A prediction method and system for energy storage frequency modulation instructions

Abstract

The present invention relates to the technical field of energy storage frequency modulation, in particular to a method and system for predicting energy storage frequency modulation commands, which includes decomposing a frequency modulation command sequence Pt by using VMD to obtain a plurality of IMF subsequences; according to the total area S of the current IMF subsequence ii and the error W of the experimental group i , use an activation function to determine whether the current IMF subsequence is a false sequence; improve the IMF subsequence determined to be false, and use the improved IMF subsequence for predicting and outputting frequency modulation commands. The beneficial effect of the present invention is that it can effectively identify and improve the false sequences that may be generated in the variational mode decomposition process. This method can accurately find and process false sequences by calculating the area enclosed by adjacent data points in the subsequence and the X-axis, reducing the prediction error and improving the prediction accuracy. In addition, the false sequences are optimized by means such as generating random sequences and performing mirror transformation, further improving the response speed and frequency modulation efficiency of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage frequency modulation, and particularly to a method and system for predicting energy storage frequency modulation commands. Background Art

[0002] In the existing hybrid energy storage assisted thermal power unit frequency modulation technology, the commonly used method is to allocate the difference between the frequency modulation command and the output of the thermal power unit to a hybrid energy storage system composed of a supercapacitor and a lithium-ion battery. Among them, the low-frequency frequency modulation task is undertaken by the lithium battery, and the high-frequency frequency modulation task is responsible for the supercapacitor. However, the signal transmission delay and the response time difference of the energy storage unit in this process affect the frequency modulation efficiency and economic benefits of the system. In addition, traditional frequency modulation command prediction methods, such as using VMD decomposition to decompose the original sequence into multiple subsequences for prediction, have problems such as mode mixing, end effects, and sensitivity to initial conditions. These problems may cause useful signals to be misjudged as noise, reduce the prediction accuracy, and thus affect the performance of the entire frequency modulation system. Therefore, there is an urgent need for a new prediction method to improve the accuracy of frequency modulation command prediction and reduce the response time difference, so as to improve the performance and benefits of the hybrid energy storage system in the frequency modulation application of thermal power units. Summary of the Invention

[0003] To solve the above technical problems, a method for predicting energy storage frequency modulation commands is proposed, including decomposing the frequency modulation command sequence Pt using VMD to obtain multiple IMF subsequences;

[0004] According to the total area S of the current IMF subsequence ii and the error W of the experimental group i , use the activation function to determine whether the current IMF subsequence is a false sequence;

[0005] Improve the IMF subsequence determined to be false, and use the improved IMF subsequence for predicting and outputting the frequency modulation command;

[0006] Among them, improving the IMF subsequence determined to be false includes:

[0007] When there are two or more false IMF subsequences, generate a random sequence to replace the false IMF subsequence; or,

[0008] When there is a single false IMF subsequence, calculate the mean value of the current false IMF subsequence, and mirror and change the position of the current false IMF subsequence in the coordinate system with the horizontal line where Y is equal to the current mean value as the axis of symmetry, and determine the changed IMF subsequence as the final IMF subsequence.

[0009] As a preferred solution of a method for predicting energy storage frequency modulation commands according to the present invention, wherein: determining the total area S of any IMF subsequence ii including:

[0010] successively determining the areas enclosed by two adjacent data points and the X-axis under the current IMF subsequence, where the enclosed areas include the magnitude of the area value and the area attribute;

[0011] summing up all the enclosed areas under the current IMF subsequence to obtain the total area S of the current IMF subsequence ii ;

[0012] The area enclosed by two adjacent data points and the X-axis is the area of the figure enclosed by the line connecting the two adjacent data points, the perpendicular connections of the two adjacent data points to the X-axis respectively, and the line connecting the coordinates of the two adjacent data points on the X-axis.

[0013] As a preferred solution of a method for predicting energy storage frequency modulation commands according to the present invention, wherein: the area attribute includes positive and negative values. When the ordinates of two adjacent data points are both greater than or equal to 0, the current area attribute is positive;

[0014] When the ordinates of two adjacent data points are both less than 0, the current area attribute is negative;

[0015] When one of the ordinates of two adjacent data points is greater than 0 and the other is less than 0, the current area attribute is determined by the area enclosed by the current two adjacent points and the X-axis.

[0016] As a preferred solution of a method for predicting energy storage frequency modulation commands according to the present invention, wherein: the current area attribute being determined by the area enclosed by the current two adjacent points and the X-axis includes:

[0017] When the area enclosed by the current two adjacent points and the X-axis is greater than or equal to 0, the current area attribute is positive;

[0018] When the area enclosed by the current two adjacent points and the X-axis is less than 0, the current area attribute is negative.

