A method, device, equipment and medium for a wind farm to participate in power market bidding

Abstract

The present invention discloses a method, device, equipment and medium for a wind farm to participate in power market bidding. This method uses the SSA-BiGRU-Attention model to predict electricity prices, and uses a wind power probability density prediction model based on the KELM-ALO semi-parametric method to characterize the uncertainty of wind power. A revenue model for the wind farm to participate in the power market is constructed, and with the maximum revenue of the wind farm as the objective function, an improved whale algorithm is used to solve the model to obtain the bidding power of the wind farm to participate in the power market. The SSA-BiGRU-Attention model solves the problem that electricity prices are not considered in the current bidding method, the KELM-ALO semi-parametric method more accurately describes the probability density of wind power, and the improved whale algorithm improves the convergence speed and accuracy of the model, and finally more accurately determines the bidding power of the wind farm to participate in the power market.

Description

Technical Field

[0001] The present invention belongs to the technical field of artificial intelligence, and particularly relates to a method, device, equipment and medium for a wind farm to participate in power market bidding. Background Art

[0002] With the continuous increase in the penetration rate of wind power, the impact of wind power on the frequency stability of the power grid becomes more serious, which will increase the power grid frequency regulation demand. At present, one way to solve this problem is to let wind power provide part of the frequency regulation capacity by itself. Therefore, setting a reasonable power market mechanism and determining the bidding power of the wind farm to participate in the energy market and the frequency regulation market will help improve the wind farm's revenue and relieve the power grid frequency regulation pressure.

[0003] In the current relevant research on wind farms participating in the power market, the mainstream methods for wind power prediction include the multi-scenario method and prior distribution assumption models such as normal distribution and Beta distribution. The multi-scenario method may have difficulties in subsequent solutions when the number of scenarios is large; when the number of scenarios is too small, it cannot effectively reflect the uncertainty of wind power. Prior distribution assumption models such as normal distribution or Beta distribution have problems with parameter estimation deviation. At the same time, the vast majority of wind power prediction powers follow the probability distribution characteristics of peaky and heavy-tailed. A single bandwidth model can hardly accurately describe this probability density characteristic at the same time. Therefore, the prediction error brought cannot accurately reflect the uncertainty of wind power, which may affect the bidding plan and revenue of the wind farm in the power market. Therefore, more in-depth research is needed on the wind power prediction method when the wind farm participates in the power market.

[0004] In the current relevant research on wind farms participating in the power market, the mainstream methods for solving the bidding power of the wind farm include the KKT condition, Leibniz rule, and mixed integer programming algorithm, etc. The KKT condition and Leibniz rule can solve relatively simple models, but when the model considers multiple factors and is relatively complex, more applicable and simple algorithms need to be further studied.

[0005] The electricity price is closely related to the bidding plan of the wind farm participating in the power market. In the current relevant research on wind farms participating in the power market, historical data or empirical values are mostly selected for the electricity price, which may not be able to reflect the revenue of the wind farm in the power market more truly and accurately. At the same time, with the continuous increase in the installed capacity of new energy, the frequency of low peaks in the electricity price will increase significantly, and the volatility of the electricity price will become more obvious, which will affect the bidding plan of the wind farm and may reduce the revenue of the wind farm. Summary of the Invention

[0006] The object of the present invention is to provide a method, device, equipment and medium for a wind farm to participate in power market bidding, so as to solve the problem in the prior art that in the current relevant research on wind farms participating in the power market, historical data or empirical values are mostly selected for electricity prices, thus being unable to more truly and accurately reflect the benefits of wind farms in the power market.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] In the first aspect of the present invention, a method for a wind farm to participate in power market bidding is provided, including the following specific steps:

[0009] Obtain electricity price influencing factor data;

[0010] Input the electricity price influencing factor data into a preset SSA - BiGRU - Attention model, and the SSA - BiGRU - Attention model outputs an electricity price prediction result;

[0011] Obtain the current numerical weather forecast data and real wind power data;

[0012] Input the current numerical weather forecast data and real wind power data into a preset KELM - ALO model, and the KELM - ALO model outputs a predicted wind power value for the next day;

[0013] Based on the predicted wind power value for the next day, use the semi - parametric method to obtain the probability density prediction function of the predicted wind power value for the next day;

[0014] Based on the electricity price prediction result and the probability density prediction function of the predicted wind power value for the next day, establish a revenue model for the wind farm to participate in the power market;

[0015] Taking the maximum wind farm revenue as the objective function, use the improved whale algorithm to solve the revenue model for the wind farm to participate in the power market, and obtain the bidding power of the wind farm in the power market.

