A method and system for predicting frequency regulation commands for energy storage based on periodic adjustment

Abstract

This invention relates to the field of power grid frequency regulation technology, specifically to a method and system for predicting energy storage frequency regulation commands based on periodic adjustment. The method includes: decomposing the frequency regulation command signal to obtain several component signals; then discretizing each component signal to obtain a subsequence corresponding to each component signal; dividing all subsequences into regular subsequences and irregular subsequences based on the periodic distribution characteristics of all data in each subsequence, and determining the optimal parameters in the regular subsequences; adjusting the data in the regular and irregular subsequences to obtain the adjusted values ​​of all data in each subsequence; predicting based on the adjusted values ​​of all data in each subsequence to obtain each predicted subsequence; and superimposing all predicted subsequences to obtain the predicted signal data of the final frequency regulation command. This invention reduces the influence of nonlinearity and improves the accuracy of prediction.

Description

Technical Field

[0001] This invention relates to the field of power grid frequency regulation technology, and in particular to a method and system for predicting energy storage frequency regulation commands based on periodic adjustment. Background Technology

[0002] Energy storage frequency regulation command prediction refers to the process of predicting changes in frequency regulation commands when using energy storage systems to perform frequency regulation tasks in a power system, thereby optimizing the scheduling and response of the energy storage system. Frequency regulation in power systems is crucial, especially with the increasing proportion of renewable energy (such as wind and solar power) and the resulting increase in frequency instability. In this context, energy storage systems are playing an increasingly important role.

[0003] Frequency regulation commands are signals issued by the power system dispatch center or automation equipment, instructing energy storage systems on how to adjust charging and discharging power to maintain grid frequency stability. Energy storage systems need to respond quickly to emergencies and faults based on these frequency regulation commands to ensure frequency stability.

[0004] Traditional methods for predicting frequency regulation commands in energy storage involve decomposing the original signal corresponding to the command using a Variational Mode Decomposition (VMD) algorithm to obtain multiple component signals. Each component is then predicted separately, and the prediction results are superimposed to obtain the final predicted frequency regulation command. However, some component signals exhibit high nonlinearity, even exceeding that of the original signal. Using highly nonlinear components for prediction increases the difficulty of prediction due to their poor regularity, thus reducing the accuracy of the prediction results. Summary of the Invention

[0005] This invention provides a method and system for predicting frequency regulation commands for energy storage based on periodic adjustment, in order to solve existing problems.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] The first aspect of this invention is to provide a method for predicting frequency regulation commands for energy storage based on periodic adjustment, comprising:

[0008] Acquire frequency modulation command signal;

[0009] The frequency modulation command signal is decomposed to obtain several component signals. Then, each component signal is discretized to obtain the subsequence corresponding to each component signal. Based on the periodic distribution characteristics of all data in each subsequence, all subsequences are divided into regular subsequences and irregular subsequences, and the optimal parameters in the regular subsequences are determined.

[0010] For regular subsequences, the data in the subsequences are adjusted using the corresponding optimal parameters. For irregular subsequences, the data in the subsequences are adjusted using mapping to obtain the adjusted values ​​of all data in each subsequence. Based on the adjusted values ​​of all data in each subsequence, a prediction is made to obtain each predicted subsequence.

[0011] All predicted subsequences are superimposed to obtain the predicted signal data for the final frequency modulation command.

[0012] Further, the process of decomposing the frequency modulation command signal to obtain several component signals, and then discretizing each component signal to obtain a subsequence corresponding to each component signal, includes:

[0013] The frequency modulation command signal is decomposed using the VMD decomposition algorithm to obtain several component signals; at a preset time interval... Each component signal is divided into segments to obtain several data points corresponding to each segment. These data points are then grouped into a sequence according to time order, which is denoted as the subsequence corresponding to each component signal.

