Bidding strategy selection method for assisting multi-day power transaction

Abstract

The invention relates to the technical field of electric quantity bidding in an electric power trading market, and discloses a bidding strategy selection method for assisting multi-day electric power trading. Comprising the following steps: S1, respectively transmitting purchased electric quantity, purchased electric quantity average price, predicted electricity consumption, predicted electricity consumption average absolute error, day-ahead clearing predicted price and day-ahead clearing predicted price average absolute error to a background server through an input module; s2, a processing module processes the purchased electric quantity, the average price of the purchased electric quantity, the predicted electricity consumption and the day-ahead predicted price to obtain a predicted quantity difference and a predicted price difference; and S3, an analysis module analyzes the predicted quantity difference and the predicted price difference, the analysis module analyzes and selects a first bidding strategy, a second bidding strategy, a third bidding strategy and a fourth bidding strategy, and outputs corresponding bidding prices and bidding electric quantities through an output module. According to the invention, automatic analysis and calculation can be realized, and a bidding strategy is given according to analysis and calculation.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity quantity bidding in an electricity trading market, and in particular to a bidding strategy selection method for assisting multi-day electricity trading. Background Art

[0002] In order to achieve efficient, stable and sustainable development of the electricity market, electricity transactions are often completed through the cooperation of the medium- and long-term electricity trading market and the multi-day electricity trading market. The medium- and long-term electricity trading market trades electricity on a time scale of years, seasons, months and ten days. In the medium- and long-term electricity trading market, power generation companies and power sales companies determine the electricity volume and electricity price in the future in advance through bilateral negotiations, centralized bidding and other methods. The multi-day electricity trading market trades electricity on a daily time scale. When there is a deviation between the predicted electricity consumption on the operating day (D day) and the contracted electricity volume in the medium- and long-term electricity trading market, the multi-day electricity trading market plays a role in regulation and supplementation. Specifically, if the predicted electricity consumption on the operating day is higher than the contracted electricity volume in the medium- and long-term electricity trading market, then it is necessary to purchase additional electricity through the multi-day electricity trading market to meet the electricity demand on the operating day; conversely, if the predicted electricity consumption on the operating day is lower than the contracted electricity volume, the excess electricity can be sold through the multi-day electricity trading market.

[0003] However, when electricity is traded in the multi-day electricity trading market, the purchase quantity and purchase price (or selling quantity and selling price) given by the power sales company are mostly decided based on the personal experience of the traders, which has high labor costs and can easily lead to a large difference between the purchase quantity and purchase price (or selling quantity and selling price) given by the power sales company and the corresponding actual electricity demand and electricity price, thereby resulting in higher overall expenditure costs or loss of profits for the power sales company. It is urgently needed to provide a method that can automatically analyze and calculate and give a selection method for bidding strategies for assisting multi-day electricity transactions based on the analysis and calculation. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a bidding strategy selection method for assisting multi-day electricity trading, which can automatically analyze and calculate and give a bidding strategy based on the analysis and calculation, thereby reducing the overall expenditure cost of the power sales company and ensuring stable income.

[0005] The technical solution adopted by the present invention is as follows: A bidding strategy selection method for assisting multi-day power trading comprises the following steps:

[0006] S1: The purchased electricity and the average price of purchased electricity in the medium and long-term electricity trading market at K time on D day, the predicted electricity consumption at K time on D day, the average absolute error of predicted electricity consumption, the day-ahead clearing predicted price, and the day-ahead clearing predicted price average absolute error are transmitted to the backend server through the input module respectively;

[0007] S2: The processing module processes the purchased electricity, the average price of purchased electricity, the predicted electricity consumption, and the predicted price of day-ahead clearing to obtain the predicted quantity difference and the predicted price difference at time K on day D;

[0008] S3: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding electricity through the output module;

[0009] When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module;

[0010] When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module;

[0011] When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module.

[0012] As a preferred embodiment of the present invention, S2 comprises the following steps:

[0013] S21: The processing module processes the average price of purchased electricity and the predicted price of day-ahead clearing to obtain the predicted price difference at time K on day D, as shown in formula (1):

[0014]

[0015] In formula (1), It represents the difference between the predicted price of the day-ahead clearing at K time on D day in the spot market predicted in advance for participating in the multi-day electricity trading market and the predicted price difference between the average price of the purchased electricity at K time on D day in the traded medium- and long-term electricity trading market. It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast. It represents the average price of purchased electricity at time K on D day in the medium and long-term electricity trading market that has been launched;

[0016] S22: The processing module processes the purchased electricity and the predicted electricity consumption to obtain the predicted quantity difference at time K on day D, as shown in formula (2):

[0017]

[0018] In formula (2), It indicates the difference between the predicted power consumption at K time on D day predicted before participating in the multi-day power trading market and the predicted power purchased at K time on D day in the completed medium- and long-term power trading market. It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It indicates the amount of electricity purchased at time K on day D in the launched medium- and long-term electricity trading market.