[0016] Further, in the step of obtaining the electricity price influencing factor data, the electricity price influencing factor data includes the following:

[0017] Historical electricity price, load, wind power generation, and photovoltaic power generation.

[0018] Further, in the step of inputting the current numerical weather forecast data and real wind power data into a preset KELM - ALO model, the obtaining method of the KELM - ALO model is as follows:

[0019] According to the historical data of numerical weather forecast data and real wind power data, use the ALO algorithm to optimize the kernel parameters of the KELM model to obtain the optimal kernel parameters;

[0020] Substitute the optimal kernel parameter into the KELM model to obtain the KELM-ALO model.

[0021] Furthermore, the KELM-ALO model is as follows:

[0022]

[0023] In the formula: N is the number of historical data; w is the wind speed; k is the kernel function; T is the predicted target value vector; I is the identity matrix; C is the kernel parameter; Ω EML is the kernel matrix.

[0024] Furthermore, the steps of obtaining the probability density prediction function of the next-day wind power prediction value by using the semi-parametric method based on the next-day wind power prediction value specifically include:

[0025] Use the Hill diagram method to determine the threshold u of the next-day wind power prediction value; the part less than u is used as the peak part for piecewise KDE to obtain the density function f u (x); the part greater than u is used as the tail for GPD to obtain the density curve g u (x);

[0026] Combine f u (x) with g u (x) to obtain the probability density curve f G (x) of the semi-parametric hybrid model;

[0027] Use the SG algorithm to smooth the discontinuous part around the threshold u of the f G (x) curve to obtain the probability density prediction function of the next-day wind power.

[0028] Furthermore, the formula of the probability density curve of the semi-parametric hybrid model is as follows:

[0029]

[0030] In the formula: F u (x) is the cumulative distribution function of fu(x).

[0031] Furthermore, the revenue model for the wind farm to participate in the electricity market is as follows:

[0032]

[0033] In the formula: E[R(P e , P r)] is the expected revenue for the wind farm to participate in the electricity market; Pt is the actual output power of the wind farm; f(P t ) is the probability density prediction function of the wind power at the next-day time t; Pt,max is the maximum value of the wind power prediction at time t.

[0034] In a second aspect of the present invention, a device for a wind farm to participate in the electricity market bidding is provided, including:

[0035] A first acquisition module for acquiring electricity price influencing factor data;

[0036] A first result prediction module for inputting the electricity price influencing factor data into a preset SSA-BiGRU-Attention model, and the SSA-BiGRU-Attention model outputs an electricity price prediction result;

[0037] A second acquisition module for acquiring current numerical weather forecast data and real wind power data;

[0038] A second result prediction module for inputting the current numerical weather forecast data and real wind power data into a preset KELM-ALO model, and the KELM-ALO model outputs a predicted value of the wind power for the next day;

[0039] A probability density prediction function acquisition module for obtaining a probability density prediction function of the predicted value of the wind power for the next day by using a semi-parametric method based on the predicted value of the wind power for the next day;

[0040] A revenue model establishment module for establishing a revenue model for the wind farm to participate in the electricity market based on the electricity price prediction result and the probability density prediction function of the predicted value of the wind power for the next day;

[0041] A solution module for taking the maximum wind farm revenue as the objective function and using an improved whale algorithm to solve the revenue model for the wind farm to participate in the electricity market to obtain the bidding power of the wind farm in the electricity market.

[0042] In a third aspect of the present invention, an electronic device is provided, including a processor and a memory, and the processor is used to execute a computer program stored in the memory to implement the above-mentioned method for a wind farm to participate in the electricity market bidding.