[0014] Furthermore, based on the periodic distribution characteristics of all data in each subsequence, the process of dividing all subsequences into regular subsequences and irregular subsequences, and determining the optimal parameters in the regular subsequences, includes:

[0015] The range of possible values ​​for the period length of each subsequence is as follows: Where N represents the number of data points in each subsequence. Indicates to Round down;

[0016] According to the preset error reference coefficient and preset error adjustment factor , determine A range of error coefficients; the first range of error coefficients is... The range of the second error coefficient is , No. The range of error coefficients is ;in, Indicates to Round down;

[0017] The system iterates through the range of period length and the number of error coefficients for each subsequence. The results of the iteration determine whether each subsequence has an optimal parameter. Based on whether an optimal parameter exists, all subsequences are divided into regular subsequences and irregular subsequences.

[0018] Further, the iterative traversal based on the range of period length and the number of error coefficient ranges for each subsequence, determining whether each subsequence has an optimal parameter based on the iteration results, and dividing all subsequences into regular and irregular subsequences based on the existence of an optimal parameter, includes:

[0019] The iterative steps for the optimal parameters are as follows:

[0020] Step 1: Determine the number of cycles based on the range of possible values ​​for the cycle length. Period length, through the first The period length divides each subsequence data into several groups. The ratios between data at the same position in all adjacent groups are calculated, and all ratios are grouped into a set, denoted as the period length. The set of ratios; judging the first ratio in order of its error coefficient range. If all data in the ratio set fall within the error coefficient range, determine the optimal parameters for each subsequence and stop iterating; otherwise, skip to step 2.

[0021] Step 2: Increase the period length by one data length, that is, for Increment by 1, then jump to step 1 to continue the iteration;

[0022] in, For iteration loop parameters;

[0023] At this point, the existence of the optimal parameter is determined through the iterative steps of the optimal parameter; if the optimal parameter exists, the subsequence is recorded as a regular subsequence; if the optimal parameter does not exist, the subsequence is recorded as an irregular subsequence.

[0024] Furthermore, the process involves adjusting the data in regular subsequences using corresponding optimal parameters, and adjusting the data in irregular subsequences using mapping, to obtain the adjusted values ​​of all data in each subsequence, including:

[0025] The adjustment formula for data in a regular subsequence is:

[0026]

[0027] In the formula, This represents the period length corresponding to the optimal parameters of each subsequence. Represents the factorial of the data. This represents the upper limit of the error coefficient range corresponding to the optimal parameters of each subsequence. This represents the lower limit of the error coefficient range corresponding to the optimal parameters of each subsequence. Indicates the first subsequence in each subsequence The value of each data point. Indicates the first subsequence in each subsequence The adjusted values ​​of the data. Indicates the first subsequence in each subsequence The sequence number of each data item. It is a sine function;

[0028] The adjustment formula for data in an irregular subsequence is:

[0029]

[0030] In the formula, It is a natural constant. For normalization function, Indicates the first subsequence in each subsequence The value of each data point. Indicates the first subsequence in each subsequence The adjusted values ​​of the data. Indicates the first subsequence in each subsequence The sequence number of each data item.

[0031] Further, the step of predicting based on the adjusted values ​​of all data in each subsequence to obtain each predicted subsequence includes:

[0032] The adjusted values ​​of all data in each subsequence are sorted in chronological order to form a new sequence, which is denoted as each recombinant subsequence. Based on each recombinant subsequence, a GRU neural network is used to make predictions, and the predicted data are arranged in chronological order to form each predicted subsequence.

[0033] Further, the step of superimposing all predicted subsequences to obtain the predicted signal data for the final frequency modulation command includes:

[0034] By using the least squares method, curve fitting is performed on each predicted subsequence to obtain the predicted fitting curve for each predicted subsequence; by using Fourier transform, the predicted fitting curves of all predicted subsequences are superimposed in the frequency domain to obtain the predicted signal data of the final frequency modulation command.

[0035] A second aspect of the present invention is to provide a frequency regulation command prediction system for energy storage based on periodic adjustment, comprising:

[0036] Data acquisition module: used to acquire frequency modulation command signals;

[0037] Periodic analysis module: used to decompose the frequency modulation command signal to obtain several component signals, then discretize each component signal to obtain the subsequence corresponding to each component signal; based on the periodic distribution characteristics of all data in each subsequence, all subsequences are divided into regular subsequences and irregular subsequences, and the optimal parameters in the regular subsequences are determined.

