A method and system for predicting energy storage frequency modulation instructions with auxiliary asymptote

Abstract

A method for predicting energy storage frequency modulation instructions with auxiliary asymptotes comprises the following steps: obtaining a frequency modulation instruction signal; decomposing the frequency modulation instruction signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm; generating corresponding upper and lower asymptotes for the modal subsequences; performing sequence decomposition on the modal subsequences using the corresponding upper and lower asymptotes to generate core sequences and auxiliary sequences; predicting the core sequences and auxiliary sequences using a neural network to obtain final prediction results; and applying the final prediction results to energy storage equipment for frequency modulation response. This method makes the prediction results of the frequency modulation instruction more accurate, predicts the size of the frequency modulation instruction in advance for energy storage regulation, and improves response accuracy and frequency modulation benefits.

Description

Technical Field

[0001] The present application relates to the technical field of power grid frequency regulation, and in particular to a method and system for predicting energy storage frequency regulation instructions using an auxiliary asymptote. Background Art

[0002] Currently, large hydropower and thermal power units (coal-fired and gas-fired) are the primary frequency regulation sources in my country's major regional power grids. Frequency regulation power output is adjusted to respond to system frequency changes. Hybrid energy storage frequency regulation command prediction involves analyzing historical frequency regulation commands and related data to develop a prediction model to predict future frequency regulation command requirements. This optimizes the control strategy of the hybrid energy storage system and improves its frequency regulation performance and efficiency. This system can enhance system performance and provide rapid response. Output power can be adjusted quickly to respond to grid frequency changes, effectively balancing supply and demand within the power system and maintaining grid frequency stability. Improved regulation accuracy allows for more precise tracking of frequency regulation commands, reducing regulation errors and improving power quality and stability. Enhanced system stability can be achieved by rationally allocating power and energy across different energy storage devices, reducing overall system risk, improving reliability and stability, and mitigating the risk of system failure due to failure or performance degradation of a single energy storage device. Prediction can also extend device life and optimize charging and discharging strategies. Based on the prediction results, a reasonable charging and discharging strategy can be formulated to avoid excessive or frequent charging and discharging of energy storage devices, thereby extending their service life. Balanced equipment usage: Rationally allocating the usage frequency and load of different energy storage devices ensures relatively even aging across all devices, reducing maintenance costs and replacement frequency. This reduces costs and improves economic efficiency. By improving frequency regulation performance and efficiency, hybrid energy storage systems can achieve greater frequency regulation benefits while reducing power outages and equipment damage costs caused by unstable grid frequency. Investment costs can also be optimized. Based on forecast results, the capacity and quantity of different energy storage devices can be rationally allocated, avoiding overinvestment and waste of resources, and improving return on investment.

[0003] The current traditional prediction methods (such as Figure 2 ), using methods such as VMD decomposition, the original frequency modulation instruction sequence is divided into a series of subsequences, which are then predicted separately and the predicted results are superimposed. Existing VMD decomposition methods have drawbacks. They may produce subsequences that are even more difficult to predict. Some subsequences may even have a higher degree of nonlinearity than the original instruction sequence, making prediction even more difficult. Inaccurate predictions can cause response deviations, affecting the accuracy of system responses and, in turn, the frequency modulation benefits of power plants. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention proposes a prediction method based on the asymptote method. First, the upper and lower asymptotes of each subsequence are made, and the parts of each subsequence that exceed the asymptotes are superimposed together to form a new sequence for prediction to obtain the final result.

[0005] An energy storage frequency modulation instruction prediction method for auxiliary asymptotes, characterized by comprising:

[0006] S101, obtaining a frequency modulation command signal;

[0007] S102, decomposing the frequency modulation command signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm;

[0008] S103. Generate corresponding upper limit asymptotes and lower limit asymptotes for a number of modal subsequences;

[0009] S104. Decomposing a plurality of modal subsequences using corresponding upper limit asymptotes and lower limit asymptotes to generate a core sequence and an auxiliary sequence.

[0010] S105. Using a neural network to predict several core sequences and several auxiliary sequences, a final prediction result is obtained; and the final prediction result is applied to an energy storage device to perform frequency modulation response.