[0043] In a fourth aspect of the present invention, a computer-readable storage medium is provided, and the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the above-mentioned method for a wind farm to participate in the electricity market bidding is implemented.

[0044] The beneficial effects of the present invention are as follows:

[0045] The method for a wind farm to participate in the electricity market bidding provided by the present invention uses the SSA-BiGRU-Attention model to predict electricity prices, combines the KELM algorithm, the ALO algorithm and the semi-parametric method to establish a wind power probability density prediction model based on the KELM-ALO-semi-parametric method to characterize the uncertainty of wind power; based on the trading rules for the wind farm to participate in the electricity market, constructs a revenue model for the wind farm to participate in the electricity market, and uses the improved whale algorithm to solve the model with the maximum revenue of the wind farm as the objective function to obtain the bidding power of the wind farm to participate in the electricity market. The SSA-BiGRU-Attention model solves the problem that electricity prices are not considered in the current bidding method, uses the KELM-ALO-semi-parametric method to more accurately describe the probability density of wind power, and at the same time adopts the improved whale algorithm to improve the convergence speed and accuracy of the model, and finally more accurately determines the bidding power of the wind farm to participate in the electricity market. Description of the Drawings

[0046] The accompanying drawings forming a part of this application are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0047] Figure 1 It is a schematic diagram of the principle of a method for a wind farm to participate in the electricity market bidding according to an embodiment of the present invention;

[0048] Figure 2 It is a flowchart of a method for a wind farm to participate in the electricity market bidding according to an embodiment of the present invention;

[0049] Figure 3 It is a structural block diagram of a device for a wind farm to participate in the electricity market bidding according to the present invention;

[0050] Figure 4 It is a structural block diagram of an electronic device according to the present invention. Detailed Embodiments

[0051] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments. It should be noted that, without conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0052] The following detailed descriptions are all exemplary descriptions, aiming to provide a further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention.

[0053] Abbreviations and Definitions of Key Terms (Explanation of abbreviations involved in the disclosure)

[0054] Karush-Kuhn-Tucker conditions: KKT conditions;

[0055] SSA: Sparrow Search Algorithm;

[0056] BiGRU: Bidirectional Gated Recurrent Unit;

[0057] Attention: Attention mechanism;

[0058] KELM: Kernel Extreme Learning Machine;

[0059] ALO: Ant Lion Optimization Algorithm;

[0060] KDE: Kernel Density Estimation;

[0061] GPD: Generalized Pareto Distribution;

[0062] SG algorithm: Savitzky-Golay algorithm.

[0063] Embodiment 1

[0064] Embodiment 1 of the present invention proposes a bidding method for a wind farm to participate in the electricity market, considering a high proportion of new energy and the uncertainty of wind power. First, a electricity price prediction model based on the SSA-BiGRU-Attention algorithm is established to achieve a more accurate prediction of the electricity price, so as to truly and effectively reflect the revenue of the wind farm when formulating the bidding plan in advance; second, aiming at the problem that a single model can hardly accurately describe the probability density characteristics of wind power, the KELM algorithm, the ALO algorithm and the semi-parametric method are combined to establish a wind power probability density prediction model based on the KELM-ALO-semi-parametric method to characterize the uncertainty of wind power; then, based on the trading rules for the wind farm to participate in the electricity market, a revenue model for the wind farm to participate in the electricity market is constructed, and with the maximum revenue of the wind farm as the objective function, the improved whale algorithm is used to solve the model to obtain the bidding power of the wind farm to participate in the electricity market. This method uses the SSA-BiGRU-Attention model to solve the problem of not considering the electricity price in the current bidding method, uses the KELM-ALO-semi-parametric method to more accurately describe the probability density of wind power, and at the same time adopts the improved whale algorithm to improve the convergence speed and accuracy of the model, and finally more accurately determines the bidding power of the wind farm to participate in the electricity market.