[0019] As a preferred solution of a method for predicting energy storage frequency modulation commands according to the present invention, wherein: determining whether the IMF subsequence is a false sequence by using an activation function includes:

[0020] If the magnitude of exp(sigmoid(W i )*S ii ) is between [0.8, 1.1], it indicates that the corresponding current sequence is a false IMF subsequence, where sigmoid() represents the activation function.

[0021] As a preferred solution of a prediction method for energy storage frequency modulation commands of the present invention, wherein: the determination of the random sequence is to use the random sequence with the smallest area enclosed by the X-axis as the final IMF subsequence.

[0022] A prediction system for energy storage frequency modulation commands includes a decomposition module for decomposing the frequency modulation command sequence Pt using VMD to obtain multiple IMF subsequences;

[0023] A determination module for, according to the total area S of the current IMF subsequence ii and the error W of the test group i , use an activation function to determine whether the current IMF subsequence is a false sequence;

[0024] An improved prediction module for improving the IMF subsequence determined to be false and using the improved IMF subsequence for predicting and outputting the frequency modulation command.

[0025] A computer device includes a memory and a processor, the memory stores a computer program, and is characterized in that the steps of a prediction method for energy storage frequency modulation commands are implemented when the processor executes the computer program.

[0026] A computer-readable storage medium stores a computer program, and is characterized in that the steps of a prediction method for energy storage frequency modulation commands are implemented when the computer program is executed by a processor.

[0027] Advantages of the present invention: It can effectively identify and improve the false sequences that may be generated during the variational mode decomposition process. This method can accurately find and process false sequences by calculating the area enclosed by adjacent data points in the subsequence and the X-axis, reducing the prediction error and improving the prediction accuracy. In addition, the false sequences are optimized by means such as generating random sequences and performing mirror transformations, further improving the response speed and frequency modulation efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0029] Figure 1 It is a schematic flowchart of a prediction method for energy storage frequency modulation commands.

[0030] Figure 2 It is a flowchart of a prediction method for energy storage frequency modulation commands.

[0031] Figure 3Schematic diagram of positive area attribute.

[0032] Figure 4 Schematic diagram of negative area attribute.

[0033] Figure 5 Schematic diagram of area attribute determination under special circumstances.

[0034] Figure 6 Schematic diagrams before and after mirroring. Detailed implementation manners

[0035] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe the detailed implementation manners of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0036] Example 1, referring to Figures 1 to 6 , which is the first embodiment of the present invention. This embodiment provides a method for predicting energy storage frequency modulation commands, including:

[0037] S1. Use VMD to decompose the frequency modulation command sequence Pt to obtain multiple IMF subsequences, where VMD is variational mode decomposition, and the IMF subsequence is a time series component with different characteristics obtained after VMD decomposes the frequency modulation command sequence Pt.

[0038] It should be noted that assume the frequency modulation command is Pt, and after VMD decomposition, it becomes IMF1, IMF2, IMF3,... IMF i .., IMF K (where: IMF i is the i-th subsequence, and IMF K is the K-th subsequence). It is precisely because there may be false subsequences in these subsequences that the prediction result is inaccurate.

[0039] Further, it should be noted that let a certain frequency modulation instruction be Pt = [X1, X2, X3,..., Xi,..., XN] (where: X1, X2, X3,..., Xi,..., XN are the values of the frequency modulation instruction at different time points). Take the integer subsequence of the first 90% length in Pt to predict the subsequence values of the last 10% length integer subsequence and the unknown group. Among them, the prediction result of the integer subsequence of the last 10% length of Pt is defined as the prediction experimental group. At the same time, since the last 10% subsequence in the current Pt is known, the accuracy of the prediction can be evaluated by the difference (all errors (MAPE) in the following embodiments of this example) between the prediction experimental group and the subsequence values corresponding to the last 10% of the current Pt. Further, what this method really wants to predict is the subsequence of the unknown group.

[0040] S2. According to the total area S of the current IMF subsequence ii and the error W of the experimental group i , use the activation function to judge whether the current IMF subsequence is a false sequence.

[0041] Further, determining the total area S of any IMF subsequence ii includes:

[0042] Successively determine the areas enclosed by two adjacent data points and the X-axis under the current IMF subsequence, where the enclosed areas include the area numerical size and the area attribute;

[0043] The area enclosed by two adjacent data points and the X-axis is the area of the figure enclosed by the line connecting two adjacent data points, the perpendicular lines connecting the two adjacent data points to the X-axis respectively, and the line connecting the coordinates of the two adjacent data points on the X-axis.