[0019] As a preferred embodiment of the present invention, S3 includes the following steps:

[0020] S31: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding quantity through the output module, as shown in formulas (3) and (4).

[0021]

[0022] In formula (3), represents the first bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0023] In formula (4), represents the first bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at time K on D-day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contracted electricity at time K on D-day in the medium- and long-term electricity trading market that is allowed to be greater than the actual electricity at time K on D-day.

[0024] As a preferred embodiment of the present invention, S3 includes the following steps:

[0025] S32: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module, as shown in formulas (5) and (6).

[0026]

[0027] In formula (5), represents the second bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0028] In formula (6), represents the second bidding quantity of electricity at time K on day D in the multi-day electricity trading market, α represents the average absolute error of the predicted electricity consumption at time K on day D in the test set multi-day electricity market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the medium- and long-term electricity trading market that has been launched, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be more than the actual amount of electricity at K time on D day.

[0029] As a preferred embodiment of the present invention, the analysis module analyzes the second bidding power. When the second bidding power is greater than zero, the second bidding power is directly output through the output module. When the second bidding power is less than zero, the second bidding power is output as zero through the output module.

[0030] As a preferred embodiment of the present invention, S3 includes the following steps:

[0031] S33: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module, as shown in formulas (7) and (8).

[0032]

[0033] In formula (7), represents the third bidding price of electricity at K time on D day in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at K time on D day in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0034] In formula (8), represents the third bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

[0035] As a preferred embodiment of the present invention, the analysis module analyzes the third bidding power. When the third bidding power is greater than zero, the third bidding power is directly output through the output module. When the third bidding power is less than zero, the third bidding power is output as zero through the output module.

[0036] As a preferred embodiment of the present invention, S3 includes the following steps:

[0037] S34: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module, as shown in formulas (9) and (10).

[0038]

[0039] In formula (9), represents the fourth bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0040] In formula (10), represents the fourth bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1) The present invention can automatically analyze and calculate the corresponding amount of electricity that needs to be bought or sold at time K on D day and the corresponding price through the input module, processing module, analysis module, output module and background server, that is, it provides a bidding strategy for multi-day electricity trading (the bidding strategy is to determine the specific amount of electricity to be bought or sold at a specific price at time K on D day before D day), which can reduce labor costs;

[0043] 2) Before conducting multi-day electricity transactions, traders often predict the actual electricity consumption on D-day and the predicted price for day-ahead clearing through the predicted electricity consumption, electricity prices, and weather-related data many days before D-day. However, the prediction itself has errors, and the existence of errors can cause traders to make wrong bidding strategies. For example, the predicted electricity consumption and the predicted electricity price will both rise. Without considering the errors, traders will make a large number of buying decisions in multi-day electricity transactions, and the prediction errors will cause the actual electricity consumption and the actual day-ahead clearing predicted price to decline. In this way, the power sales company is forced to sell the excess electricity at the real lower day-ahead clearing predicted price on the day-ahead spot market, resulting in losses for the power sales company, thereby increasing the expenditure costs of the power sales company. The present invention comprehensively considers the predicted electricity difference at the K moment of D-day and the predicted electricity price difference at the K moment of D-day on the day-ahead spot market, ensuring the provision of accurate and reasonable bidding strategies, improving the success rate of transactions, thereby reducing the overall expenditure costs of the power sales company and ensuring stable income. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a flow chart of a bidding strategy selection method for assisting multi-day electricity trading. DETAILED DESCRIPTION

[0045] Typical embodiments that embody the features and advantages of the present invention will be described in detail in the following description. It should be understood that the present invention can have various changes in different embodiments without departing from the scope of the present invention, and the descriptions and illustrations therein are essentially used for illustration purposes rather than for limiting the present invention.