[0065] As Figure 1 and Figure 2 shown, a bidding method for a wind farm to participate in the electricity market includes the following specific steps:

[0066] Step S1: Obtain electricity price influencing factor data; among them, the electricity price influencing factor data may include sample data such as historical electricity price, load, wind power generation, and photovoltaic power generation.

[0067] Step S2: Input the sample data obtained in Step S1, such as historical electricity prices, loads, wind power generation, photovoltaic power generation and other sample data, into the SSA-BiGRU-Attention model to predict the electricity price.

[0068] Specifically, Step S2 includes the following steps:

[0069] Step S21: Use the Attention mechanism to analyze the correlation between each electricity price influencing factor data and the electricity price, and calculate the specific weights of each electricity price influencing factor data;

[0070] Step S22: Adopt the SSA search algorithm to intelligently optimize the parameters of the BiGRU model, such as batch size, learning rate, number of hidden layers, and number of neurons in each layer, to obtain the optimal parameters, that is, the optimized SSA-BiGRU model;

[0071] Step S23: According to the specific weights of each electricity price influencing factor data obtained in Step S21, weight the electricity price influencing factor data, input the weighted sample data into the optimized SSA-BiGRU model for electricity price prediction, and obtain the electricity price prediction result.

[0072] Step S3: Obtain the current numerical weather forecast data and real wind power data.

[0073] Step S4: According to the historical data of the numerical weather forecast data and the real wind power data, use the ALO algorithm to optimize the kernel parameters of the KELM model, substitute the optimal kernel parameters into the KELM to obtain the KELM-ALO model, and input the current numerical weather forecast data and the real wind power data into the KELM-ALO model, and use this KELM-ALO model to perform deterministic prediction of the wind power, and obtain the predicted value of the wind power for the next day and the historical prediction error.

[0074] Specifically, according to Equation (1), the predicted value of the wind power for the next day and the historical prediction error can be obtained:

[0075]

[0076] In the formula: Φ is the predicted wind power obtained through KELM-ALO; N is the number of historical data; w is the wind speed; k is the kernel function; T is the predicted target value vector; I is the identity matrix; C is the kernel parameter; Ω EML is the kernel matrix.

[0077] Step S5: Based on the prediction result obtained by the KELM-ALO model, obtain the probability density prediction function of the wind power for the next day based on the semi-parametric method.

[0078] Specifically, step S5 includes the following steps:

[0079] Step S51: Use the Hill diagram method to determine the threshold u of the predicted wind power value for the next day; the part less than u is used as the peak part, and segmented KDE is performed according to Equation (2) to obtain the density function f u (x); the part greater than u is used as the tail part, and GPD is performed according to Equation (3) to obtain the density curve g u (x);

[0080] Step S52: Combine f u (x) and g u (x) according to Equation (4) to obtain the probability density f G (x) of the semi-parametric hybrid model, and use the SG algorithm to smooth the discontinuous part around the threshold u of the f G (x) curve, so as to obtain the probability density prediction function of the wind power for the next day.

[0081] Specifically, the formula of the density function f u (x) is as follows:

[0082]

[0083] In the formula: x is the wind power; h is the bandwidth; n is the number of samples. x i is the test sample of the wind power determined by h and x.

[0084] Specifically, the formula of the density curve g u (x) is as follows:

[0085]

[0086] In the formula: x is the wind power; v is the shape parameter; σ is the scale parameter.

[0087] Specifically, the formula of the probability density curve of the semi-parametric hybrid model is as follows:

[0088]

[0089] In the formula: F u (x) is the cumulative distribution function of fu(x).

[0090] Step S6: Based on the predicted electricity price and the probability density prediction function of the wind power for the next day, establish the revenue model of the wind farm participating in the electricity market according to Equation (6).

[0091] Specifically, referring to the domestic electricity market mechanism and the relatively mature electricity markets abroad, determine the revenue mechanism of the wind farm participating in the electricity market. In this solution, the total revenue R of the wind farm participating in the energy market and the frequency regulation market is as shown in Equation (5):

[0092] R = R e + R r = P e · L e + f e + P r · L r + f r (5)

[0093] Where: R e and R r are the revenues of the wind farm in the energy market and the frequency regulation market respectively; P e and P r are the bid powers of the wind farm in the day-ahead energy market and the frequency regulation market respectively; L e and L r are the prices of wind power in the day-ahead energy market and the frequency regulation market respectively; f e and f r are the penalty revenues of the wind farm in the two markets respectively.

