Partial frequency domain super-capacity energy storage frequency modulation instruction prediction method and device

Through partially frequency domain-based supercapacitor energy storage frequency modulation instruction prediction method, the frequency disorder and sequence similarity are calculated using variational modulation algorithms, which solves the problem of low accuracy of frequency modulation instruction prediction in the existing technology, and improves the The frequency modulation quality and grid frequency stability are achieved.

CN118971018APending Publication Date: 2024-11-15HUANENG TAICANG POWER GENERATION CO LTD +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411004005.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing frequency modulation command has low accuracy in prediction, resulting in poor frequency modulation quality and affecting the stability of the power grid frequency.

Method used

The partially frequency domain-based supercapacitor energy storage frequency modulation instruction prediction method is used to obtain subsequences through variational modal decomposition, and frequency and time domain decomposition are performed. The frequency chaos and sequence similarity are calculated by combining the preset algorithm, and the prediction model is used to perform the Comprehensive prediction to obtain accurate frequency modulation instructions.

Benefits of technology

The prediction accuracy of frequency modulation instructions is improved, the frequency modulation quality is improved, and the stability of the power grid frequency is ensured.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118971018A_ABST
    Figure CN118971018A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a partial-frequency-domain super-capacity energy storage frequency modulation instruction prediction method and device, and the method comprises the steps: obtaining a historical frequency modulation instruction and a decomposition layer number, and carrying out the variational mode decomposition of the historical frequency modulation instruction and the decomposition layer number, and obtaining a sub-sequence of the decomposition layer number; for each cycle, performing partial frequency domain processing on the sub-sequence based on the cycle number i, performing a first preset algorithm on the frequency domain sequence to obtain a frequency confusion degree, and performing a second preset algorithm on the time domain sequence to obtain a sequence similarity; the sum of the frequency confusion degree and the sequence similarity of the current cycle is determined as the ith total value; determining the frequency domain sequence and the time domain sequence of the ith cycle corresponding to the minimum value in all the total values as a first sequence; and based on the first sequence, utilizing a third preset algorithm to obtain a comprehensive prediction result as the current frequency modulation instruction. Super-capacity energy storage is used for assisting frequency modulation, a new algorithm is used for predicting a frequency modulation instruction, the prediction accuracy is improved, and then the frequency modulation quality is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the technical field of power grid frequency regulation, and in particular to a method and device for predicting frequency regulation instructions of ultra-capacity energy storage in a partial frequency domain. Background Art

[0002] Vigorously developing new energy power has become an important development content in today's era. Among them, the emergence of supercapacitors, a new type of auxiliary frequency regulation energy storage device, provides a new option for new energy power frequency regulation. However, the uncertainty and volatility of new energy power lead to instability on the power grid side. Maintaining frequency stability is an important aspect to ensure the stable operation of the power grid. New energy power frequency regulation control is a comprehensive control process involving many technical indicators. Only by effectively coordinating with all parties can the ideal effect be achieved. The frequency regulation control function is closely related to the coordination of the unit AGC. Once effective coordination fails, the unit frequency regulation function cannot operate, which seriously affects the stability of the power grid frequency. Therefore, while considering new options for energy storage devices, further optimizing the frequency regulation control strategy is an important aspect to improve the quality of frequency regulation. Summary of the invention

[0003] The disclosed embodiments provide a method and device for predicting frequency modulation instructions of a partially frequency-domain ultra-capacitor energy storage system, which are used to solve the problems of low accuracy in current frequency modulation instruction prediction and poor frequency modulation quality.

[0004] Based on the above problems, in the first aspect, a method for predicting frequency modulation instructions of ultra-capacity energy storage in a partial frequency domain is provided, comprising:

[0005] Obtain historical frequency modulation instructions and decomposition levels, and perform variational mode decomposition on them to obtain subsequences corresponding to the number of decomposition levels;

[0006] Perform a decomposition layer number cycle, for each cycle, perform partial frequency domain conversion on the decomposition layer number subsequences based on the cycle number i corresponding to this cycle, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; sum the frequency disorder degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding cycle number;

[0007] Determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the minimum value as the first sequence;

[0008] A comprehensive prediction result is obtained based on the first sequence using a third preset algorithm, and the comprehensive prediction result is determined to be the current frequency modulation instruction.