[0046] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0047] Explanation: The multi-day electricity trading market is between the medium- and long-term electricity trading market and the spot market. The spot market includes the day-ahead spot market and the real-time spot market. The day-ahead clearing forecast price at K moment on D-day needs to be determined in the day-ahead (D-1) spot market. The multi-day electricity trading market of the present invention connects the medium- and long-term electricity trading market with the day-ahead spot market. During multi-day electricity trading, if the predicted electricity consumption at K moment on D-day and the day-ahead clearing forecast electricity price at K moment on D-day in the day-ahead spot market can be predicted in advance, the formulated bidding strategy will be more effective.

[0048] A bidding strategy selection system for assisting multi-day power trading includes a backend server and a user terminal connected to the backend server in communication. The backend server includes a storage module, a processing module, and an analysis module. The user terminal includes an input module and an output module. The storage module is used to store four bidding strategies. The processing module is used to process data such as purchased electricity, average price of purchased electricity, predicted electricity consumption, and predicted price for day-ahead clearing, and obtain the predicted quantity difference and predicted price difference at K time on D day. The analysis module is used to analyze the predicted quantity difference and predicted price difference and derive specific bidding strategies. The input module is used to edit information on purchased electricity, average price of purchased electricity, predicted electricity consumption, average absolute error of predicted electricity consumption, predicted price for day-ahead clearing, and average absolute error of predicted price for day-ahead clearing, and transmit the edited information on purchased electricity, average price of purchased electricity, predicted electricity consumption, average absolute error of predicted electricity consumption, predicted price for day-ahead clearing, and average absolute error of predicted price for day-ahead clearing to the backend server. The output module is used to output the bidding strategy given by the analysis module.

[0049] When selling electricity, power sales companies first purchase electricity in the medium- and long-term electricity trading market, and then purchase or sell electricity in the multi-day electricity trading market based on the predicted electricity consumption on the operating days. Specifically, they purchase the total electricity for the year on an annual basis, the total electricity for the month on a monthly basis, and the total electricity for the ten-day period on a ten-day basis. They then purchase or sell electricity through the multi-day electricity trading market a few days before the operating day.

[0050] In this embodiment, the purchased electricity volume is the sum of the decomposed electricity volume of the annual, monthly, and ten-day medium- and long-term electricity trading markets that have been carried out at the time K on the D day, and the average price of the purchased electricity volume is the average transaction price of the annual, monthly, and ten-day medium- and long-term electricity trading markets that have been carried out at the time K on the D day, as shown in formulas (11) and (12):

[0051]

[0052] In formula (11), It indicates the purchased electricity volume at time K on D day in the medium and long-term electricity trading market that has been launched. They are the decomposed electricity volume of the annual, monthly and ten-day medium- and long-term electricity trading markets that have been launched at the K time on the D day;

[0053] In formula (12), It represents the average price of purchased electricity at time K on D day in the medium and long-term electricity trading market that has been launched. They are respectively the transaction prices at K time on D day of the annual, monthly and ten-day medium- and long-term electricity trading markets that have been launched.

[0054] The predicted electricity consumption at time K on day D, the average absolute error of predicted electricity consumption, the predicted price for day-ahead clearing, and the average absolute error of the predicted price for day-ahead clearing are all known data.

[0055] In this embodiment, a data set consisting of historical day-ahead transaction prices and day-ahead inter-system transmission curve forecasts, day-ahead network-wide electricity consumption forecasts, day-ahead non-market unit output plans and other information is divided into a training set and a test set. An XGBoost, LSTM and other algorithms are used on the training set to establish a 24-period clearing electricity price forecast model for the day-ahead spot market. The day-ahead spot market 24-period clearing electricity price forecast model is used to perform predictions on the test set, and the predicted values ​​are compared with the original values ​​of the test set to obtain the average absolute error of the day-ahead clearing forecast price at K time on D day in the day-ahead spot market. The trained day-ahead spot market 24-period clearing electricity price forecast model is used to predict the day-ahead clearing forecast price of electricity at K time on D day in the day-ahead spot market.

[0056] In this embodiment, a data set consisting of historical electricity consumption, weather and other information is divided into a training set and a test set. A multi-day 24-period electricity consumption prediction model for the power sales company is established on the training set using machine learning algorithms such as support vector machines and random forests. The multi-day 24-period electricity consumption prediction model is used to perform predictions on the test set, and the predicted values ​​are compared with the original values ​​of the test set to obtain the average absolute error of the predicted electricity consumption at time K on day D. The trained multi-day 24-period electricity consumption prediction model is used to predict the actual demand electricity at time K on day D, that is, the predicted electricity consumption at time K on day D.