[0094] The revenue model of the wind farm participating in the power market is as follows in formula (6):

[0095]

[0096] Where: E[R(P e , P r )] is the expected revenue of the wind farm participating in the power market; P t is the actual output power of the wind farm; f(P t ) is the predicted function of the probability density of wind power at the next day's t moment; P t,max is the maximum value of the predicted wind power at the t moment.

[0097] Step S7: Taking the maximum revenue of the wind farm as the objective function, use the improved whale algorithm to solve the above model formula (6) to obtain the bid power of the wind farm in the power market.

[0098] The improved whale algorithm in this solution can integrate Tent chaotic mapping, non-linear parameters, fitness control, and siege mechanism, etc., to ensure the convergence accuracy and speed.

[0099] A bidding method for a wind farm to participate in the electricity market proposed in this solution takes into account the high proportion of new energy and the uncertainty of wind power, including the SSA-BiGRU-Attention electricity price prediction model considering the high proportion of new energy, the wind power probability density prediction model based on the KELM-ALO semi-parameter method, etc. When formulating the day-ahead bidding method for the wind farm, the electricity price factor is considered, and the impact of the high proportion of new energy on the electricity price is considered. By using the SSA-BiGRU-Attention model to predict the electricity price, a more accurate prediction of the electricity price can be achieved, so as to truly reflect the revenue of the wind farm during day-ahead bidding and lay a foundation for determining the bidding power of the wind farm.

[0100] This solution combines the KELM-ALO model with the semi-parameter method to characterize the uncertainty of wind power, which can more accurately describe the probability density of wind power and avoid the poor prediction effect caused by problems such as complex models, the need for prior distribution assumptions, and deviation in estimated parameters in traditional power probability prediction methods.

[0101] This solution uses an improved whale algorithm to solve the model. This algorithm has good global search and local optimization capabilities and can obtain the bidding power of the wind farm participating in the electricity market more efficiently and accurately.

[0102] Embodiment 2

[0103] As Figure 3 shown, a device for a wind farm to participate in the electricity market bidding includes:

[0104] The first acquisition module is used to acquire electricity price influencing factor data.

[0105] In the first acquisition module, the electricity price influencing factor data includes: historical electricity price, load, wind power generation, and photovoltaic power generation.

[0106] The first result prediction module is used to input the electricity price influencing factor data into a preset SSA-BiGRU-Attention model, and the SSA-BiGRU-Attention model outputs an electricity price prediction result.

[0107] The second acquisition module is used to acquire current numerical weather forecast data and real wind power data.

[0108] The second result prediction module is used to input the current numerical weather forecast data and real wind power data into a preset KELM-ALO model, and the KELM-ALO model outputs a predicted value of the next day's wind power.

[0109] In the second result prediction module, the acquisition method of the KELM-ALO model is as follows:

[0110] Based on the historical data of numerical weather prediction data and real wind power data, the kernel parameters of the KELM model are optimized using the ALO algorithm to obtain the optimal kernel parameters; the optimal kernel parameters are substituted into the KELM model to obtain the KELM-ALO model.

[0111] The KELM-ALO model is as follows:

[0112]

[0113] In the formula: Φ is the predicted wind power obtained through KELM-ALO; N is the number of historical data; w is the wind speed; k is the kernel function; T is the predicted target value vector; I is the identity matrix; C is the kernel parameter; Ω EML is the kernel matrix.

[0114] The probability density prediction function acquisition module is used to obtain the probability density prediction function of the predicted wind power value for the next day using the semi-parametric method based on the predicted wind power value for the next day.

