Super capacitor energy storage frequency modulation method and device, medium and program product

Abstract

The invention discloses a super capacitor energy storage frequency modulation method and device, a medium and a program product, and belongs to the technical field of wind power generation. According to the super capacitor energy storage frequency modulation method provided by the invention, wind speed data are collected, an original wind speed sequence is established, the original wind speed sequence is divided into a first prediction set and a first verification set, and a reference error is predicted and calculated by using the first prediction set; performing position pseudo-random interchange on the subsequences in the first prediction set to generate a plurality of transformed wind speed sequences; dividing the transformed wind speed sequence into a second prediction set and a second verification set, and predicting and calculating a prediction error parameter by using the second prediction set; according to the method, the inherent rules and characteristics of the wind speed data can be fully utilized, the accuracy and reliability of wind speed prediction are improved through the position pseudo-random interchange technology, and more accurate data support is provided for grid-connected dispatching of a wind power plant.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and in particular to a supercapacitor energy storage frequency modulation method, equipment, medium, and program product. Background Art

[0002] Wind power generation, as a clean, environmentally friendly and sustainable renewable energy technology, has attracted much attention and has developed rapidly in recent years. With the rapid development of technology and the continuous reduction of costs, wind power generation has gradually become an indispensable part of the energy structure of many countries.

[0003] However, the natural randomness of wind speed poses a severe challenge to the grid connection of large-scale wind farms. This randomness causes the output power of wind farms to show volatility and uncertainty, which brings great difficulty to the grid dispatching agency in prediction and arrangement. This not only increases the operating burden of the grid, but also may pose a potential threat to the stability of the system. As one of the core technologies in the field of wind power generation, the accuracy and reliability of wind speed prediction are crucial. Wind farm scheduling is the key to ensuring the coordinated operation between wind farms and power grids. It involves intelligent control and optimization of the output power and operating status of wind farms to balance the output power of wind farms and the actual needs of the power grid, and plays an important role in wind farm scheduling.

[0004] Traditional wind speed prediction methods generally put the original wind speed sequence directly into the neural network for prediction, and then calculate the power sent by the wind farm to the grid based on the predicted wind speed value. However, this method fails to fully utilize the inherent laws and characteristics of wind speed data, resulting in large prediction errors. This not only increases the operating burden of the grid, but also poses a potential threat to the stability of the system.

[0005] Therefore, how to more effectively utilize the inherent laws and characteristics of wind speed data to improve the accuracy and reliability of prediction has become a key technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0006] The purpose of the present invention is to provide a supercapacitor energy storage frequency modulation method, device, medium, and program product, aiming to fundamentally solve the problem of insufficient wind speed prediction accuracy and reliability in traditional methods.

[0007] The present invention solves the above technical problems through the following technical solutions: A supercapacitor energy storage frequency modulation method comprises the following steps: Step 1: Collect wind speed data, establish an original wind speed sequence, and divide the original wind speed sequence into a first prediction set and a first verification set according to a sequence and a first preset ratio; input the first prediction set into the trained neural network model in sequence for prediction to obtain a first prediction value, and calculate a benchmark prediction error W based on the first prediction value and the actual value in the first verification set; Step 2: Let the first prediction set be X0(t)=[X 01 ,X 02 ,X 03 ,...,X 0N ], N is the number of subsequences in X0(t), X 01 ,X 02 ,X 03 ,...,X 0N are the 1st, 2nd, 3rd, …, Nth subsequences in the first prediction set respectively; the subsequences in X0(t) are pseudo-randomly swapped for input positions to obtain N wind speed sequences after pseudo-randomly swapping input positions. Let X i (t) is the i-th wind speed sequence after the positions of N! input quantities are pseudo-randomly swapped, i∈{1~N!}, let i=1, and execute step 3; Step 3: Convert wind speed sequence X i (t) Dividing the set into a second prediction set and a second verification set according to the order of precedence and the second preset ratio, inputting the second prediction set into the trained neural network model in order for prediction, obtaining a second prediction value, and calculating the prediction error parameter A based on the second prediction value and the actual value in the second verification set. i ; Step 4: Determine whether i is equal to N! If not, set i=i+1 and execute step 3; if yes, execute step 5; Step 5: Obtain prediction error parameter A i The minimum value A in min , and get the corresponding wind speed sequence X min (t), based on the baseline prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence, input the final wind speed sequence into the trained neural network model in sequence for prediction, and obtain the final wind speed prediction result; Step 6. Calculate the predicted power sent by the wind farm to the grid based on the final wind speed prediction result, obtain the actual power sent by the wind farm to the grid, and calculate the difference between the predicted power and the actual power. When the difference is positive, adjust the super-capacity energy storage system connected to the wind farm exit to the charging state; when the difference is 0, the super-capacity energy storage system connected to the wind farm exit maintains the current state unchanged; when the difference is negative, adjust the super-capacity energy storage system connected to the wind farm exit to the discharging state.