[0057] In this embodiment, the multi-day electricity trading market refers to the electricity energy market that determines the unit combination status and power generation plan on the operation day (D-day) a few days before the operation day (D-day), which can be expressed as the DN electricity trading market, which conducts electricity trading N days before the D-day, specifically, the D-3 electricity trading market, which conducts electricity trading three days before the D-day. The day-ahead electricity market refers to the electricity energy market that determines the unit combination status and power generation plan on the operation day (D-day) one day in advance (D-1 day) of the operation day (D-day).

[0058] like Figure 1 As shown, the bidding strategy selection method for assisting multi-day power trading includes the following steps:

[0059] S1: The purchased electricity and the average price of purchased electricity in the medium and long-term electricity trading market at K time on D day, the predicted electricity consumption at K time on D day, the average absolute error of predicted electricity consumption, the day-ahead clearing predicted price, and the day-ahead clearing predicted price average absolute error are transmitted to the backend server through the input module respectively;

[0060] S2: The processing module processes the purchased electricity, the average price of purchased electricity, the predicted electricity consumption, and the predicted price of day-ahead clearing to obtain the predicted quantity difference and the predicted price difference at K time on D day;

[0061] S2 includes the following steps:

[0062] S21: The processing module processes the average price of purchased electricity and the predicted price of day-ahead clearing to obtain the predicted price difference at time K on day D, as shown in formula (1):

[0063]

[0064] In formula (1), It represents the difference between the predicted price of the day-ahead clearing at K time on D day in the spot market predicted in advance for participating in the multi-day electricity trading market and the predicted price difference between the average price of the purchased electricity at K time on D day in the traded medium- and long-term electricity trading market. It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast. It represents the average price of purchased electricity at time K on D day in the medium and long-term electricity trading market that has been launched;

[0065] S22: The processing module processes the purchased electricity and the actual predicted electricity consumption to obtain the predicted quantity difference at time K on day D, as shown in formula (2):

[0066]

[0067] In formula (2), It indicates the difference between the predicted power consumption at K time on D day predicted before participating in the multi-day power trading market and the predicted power purchased at K time on D day in the completed medium- and long-term power trading market. It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It indicates the amount of electricity purchased at time K on day D in the launched medium- and long-term electricity trading market.

[0068] S3: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding electricity through the output module;

[0069] When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module;

[0070] When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module;

[0071] When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module.

[0072] S3 includes the following steps:

[0073] S31: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding quantity through the output module, as shown in formulas (3) and (4).

[0074]

[0075] In formula (3), represents the first bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0076] In formula (4), represents the first bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at time K on D-day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contracted electricity at time K on D-day in the medium- and long-term electricity trading market that is allowed to be greater than the actual electricity at time K on D-day.

[0077] In this embodiment, the predicted price difference is greater than zero, that is, the predicted day-ahead clearing price at K time on D-day in the day-ahead spot market is higher than the average price of electricity purchased by the power sales company at K time on D-day in the medium- and long-term power trading market that has been carried out. Considering the average absolute error of the day-ahead clearing price forecast, the actual clearing price of electricity at K time on D-day will be In this embodiment, because the predicted quantity difference is greater than zero, it is necessary to buy electricity. It is recommended to buy at a price not higher than the first bidding price. Buying electricity means buying it at a relatively low price.

[0078] Prediction measurement difference Greater than zero, that is, the predicted power consumption at K time on D day is greater than the power sales company's purchased power at K time on D day in the medium- and long-term power trading market that has been carried out. The power sales company needs to buy power in the multi-day power trading market. Considering the average absolute error of the predicted power consumption, the actual power consumption at K time on D day will be Within the range, since the predicted price for clearing at K time on D-day in the day-ahead spot market is expected to rise, and the first bidding price may be the minimum value of the predicted price for clearing at K time on D-day, if the transaction can be completed, it is recommended to buy as much as possible. Even if you buy too much, it will cause the net contracted electricity at K time on D-day in the medium- and long-term electricity trading market (the net contracted electricity is the sum of the purchased electricity at K time on D-day in the medium- and long-term electricity trading market that has been launched and the bid electricity at K time on D-day in the multi-day electricity trading market) to exceed the actual electricity at K time on D-day. In the day-ahead spot market, it can also be sold at a price not lower than the first bidding price, thereby making a profit. In order to prevent the risks caused by forecast errors, it is recommended to buy as much as possible. In addition, within the scope of electricity trading rules, the net contracted electricity at K time on D-day in the medium- and long-term electricity trading market is allowed to be more than θ than the actual electricity consumption at K time on D-day. In this embodiment, the first bidding electricity is used. Buy the corresponding amount of electricity.