[0115] In the probability density prediction function acquisition module, it specifically includes:

[0116] Use the Hill diagram method to determine the threshold u of the predicted wind power value for the next day; the part less than u is used as the peak part for piecewise KDE to obtain the density function f u (x); the part greater than u is used as the tail for GPD to obtain the density curve g u (x);

[0117] Combine f u (x) with g u (x) to obtain the probability density f of the semi-parametric hybrid model G (x);

[0118] Use the SG algorithm to smooth the discontinuous part around the threshold u of the f G (x) curve to obtain the probability density prediction function of the predicted wind power value for the next day.

[0119] The probability density formula of the semi-parametric hybrid model is as follows:

[0120]

[0121] In the formula: F u (x) is the cumulative distribution function of fu(x).

[0122] The revenue model establishment module is used to establish a revenue model for the wind farm to participate in the electricity market based on the electricity price prediction result and the probability density prediction function of the predicted wind power value for the next day.

[0123] In the revenue model establishment module, the revenue model for the wind farm to participate in the electricity market is as follows:

[0124]

[0125] Where: E[R(P e ,P r ) is the expected revenue of the wind farm participating in the electricity market; P t is the actual output power of the wind farm; f(P t ) is the predicted probability density function of wind power at the next day's time t; P t,max is the maximum value of the predicted wind power at time t.

[0126] The solving module is used to take the maximum revenue of the wind farm as the objective function, and use the improved whale algorithm to solve the revenue model of the wind farm participating in the electricity market, so as to obtain the bidding power of the wind farm in the electricity market.

[0127] Embodiment 3

[0128] As Figure 4 shown, the present invention also provides an electronic device 100 for implementing the method for the wind farm to participate in the electricity market bidding in the above embodiment; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on at least one processor 102, and at least one communication bus 104. The memory 101 can be used to store the computer program 103. The processor 102 realizes the steps of the method for the wind farm to participate in the electricity market bidding in Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. Among them, the program storage area may store an operating system, application programs required for at least one function (such as a sound playing function, an image playing function, etc.); the data storage area may store data created according to the use of the electronic device 100 (such as audio data, etc.). In addition, the memory 101 may include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0129] At least one processor 102 can be a Central Processing Unit (CPU), or can also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 can be a microprocessor or the processor 102 can also be any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and connects various parts of the entire electronic device 100 through various interfaces and lines.

[0130] The memory 101 in the electronic device 100 stores multiple instructions to implement a method for a wind farm to participate in the electricity market bidding. The processor 102 can execute the multiple instructions to implement:

[0131] Obtain electricity price influencing factor data;

[0132] Input the electricity price influencing factor data into a preset SSA-BiGRU-Attention model, and the SSA-BiGRU-Attention model outputs an electricity price prediction result;

[0133] Obtain the current numerical weather forecast data and actual wind power data;

[0134] Input the current numerical weather forecast data and actual wind power data into a preset KELM-ALO model, and the KELM-ALO model outputs a predicted value of the next day's wind power;

[0135] Based on the predicted value of the next day's wind power, use the semi-parametric method to obtain a probability density prediction function of the predicted value of the next day's wind power;

[0136] Based on the electricity price prediction result and the probability density prediction function of the predicted value of the next day's wind power, establish a revenue model for the wind farm to participate in the electricity market;

[0137] Taking the maximum wind farm revenue as the objective function, use the improved whale algorithm to solve the revenue model for the wind farm to participate in the electricity market, and obtain the bidding power of the wind farm in the electricity market.

[0138] Embodiment 4

[0139] If the integrated module / unit of the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, and read-only memory (ROM, Read-Only Memory).

[0140] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer-usable program code.

[0141] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0142] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or boxes Figure 1 one process or a plurality of processes and / or boxes Figure 1 steps of the functions specified in one box or a plurality of boxes.

