4. The method according to claim 3, wherein The determining the optimal contract quantity function on the power supply side based on the revenue function on the power supply side and the determining the optimal contract quantity function on the power demand side based on the revenue function on the power demand side include: Taking the maximum function value of the revenue function on the power supply side as the target, optimizing and solving the revenue function on the power supply side to obtain the optimal contract quantity function on the power supply side; Taking the maximum function value of the revenue function on the power demand side as the target, optimizing and solving the revenue function on the power demand side to obtain the optimal contract quantity function on the power demand side.

5. The method according to claim 4, wherein Taking the maximum value of the function value of the revenue function on the power supply side as the objective, optimizing and solving the revenue function on the power supply side to obtain the optimal contract quantity function on the power supply side, including: Taking the power supply side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function on the power supply side, constructing a first optimization function of the revenue function on the power supply side; wherein, the variables of the first optimization function include the power supply side contract quantity; Taking the maximum value of the function value of the first optimization function as the objective, solving the first optimization function to obtain the optimal contract quantity function on the power supply side.

6. The method according to claim 4, wherein Taking the maximum value of the function value of the revenue function on the power demand side as the objective, optimizing and solving the revenue function on the power demand side to obtain the optimal contract quantity function on the power demand side, including: Taking the power demand side revenue as a variable, based on the expected value function and variance value function corresponding to the revenue function on the power demand side, constructing a second optimization function of the revenue function on the power demand side; wherein, the variables of the second optimization function include the power demand side contract quantity; Taking the maximum value of the function value of the second optimization function as the objective, solving the second optimization function to obtain the optimal contract quantity function on the power demand side.

7. The method according to any one of claims 1-6, characterized in that, The obtaining of the historical demand parameters and historical supply parameters for the target period includes: According to the demand window period and supply window period of the historical same period corresponding to the target period, determining the first window period and the second window period of the target period; Obtaining the historical demand parameters of the first window period and the historical supply parameters of the second window period; Wherein, the power supply prediction result of the historical same period predicted based on the supply window period satisfies the supply prediction condition with the actual power supply value of the historical same period, and the power remaining demand prediction result of the historical same period predicted based on the demand window period satisfies the demand prediction condition with the actual power remaining demand value of the historical same period.

8. The method according to claim 7, wherein The predicting of the power remaining demand for the target period based on the historical demand parameters to obtain a first prediction result, and the predicting of the power spot price for the target period based on the first prediction result and the historical supply parameters to obtain a second prediction result, including: Inputting the historical demand parameters into a demand prediction model to obtain the prediction result of the power remaining demand for the target period as the first prediction result; Determining the power supply prediction result for the target period according to the historical supply parameters; Predicting the power spot price for the target period based on the first prediction result and the power supply prediction result for the target period to obtain a second prediction result.

9. An electric quantity prediction device, characterized in that, The device includes: A parameter acquisition module for acquiring historical demand parameters and historical supply parameters for a target period; wherein, the target period is the next period of the current period; A result prediction module, configured to predict the remaining power demand in the target period based on the historical demand parameters to obtain a first prediction result, and predict the spot power price in the target period based on the first prediction result and the historical supply parameters to obtain a second prediction result; A power quantity determination module, configured to determine the power demand quantity of the target power demander in the current period according to the first prediction result, the second prediction result, the global power parameters, and the local power parameters of the target power demander; Wherein, the global power parameters include the global risk aversion level on the power supply side, the global risk aversion level on the power demand side, and the global power consumption price on the power consumption side, and the local power parameters of the target power demander include the local risk aversion level, the local power consumption price, and the demand quantity prediction ratio of the target power demander.

10. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images