[0079] The first bidding strategy is to buy in large quantities at a relatively low price. If the power sales company buys electricity at a higher price, the cost of the power sales company will increase. In this embodiment, a large amount of electricity is purchased at the first bidding price. The first bidding price may be the minimum of the actual clearing price at K time on D day. Even if the actual electricity price fluctuates, the cost of the power sales company can be kept at a relatively low level, avoiding the cost risk caused by inaccurate electricity price forecasts.

[0080] The final power sales company may have more electricity than the actual electricity demand at K time on D day, and can purchase more electricity in the day-ahead spot market. Selling at a price within the range, that is, not lower than the first bid price, retains the arbitrage opportunity and increases economic income.

[0081] S32: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module, as shown in formulas (5) and (6).

[0082]

[0083] In formula (5), represents the second bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0084] In formula (6), represents the second bidding quantity of electricity at time K on day D in the multi-day electricity trading market, α represents the average absolute error of the predicted electricity consumption at time K on day D in the test set multi-day electricity market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the medium- and long-term electricity trading market that has been launched, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be more than the actual amount of electricity at K time on D day.

[0085] In this embodiment, the analysis module analyzes the second bidding power. When the second bidding power is greater than zero, the second bidding power is directly output through the output module. When the second bidding power is less than zero, the second bidding power is output as zero through the output module.

[0086] In this embodiment, the predicted price difference is greater than zero, that is, the predicted day-ahead clearing price at K time on D-day in the day-ahead spot market is higher than the average price of electricity purchased by the power sales company at K time on D-day in the medium- and long-term power trading market that has been carried out. Considering the average absolute error of the day-ahead clearing price forecast, the actual clearing price of electricity at K time on D-day will be In this embodiment, because the predicted quantity difference is less than zero, it is necessary to sell the electricity. It is recommended to bid at a price not lower than the second bidding price. Selling electricity means selling it at a relatively low price to increase the success rate of the transaction.

[0087] Prediction measurement difference Less than zero, that is, the predicted power consumption at K time on D day is less than the power sales company's purchased power at K time on D day in the medium- and long-term power trading market that has been carried out. The power sales company needs to sell power in the multi-day power trading market. Considering the average absolute error of the predicted power consumption, the actual power consumption at K time on D day will be Within the range, assuming that the prediction error is not considered, it is profitable even if no operation is performed in the multi-day power trading market, because the net contract electricity at D-day K-time in the medium- and long-term power trading market has exceeded the actual electricity consumption at D-day K-time, and the predicted price of the day-ahead clearing at D-day K-time in the day-ahead spot market will rise. In this way, the excess electricity at D-day K-time in the medium- and long-term power trading market that has been launched will be automatically sold by the system in the day-ahead spot market, thereby making a profit. However, the trading rules require that the second net contract electricity at D-day K-time needs to tend to the actual electricity consumption at D-day K-time, so it is generally necessary to sell. Since the predicted price of the day-ahead clearing at D-day K-time in the day-ahead spot market has risen, in order to increase the transaction rate of selling, it is recommended to sell at the lowest possible price. Because of the existence of electricity prediction errors, selling at the minimum value of the predicted electricity consumption at D-day K-time will lead to overselling, and then being forced to buy back in the day-ahead spot market, and the predicted price of the day-ahead clearing at D-day K-time in the day-ahead spot market will rise, which poses a risk of loss. Therefore, in the multi-day power trading market, as little as possible is sold at a lower price, and the remaining excess electricity is automatically sold by the system at a possible high price in the day-ahead spot market. Therefore, in order to prevent the risk caused by prediction errors, it is recommended to sell as little as possible. In addition, within the scope of power trading rules, the net contract electricity at K time on D day in the medium and long-term power trading market is allowed to be more than the actual electricity at K time on D day by θ. In this embodiment, the second bidding electricity Sell ​​the corresponding amount of electricity.

[0088] The second bidding strategy is to sell a small amount at a relatively low price. In the end, the power sales company may have more electricity than the actual electricity demand at K time on D day, and can sell it at a lower price in the spot market. Selling at a price within the range, that is, selling at a price not lower than the second bid price, retaining the arbitrage opportunity and increasing economic income.