[0038] The adjustment and prediction module is used to adjust the data in regular subsequences using corresponding optimal parameters, and to adjust the data in irregular subsequences through mapping, thereby obtaining the adjusted values ​​of all data in each subsequence; and to make predictions based on the adjusted values ​​of all data in each subsequence, thereby obtaining each predicted subsequence.

[0039] Superposition module: Used to superimpose all predicted subsequences to obtain the predicted signal data of the final frequency modulation command.

[0040] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the aforementioned method for predicting energy storage frequency modulation commands based on periodic adjustment.

[0041] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for predicting energy storage frequency modulation commands based on periodic adjustment.

[0042] Compared with existing technologies, the beneficial effects of this invention are as follows: The frequency modulation command signal is decomposed to obtain several component signals; each component signal is then discretely divided to obtain a subsequence corresponding to each component signal, which is used for subsequent analysis of the data distribution of each component; based on the periodic distribution characteristics of all data in each subsequence, all subsequences are divided into regular subsequences and irregular subsequences, and the optimal parameters in the regular subsequences are determined, improving the accuracy of the periodic regularity analysis of each subsequence; the data in the regular subsequences are adjusted using the corresponding optimal parameters, and the data in the irregular subsequences are adjusted through mapping, obtaining the adjusted values ​​of all data in each subsequence, reducing the influence of nonlinearity in the components; prediction is performed based on the adjusted values ​​of all data in each subsequence to obtain each predicted subsequence, improving the accuracy of component signal prediction; all predicted subsequences are superimposed to obtain the predicted signal data of the final frequency modulation command, improving the response speed to sudden events and enhancing the stability of the system. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating the steps of a method for predicting frequency regulation commands for energy storage based on periodic adjustment, as provided in this invention.

[0045] Figure 2 This invention provides a schematic diagram of the module flow of an energy storage frequency regulation command prediction system based on periodic adjustment. Detailed Implementation

[0046] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0047] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0048] To address the problems existing in the background technology, a method and system for predicting frequency regulation commands for energy storage based on periodic adjustment has been developed, which has significant practical implications.

[0049] like Figure 1 As shown, the first aspect of the present invention is to provide a method for predicting frequency regulation commands for energy storage based on periodic adjustment, comprising the following steps:

[0050] Step S001: Acquire the frequency modulation command signal.

[0051] It should be noted that when there is a deviation in the balance between power supply and demand (such as load changes, fluctuations in power generation, etc.), the power system's dispatch center or automation system generates frequency regulation commands based on the system's frequency fluctuations, load demand, and power generation. These commands maintain frequency stability, enabling responses to emergencies and reducing the likelihood of failures. To respond even more quickly to emergencies, frequency regulation commands can be predicted. Therefore, collecting frequency regulation commands for prediction allows for a faster response to emergencies or failures.

[0052] Specifically, obtain the preset duration before the current moment. The frequency regulation command signal from the power system dispatch center every hour, wherein, in this embodiment, the preset duration is... In this embodiment, a preset duration is used. No specific restrictions are imposed; implementers can decide based on the specific circumstances.

[0053] At this point, the frequency modulation command signal is obtained.

[0054] Step S002: Decompose the frequency modulation command signal to obtain several component signals, then discretize each component signal to obtain the subsequence corresponding to each component signal; according to the periodic distribution characteristics of all data in each subsequence, divide all subsequences into regular subsequences and irregular subsequences, and determine the optimal parameters in the regular subsequences.

[0055] It should be noted that some component signals have a high degree of nonlinearity, resulting in poor prediction results. Therefore, to address this issue, since periodic component signals have strong regularity, the stronger the regularity, the more accurate the prediction results. Thus, the component signals are distinguished by their periodicity, and then different adjustments are made to the component signals according to their strength, thereby improving the quality of the component signal data and reducing the impact of the nonlinearity of the component signals on the prediction results.

[0056] It should be further explained that, in order to adjust the component signals by the strength of the periodicity, each component signal is discretized to obtain discrete data points. The periodicity is determined by the distribution of the data points, and the discrete data points are then adjusted based on the periodicity.