[0044] Further, the area attribute includes positive and negative values. When the ordinates of two adjacent data points are both greater than or equal to 0, the current area attribute is positive;

[0045] When the ordinates of two adjacent data points are both less than 0, the current area attribute is negative;

[0046] When one of the ordinates of two adjacent data points is greater than 0 and the other is less than 0, the current area attribute is determined by the area enclosed by the current two adjacent points and the X-axis.

[0047] It should be noted that let a subsequence IMF of a certain VMD decomposition i = [X i1 , X i2 , X i3 ,... X ii ,... X iN (where: X ii is the value of the i-th IMF subsequence at the i-th time point, XiN (where \(x_{i,N}\) is the value of the \(i\)-th IMF subsequence at the \(N\)-th time point), and the experimental group error of the current sequence is \(W\). i Calculate the areas \(S_1, S_2, \cdots, S\) enclosed by adjacent data points in sequence. N-1 Furthermore, the ordinate corresponding to each data point in the coordinate system represents the signal feature. There may be positive / negative and numerical differences between signal features. The abscissa corresponding to each data point represents the sampling times (i.e., the subscript).

[0048] Furthermore, it should be noted that Figure 3 is a schematic diagram of the case where "when the ordinates of two adjacent data points are both greater than or equal to 0, the current area attribute is positive"; Figure 4 is a schematic diagram of the case where "when the ordinates of two adjacent data points are both less than 0, the current area attribute is negative".

[0049] Furthermore, the current area attribute is determined by the area enclosed by the current two adjacent points and the \(X\)-axis, including: when the area enclosed by the current two adjacent points and the \(X\)-axis is greater than or equal to 0, the current area attribute is positive;

[0050] when the area enclosed by the current two adjacent points and the \(X\)-axis is less than 0, the current area attribute is negative.

[0051] It should be noted that when one of the ordinates of two adjacent data points is greater than 0 and the other is less than 0, the specific determination process is as follows: First, determine the areas of the two triangles enclosed by the current two adjacent data points and the \(X\)-axis (refer to Figure 5 ). The current area also includes the numerical value and the area attribute (positive / negative value) of the area. Calculate the area difference between the two triangles (i.e., \(C_1 - C_2\). Furthermore, the numerical value of the current area difference is the numerical value of the area enclosed by the current group of adjacent points and the \(X\)-axis). Finally, compare the current area difference with 0. If the area difference is greater than or equal to 0, it is determined that the current area attribute is positive; if the area difference is less than 0, it is determined that the current area attribute is negative.

[0052] Sum up all the enclosed areas of the current IMF subsequence to obtain the total area \(S\) of the current IMF subsequence ii ;

[0053] It should be noted that the sum of the corresponding areas of the subsequence IMF i is \(S\) ii = \(S_1 + S_2+\cdots+S\) N-1 . Execute the above steps until the sum of the areas \(S\) K corresponding to all subsequences \(IMF_1, IMF_2, IMF_3, \cdots, IMF\) 11 , \(S\) 22 , \(S\) are determined.33 ,...,S KK。

[0054] Judging whether the IMF subsequence is a spurious sequence by using an activation function includes:

[0055] exp(sigmoid(W i )*S ii If the value of ) is between [0.8, 1.1], it indicates that the corresponding current sequence is a spurious IMF subsequence, where sigmoid() represents the activation function.

[0056] It should be noted that for a specific subsequence IMF i The area is S ii The corresponding error of the test group is Wi. When the value of exp(sigmoid(W i )*S ii ) is between [0.8, 1.1], it is determined that the current sequence is a spurious sequence and needs to be improved.

[0057] S3. Improve the IMF subsequence determined to be spurious, and use the improved IMF subsequence to predict and output the frequency modulation command. The prediction and output of the frequency modulation command for the improved IMF subsequence can be realized by conventional technical means (GRU), which will not be elaborated here.

[0058] Among them, improving the IMF subsequence determined to be spurious includes:

[0059] When there are two or more spurious IMF subsequences, generate a random sequence to replace the spurious IMF subsequence. Among them, the random sequence is determined by taking the random sequence with the smallest area enclosed with the X-axis as the final IMF subsequence; or,

[0060] When there is a single spurious IMF subsequence, calculate the mean value of the current spurious IMF subsequence, and mirror and change the position of the current spurious IMF subsequence in the coordinate system with the horizontal line where Y is equal to the current mean value as the symmetry axis, and determine the changed IMF subsequence as the final IMF subsequence.