[0009] In combination with the first aspect, in a possible implementation manner, the partially frequency-domainizing the subsequences of the decomposition layer based on the value of i in this cycle includes:

[0010] Frequency domain conversion is performed on a part of subsequences from the 1st subsequence to the ith subsequence in the decomposition layer number of subsequences.

[0011] In combination with the first aspect, in a possible implementation manner, the following formula is used to perform partial frequency domain conversion on the subsequences of the decomposition layer:

[0012]

[0013] Among them, X(k) represents the frequency component in the frequency domain, x[n] represents the signal sequence in the time domain, N represents the signal length, k represents the frequency index in the frequency domain, and j is an imaginary unit.

[0014] In combination with the first aspect, in a possible implementation manner, the frequency domain sequence is subjected to a first preset algorithm using the following formula to obtain a frequency chaos degree:

[0015]

[0016] Among them, corrl represents the similarity transformation function, ω (i+1)-(1) Characterizes the aliasing width between adjacent subsequences, ω (i+1)max -ω (i)min Characterizes the frequency band occupied by two adjacent subsequences.

[0017] In combination with the first aspect, in a possible implementation manner, performing a second preset algorithm on the time domain sequence to obtain sequence similarity includes:

[0018] For the time domain sequence, two subsequences are selected multiple times to perform the following operation to obtain the subsequence similarity;

[0019]

[0020] Where A represents one of the two optional subsequences, B represents the other of the two optional subsequences, A i represents the i-th value in sequence A, B i Represents the i-th value in sequence B;

[0021] The sequence similarity is obtained by summing up all the subsequence similarities obtained by the operation.

[0022] In combination with the first aspect, in a possible implementation manner, obtaining a comprehensive prediction result based on the first sequence by using a preset algorithm includes:

[0023] For the frequency domain sequence and the time domain sequence in the first sequence, use the pre-trained first prediction model and the second prediction model to predict the frequency domain sequence and the time domain sequence respectively, to obtain a frequency domain prediction result corresponding to the frequency domain sequence and a first prediction result corresponding to the time domain sequence;

[0024] Using inverse Fourier transform to convert the frequency domain prediction result corresponding to the frequency domain sequence in the first sequence into the time domain, to obtain the time domain prediction result corresponding to the frequency domain sequence, and determine it as the second prediction result;

[0025] The first prediction result and the second prediction result are summed to obtain a comprehensive prediction result.

[0026] In combination with the first aspect, in a possible implementation manner, the first prediction model takes the frequency domain sequence in the first sequence corresponding to any time period in the historical data as input, the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed first prediction model;

[0027] The second prediction model takes the time domain sequence in the first sequence corresponding to any time period in the historical data as input, and the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed second prediction model.

[0028] In a second aspect, a partially frequency-domain ultra-capacitor energy storage frequency modulation instruction prediction device is provided, comprising:

[0029] The signal decomposition module is used to obtain the historical frequency modulation instructions and the number of decomposition layers, and perform VMD decomposition on them to obtain subsequences corresponding to the number of decomposition layers;

[0030] The partial frequency domainization module is used to perform a decomposition layer number cycle, and for each cycle, based on the cycle number i corresponding to this cycle, perform partial frequency domainization on the decomposition layer number subsequences, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; sum the frequency disorder degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding cycle number;

[0031] A target sequence determination module is used to determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the ith cycle corresponding to the minimum value as the first sequence;

[0032] A frequency modulation instruction prediction module is used to obtain a comprehensive prediction result based on the first sequence using a third preset algorithm, and determine that the comprehensive prediction result is the current frequency modulation instruction.

[0033] In combination with the second aspect, in a possible implementation manner, in the partial frequency domain conversion module, performing partial frequency domain conversion on the subsequences of the decomposition layer based on the i value of this cycle includes:

[0034] Frequency domain conversion is performed on a part of subsequences from the 1st subsequence to the ith subsequence in the decomposition layer number of subsequences.