[0011] A system for predicting energy storage frequency modulation instructions for auxiliary asymptotes is also provided, which is characterized by comprising:

[0012] An instruction acquisition module is used to acquire a frequency modulation instruction signal;

[0013] An instruction decomposition module, configured to decompose the frequency modulation instruction signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm;

[0014] An asymptote generation module, used for generating corresponding upper asymptotes and lower asymptotes for a number of modal subsequences;

[0015] The subsequence decomposition module is used to perform sequence decomposition on several modal subsequences using corresponding upper and lower asymptotes to generate core sequences and auxiliary sequences;

[0016] The prediction processing module is used to use a neural network to predict several core sequences and several auxiliary sequences to obtain a final prediction result; and the final prediction result is applied to the energy storage device for frequency modulation response.

[0017] The beneficial effects of the present invention are:

[0018] Reducing the nonlinearity of the subsequences after VMD decomposition makes each subsequence of VMD smoother, greatly reducing the prediction difficulty and improving the accuracy of frequency regulation instruction prediction, thereby improving the response accuracy of the power plant and the frequency regulation benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a step diagram of the application method.

[0020] Figure 2 Schematic diagram of the traditional algorithm.

[0021] Figure 3 Schematic diagram of the data processing coordinate system.

[0022] Figure 4 To generate asymptote diagram. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0024] It should be understood that the step numbers used herein are only for convenience of description and are not intended to limit the order in which the steps are to be executed.

[0025] It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0026] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0027] The term "and / or" refers to and includes any and all possible combinations of one or more of the associated listed items.

[0028] The embodiment of the present disclosure provides a method for predicting energy storage frequency modulation instructions of auxiliary asymptotes, such as Figure 1 Shown, including:

[0029] S101, obtaining a frequency modulation command signal;

[0030] S102, decomposing the frequency modulation command signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm;

[0031] S103. Generate corresponding upper limit asymptotes and lower limit asymptotes for a number of modal subsequences;

[0032] S104. Decomposing a plurality of modal subsequences using corresponding upper limit asymptotes and lower limit asymptotes to generate a core sequence and an auxiliary sequence.

[0033] S105. Using a neural network to predict several core sequences, and using a neural network to predict several auxiliary sequences after mixing and superimposing them, to obtain a final prediction result; and applying the final prediction result to the energy storage device for frequency modulation response.

[0034] The frequency modulation command prediction method provided by the disclosed embodiments uses novel supercapacitors and lithium batteries as energy storage devices to modulate the frequency of a power plant's power grid when its frequency fluctuates. The method monitors the power plant's power grid in real time and generates corresponding frequency modulation command signals when the power supply frequency fluctuates. After acquiring this original signal, the method then predicts future signals to implement frequency modulation of the power plant's power grid.

[0035] Adjust the power output in real time based on the prediction results, such as the active power distribution of energy storage units under different working conditions, and coordinate the operation of each unit to smooth out power fluctuations and improve the frequency regulation performance of the power grid.

[0036] In another embodiment provided by the present disclosure, the above step “S102, decomposing the frequency modulation instruction signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm” includes:

[0037] Receive a frequency modulation command signal; assume that the frequency modulation command signal is Pt; Pt is a function of a time parameter t.

[0038] The frequency modulation command signal is decomposed based on a preset algorithm.

[0039] Determine the number of decomposition layers and use VMD decomposition to decompose the frequency modulation command signal; set the number of VMD decomposition layers to K, and decompose the frequency modulation command signal into K subsequences IMF1, IMF2, IMF3, ..., IMF K .

[0040] In another embodiment provided by the present disclosure, the above step “S103, generating corresponding upper limit asymptotes and lower limit asymptotes for a number of modal subsequences” includes:

[0041] Assume that the frequency modulation command signal Pt is decomposed into IMF1, IMF2, IMF3, ..., IMF i ,…,IMF K . Among them, the subsequence IMF i =[X i1 ,X i2 ,Xi3 ,…,X ii ,…,X iN ].

[0042] Step 1: Obtain the intersection sequence of each subsequence.