[0008] A further improvement of the present invention is that the first preset ratio is 8:2 or 9:1.

[0009] A further improvement of the present invention is that the second preset ratio is equal to the first preset ratio.

[0010] A further improvement of the present invention is that the reference prediction error W is specifically:

[0011] Among them, y i is the i-th first prediction value, i is the ith actual value in the first validation set, and n is the number of subsequences in the first validation set.

[0012] The present invention is further improved in that: based on the reference prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence X f (t) Specifically: Assume that the wind speed sequence X min (t) is [X1,X2,X3,...,X N ], Final wind speed sequence X f (t)=[X 01 +sigmoid(X1)×X1× , X 02 +sigmoid(X 02 )×X2× , X 03 +sigmoid(X3)×X3× , ..., X 0N +sigmoid(X N )×X N × ]; Among them, X1,X2,X3,...,X N is the wind speed sequence X min The 1st, 2nd, 3rd, …, Nth subsequence in (t); sigmoid() is the activation function; tanh() is the activation function.

[0013] A further improvement of the present invention is that after collecting the wind speed data, the method further includes the following steps: deleting and supplementing abnormal data.

[0014] Furthermore, the neural network model is LSTM (Long Short-Term Memory).

[0015] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0016] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0017] A computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0018] Compared with the prior art, the positive and progressive effects of the present invention are: The supercapacitor energy storage frequency modulation method provided by the present invention collects wind speed data and establishes an original wind speed sequence, divides the original wind speed sequence into a first prediction set and a first verification set, uses the first prediction set to predict and calculate the benchmark error; then performs pseudo-random position swapping on the subsequences in the first prediction set to generate multiple transformed wind speed sequences; divides the transformed wind speed sequence into a second prediction set and a second verification set, uses the second prediction set to predict and calculate the prediction error parameter; selects the wind speed sequence with the smallest prediction error parameter to construct a final wind speed sequence. The method can make full use of the inherent laws and characteristics of wind speed data, improves the accuracy and reliability of wind speed prediction through the pseudo-random position swapping technology, and provides more accurate data support for the grid-connected scheduling of wind farms.

[0019] Furthermore, a reasonable proportion division ensures the adequacy of the training data (the first prediction set), while the validation data (the first validation set) can also provide sufficient samples for error evaluation, which helps to improve the generalization ability of the prediction model.

[0020] Furthermore, keeping the ratio of the two data sets consistent helps maintain the consistency of the model's prediction performance at different stages and avoids prediction bias caused by uneven data division.

[0021] Furthermore, by comprehensively considering multiple factors, including the first prediction set, the benchmark prediction error, the minimum prediction error parameter, and the corresponding wind speed sequence, the constructed final wind speed sequence can more accurately reflect the changing trend of wind speed and improve the accuracy of the prediction; the use of sigmoid and tanh activation functions for nonlinear transformation helps to further improve the accuracy of the prediction model.

[0022] Furthermore, by deleting abnormal data, erroneous data points caused by equipment failure, measurement errors, or data transmission errors can be eliminated, thereby improving the overall quality of the wind speed dataset and improving the accuracy and credibility of the prediction results; at the same time, reducing the model's dependence on abnormal data, thereby enhancing the model's generalization ability and enabling it to better adapt to wind speed prediction tasks in different scenarios.

[0023] Furthermore, the use of LSTM neural network can effectively process the noise and outliers in the input data, thereby improving the robustness of the prediction model, adapting to complex and changeable wind speed changes, and optimizing the model training process. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings in the specification are used to provide further understanding of the present invention and constitute a part 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 improper limitations on the present invention.

[0025] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] In the description of the present invention, it should be understood that the terms “include” and “comprises” indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0028] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0029] It should be understood that, although the terms first, second, third, etc. may be used to describe preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0030] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0031] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0032] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments, which are intended to explain the present invention rather than to limit it.