[0035] In combination with the second aspect, in a possible implementation manner, in the partial frequency domain conversion module, the following formula is used to perform partial frequency domain conversion on the subsequences of the decomposition layer:

[0036]

[0037] Among them, X(k) represents the frequency component in the frequency domain, x[n] represents the signal sequence in the time domain, N represents the signal length, k represents the frequency index in the frequency domain, and j is an imaginary unit.

[0038] In combination with the second aspect, in a possible implementation, in the partial frequency domain conversion module, the frequency domain sequence is subjected to a first preset algorithm using the following formula to obtain a frequency chaos degree:

[0039]

[0040] Among them, corrl represents the similarity transformation function, ω (i+1)-(1) Characterizes the aliasing width between adjacent subsequences, ω (i+1)max -ω (i)min Characterizes the frequency band occupied by two adjacent subsequences.

[0041] In combination with the second aspect, in a possible implementation, in the partial frequency domain conversion module, performing a second preset algorithm on the time domain sequence to obtain sequence similarity includes:

[0042] For the time domain sequence, two subsequences are selected multiple times to perform the following operation to obtain the subsequence similarity;

[0043]

[0044] Where A represents one of the two optional subsequences, B represents the other of the two optional subsequences, A i represents the i-th value in sequence A, B i Represents the i-th value in sequence B;

[0045] The sequence similarity is obtained by summing up all the subsequence similarities obtained by the operation.

[0046] In combination with the second aspect, in a possible implementation manner, in the frequency modulation instruction prediction module, obtaining a comprehensive prediction result based on the first sequence using a preset algorithm includes:

[0047] For the frequency domain sequence and the time domain sequence in the first sequence, use the pre-trained first prediction model and the second prediction model to predict the frequency domain sequence and the time domain sequence respectively, to obtain a frequency domain prediction result corresponding to the frequency domain sequence and a first prediction result corresponding to the time domain sequence;

[0048] Using inverse Fourier transform to convert the frequency domain prediction result corresponding to the frequency domain sequence in the first sequence into the time domain, to obtain the time domain prediction result corresponding to the frequency domain sequence, and determine it as the second prediction result;

[0049] The first prediction result and the second prediction result are summed to obtain a comprehensive prediction result.

[0050] In combination with the second aspect, in a possible implementation manner, the first prediction model takes the frequency domain sequence in the first sequence corresponding to any time period in the historical data as input, the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed first prediction model;

[0051] The second prediction model takes the time domain sequence in the first sequence corresponding to any time period in the historical data as input, and the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed second prediction model.

[0052] In a third aspect, a computer device is provided, comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, and when the computer device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the partial frequency-domain super-capacity energy storage frequency modulation instruction prediction method as described in the first aspect, or in combination with any possible implementation manner of the first aspect are performed.

[0053] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the partial frequency-domain ultra-capacity energy storage frequency modulation instruction prediction method as described in the first aspect or in combination with any possible implementation of the first aspect are executed.

[0054] The beneficial effects of the embodiments of the present disclosure include:

[0055] The disclosed embodiment provides a method and device for predicting frequency modulation instructions of supercapacitor energy storage in partial frequency domain, which is applied to the frequency modulation of the power grid during the actual operation of the power plant, using supercapacitors as new auxiliary frequency modulation energy storage devices, and using a new partial frequency domain algorithm to complete the prediction of frequency modulation instructions; including: obtaining historical frequency modulation instructions and decomposition layers, and performing variational modal decomposition on them to obtain subsequences corresponding to the number of decomposition layers; performing decomposition layer number cycles, for each cycle, performing partial frequency domain on the number of decomposition layer number subsequences based on the number of cycles i corresponding to this cycle, and dividing all subsequences into frequency domain sequences and time domain sequences; using the frequency domain sequence to obtain the frequency chaos degree, and using the time domain sequence to obtain the sequence similarity; summing the frequency chaos degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding number of cycles; taking the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the minimum value of all the total values ​​obtained in the cycle as the first sequence; based on the first sequence, using the third preset algorithm to obtain a comprehensive prediction result, and determining the comprehensive prediction result as the frequency modulation instruction of this time. The method disclosed in the present invention improves the prediction accuracy, thereby improving the frequency modulation quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flow chart of a method for predicting frequency modulation instructions of ultra-capacitor energy storage in a partial frequency domain provided in an embodiment of the present disclosure;