[0043] Subsequence IMF i Each numerical point in is projected onto Figure 3 The data processing coordinate system is shown.

[0044] like Figure 3 As shown in the coordinate system, the horizontal axis represents the time parameter t, and the vertical axis is the signal value X i1 、X i2 、X i3 ,…,X ii ,…,X iN . Establish a connecting line segment between two adjacent points and find the perpendicular bisector of the connecting line segment;

[0045] The intersection points of two adjacent perpendicular bisectors in the coordinate system are O1, O2, O3, ..., O N-1 , there are N-1 intersections, forming an intersection sequence [O1, O2, O3, ..., O N-1 ].

[0046] Step 2: Obtain the upper limit point sequence and generate the upper limit asymptote.

[0047] Select the upper limit point from the intersection sequence, the upper limit point is defined as O m >X i(m+1) , then the intersection point is the upper limit point.

[0048] like Figure 3 As shown in i2 , O1 is the upper limit point.

[0049] All selected upper limit points form an upper limit point sequence [O S1 ,O S2 ,O S3 ,…,O Sq ].

[0050] like Figure 4 As shown, a quadratic function curve is established through two adjacent points in the upper limit point sequence, so that the area enclosed by the quadratic function curve and the line between the two adjacent points is minimized.

[0051] In the specific process,

[0052] Pass O S1 and O S2 Generate an upward-opening quadratic function X=a1t 2 +b1t+c1, this function is the same as O S1 and OS2 The enclosed area is S1. Adjust a1, b1, and c1 to minimize the area of ​​S1.

[0053] Continue to pass O S2 and O S3 Generate an upward-opening quadratic function X=a2t 2 +b2t+c2, this function is the same as O S2 and O S3 The enclosed area is S2. Adjust a2, b2, and c2 to minimize the area of ​​S2.

[0054] Follow this process in order, passing point O Sq-1 and O Sq Generate an upward-opening quadratic function X=a q-1 t 2 +b q-1 t+c q-1 , this function is the same as O Sq-1 and O Sq The enclosed area is S q-1 , adjust a q-1 ,b q-1 ,c q-1 Make S q-1 The area is the smallest.

[0055] Through the above process, S1+S2+…+S q-1 The area and minimum.

[0056] The curve connecting the above q-1 upward-opening quadratic function curves serves as the upper asymptote.

[0057] Step 3: Obtain the lower limit point sequence and generate the lower limit asymptote.

[0058] Select the lower limit point from the intersection sequence, the lower limit point is defined as O m <X i(m+1) , then the intersection point is the upper limit point.

[0059] like Figure 3 As shown in <X i3 , O2 is the lower limit point.

[0060] All selected lower limit points form a lower limit point sequence [O S1 、 ,O S2 、 ,O S3 、 ,…,O Sq 、 ].

[0061] A quadratic function curve is established through two adjacent points in the lower limit point sequence so that the area enclosed by the quadratic function curve and the line between the two adjacent points is minimized.

[0062] In the specific process,

[0063] Pass O S1 、 and O S2 、 Generate a quadratic function X=a1t that opens downward 2 +b1t+c1, this function is the same as O S1 、 and O S2 、 The enclosed area is S1 、 , adjust a1, b1, c1 so that S1 、 The area is the smallest.

[0064] Continue to pass O S2 、 and O S3 、 Generate a downward-opening quadratic function X=a2t 2 +b2t+c2, this function is the same as O S2 、 and O S3 、 The enclosed area is S2 、 , adjust a2, b2, c2 so that S2 、 The area is the smallest.

[0065] Follow this process in order, passing point O Sq-1 、 and O Sq 、 Generate an upward-opening quadratic function X=a q-1 t 2 +b q-1 t+c q-1 , this function is the same as O Sq-1 、 and O Sq 、 The enclosed area is S q-1 、 , adjust a q-1 ,b q-1 ,c q-1 Make S q-1 、 The area is the smallest.

[0066] Through the above process, S1 、 +S2 、 +…+S q-1 、 The area and minimum.

[0067] The curve connecting the above q-1 downward-opening quadratic function curves is used as the lower limit asymptote.