[0089] S33: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module, as shown in formulas (7) and (8).

[0090]

[0091] In formula (7), represents the third bidding price of electricity at K time on D day in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at K time on D day in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0092] In formula (8), represents the third bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

[0093] In this embodiment, the analysis module analyzes the third bidding power. When the third bidding power is greater than zero, the third bidding power is directly output through the output module. When the third bidding power is less than zero, the third bidding power is output as zero through the output module.

[0094] In this embodiment, the predicted price difference Less than zero, that is, the predicted day-ahead clearing price at K time on D-day in the day-ahead spot market is lower than the average price of electricity purchased by the power sales company at K time on D-day in the medium- and long-term power trading market that has been carried out. Considering the average absolute error of the day-ahead clearing price forecast, the actual clearing price of electricity at K time on D-day will be In this embodiment, because the predicted quantity difference is greater than zero, it is necessary to buy electricity. It is recommended to bid at a price not higher than the third bidding price. Buying electricity means buying at a relatively high price to increase the success rate of the transaction.

[0095] Prediction measurement difference is greater than zero, that is, the actual electricity demand predicted at K time on D day is greater than the electricity purchased by the power sales company at K time on D day in the medium- and long-term electricity trading market that has been carried out. The power sales company needs to buy electricity in the multi-day electricity trading market. Considering the average absolute error of the predicted electricity consumption, the actual electricity consumption at K time on D day will be Within the range, because of the existence of electricity forecast errors, buying with the maximum electricity consumption predicted at K moment on D day will lead to overbuying, and then being forced to sell in the day-ahead spot market, and the day-ahead clearing forecast price at K moment on D day predicted in the day-ahead spot market will drop, which poses a risk of losing money. Therefore, it is recommended to buy less in the multi-day electricity trading market, that is, to buy with the minimum electricity consumption predicted at K moment on D day, and to buy with the upper limit of the day-ahead clearing forecast price at K moment on D day predicted in the day-ahead spot market (also to increase the transaction rate). Even if you buy less, you can buy it in the day-ahead spot market at a price not higher than the third bidding electricity price in the multi-day electricity trading market. Therefore, in order to prevent risks caused by forecast errors, it is recommended to buy as little as possible. In addition, within the scope of electricity trading rules, the medium- and long-term electricity trading market is allowed to have a net contract electricity quantity at K moment on D day that is θ less than the actual electricity quantity at K moment on D day. In this embodiment, the third bidding electricity quantity is used. Buy the corresponding amount of electricity.

[0096] The third bidding strategy is to buy a small amount at a relatively high price. In the end, the power sales company may have less electricity than the actual electricity demand at K time on D day, and can buy it at a higher price in the spot market. Buying at a price within the range, that is, buying at a price not higher than the third bidding price, retains the opportunity to short sell.

[0097] S34: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module, as shown in formulas (9) and (10).

[0098]

[0099] In formula (9), represents the fourth bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast;

[0100] In formula (10), represents the fourth bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

[0101] In this embodiment, the predicted price difference Less than zero, that is, the predicted clearing price at K time on D-day in the day-ahead spot market is lower than the average price of electricity purchased by the power sales company at K time on D-day in the medium- and long-term power trading market that has been carried out. Considering the average absolute error of the day-ahead clearing price forecast, the actual clearing price of electricity at K time on D-day will be In this embodiment, because the predicted quantity difference is less than zero, it is necessary to sell the electricity. It is recommended to bid at a price not lower than the fourth bidding price. Selling electricity means selling it at a relatively high price.

[0102] Prediction measurement difference Less than zero, that is, the predicted power consumption at K time on D day is less than the power sales company's purchased power at K time on D day in the medium- and long-term power trading market that has been carried out. The power sales company needs to sell power in the multi-day power trading market. Considering the average absolute error of the predicted power consumption, the actual power consumption at K time on D day will be Within the range, it is recommended to sell as much as possible to prevent the risk caused by the forecast error. Because, considering the existence of the electricity forecast error, selling with the minimum electricity consumption predicted at K time on D day will lead to selling less, and then being forced to continue selling in the day-ahead spot market, and the predicted day-ahead clearing forecast price at K time on D day in the predicted day-ahead spot market will drop, which may cause loss of profit. Therefore, it is recommended to sell more in the multi-day electricity trading market, that is, to sell with the maximum electricity consumption predicted at K time on D day, and the day-ahead clearing forecast price at K time on D day in the predicted day-ahead spot market will drop, so it is recommended to sell at the upper limit of the day-ahead clearing forecast price (profitable in the multi-day electricity trading market), even if too much is sold, it can be bought back in the day-ahead spot market at a price not lower than the fourth bidding electricity price in the multi-day electricity trading market, thereby avoiding the additional risk caused by inaccurate electricity price forecasts. In addition, within the scope of the electricity trading rules, it is allowed that the net contract electricity at K time on D day in the already launched medium- and long-term electricity trading market is less than the actual electricity at K time on D day by θ. In this embodiment, the fourth bidding electricity Sell ​​the corresponding amount of electricity.