[0144] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A method for a wind farm to participate in power market bidding, characterized in that, It includes the following steps: Obtain data on factors affecting electricity prices; Input the data on factors affecting electricity prices into a preset SSA-BiGRU-Attention model, and the SSA-BiGRU-Attention model outputs an electricity price prediction result; Obtain current numerical weather forecast data and actual wind power data; Input the current numerical weather forecast data and actual wind power data into a preset KELM-ALO model, and the KELM-ALO model outputs a predicted value of the next-day wind power; Based on the predicted value of the next-day wind power, use the semi-parametric method to obtain a probability density prediction function of the predicted value of the next-day wind power; Based on the electricity price prediction result and the probability density prediction function of the predicted value of the next-day wind power, establish a revenue model for the wind farm to participate in the electricity market; Taking the maximum revenue of the wind farm as the objective function, use the improved whale algorithm to solve the revenue model for the wind farm to participate in the electricity market, and obtain the bidding power of the wind farm in the electricity market; In the step of inputting the current numerical weather forecast data and actual wind power data into a preset KELM-ALO model, the acquisition method of the KELM-ALO model is as follows: According to the historical data of numerical weather forecast data and actual wind power data, use the ALO algorithm to optimize the kernel parameters of the KELM model to obtain the optimal kernel parameters; Substitute the optimal kernel parameters into the KELM model to obtain the KELM-ALO model; The KELM-ALO model is as follows: Where: N is the number of historical data; w is the wind speed; is the kernel function; T is the predicted target value vector; I is the identity matrix; C is the kernel parameter; is the kernel matrix; The step of obtaining the probability density prediction function of the predicted value of the next-day wind power by using the semi-parametric method based on the predicted value of the next-day wind power specifically includes: Determine the threshold value of the predicted wind power for the next day using the Hill diagram method u ; Less than u The part is segmented by KDE as the peak part to obtain the density function ; Greater than u The part is used for GPD as the tail part to obtain the density curve ; Combine with to obtain the probability density curve of the semi-parametric hybrid model ; Smoothing using the SG algorithm Curve threshold u For the discontinuous parts of the periphery, obtain the probability density prediction function of the next-day wind power 2. The method for a wind farm to participate in power market bidding according to claim 1, characterized in that, In the step of obtaining data on factors affecting electricity prices, the data on factors affecting electricity prices include the following: Historical electricity prices, loads, wind power generation, and photovoltaic power generation.

3. The method for a wind farm to participate in power market bidding according to claim 1, wherein The probability density formula of the semi-parametric hybrid model is as follows: In the formula: F u ( x ) is fu ( x )'s cumulative distribution function.

4. The method for a wind farm to participate in electricity market bidding according to claim 1, wherein The revenue model for the wind farm to participate in the electricity market is as follows: Wherein: is the expected revenue of the wind farm participating in the electricity market; is the actual output power of the wind farm; is the next day t wind power probability density prediction function at time; is t the maximum value of wind power prediction at time.

5. A device for a wind farm to participate in power market bidding, which is used to implement the method for a wind farm to participate in power market bidding as described in claim 1, characterized in that It includes: A first acquisition module for obtaining data on factors affecting electricity prices; A first result prediction module for inputting the data on factors affecting electricity prices into a preset SSA-BiGRU-Attention model, and the SSA-BiGRU-Attention model outputs an electricity price prediction result; A second acquisition module for obtaining current numerical weather forecast data and actual wind power data; A second result prediction module for inputting the current numerical weather forecast data and actual wind power data into a preset KELM-ALO model, and the KELM-ALO model outputs a predicted value of the next-day wind power; A probability density prediction function acquisition module for obtaining a probability density prediction function of the predicted value of the next-day wind power by using the semi-parametric method based on the predicted value of the next-day wind power; A revenue model establishment module for establishing a revenue model for the wind farm to participate in the electricity market based on the electricity price prediction result and the probability density prediction function of the predicted value of the next-day wind power; A solution module, which is used to take the maximum benefit of the wind farm as the objective function, and use the improved whale algorithm to solve the benefit model of the wind farm participating in the electricity market, so as to obtain the bidding power of the wind farm in the electricity market.

6. An electronic device, characterized in that, It includes a processor and a memory. The processor is used to execute the computer program stored in the memory to implement the method for the wind farm to participate in the electricity market bidding as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, it implements the method for the wind farm to participate in the electricity market bidding as described in any one of claims 1 to 4.

Images