[0068] In another embodiment provided by the present disclosure, the above step “S104, performing sequence decomposition on a plurality of modal subsequences using corresponding upper limit asymptotes and lower limit asymptotes to generate core sequences and auxiliary sequences” includes:

[0069] The upper and lower asymptotes are used as threshold boundaries to transform the subsequence IMF i Decomposition into core sequence IMF i 、 and auxiliary sequence IMF i 、、 .

[0070] Subsequence IMF i =[X i1 ,X i2 ,X i3 ,…,X ii ,…,X iN ] are compared with the upper and lower asymptotes, and the numerical points between the upper and lower asymptotes are retained, and the other numerical points are reset to zero to form a new core sequence IMF. i 、 .

[0071] In addition, the subsequence IMF i =[X i1 ,X i2 ,X i3 ,…,X ii ,…,X iN ] are compared with the upper asymptote and the lower asymptote. The numerical points outside the upper asymptote and the lower asymptote are retained, and the other numerical points are reset to zero to form a new auxiliary sequence IMF. i 、、 .

[0072] Such as IMF i =[ X i1 ,X i2 ,X i3 ,X i4 ,X i5 ,X i6 ,X i7 ,X i8 ,X i9 ], where the numerical point X i4 and X i7 Outside the upper and lower asymptotes,

[0073] Core Sequence IMFi 、 =[ X i1 ,X i2 ,X i3 ,0,X i5 ,X i6 ,0,X i8 ,X i9 ],

[0074] Auxiliary sequence IMF i 、、 =[0,0,0,X i4 ,0,0,X i7 ,0,0].

[0075] The above step "S105, predicting several core sequences using a neural network, and mixing and superimposing several auxiliary sequences and then predicting using a neural network to obtain a final prediction result" includes:

[0076] Step 1: Core sequence IMF1 、 ,IMF2 、 ,IMF3 、 ,…,IMF i 、 ,…,IMF K 、 The GRU (Gated Recurrent Network) network is used for prediction.

[0077] Step 2: Auxiliary sequence IMF1 、、 ,IMF2 、、 ,IMF3 、、 ,…,IMF i 、、 ,…,IMF K 、、 First, perform mixed superposition, and then use the GRU (gated recurrent network) network for prediction.

[0078] Assume that each auxiliary sequence IMF i 、、 The non-zero values ​​of L are i , zero value has B i The algorithm for mixed superposition is:

[0079] exp(sigmoid(L1 / B1)) IMF1 、、 +exp(sigmoid(L2 / B2)) IMF2 、、 +exp(sigmoid(L3 / B3))IMF3 、、 +…exp(sigmoid(L i / B i IMF i、、 +…+exp(sigmoid(L k / B k IMF K 、、 , where sigmoid() is the activation function, and after mixed superposition, the GRU (gated recurrent network) network is used for prediction.

[0080] The core sequence and auxiliary sequence are fed into the GRU network for prediction, and the prediction results are then superimposed to obtain the final prediction result. The final prediction result is applied to the power plant grid for frequency regulation response.

[0081] According to the prediction results, the output of the power source is adjusted in real time, the active power of the energy storage unit is distributed under different working conditions, and the operation of each unit is coordinated to smooth out power fluctuations and improve the frequency regulation performance of the power grid.

[0082] With the above Figure 1 Corresponding to the method shown, the embodiment of the present disclosure further provides an energy storage frequency modulation instruction prediction system for auxiliary asymptotes, comprising:

[0083] An instruction acquisition module is used to acquire a frequency modulation instruction signal;

[0084] An instruction decomposition module, configured to decompose the frequency modulation instruction signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm;

[0085] An asymptote generation module, used for generating corresponding upper asymptotes and lower asymptotes for a number of modal subsequences;

[0086] The subsequence decomposition module is used to perform sequence decomposition on several modal subsequences using corresponding upper and lower asymptotes to generate core sequences and auxiliary sequences;

[0087] The prediction processing module is used to use a neural network to predict several core sequences and several auxiliary sequences to obtain a final prediction result; and the final prediction result is applied to the energy storage device for frequency modulation response.