[0103] The fourth bidding strategy is to sell a large amount of electricity at a relatively high price. In the end, the power sales company may have less electricity than the actual electricity demand at K time on D day, and can sell it at a higher price in the spot market. Buying at a price within the range, that is, buying at a price not lower than the fourth bidding price, retains the opportunity to short sell.

[0104] The following is an example of trading in the D-3 power trading market:

[0105] S1: The purchased electricity and average price of purchased electricity in the medium and long-term electricity trading market at K time on D day, the predicted electricity consumption, the average absolute error of predicted electricity consumption, the day-ahead clearing predicted price, and the average absolute error of the day-ahead clearing predicted price at K time on D day are transmitted to the background server through the input module respectively;

[0106] S2: The processing module processes the purchased electricity, the average price of purchased electricity, the predicted electricity consumption, and the predicted price of day-ahead clearing to obtain the predicted quantity difference and the predicted price difference at K time on D day;

[0107] S21: The processing module processes the average price of purchased electricity and the day-ahead clearing forecast price to obtain the forecast price difference at time K on day D, as shown in formula (1):

[0108]

[0109] In formula (1), It represents the predicted price difference between the day-ahead clearing forecast price at K time on D-day in the D-1 spot market predicted before the D-3 power market transaction and the average price of purchased electricity at K time on D-day in the traded medium- and long-term power trading market. It indicates the day-ahead clearing forecast price at K time on D-day in the D-1 spot market that is predicted before the D-3 power market transaction. It represents the average price of purchased electricity at time K on D day in the medium and long-term electricity trading market that has been launched;

[0110] S22: The processing module processes the purchased electricity and the predicted electricity consumption to obtain the predicted quantity difference at time K on day D, as shown in formula (2):

[0111]

[0112] In formula (2), It represents the difference between the predicted power consumption at K time on D day predicted before participating in the D-3 power market transaction and the predicted power purchased at K time on D day in the completed medium- and long-term power trading market. It indicates the predicted electricity consumption at K time on D day before participating in D-3 electricity market transaction. It indicates the electricity volume at time K on day D in the launched medium- and long-term electricity trading market.

[0113] S3: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding quantity through the output module, as shown in formulas (3) and (4).

[0114]

[0115] In formula (3), represents the first bidding price of electricity at K time on D day in the D-3 power trading market, β represents the average absolute error of the day-ahead clearing forecast price at K time on D day in the D-1 spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D-day in the D-1 spot market that is forecasted before the D-3 power market transaction;

[0116] In formula (4), represents the first bidding power at K time on D-day in the D-3 power trading market, α represents the average absolute error of the predicted power consumption at K time on D-day in the D-3 power market in the test set, It indicates the predicted electricity consumption at K time on D day before participating in D-3 electricity market transaction. It represents the amount of electricity purchased at K time on D day in the medium- and long-term electricity trading market that has been launched, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be more than the actual amount of electricity at K time on D day.

[0117] In this embodiment, the bid price is no higher than the first bidding price in the D-3 power trading market. Buy the first bidding electricity of electricity.

[0118] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. A bidding strategy selection method for assisting multi-day power trading, characterized in that: The following steps are involved: S1: The purchased electricity and the average price of purchased electricity in the medium and long-term electricity trading market at K time on D day, the predicted electricity consumption at K time on D day, the average absolute error of predicted electricity consumption, the day-ahead clearing predicted price, and the day-ahead clearing predicted price average absolute error are transmitted to the backend server through the input module respectively; S2: The processing module processes the purchased electricity, the average price of purchased electricity, the predicted electricity consumption, and the predicted price of day-ahead clearing to obtain the predicted quantity difference and the predicted price difference at time K on day D; S3: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding electricity through the output module; When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module; When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module; When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module.