[0033] See also Figure 1 The present invention provides a supercapacitor energy storage frequency modulation method, comprising the following steps: Step 1: Collect wind speed data, establish an original wind speed sequence, and divide the original wind speed sequence into a first prediction set and a first verification set according to a sequence and a first preset ratio; input the first prediction set into the trained neural network model in sequence for prediction to obtain a first prediction value, and calculate a benchmark prediction error W based on the first prediction value and the actual value in the first verification set; Step 2: Let the first prediction set be X0(t)=[X 01 ,X 02 ,X 03 ,...,X 0N ], N is the number of subsequences in X0(t), X 01 ,X 02 ,X 03 ,...,X 0N are the 1st, 2nd, 3rd, …, Nth subsequences in the first prediction set respectively; the subsequences in X0(t) are pseudo-randomly swapped for input positions to obtain N wind speed sequences after pseudo-randomly swapping input positions. Let X i (t) is the i-th wind speed sequence after the positions of N! input quantities are pseudo-randomly swapped, i∈{1~N!}, let i=1, and execute step 3; Step 3: Convert wind speed sequence X i(t) Dividing the set into a second prediction set and a second verification set according to the order of precedence and the second preset ratio, inputting the second prediction set into the trained neural network model in order for prediction, obtaining a second prediction value, and calculating the prediction error parameter A based on the second prediction value and the actual value in the second verification set. i ; Step 4: Determine whether i is equal to N! If not, set i=i+1 and execute step 3; if yes, execute step 5; Step 5: Obtain prediction error parameter A i The minimum value A in min , and get the corresponding wind speed sequence X min (t), based on the baseline prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence, input the final wind speed sequence into the trained neural network model in sequence for prediction, and obtain the final wind speed prediction result; Step 6. Calculate the predicted power sent by the wind farm to the grid based on the final wind speed prediction result, obtain the actual power sent by the wind farm to the grid, and calculate the difference between the predicted power and the actual power. When the difference is positive, adjust the super-capacity energy storage system connected to the wind farm exit to the charging state; when the difference is 0, the super-capacity energy storage system connected to the wind farm exit maintains the current state unchanged; when the difference is negative, adjust the super-capacity energy storage system connected to the wind farm exit to the discharging state.

[0034] By collecting wind speed data and establishing an original wind speed sequence, the original wind speed sequence is divided into a first prediction set and a first verification set, and the first prediction set is used to predict and calculate the benchmark error; then the subsequences in the first prediction set are pseudo-randomly interchanged to generate multiple transformed wind speed sequences; the transformed wind speed sequence is divided into a second prediction set and a second verification set, and the second prediction set is used to predict and calculate the prediction error parameter; the wind speed sequence with the smallest prediction error parameter is selected to construct the final wind speed sequence. This method can make full use of the inherent laws and characteristics of wind speed data, improve the accuracy and reliability of wind speed prediction through the position pseudo-random interchange technology, and provide more accurate data support for the grid-connected scheduling of wind farms.

[0035] Specifically, the first preset ratio is 8:2 or 9:1; reasonable ratio division ensures the adequacy of training data (first prediction set), while the verification data (first verification set) can also provide sufficient samples for error evaluation, which helps to improve the generalization ability of the prediction model.

[0036] Specifically, the second preset ratio is equal to the first preset ratio; keeping the ratios of the two data sets consistent helps to maintain the consistency of the prediction performance of the model at different stages and avoid prediction bias caused by uneven data division.

[0037] Specifically, the benchmark prediction error W is:

[0038] Among them, y i is the i-th first prediction value, i is the ith actual value in the first validation set, and n is the number of subsequences in the first validation set.

[0039] Specifically, based on the baseline prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence X f (t) Specifically: Assume that the wind speed sequence X min (t) is [X1,X2,X3,...,X N ], the final wind speed sequence X f (t)=[X 01 +sigmoid(X1)×X1× ,X 02 +sigmoid(X 02 )×X2× ,X 03 +sigmoid(X3)×X3× ,...,X 0N +sigmoid(X N )×X N × ]; Among them, X1,X2,X3,...,X N is the wind speed sequence X min The 1st, 2nd, 3rd, …, Nth subsequence in (t); sigmoid() is the activation function; tanh() is the activation function.

[0040] By comprehensively considering multiple factors, including the first prediction set, the benchmark prediction error, the minimum prediction error parameter, and the corresponding wind speed sequence, the final wind speed sequence constructed can more accurately reflect the changing trend of wind speed and improve the accuracy of the prediction; the use of sigmoid and tanh activation functions for nonlinear transformation helps to further improve the accuracy of the prediction model.

[0041] Specifically, after collecting wind speed data, the following steps are also included: deleting and supplementing abnormal data; by deleting abnormal data, erroneous data points caused by equipment failure, measurement errors or data transmission errors can be eliminated, thereby improving the overall quality of the wind speed data set and improving the accuracy and credibility of the prediction results; at the same time, reducing the model's dependence on abnormal data, thereby enhancing the model's generalization ability and enabling it to better adapt to wind speed prediction tasks in different scenarios.