[0088] In addition, the prediction processing module includes:

[0089] A core sequence prediction module is used to predict several core sequences using a neural network;

[0090] The auxiliary sequence prediction module is used to mix and superimpose several auxiliary sequences and then use neural network to predict.

[0091] In order to further verify the advantages of the present invention, the present invention uses the method of the present invention and the GRU prediction method to predict the frequency modulation sequence, and the results are as follows.

[0092] The experimental frequency modulation instructions were obtained from a power plant in Gansu. The acquisition time for frequency modulation sequence 1 was from 0:00 to 15:00 on November 1, 2022, with one data point collected every 1 second. The acquisition time for frequency modulation sequence 2 was from 0:00 to 15:00 on November 2, 2022, with one data point collected every 1 second.

[0093] The performance evaluation is shown in Table 1 below:

[0094] Table 1

[0095]

[0096] The four evaluation indicators are shown in Table 2:

[0097] Table 2

[0098]

[0099] N represents the sample size, and Represent the actual value and predicted value at time n respectively.

[0100] From the experimental results, we can see that all four evaluation indicators have been reduced, which shows that the proposed model can well improve the prediction accuracy.

[0101] An embodiment of the present disclosure provides a computer device, comprising: 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 and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps provided in any embodiment of the present disclosure are performed.

[0102] The computer device provided in the embodiment of the present application includes a processor, a memory, and a bus. The memory is used to store and execute instructions, and includes internal memory and external memory. The internal memory here is also called internal memory, which is used to temporarily store the calculation 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:

[0103] Obtaining frequency modulation command signal;

[0104] Decomposing the frequency modulation command signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm;

[0105] Generate corresponding upper limit asymptotes and lower limit asymptotes for several modal subsequences;

[0106] For several modal subsequences, the corresponding upper limit asymptotes and lower limit asymptotes are used to perform sequence decomposition to generate core sequences and auxiliary sequences;

[0107] A number of core sequences and a number of auxiliary sequences are predicted using a neural network to obtain a final prediction result; the final prediction result is applied to the energy storage device for frequency modulation response.

[0108] The present disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, executes the steps provided in any embodiment of the present disclosure. The storage medium may be a volatile or non-volatile computer-readable storage medium.

[0109] Through the above description of the embodiments, those skilled in the art will clearly understand that the embodiments of the present disclosure can be implemented through hardware or through software plus the necessary general hardware platform. Based on this understanding, the technical solutions 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 (such as a read-only optical disk, a USB flash drive, a mobile hard disk, etc.) and includes a number of instructions for causing a computer device (such as a personal computer, a server, or a network device) to execute the methods described in the various embodiments of the present disclosure.

[0110] Those skilled in the art will understand 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.

[0111] Those skilled in the art will appreciate that the modules in the devices of the embodiments may be distributed in the devices of the embodiments as described in the embodiments, or may be located in one or more devices different from the embodiments with corresponding changes. The modules of the above embodiments may be combined into one module or further split into multiple submodules.

[0112] The serial numbers of the above-mentioned embodiments of the present disclosure are for description only and do not represent the advantages or disadvantages of the embodiments.

[0113] Obviously, those skilled in the art may 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 intended to include these modifications and variations.