[0057] Figure 2 One of the specific flow charts of the partial frequency domain super-capacity energy storage frequency modulation instruction prediction method provided in the embodiment of the present disclosure;

[0058] Figure 3 The second specific flow chart of the partial frequency domain super-capacity energy storage frequency modulation instruction prediction method provided by the embodiment of the present disclosure;

[0059] Figure 4 A structural diagram of a partially frequency-domain ultra-capacitor energy storage frequency modulation instruction prediction device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0060] The embodiment of the present disclosure provides a method for predicting frequency modulation instructions of ultra-capacitor energy storage in a partial frequency domain. The preferred embodiments of the present disclosure are described below in conjunction with the drawings of the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure, and are not used to limit the present disclosure. In addition, the embodiments and features in the embodiments of the present application can be combined with each other if there is no conflict.

[0061] The disclosed embodiment provides a partial frequency domain super-capacitor energy storage frequency modulation instruction prediction method, such as Figure 1 As shown, including:

[0062] S101, obtaining historical frequency modulation instructions and decomposition levels, and performing variational mode decomposition on them to obtain subsequences corresponding to the number of decomposition levels;

[0063] S102, performing a decomposition layer number cycle, for each cycle, performing partial frequency domain conversion on the decomposition layer number subsequences based on the cycle number i corresponding to this cycle, and dividing all subsequences into frequency domain sequences and time domain sequences; performing a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and performing a second preset algorithm on the time domain sequence to obtain a sequence similarity; summing the frequency disorder degree and sequence similarity obtained in this cycle to obtain an i-th total value corresponding to the cycle number;

[0064] S103, determining the minimum value among all the total values ​​obtained in the above cycle, and determining the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the minimum value as the first sequence;

[0065] S104: Obtain a comprehensive prediction result based on the first sequence using a third preset algorithm, and determine that the comprehensive prediction result is the current frequency modulation instruction.

[0066] In the disclosed embodiment, the partial frequency domain super-capacity energy storage frequency modulation instruction prediction method uses variational mode decomposition to decompose the historical frequency modulation instructions according to the above steps to obtain subsequences of the corresponding decomposition layers, and then frequency domainizes some of the subsequences to obtain frequency domain sequences and time domain sequences and respectively obtain the corresponding frequency chaos and sequence similarity; sum the frequency chaos and sequence similarity of this cycle to obtain the i-th total value of the corresponding number of cycles; screen to obtain the minimum total value and determine the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the total value as the first sequence; use the first sequence to obtain a comprehensive prediction result and determine it as the current frequency modulation instruction. The disclosed method uses historical frequency modulation instructions to predict this frequency modulation instruction, and the prediction accuracy is higher.

[0067] In the disclosed embodiment, the variational mode decomposition (VMD) used is an adaptive, completely non-recursive method of modal variation and signal processing. This technology has the advantage of being able to determine the number of modal decompositions, and its adaptability is reflected in determining the number of modal decompositions of a given sequence according to actual conditions. In the subsequent search and solution process, the optimal center frequency and limited bandwidth of each mode can be adaptively matched, and the effective separation of the intrinsic modal component (IMF) and the frequency domain division of the signal can be achieved, thereby obtaining the effective decomposition components of the given signal, and finally obtaining the optimal solution to the variational problem. The VMD method overcomes the problems of endpoint effects and aliasing of modal components in the EMD method, and has a more solid mathematical theoretical foundation, which can reduce the non-stationarity of time series with high complexity and strong nonlinearity, decompose to obtain relatively stable subsequences containing multiple different frequency scales, and is suitable for non-stationary sequences. The core idea of ​​VMD is to construct and solve variational problems.