[0061] It should be noted that let IMF i =[X i1 ,X i2 ,X i3 ,.X ii ..,X iN and IMF j =[X j1 ,X j2 ,X j3 ,.X ji ..,X jNThere are 2 false sequences, and at this time, a random sequence IMF needs to be generated ij =[X ij1 ,X ij2 ,X ij3 ,.X iji ..,X ijN is used to replace these two false subsequences, and each value The generated random range is:

[0062] .

[0063] In the formula, represents the lower limit of the generated random value of and are the data point values (signal feature values) at the corresponding positions in sequences IMF i and IMF j respectively, while and are the factorial factors of these two sequences respectively, used to adjust the magnitude of the data point values. Further, and the smaller value in

[0064] represents the lower limit of the random value; and are used to increase randomness or consider the volatility of the data.

[0065] Furthermore, first generate a random IMF ij =[X ij1 ,X ij2 ,X ij3 ,.X iji ..,X ijN (where: X iji is the value of the current subsequence at the i-th time point, and X ijN is the value of the current subsequence at the N-th time point), and calculate the area enclosed by this sequence and the X-axis at this time.

[0066] Then generate another random IMF ij =[X ij1 ,X ij2 ,X ij3 ,.X iji ..,X ijN and calculate the area enclosed by this sequence and the X-axis at this time.

[0067] Repeat this process 20 times randomly.

[0068] Among these 20 outputs, the IMF corresponding to the smallest enclosed area under the corresponding randomness is ij determined as the final sequence. It should be noted that the combination of false sequences is based on the principle of proximity. For example, if IMF1, IMF4, IMF7, IMF9, and IMF10 are false sequences, then IMF1 and IMF4 are combined first, IMF7 and IMF9 are combined, and IMF10 is processed according to the case of a single false sequence.

[0069] Preferably, by randomly generating each value of the IMF sequence within the determined upper and lower limits, it can be ensured that the newly generated sequence neither completely deviates from the characteristics of the original sequence nor can introduce a certain degree of randomness, which helps to eliminate the influence of false sequences. ij Each value of the sequence can ensure that the newly generated sequence neither completely deviates from the characteristics of the original sequence nor can introduce a certain degree of randomness, which helps to eliminate the influence of false sequences.

[0070] It should be further noted that when there is only one false sequence in a certain sequence, or the situation of IMF10 as described above occurs, refer to Appendix Figure 6 and process it in the following way:

[0071] Let IMF i = [X i1 , X i2 , X i3 ,.X ii ,.., X iN be the false sequence that appears alone. Calculate the mean avg[X i1 , X i2 , X i3 ,.X ii ,.., X iN of this sequence. With the horizontal line of the mean Y = avg[X i1 , X i2 , X i3 ,.X ii ,.., X iN as the axis of symmetry, mirror and change the position of the sequence, and determine the changed sequence as the final sequence.

[0072] It should be noted that the mean is the average of all data points in the sequence, reflecting the central tendency of the sequence. Using the calculated mean Y as the axis of symmetry, perform a mirror transformation on the sequence.

[0073] Preferably, through this mirror transformation, the data points in the false sequence are redistributed so that they are symmetrically distributed around the mean Y, which helps to eliminate the outliers or noise in the false sequence, making the sequence more in line with the characteristics of the true signal, thereby improving the accuracy and robustness of the prediction model.

[0074] In summary, the effective effect of a prediction method for energy storage frequency modulation instructions is that it can effectively identify and improve the false sequences that may be generated during the variational mode decomposition process. Further, this method accurately locates and processes false sequences by calculating the area enclosed by adjacent data points and the X-axis in the subsequence, reducing the prediction error and improving the prediction accuracy. In addition, by means of generating random sequences and performing mirror transformations, etc., the false sequences are optimized, further enhancing the system's response speed and frequency modulation efficiency, thus bringing higher economic benefits and better operating performance to thermal power units.

[0075] Example 2 is the second example of the present invention. In order to verify the advantages of the present invention, the frequency modulation sequence is predicted by adopting the method of the present invention and the prediction method of VMD-GRU, and the results are shown in Table 1:

[0076] Table 1 Comparison table of prediction results

[0077] ,

[0078] It should be noted that the calculation formulas and definitions of the above four evaluation indexes are shown in Table 2.

[0079] Table 2 Operation formulas and definitions of each index

[0080] ,

[0081] Among them, N represents the sample size, and respectively represent the actual value and the predicted value at time n.