[0114] Finally, it should be noted that the foregoing description is merely an explanation of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail, those skilled in the art will be able to modify the aforementioned technical solutions or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for predicting energy storage frequency modulation instructions with auxiliary asymptotes, characterized in that: include, S101, obtaining a frequency modulation command signal; S102, decomposing the frequency modulation command signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm; S103. Generate corresponding upper limit asymptotes and lower limit asymptotes for a number of modal subsequences; S104. Decomposing a plurality of modal subsequences using corresponding upper limit asymptotes and lower limit asymptotes to generate a core sequence and an auxiliary sequence. S105, using a neural network to predict several core sequences and several auxiliary sequences to obtain a final prediction result; applying the final prediction result to the energy storage device to perform frequency modulation response; In step S102, the number of decomposition layers is determined, and the frequency modulation command signal is decomposed by VMD decomposition; the number of VMD decomposition layers is set to K, and the frequency modulation command signal is decomposed into K subsequences IMF1, IMF2, IMF3, ..., IMF K ; Step S103 includes: IMF of each subsequence decomposed by VMD i =[ X i1 ,X i2 ,X i3 ,…,X ii ,…,X iN ]; S1031, obtaining the intersection sequence of each subsequence; S1032, obtaining an upper limit point sequence based on the intersection point sequence, and generating an upper limit asymptote; S1033, obtaining a lower limit point sequence based on the intersection point sequence, and generating a lower limit asymptote; In S1031, Subsequence IMF i Each numerical point in is projected onto the data processing coordinate system; a connecting line segment between two adjacent points is established, and the perpendicular bisector of the connecting line segment is obtained; two adjacent perpendicular bisectors generate an intersection point in the coordinate system, and there are N-1 intersection points in total, forming an intersection sequence [O1, O2, O3, ..., O N-1 ]; In S1032, Select the upper limit point from the intersection sequence, the upper limit point is defined as O m >X i(m+1) , then the intersection point is the upper limit point; All selected upper limit points form an upper limit point sequence [O S1 ,O S2 ,O S3 ,…,O Sq ]; Establish an upward-opening quadratic function curve through two adjacent points in the upper limit point sequence, so that the area enclosed by the quadratic function curve and the line connecting the two adjacent points is minimized; continue in this order to obtain a curve connecting q-1 quadratic function curves as the upper limit asymptote; In S1033, Select the lower limit point from the intersection sequence, the lower limit point is defined as O m <X i(m+1) , then the intersection point is the lower limit point; All selected lower limit points form a lower limit point sequence [O S1 、 ,O S2 、 ,O S3 、 ,…,O Sq 、 ]; Establish a downward-opening quadratic function curve through two adjacent points in the lower limit point sequence, so that the area enclosed by the quadratic function curve and the line connecting the two adjacent points is minimized; execute in this order to obtain a curve connecting q-1 quadratic function curves as the lower limit asymptote; In step S104, The upper and lower asymptotes are used as threshold boundaries to transform the subsequence IMF i Decomposition into core sequence IMF i 、 and auxiliary sequence IMF i 、、 ; Subsequence IMF i =[ X i1 ,X i2 ,X i3 ,…,X ii ,…,X iN ] are compared with the upper asymptote and the lower asymptote. The numerical points between the upper asymptote and the lower asymptote are retained, and the other numerical points are reset to zero to form the core sequence IMF. i 、 ; subsequenceIMF i =[ X i1 ,X i2 ,X i3 ,…,X ii ,…,X iN ] are compared with the upper and lower asymptotes, and the numerical points outside the upper and lower asymptotes are retained, and the other numerical points are reset to zero to form the auxiliary sequence IMF. i 、、 .

2. The energy storage frequency modulation instruction prediction method of the auxiliary asymptote according to claim 1 is characterized in that: In step S105, a plurality of core sequences are predicted using a neural network, and a plurality of auxiliary sequences are mixed and superimposed and then predicted using a neural network to obtain a final prediction result.

3. A prediction system based on the energy storage frequency modulation instruction prediction method of the auxiliary asymptote according to claim 1, characterized in that: include, An instruction acquisition module is used to acquire a frequency modulation instruction signal; An instruction decomposition module, configured to decompose the frequency modulation instruction signal into a number of modal subsequences corresponding to the number of decomposition layers based on a preset algorithm; An asymptote generation module, used for generating corresponding upper asymptotes and lower asymptotes for a number of modal subsequences; The subsequence decomposition module is used to perform sequence decomposition on several modal subsequences using corresponding upper and lower asymptotes to generate core sequences and auxiliary sequences; The prediction processing module is used to use a neural network to predict several core sequences and several auxiliary sequences to obtain a final prediction result; and the final prediction result is applied to the energy storage device for frequency modulation response.

4. The prediction system according to claim 3, characterized in that: The prediction processing module includes: A core sequence prediction module is used to predict several core sequences using a neural network; The auxiliary sequence prediction module is used to mix and superimpose several auxiliary sequences and then use neural network to predict.

Images