[0057] Specifically, the frequency modulation command signal is decomposed using the VMD decomposition algorithm to obtain several component signals; at a preset time interval... The signal component is divided into segments, yielding several data points corresponding to each segment. These data points are then grouped into a sequence according to time order, denoted as the subsequence corresponding to each signal component. The VMD decomposition algorithm is a well-known technique and will not be described in detail here. In this embodiment, a preset time interval is used. seconds, where the preset time interval is... No specific restrictions are imposed; implementers can decide based on the specific circumstances.

[0058] At this point, the subsequence corresponding to each component signal is obtained.

[0059] It should be noted that due to noise and other factors during actual data acquisition, there are certain differences between the data corresponding to the same time in different periods. Therefore, by using an error coefficient and a given period range, the final period and error coefficient of each subsequence are obtained through traversal. Based on the traversal results, it is determined whether the subsequence has a regularity. Based on the final period and error coefficient of each subsequence, adjustments are made to different subsequences based on whether the subsequence has a regularity.

[0060] It should be further noted that when performing periodic analysis, since the periodic characteristics of each subsequence need to be analyzed based on the differences between the corresponding data of different groups, when dividing the subsequence according to a certain period length, at least two groups must be included, and at least two groups are required for difference analysis.

[0061] Specifically, let N be the number of data points in each subsequence. When performing periodic analysis, the maximum number of data points contained in one period of each subsequence is then... ;in, Indicates to Rounding down, the range of values ​​for the period length is: .

[0062] Preset error reference coefficient By setting a reference coefficient for error Determine the adjustment range of the error coefficient. In this embodiment, a preset error reference coefficient is used. In this embodiment, a preset error reference coefficient is used. No specific restrictions are imposed; implementers can decide based on the specific circumstances.

[0063] Preset error adjustment factor In this embodiment, a preset error adjustment factor is used. In this embodiment, a preset error adjustment factor is used. No specific restrictions are imposed; implementers can decide based on the specific circumstances.

[0064] According to the preset error reference coefficient and preset error adjustment factor , determine A range of error coefficients; the first range of error coefficients is... The range of the second error coefficient is , No. The range of error coefficients is .in, Indicates to Round down to the nearest integer.

[0065] Based on the period length and error coefficient range, determine whether each subsequence has an optimal parameter; the iterative steps for determining the optimal parameter are as follows:

[0066] Step 1: Based on the range of values ​​for the period length, determine a period length, denoted as the th period. Period length, through the first The period length divides each subsequence data into several groups. The ratios between data at the same position in all adjacent groups are calculated, and all ratios are grouped into a set, denoted as the period length. The set of ratios; judging the first ratio in order of its error coefficient range. If all data in the ratio set fall within the error coefficient range, determine the optimal parameters for each subsequence and stop iterating; otherwise, skip to step 2.

[0067] Step 2: Increase the period length by one data length, that is, for Increment by 1, then jump to step 1 to continue the iteration;

[0068] At this point, the existence of the optimal parameter is determined through the iterative steps of the optimal parameter; if the optimal parameter exists, the subsequence is recorded as a regular subsequence; if the optimal parameter does not exist, the subsequence is recorded as an irregular subsequence.

[0069] The specific process for determining whether each subsequence exhibits aperiodic regularity based on the period length and error coefficient range is as follows:

[0070] Based on the range of period length values, the minimum period length is determined and designated as the first period length. Each subsequence data is then divided into several groups based on this first period length (the number of groups is...). Calculate the ratio between data at the same position in all adjacent groups within several groups, and group all ratios into a set, denoted as the first ratio set. Determine whether all data in the first ratio set belongs to the first error coefficient range. If it does, use the first period length and the first error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, determine whether all data in the first ratio set belongs to the second error coefficient range. If it does, use the first period length and the second error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, determine whether all data in the first ratio set belongs to the third error coefficient range. If it does, use the first period length and the third error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, sequentially determine whether all data in the first ratio set belongs to subsequent error coefficient ranges until all subsequent error coefficient ranges have been iterated. If no optimal parameter is found after all subsequent error coefficient ranges have been iterated, it indicates that the first period length is not the optimal parameter for each subsequence, and the next period length is iterated.