[0082] In summary, according to the experimental data, the errors of the prediction using this method are all smaller than the results of the VMD-GRU prediction, which fully shows that the effect of the prediction using the present invention is better.

[0083] Example 3 is the third example of the present invention. The difference from the previous two examples is:

[0084] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the current technical solution can be embodied in the form of a software product. The current computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0085] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0086] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion (electronic device) having one or more wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, a computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other appropriate processing as necessary, and then stored in a computer memory.

[0087] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0088] Embodiment 4 is the fourth embodiment of the present invention. This embodiment provides a prediction system for energy storage frequency modulation instructions, including a decomposition module for decomposing the frequency modulation instruction sequence Pt using VMD to obtain multiple IMF subsequences.

[0089] A determination module for judging whether the current IMF subsequence is a false sequence by using an activation function according to the total area S ii of the current IMF subsequence and the error W i of the test group.

[0090] An improved prediction module for improving the IMF subsequences determined to be false and using the improved IMF subsequences for predicting and outputting frequency modulation instructions.

[0091] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for predicting energy storage frequency modulation instructions, characterized in that: Including, Decompose the frequency modulation command sequence Pt using VMD to obtain multiple IMF subsequences; According to the total area S of the current IMF subsequence ii and the error W of the experimental group i , use the activation function to determine whether the current IMF subsequence is a spurious sequence; Improve the IMF subsequences determined to be false, and use the improved IMF subsequences for the prediction output of the frequency modulation command; Among them, improving the IMF subsequences determined to be false includes: When there are two or more false IMF subsequences, generate a random sequence to replace the false IMF subsequences; or, When there is a single false IMF subsequence, calculate the mean value of the current false IMF subsequence, and mirror and change the position of the current false IMF subsequence in the coordinate system with the horizontal line where Y is equal to the current mean value as the axis of symmetry, and determine the changed IMF subsequence as the final IMF subsequence; Determine the total area S of any IMF subsequence ii Including: Successively determine the areas enclosed by two adjacent data points and the X-axis under the current IMF subsequence, where the enclosed areas include the magnitude of the area value and the area attribute; Sum the areas enclosed by all the current IMF subsequences to obtain the total area S of the current IMF subsequence ii ; The area enclosed by two adjacent data points and the X-axis is the area of the figure enclosed by the line connecting the two adjacent data points, the perpendicular lines from the two adjacent data points to the X-axis respectively, and the line connecting the coordinates of the two adjacent data points on the X-axis.

2. The prediction method of an energy storage frequency modulation command according to claim 1, characterized in that: The area attribute includes positive and negative values. When the ordinates of two adjacent data points are both greater than or equal to 0, the current area attribute is positive; When the ordinates of two adjacent data points are both less than 0, the current area attribute is negative; When one of the ordinates of two adjacent data points is greater than 0 and the other is less than 0, the current area attribute is determined by the area enclosed by the current two adjacent points and the X-axis.

3. The prediction method of an energy storage frequency modulation command according to claim 2, characterized in that: The current area attribute being determined by the area enclosed by the current two adjacent points and the X-axis includes: When the area enclosed by the current two adjacent points and the X-axis is greater than or equal to 0, the current area attribute is positive; When the area enclosed by the current two adjacent points and the X-axis is less than 0, the current area attribute is negative.

4. The prediction method of an energy storage frequency modulation command according to claim 3, wherein: Judging whether the IMF subsequence is a false sequence using an activation function includes: exp(sigmoid(W i )*S ii ) is between [0.8, 1.1], it indicates that the corresponding current sequence is a false IMF subsequence, where sigmoid() represents the activation function.

5. The prediction method of an energy storage frequency modulation command according to claim 4, wherein: The determination of the random sequence is to use the random sequence with the smallest area enclosed by the X-axis as the final IMF subsequence.

6. A system adopting a prediction method for an energy storage frequency modulation command as described in any one of claims 1 to 5, characterized in that, Including: A decomposition module for decomposing the frequency modulation command sequence Pt using VMD to obtain multiple IMF subsequences; A determination module, which is used to determine whether the current IMF subsequence is a spurious sequence by using an activation function according to the total area S of the current IMF subsequence ii and the error W of the test group i ​ An improvement and prediction module for improving the IMF subsequences determined to be false and using the improved IMF subsequences for the prediction output of the frequency modulation command.

7. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of a method for predicting a frequency modulation command of an energy storage as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of a method for predicting a frequency modulation command of an energy storage as described in any one of claims 1 to 5.

Images