[0042] Specifically, the neural network model is LSTM; the use of LSTM neural network can effectively process noise and outliers in input data, thereby improving the robustness of the prediction model, adapting to complex and changeable wind speed changes, and optimizing the model training process.

[0043] In a specific embodiment of the present invention, the pseudo-random exchange of input positions specifically includes the following steps: assuming that the first prediction set X0(t)=[a,b,c,d,e], a,b,c,d,e are 5 subsequences of the first prediction set; performing pseudo-random exchange of input positions on the subsequences in X0(t) to obtain 5 wind speed sequences after pseudo-random exchange of input positions, The wind speed sequence after pseudo-randomly swapping the positions of the 5 input quantities is: Wind speed sequence after pseudo-randomly swapping the positions of 24 input quantities starting with "a": abcde, abced, abdce, abdec, abecd, abedc, acbde, acbed, acdbe, acdeb, acebd, acedb, adbce, adbec, adcbe, adceb, adebc, adecb, aebcd, aebdc, aecbd, aecdb, aedbc, aedcb; Wind speed sequence after pseudo-randomly swapping the positions of 24 input quantities starting with "b": bacde, baced, badce, badec, baecd, baedc, bcade, bcaed, bcdae, bcdea, bcead, bceda, bdace, bdaec, bdcae, bdcea, bdeac, bdeca, beacd, beadc, becad, becda, bedac, bedca; Wind speed sequence after pseudo-randomly swapping the positions of 24 input quantities starting with "c": cabde, cabed, cadbe, cadeb, caebd, caedb, cbade, cbaed, cbdae, cbdea, cbead, cbeda, cdabe, cdaeb, cdbae, cdbea, cdeab, cdeba, ceabd, ceadb, cebad, cebda, cedab, cedba; The wind speed sequence after the positions of the 24 input quantities starting with "d" are pseudo-randomly swapped: dabce, dabec, dacbe, daceb, daebc, daecb, dbace, dbaec, dbcae, dbcea, dbeac, dbeca, dcabe, dcaeb, dcbae, dcbea, dceab, dceba, deabc, deacb, debac, debc a, decab, decba; Wind speed sequence after pseudo-randomly swapping the positions of 24 input quantities starting with "e": eabcd, eabdc, eacbd, eacdb, eadbc, eadcb, ebacd, ebadc, ebcad, ebcda, ebdac, ebdca, ecabd, ecadb, ecbad, ecbda, ecdab, ecdba, edabc, edacb, edbac, edbca, edcab, edcba.

[0044] The first preset ratio is 8:2, let i=1, and perform the following steps; The wind speed sequence X i (t) dividing into a second prediction set and a second verification set according to the order and the same preset ratio, wherein the second prediction set is the first 4 subsequences and the second verification set is the last 1 subsequence; sequentially inputting the above second prediction set into the trained neural network model for prediction to obtain a second prediction value, and calculating the prediction error parameter A based on the second prediction value and the actual value in the second verification set i .

[0045] Determine whether i is equal to 5! If the judgment is no, set i=i+1 and continue to execute the above steps. If the judgment is yes, obtain the prediction error parameter A i The minimum value A in min , assuming that the above prediction error parameter A i The minimum value of A 106 , and get the corresponding wind speed sequence X min (t) = [e, b, c, d, a], where the second prediction set is [e, b, c, d] and the second validation set is [a]. Based on the benchmark prediction error W and the minimum value Amin= A 106 , wind speed sequence X min (t)=[e, b, c, d, a] and the first prediction set X0(t)=[a, b, c, d, e], construct the final wind speed sequence, Final wind speed sequence X f (t)=[a+sigmoid(e)×e× ,b+sigmoid(b)×b× ,c+sigmoid(c)×c× , d+sigmoid(d)×d× ],e+sigmoid(a)×a× ].

[0046] In order to further verify the advantages of this method, the wind speed is predicted by using the method of this application and the traditional prediction method VMD (Variational Mode Decomposition). In the experiment of this application, MAPE (Mean Absolute Percentage Error) is selected as the evaluation standard of each model. The prediction results (i.e., MAPE values) are as follows:

[0047] It can be seen that the wind speed prediction using this method is more accurate and reliable.

[0048] Based on the same inventive concept, an embodiment of the present application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the supercapacitor energy storage frequency modulation method are implemented, wherein the memory may include a memory, such as a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk memory, etc.; the processor, the network interface, and the memory are interconnected through an internal bus, and the internal bus may be an industrial standard architecture bus, a peripheral component interconnection standard bus, an extended industrial standard structure bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory is used to store programs. Specifically, the program may include a program code, and the program code includes computer operation instructions. The memory may include a memory and a non-volatile memory, and provide instructions and data to the processor.