[0071] Let the second longest period length be denoted as the second period length. Divide each subsequence data into several groups based on the second period length (the number of groups is...). Calculate the ratio between data at the same position in all adjacent groups within several groups, and group all ratios into a set, denoted as the second ratio set. Determine whether all data in the second ratio set belongs to the first error coefficient range. If it does, use the second period length and the first error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, determine whether all data in the second ratio set belongs to the second error coefficient range. If it does, use the second period length and the second error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, determine whether all data in the second ratio set belongs to the third error coefficient range. If it does, use the second period length and the third error coefficient range as the optimal parameters for each subsequence, and stop iterating. If it does not, sequentially determine whether all data in the second ratio set belongs to subsequent error coefficient ranges until all subsequent error coefficient ranges have been iterated. If no optimal parameter is found after all subsequent error coefficient ranges have been iterated, it indicates that the second period length is not the optimal parameter for each subsequence, and the next period length is iterated.

[0072] Similarly, the analysis determines whether an optimal parameter exists between the third period length and the maximum period length. If an optimal parameter exists, it indicates a periodic regularity in the subsequence, meaning there is a suitable period. If an optimal parameter does not exist, it indicates no periodic regularity in the subsequence, meaning there is no suitable period, and therefore no regularity. Indicates to Round down. Indicates to Round down to the nearest integer.

[0073] In this process, the cycle length increases by only one data length each time.

[0074] At this point, the optimal parameters for each subsequence are obtained.

[0075] Subsequences with optimal parameters are denoted as regular subsequences, and subsequences without optimal parameters are denoted as irregular subsequences.

[0076] This concludes the determination of whether each subsequence has a regularity.

[0077] Step S003: Adjust the data in the regular subsequence using the corresponding optimal parameters, and adjust the data in the irregular subsequence using mapping to obtain the adjusted values ​​of all data in each subsequence; make predictions based on the adjusted values ​​of all data in each subsequence to obtain each predicted subsequence.

[0078] It should be noted that when each subsequence has a periodic pattern, adjustments can be made based on the periodic characteristic parameters of each subsequence; however, when each subsequence does not have a periodic pattern, adjustments are made directly based on the order of the data.

[0079] Specifically, the adjustment formula for the data in the regular subsequence is:

[0080]

[0081] In the formula, This represents the period length corresponding to the optimal parameters of each subsequence. Represents the factorial of the data. This represents the upper limit of the error coefficient range corresponding to the optimal parameters of each subsequence. This represents the lower limit of the error coefficient range corresponding to the optimal parameters of each subsequence. Indicates the first subsequence in each subsequence The value of each data point. Indicates the first subsequence in each subsequence The adjusted values ​​of the data. Indicates the first subsequence in each subsequence The sequence number of each data item. It is a sine function.

[0082] The adjustment formula for data in an irregular subsequence is:

[0083]

[0084] In the formula, It is a natural constant. For normalization function, Indicates the first subsequence in each subsequence The value of each data point. Indicates the first subsequence in each subsequence The adjusted values ​​of the data. Indicates the first subsequence in each subsequence The sequence number of each data item. It is a sine function.

[0085] The adjusted values ​​of all data in each subsequence are sorted in chronological order to form a new sequence, which is denoted as each recombinant subsequence. Based on each recombinant subsequence, a GRU neural network is used to make predictions, and the predicted data are arranged in chronological order to form each predicted subsequence.

[0086] When using the GRU neural network for prediction, the prediction error is measured by the metrics MAE (Mean Absolute Error), SSE (Sum of Squared Errors), RMSE (Root Mean Square Error), and MAPE (Mean Absolute Percentage Error). These metrics are used to evaluate the prediction results before and after adjustment of the component signals. The loss function in the GRU neural network is the cross-entropy loss function.

[0087] Among them, GRU (Gated Recurrent Unit) neural network, MAE, SSE, RMSE, MAPE and cross-entropy loss function are all well-known techniques, and will not be described in detail here.

[0088] Step S004: Superimpose all predicted subsequences to obtain the predicted signal data of the final frequency modulation command.

[0089] The least squares method is used to perform curve fitting on each predicted subsequence to obtain the predicted fitting curve for each predicted subsequence. The predicted fitting curves of all predicted subsequences are then superimposed in the frequency domain using Fourier transform to obtain the predicted signal data for the final frequency modulation command.