[0068] In another embodiment provided by the present disclosure, in the above step 102, partially frequency-domainizing the subsequences of the decomposition layer based on the value of i in this cycle includes:

[0069] Frequency domain conversion is performed on a part of subsequences from the 1st subsequence to the ith subsequence in the decomposition layer number of subsequences.

[0070] In the disclosed embodiment, taking the first cycle as an example, this cycle performs frequency domain conversion on the first subsequence, converts the subsequence into a frequency domain sequence, and the other subsequences remain unchanged; taking the second cycle as an example, this cycle performs frequency domain conversion on the first subsequence and the second subsequence, converts the two subsequences into frequency domain sequences, and the other subsequences remain unchanged; and so on, the i-th cycle performs frequency domain conversion on part of the subsequences from the first subsequence to the i-th subsequence.

[0071] In another embodiment provided by the present disclosure, in the above step 102, the following formula is used to perform partial frequency domain conversion on the subsequences of the decomposition layer:

[0072]

[0073] Among them, X(k) represents the frequency component in the frequency domain, x[n] represents the signal sequence in the time domain, N represents the signal length, k represents the frequency index in the frequency domain, and j is an imaginary unit.

[0074] In the embodiment of the present disclosure, the formula used for the above partial frequency domain conversion is the application of Fourier transform in the present method. The formula provided by the present disclosure is used to process the subsequence to be frequency-domain converted to obtain the required frequency domain sequence.

[0075] In another embodiment provided by the present disclosure, in the above step 102, the frequency domain sequence is subjected to a first preset algorithm using the following formula to obtain the frequency chaos degree:

[0076]

[0077] Among them, corrl represents the similarity transformation function, ω (i+1)-(1) Characterizes the aliasing width between adjacent subsequences, ω (i+1)max -ω (i)min Characterizes the frequency band occupied by two adjacent subsequences.

[0078] In the disclosed embodiment, the frequency chaos is an important indicator that reflects the decomposition effect of each subsequence and affects the prediction result. The smaller the frequency chaos, the better the decomposition effect. Therefore, using the cyclic subsequence corresponding to the minimum frequency chaos can make the prediction result more accurate.

[0079] In another embodiment provided by the present disclosure, in the above step 102, performing a second preset algorithm on the time domain sequence to obtain sequence similarity includes:

[0080] For the time domain sequence, two subsequences are selected multiple times to perform the following operation to obtain the subsequence similarity;

[0081]

[0082] Where A represents one of the two optional subsequences, B represents the other of the two optional subsequences, A i represents the i-th value in sequence A, B i Represents the i-th value in sequence B;

[0083] The sequence similarity is obtained by summing up all the subsequence similarities obtained by the operation.

[0084] In the embodiment of the present disclosure, the time series similarity refers to the degree to which the increase and decrease change patterns of the time series points are the same, that is, the increase or decrease at the same time point, and the two time series show a certain degree of mutual parallelism. This is generally measured by the Minkowski distance. The present disclosure refers to the Minkowski distance to obtain the sequence similarity through the above formula. The smaller the sequence similarity, the worse the similarity of the sequence, the smaller the correlation between different time domain sequences, and the more accurate the corresponding prediction result.

[0085] In another embodiment provided by the present disclosure, in the above step 104, obtaining a comprehensive prediction result based on the first sequence using a preset algorithm includes:

[0086] Step 1: for the frequency domain sequence and the time domain sequence in the first sequence, use the pre-trained first prediction model and the second prediction model to predict the frequency domain sequence and the time domain sequence respectively, and obtain a frequency domain prediction result corresponding to the frequency domain sequence and a first prediction result corresponding to the time domain sequence;

[0087] Step 2: using inverse Fourier transform to convert the frequency domain prediction result corresponding to the frequency domain sequence in the first sequence into the time domain, to obtain the time domain prediction result corresponding to the frequency domain sequence, and determine it as the second prediction result;

[0088] Step three: sum the first prediction result and the second prediction result to obtain a comprehensive prediction result.