2. The bidding strategy selection method for assisting multi-day power trading according to claim 1 is characterized in that: S2 includes the following steps: S21: The processing module processes the average price of purchased electricity and the predicted price of day-ahead clearing to obtain the predicted price difference at time K on day D, as shown in formula (1): In formula (1), It represents the predicted price difference between the day-ahead clearing forecast price at K time on D day in the day-ahead spot market predicted in advance for participating in the multi-day power trading market and the average price of purchased electricity at K time on D day in the traded medium- and long-term power trading market. It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast. It represents the average price of purchased electricity at time K on D day in the medium and long-term electricity trading market that has been launched; S22: The processing module processes the purchased electricity and the predicted electricity consumption to obtain the predicted quantity difference at time K on day D, as shown in formula (2): In formula (2), It indicates the difference between the predicted power consumption at K time on D day predicted before participating in the multi-day power trading market and the predicted power purchased at K time on D day in the completed medium- and long-term power trading market. It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It indicates the amount of electricity purchased at time K on day D in the launched medium- and long-term electricity trading market.

3. The bidding strategy selection method for assisting multi-day power trading according to claim 1 is characterized in that: S3 includes the following steps: S31: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the first bidding strategy, and outputs the first bidding price and the first bidding quantity through the output module, as shown in formulas (3) and (4). In formula (3), represents the first bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast; In formula (4), represents the first bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at time K on D-day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contracted electricity at time K on D-day in the medium- and long-term electricity trading market that is allowed to be greater than the actual electricity at time K on D-day.

4. The bidding strategy selection method for assisting multi-day power trading according to claim 1 is characterized in that: S3 includes the following steps: S32: The analysis module analyzes the predicted quantity difference and the predicted price difference of the electricity. When the predicted quantity difference is less than zero and the predicted price difference is greater than zero, the analysis module analyzes and selects the second bidding strategy, and outputs the second bidding price and the second bidding quantity through the output module, as shown in formulas (5) and (6). In formula (5), represents the second bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast; In formula (6), represents the second bidding quantity of electricity at time K on day D in the multi-day electricity trading market, α represents the average absolute error of the predicted electricity consumption at time K on day D in the test set multi-day electricity market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the medium- and long-term electricity trading market that has been launched, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be more than the actual amount of electricity at K time on D day.

5. The bidding strategy selection method for assisting multi-day power trading according to claim 4 is characterized in that: The analysis module analyzes the second bidding power. When the second bidding power is greater than zero, the second bidding power is directly output through the output module. When the second bidding power is less than zero, the second bidding power is output as zero through the output module.

6. The bidding strategy selection method for assisting multi-day power trading according to claim 1 is characterized in that: S3 includes the following steps: S33: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is greater than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the third bidding strategy, and outputs the third bidding price and the third bidding quantity through the output module, as shown in formulas (7) and (8). In formula (7), represents the third bidding price of electricity at K time on D day in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at K time on D day in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast; In formula (8), represents the third bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

7. The bidding strategy selection method for assisting multi-day power trading according to claim 6 is characterized in that: The analysis module analyzes the third bidding power. When the third bidding power is greater than zero, the third bidding power is directly output through the output module. When the third bidding power is less than zero, the third bidding power is output as zero through the output module.

8. The bidding strategy selection method for assisting multi-day power trading according to claim 1 is characterized in that: S3 includes the following steps: S34: The analysis module analyzes the predicted quantity difference and the predicted price difference. When the predicted quantity difference is less than zero and the predicted price difference is less than zero, the analysis module analyzes and selects the fourth bidding strategy, and outputs the fourth bidding price and the fourth bidding quantity through the output module, as shown in formulas (9) and (10). In formula (9), represents the fourth bidding price of electricity at time K on day D in the multi-day electricity trading market, β represents the average absolute error of the day-ahead clearing forecast price at time K on day D in the day-ahead spot market in the test set, It indicates the day-ahead clearing forecast price at K time on D day in the day-ahead spot market participating in the multi-day electricity trading market forecast; In formula (10), represents the fourth bidding power at time K on day D in the multi-day power trading market, α represents the average absolute error of the predicted power consumption at time K on day D in the test set multi-day power market, It indicates the predicted electricity consumption at K time on D day before participating in the multi-day electricity trading market. It represents the amount of electricity purchased at K time on D day in the launched medium- and long-term electricity trading market, and θ represents the percentage of net contract at K time on D day in the medium- and long-term electricity trading market that is allowed to be less than the actual amount of electricity at K time on D day.

Images