[0090] The prediction error analysis was performed on the predicted sequence of the frequency modulation command signal and the predicted signal data of the final frequency modulation command using the four indicators mentioned above. The error indices corresponding to the predicted sequence of the frequency modulation command signal were MAE 8.22, SSE 98.12, RMSE 0.97, and MAPE 16.02%. The error indices corresponding to each predicted subsequence after adjustment were MAE 0.56, SSE 10.01, RMSE 0.25, and MAPE 4.13%. The results show that the prediction error is significantly reduced, that is, the prediction accuracy is significantly improved.

[0091] The least squares method and Fourier transform are well-known techniques, and will not be elaborated on here.

[0092] like Figure 2 As shown, a second aspect of the present invention is to provide a frequency regulation command prediction system for energy storage based on periodic adjustment, comprising the following modules:

[0093] Data acquisition module 101: used to acquire frequency modulation command signals;

[0094] Periodic analysis module 102: is used to decompose the frequency modulation command signal to obtain several component signals, then to discretize each component signal to obtain the subsequence corresponding to each component signal; according to the periodic distribution characteristics of all data in each subsequence, all subsequences are divided into regular subsequences and irregular subsequences, and the optimal parameters in the regular subsequences are determined.

[0095] The adjustment and prediction module 103 is used to adjust the data in the regular subsequences using the corresponding optimal parameters, and to adjust the data in the irregular subsequences using mapping, so as to obtain the adjusted values ​​of all data in each subsequence; and to make predictions based on the adjusted values ​​of all data in each subsequence to obtain each predicted subsequence.

[0096] Superposition module 104: used to superimpose all predicted subsequences to obtain the predicted signal data of the final frequency modulation command.

[0097] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for predicting energy storage frequency modulation commands based on periodic adjustment.

[0098] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for predicting energy storage frequency modulation commands based on periodic adjustment.

[0099] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0100] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, systems, and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0101] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0102] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for predicting frequency regulation instructions based on period adjustment of energy storage, characterized in that, The method comprises the following steps: obtaining a frequency modulation instruction signal; decomposing the frequency modulation instruction signal to obtain a plurality of component signals, then discretely dividing each component signal to obtain a subsequence corresponding to each component signal; according to the period distribution characteristics of all data in each subsequence, dividing all subsequences into regular subsequences and irregular subsequences, and determining the optimal parameters in the regular subsequences; adjusting the data in the regular subsequences through the corresponding optimal parameters, and adjusting the data in the irregular subsequences through mapping to obtain the adjusted values of all data in each subsequence; predicting according to the adjusted values of all data in each subsequence to obtain each predicted subsequence; superimposing all predicted subsequences to obtain the predicted signal data of the final frequency modulation instruction.

2. The method of claim 1, wherein, The step of decomposing the frequency modulation instruction signal to obtain a plurality of component signals, then discretely dividing each component signal to obtain a subsequence corresponding to each component signal comprises the following steps: The frequency modulation command signal is decomposed by a VMD decomposition algorithm to obtain a plurality of component signals Each component signal is divided at preset time intervals to obtain a plurality of data corresponding to each component signal after division, and a plurality of data are sequentially grouped into a sequence, denoted as a sub-sequence corresponding to each component signal.

3. The method of claim 1, wherein, The step of dividing all subsequences into regular subsequences and irregular subsequences according to the period distribution characteristics of all data in each subsequence, and determining the optimal parameters in the regular subsequences comprises the following steps: The value range of the period length of each subsequence is wherein N represents the number of data in each subsequence, represents rounding down of to an integer. According to the preset error reference coefficient and preset error adjustment factor , determine A range of error coefficients; the first range of error coefficients is... The range of the second error coefficient is , No. The range of error coefficients is ;in, Indicates to Round down; iterating according to the value range of the period length of each subsequence and the number of error coefficient ranges, judging whether there is an optimal parameter in each subsequence through the iteration result, and dividing all subsequences into regular subsequences and irregular subsequences according to whether there is an optimal parameter.