[0089] In the embodiment of the present disclosure, in the above step 2, the inverse Fourier transform is actually an operation to find the corresponding original function f(t) for a given Fourier transform F(jω).

[0090] In another embodiment provided by the present disclosure, the first prediction model takes the frequency domain sequence in the first sequence corresponding to any time period in the historical data as input, the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed first prediction model;

[0091] The second prediction model takes the time domain sequence in the first sequence corresponding to any time period in the historical data as input, and the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed second prediction model.

[0092] In the embodiment of the present disclosure, the first prediction model and the second prediction model may select a variety of neural network models including but not limited to convolutional neural networks, recurrent neural networks, and long short-term memory networks.

[0093] In order to facilitate a deeper understanding of the method disclosed in the present invention, a specific flow chart of the partial frequency domain super-capacity energy storage frequency modulation instruction prediction method is provided as follows: Figure 2 , Figure 3 As shown,

[0094] Based on the same disclosed concept, the disclosed embodiment also provides a partially frequency-domain super-capacitor energy storage frequency modulation instruction prediction device. Since the principle of the problem solved by the device is similar to the aforementioned partially frequency-domain super-capacitor energy storage frequency modulation instruction prediction method, the implementation of the device can refer to the implementation of the aforementioned method, and the repeated parts will not be repeated.

[0095] With the above Figure 1Corresponding to the method shown, the embodiment of the present disclosure also provides a partial frequency domain super-capacitor energy storage frequency modulation instruction prediction device, such as Figure 4 As shown, including:

[0096] The signal decomposition module 601 is used to obtain the historical frequency modulation instructions and the number of decomposition layers, and perform VMD decomposition on them to obtain subsequences corresponding to the number of decomposition layers;

[0097] The partial frequency domainization module 602 is used to perform a decomposition layer number cycle, and for each cycle, perform partial frequency domainization on the decomposition layer number subsequences based on the cycle number i corresponding to this cycle, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; sum the frequency disorder degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding cycle number;

[0098] The target sequence determination module 603 is used to determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the ith cycle corresponding to the minimum value as the first sequence;

[0099] The frequency modulation instruction prediction module 604 is used to obtain a comprehensive prediction result based on the first sequence using a third preset algorithm, and determine that the comprehensive prediction result is the current frequency modulation instruction.

[0100] In another embodiment provided by the present disclosure, in the partial frequency domain conversion module 602, partial frequency domain conversion is performed on the subsequences of the decomposition layer number based on the i value of this cycle, including:

[0101] Frequency domain conversion is performed on a part of subsequences from the 1st subsequence to the ith subsequence in the decomposition layer number of subsequences.

[0102] In combination with the second aspect, in a possible implementation manner, in the partial frequency domain conversion module 602, the following formula is used to perform partial frequency domain conversion on the subsequences of the decomposition layer:

[0103]

[0104] Among them, X(k) represents the frequency component in the frequency domain, x[n] represents the signal sequence in the time domain, N represents the signal length, k represents the frequency index in the frequency domain, and j is an imaginary unit.

[0105] In another embodiment provided by the present disclosure, in the partial frequency domain conversion module 602, the frequency domain sequence is subjected to a first preset algorithm using the following formula to obtain a frequency disorder degree:

[0106]

[0107] Among them, corrl represents the similarity transformation function, ω (i+1)-(1) Characterizes the aliasing width between adjacent subsequences, ω (i+1)max -ω (i)min Characterizes the frequency band occupied by two adjacent subsequences.

[0108] In another embodiment provided by the present disclosure, in the partial frequency domain conversion module 602, performing a second preset algorithm on the time domain sequence to obtain sequence similarity includes:

[0109] For the time domain sequence, two subsequences are selected multiple times to perform the following operation to obtain the subsequence similarity;

[0110]

[0111] Where A represents one of the two optional subsequences, B represents the other of the two optional subsequences, A i represents the i-th value in sequence A, B i Represents the i-th value in sequence B;

[0112] The sequence similarity is obtained by summing up all the subsequence similarities obtained by the operation.