[0049] Based on the same inventive concept, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the supercapacitor energy storage frequency modulation method are implemented. Specifically, the computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may include RAM (Random Access Memory) and / or cache memory (cache), etc. The non-volatile memory may include ROM (Read-Only Memory), hard disk, flash memory, optical disk, magnetic disk, etc.

[0050] Based on the same inventive concept, an embodiment of the present application provides a computer program product, which includes a computer program stored on a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer device, the computer device executes the steps of the above-mentioned supercapacitor energy storage frequency modulation method.

[0051] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM (Compact Disc Read-Only Memory), optical storage, etc.) containing computer-usable program codes.

[0052] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (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, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer device or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0053] These computer program instructions may also be stored in a computer readable memory capable of directing a computer device or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0054] These computer program instructions can also be loaded onto a computer device or other programmable data processing device so that a series of operation steps are executed on the computer device or other programmable device to produce a process implemented by the computer device, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0055] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0056] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A supercapacitor energy storage frequency modulation method, characterized in that: The following steps are involved: Step 1: Collect wind speed data, establish an original wind speed sequence, and divide the original wind speed sequence into a first prediction set and a first verification set according to a sequence and a first preset ratio; input the first prediction set into the trained neural network model in sequence for prediction to obtain a first prediction value, and calculate a benchmark prediction error W based on the first prediction value and the actual value in the first verification set; Step 2: Let the first prediction set be X0(t)=[X 01 ,X 02 ,X 03 ,...,X 0N ], N is the number of subsequences in X0(t), X 01 ,X 02 ,X 03 ,...,X 0N are the 1st, 2nd, 3rd, …, Nth subsequences in the first prediction set respectively; the subsequences in X0(t) are pseudo-randomly swapped for input positions to obtain N wind speed sequences after pseudo-randomly swapping input positions. Let X i (t) is the i-th wind speed sequence after the positions of N! input quantities are pseudo-randomly swapped, i∈{1~N!}, let i=1, and execute step 3; Step 3: Convert wind speed sequence X i (t) Dividing the set into a second prediction set and a second verification set according to the order of precedence and the second preset ratio, inputting the second prediction set into the trained neural network model in order for prediction, obtaining a second prediction value, and calculating the prediction error parameter A based on the second prediction value and the actual value in the second verification set. i ; Step 4: Determine whether i is equal to N! If not, set i=i+1 and execute step 3; if yes, execute step 5; Step 5: Obtain prediction error parameter A i The minimum value A in min , and get the corresponding wind speed sequence X min (t), based on the baseline prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence, input the final wind speed sequence into the trained neural network model in sequence for prediction, and obtain the final wind speed prediction result; Step 6. Calculate the predicted power sent by the wind farm to the grid based on the final wind speed prediction result, obtain the actual power sent by the wind farm to the grid, and calculate the difference between the predicted power and the actual power. When the difference is positive, adjust the super-capacity energy storage system connected to the wind farm exit to the charging state; when the difference is 0, the super-capacity energy storage system connected to the wind farm exit maintains the current state unchanged; when the difference is negative, adjust the super-capacity energy storage system connected to the wind farm exit to the discharging state.

2. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: The first preset ratio is 8:2 or 9:

1.

3. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: The second preset ratio is equal to the first preset ratio.

4. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: The benchmark prediction error W is specifically: Among them, y i is the i-th first prediction value, i is the ith actual value in the first validation set, and n is the number of subsequences in the first validation set.

5. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: Based on the benchmark prediction error W, the minimum value A min , wind speed sequence X min (t) and the first prediction set X0(t), construct the final wind speed sequence X f (t) Specifically: Assume that the wind speed sequence X min (t) is [X1,X2,X3,...,X N ], Final wind speed sequence X f (t)=[X 01 +sigmoid(X1)×X1× , X 02 +sigmoid(X 02 )×X2× , X 03 +sigmoid(X3)×X3× , ..., X 0N +sigmoid(X N )×X N × ]; Among them, X1,X2,X3,...,X N is the wind speed sequence X min The 1st, 2nd, 3rd, …, Nth subsequence in (t); sigmoid() is the activation function; tanh() is the activation function.

6. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: After collecting wind speed data, the following steps are also included: deleting and supplementing abnormal data.

7. A supercapacitor energy storage frequency modulation method according to claim 1, characterized in that: The neural network model is LSTM.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images