4. The method of claim 3, wherein, The step of iterating according to the value range of the period length of each subsequence and the number of error coefficient ranges, judging whether there is an optimal parameter in each subsequence through the iteration result, and dividing all subsequences into regular subsequences and irregular subsequences according to whether there is an optimal parameter comprises the following steps: The iteration step of the optimal parameter is as follows: Step 1: according to the value range of the cycle length, determine the first cycle length, divide each sub-sequence data into several groups by the first cycle length, calculate the ratio between the data at the same position between any two adjacent groups in the several groups, and form a set of all the ratios, denoted as the first ratio set; according to the order of the error coefficient range, sequentially determine whether all the data in the first ratio set belongs to the error coefficient range, and if yes, determine the optimal parameters of each sub-sequence and stop iteration; if not, jump to step 2. Step 2: Increment the cycle length by one data length, i.e., to Add 1 and then jump to Step 1 to continue the iteration; wherein is an iteration loop parameter; At this point, whether the optimal parameter exists is judged through the iteration step of the optimal parameter; if the optimal parameter exists, the subsequence is recorded as a regular subsequence; if the optimal parameter does not exist, the subsequence is recorded as an irregular subsequence.

5. The method of claim 4, wherein, The step of adjusting the data in the regular subsequences through the corresponding optimal parameters, and adjusting the data in the irregular subsequences through mapping to obtain the adjusted values of all data in each subsequence comprises the following steps: The adjustment formula of the data in the regular subsequence is as follows: In the formula, denotes the length of the period corresponding to the optimal parameter of each sub-sequence, denotes the factorial of the data, denotes the upper limit value in the error coefficient range corresponding to the optimal parameter of each sub-sequence, denotes the lower limit value in the error coefficient range corresponding to the optimal parameter of each sub-sequence, denotes the value of the th data in each sub-sequence, denotes the adjusted value of the th data in each sub-sequence, denotes the serial number of the th data in each sub-sequence, is a sine function; The adjustment formula of the data in the irregular subsequence is as follows: wherein, is a natural constant, is a normalization function, denotes the value of the th data in each sub-sequence, denotes the adjusted value of the th data in each sub-sequence, denotes the sequence number of the th data in each sub-sequence.

6. The method of claim 1, wherein, The step of predicting according to the adjusted values of all data in each subsequence to obtain each predicted subsequence comprises the following steps: The adjusted values of all data in each subsequence are sorted in time sequence to form a new sequence, recorded as each reorganized subsequence; each reorganized subsequence is predicted using a GRU neural network, and the predicted data is sorted in time sequence to form each predicted subsequence.

7. The method of claim 1, wherein the method is a periodic adjustment based energy modulation command prediction method. The step of superimposing all predicted subsequences to obtain the predicted signal data of the final frequency modulation instruction comprises the following steps: The prediction fitting curve of each prediction sub-sequence is obtained by curve fitting of each prediction sub-sequence through the least square method; and the prediction signal data of the final frequency modulation instruction is obtained by superimposing the prediction fitting curves of all prediction sub-sequences in the frequency domain through Fourier transform.

8. A cycle adjustment based energy storage frequency regulation instruction prediction system, characterized in that, The method comprises: a data acquisition module for acquiring the frequency modulation instruction signal; a period analysis module for decomposing the frequency modulation instruction signal to obtain a plurality of component signals, then discretely dividing each component signal to obtain a sub-sequence corresponding to each component signal, dividing all sub-sequences into regular sub-sequences and irregular sub-sequences according to the period distribution characteristics of all data in each sub-sequence, and determining the optimal parameters in the regular sub-sequences; an adjustment prediction module for adjusting the data in the regular sub-sequences through the corresponding optimal parameters, adjusting the data in the irregular sub-sequences through mapping, and obtaining the adjusted values of all data in each sub-sequence; predicting according to the adjusted values of all data in each sub-sequence to obtain each prediction sub-sequence; a superimposition module for superimposing all prediction sub-sequences to obtain the prediction signal data of the final frequency modulation instruction.

9. An electronic device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the energy storage frequency modulation instruction prediction method based on period adjustment according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the energy storage frequency modulation instruction prediction method based on period adjustment according to any one of claims 1-7.

Images