[0113] In another embodiment provided by the present disclosure, in the frequency modulation instruction prediction module 604, a comprehensive prediction result is obtained based on the first sequence using a preset algorithm, including:

[0114] For the frequency domain sequence and the time domain sequence in the first sequence, use the pre-trained first prediction model and the second prediction model to predict the frequency domain sequence and the time domain sequence respectively, to obtain a frequency domain prediction result corresponding to the frequency domain sequence and a first prediction result corresponding to the time domain sequence;

[0115] Using inverse Fourier transform to convert the frequency domain prediction result corresponding to the frequency domain sequence in the first sequence into the time domain, to obtain the time domain prediction result corresponding to the frequency domain sequence, and determine it as the second prediction result;

[0116] The first prediction result and the second prediction result are summed to obtain a comprehensive prediction result.

[0117] In another embodiment provided by the present disclosure, the first prediction model takes the frequency domain sequence in the first sequence corresponding to any time period in the historical data as input, the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed first prediction model;

[0118] The second prediction model takes the time domain sequence in the first sequence corresponding to any time period in the historical data as input, and the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed second prediction model.

[0119] An embodiment of the present disclosure provides a computer device, including: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the steps of the partial frequency-domain super-capacity energy storage frequency modulation instruction prediction method provided in any embodiment of the present disclosure are executed.

[0120] The computer device provided by the embodiment of the present disclosure includes a processor, a memory, and a bus. The memory is used to store execution instructions, including internal memory and external memory; the internal memory here is also called internal memory, which is used to temporarily store the operation data in the processor and the data exchanged with the external memory such as the hard disk. The processor exchanges data with the external memory through the internal memory. When the electronic device is running, the processor and the memory communicate through the bus, so that the processor executes the following instructions:

[0121] Obtain historical frequency modulation instructions and decomposition levels, and perform variational mode decomposition on them to obtain subsequences corresponding to the number of decomposition levels;

[0122] Perform a decomposition layer number cycle, for each cycle, perform partial frequency domain conversion on the decomposition layer number subsequences based on the cycle number i corresponding to this cycle, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; sum the frequency disorder degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding cycle number;

[0123] Determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the minimum value as the first sequence;

[0124] A comprehensive prediction result is obtained based on the first sequence using a third preset algorithm, and the comprehensive prediction result is determined to be the current frequency modulation instruction.

[0125] The embodiment of the present disclosure provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for predicting the frequency modulation instructions of the partially frequency-domain super-capacitor energy storage provided in any embodiment of the present disclosure are executed. The storage medium may be a volatile or non-volatile computer-readable storage medium.

[0126] Through the description of the above implementation methods, those skilled in the art can clearly understand that the embodiments of the present disclosure can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present disclosure.

[0127] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present disclosure.

[0128] Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description of the embodiment, or can be changed accordingly and located in one or more devices different from the present embodiment. The modules in the above embodiment can be combined into one module, or can be further divided into multiple sub-modules.

[0129] The serial numbers of the above-mentioned embodiments of the present disclosure are only for description and do not represent the advantages or disadvantages of the embodiments.

[0130] Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A partial frequency domain super-capacitor energy storage frequency modulation instruction prediction method, characterized in that: include: Obtain historical frequency modulation instructions and decomposition levels, and perform variational mode decomposition on them to obtain subsequences corresponding to the number of decomposition levels; Perform a decomposition layer number cycle, for each cycle, perform partial frequency domain conversion on the decomposition layer number subsequences based on the cycle number i corresponding to this cycle, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency chaos degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; The frequency disorder degree and sequence similarity obtained in this cycle are summed to obtain the i-th total value of the corresponding cycle number; Determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the i-th cycle corresponding to the minimum value as the first sequence; A comprehensive prediction result is obtained based on the first sequence using a third preset algorithm, and the comprehensive prediction result is determined to be the current frequency modulation instruction.

2. The method according to claim 1, characterized in that The partially frequency-domainizing the subsequences of the decomposition layer based on the value of i in this cycle includes: Frequency domain conversion is performed on a part of subsequences from the 1st subsequence to the ith subsequence in the decomposition layer number of subsequences.

3. The method according to claim 1, characterized in that The following formula is used to perform partial frequency domain conversion on the subsequences of the decomposition layer: Among them, X(k) represents the frequency component in the frequency domain, x[n] represents the signal sequence in the time domain, N represents the signal length, k represents the frequency index in the frequency domain, and j is an imaginary unit.

4. The method according to claim 1, characterized in that The frequency domain sequence is subjected to a first preset algorithm using the following formula to obtain the frequency disorder degree: Among them, corrl represents the similarity transformation function, ω (i+1)-(1) Characterizes the aliasing width between adjacent subsequences, ω (i+1)max -ω (i)min Characterizes the frequency band occupied by two adjacent subsequences.

5. The method according to claim 1, characterized in that Performing a second preset algorithm on the time domain sequence to obtain sequence similarity includes: For the time domain sequence, two subsequences are selected multiple times to perform the following operation to obtain the subsequence similarity; Where A represents one of the two optional subsequences, B represents the other of the two optional subsequences, A i represents the i-th value in sequence A, B i Represents the i-th value in sequence B; The sequence similarity is obtained by summing up all the subsequence similarities obtained by the operation.

6. The method according to claim 1, characterized in that The obtaining a comprehensive prediction result based on the first sequence by using a preset algorithm includes: For the frequency domain sequence and the time domain sequence in the first sequence, use the pre-trained first prediction model and the second prediction model to predict the frequency domain sequence and the time domain sequence respectively, to obtain a frequency domain prediction result corresponding to the frequency domain sequence and a first prediction result corresponding to the time domain sequence; Using inverse Fourier transform to convert the frequency domain prediction result corresponding to the frequency domain sequence in the first sequence into the time domain, to obtain the time domain prediction result corresponding to the frequency domain sequence, and determine it as the second prediction result; The first prediction result and the second prediction result are summed to obtain a comprehensive prediction result.

7. The method according to claim 6, characterized in that The first prediction model uses the frequency domain sequence in the first sequence corresponding to any time period in the historical data as input, the frequency modulation instruction of the next time period adjacent to the time period as output, and the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed first prediction model; The second prediction model takes the time domain sequence in the first sequence corresponding to any time period in the historical data as input, and the frequency modulation instruction of the next time period adjacent to the time period as output, and takes the frequency modulation instruction of the next time period adjacent to the time period in the historical data as the true value to train the constructed second prediction model.

8. A partially frequency-domain super-capacitor energy storage frequency modulation instruction prediction device, characterized in that: include: The signal decomposition module is used to obtain the historical frequency modulation instructions and the number of decomposition layers, and perform VMD decomposition on them to obtain subsequences corresponding to the number of decomposition layers; The partial frequency domainization module is used to perform a decomposition layer number cycle, and for each cycle, based on the cycle number i corresponding to this cycle, perform partial frequency domainization on the decomposition layer number subsequences, and divide all subsequences into frequency domain sequences and time domain sequences; perform a first preset algorithm on the frequency domain sequence to obtain a frequency disorder degree, and perform a second preset algorithm on the time domain sequence to obtain a sequence similarity; sum the frequency disorder degree and sequence similarity obtained in this cycle to obtain the i-th total value of the corresponding cycle number; A target sequence determination module is used to determine the minimum value among all the total values ​​obtained in the above cycle, and determine the frequency domain sequence and time domain sequence of the ith cycle corresponding to the minimum value as the first sequence; A frequency modulation instruction prediction module is used to obtain a comprehensive prediction result based on the first sequence using a third preset algorithm, and determine that the comprehensive prediction result is the current frequency modulation instruction.

9. A computer device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of a partial frequency-domain super-capacity energy storage frequency modulation instruction prediction method as described in any one of claims 1 to 7 are performed.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of a method for predicting frequency modulation instructions of a partially frequency-domain ultra-capacitor energy storage system as claimed in any one of claims 1 to 7.

Citation Information

Cited By

  • Energy storage frequency modulation instruction prediction method and system of centripetal force method

    CN119848519A