Energy storage frequency modulation instruction prediction method and system of distance optimization method

Abstract

The application discloses a distance optimization method for predicting energy storage frequency modulation instruction, obtains an original frequency modulation instruction sequence; calculates the nonlinearity degree value of each signal in the original frequency modulation instruction sequence, and forms a nonlinearity degree value sequence; corrects the signal in the original frequency modulation instruction sequence by using the nonlinearity degree value sequence, and obtains a corrected frequency modulation instruction sequence; and uses a neural network to predict the corrected frequency modulation instruction sequence, and obtains a final prediction result; the prediction result of the frequency modulation instruction is more accurate, the size of the frequency modulation instruction is predicted in advance for energy storage regulation, and the response accuracy and frequency modulation benefit are improved.

Description

Technical Field

[0001] This application relates to the field of power grid frequency regulation technology, and in particular to a distance-optimized method and system for predicting energy storage frequency regulation commands. Background Technology

[0002] Currently, large-scale hydropower and thermal power units (coal-fired / gas-fired units) are mainly used as frequency regulation power sources in my country's major regional power grids. The output of these power sources is adjusted to respond to changes in system frequency. Hybrid energy storage frequency regulation command prediction refers to establishing a predictive model by analyzing historical frequency regulation commands and related data to predict future frequency regulation command demands. This optimizes the control strategy of the hybrid energy storage system, improving its frequency regulation performance and efficiency. It can enhance system performance, provide rapid response, quickly adjust output power in a short time, respond swiftly to changes in grid frequency, effectively balance supply and demand differences in the power system, and maintain grid frequency stability. It improves regulation accuracy, enabling more precise tracking of frequency regulation commands, reducing regulation errors, and improving the power supply quality and stability of the grid. It enhances system stability by rationally allocating the power and energy of different energy storage devices, reducing overall system risk, improving system reliability and stability, and reducing the risk of system failure due to the failure or performance degradation of a single energy storage device. It extends equipment lifespan and optimizes charging and discharging strategies: based on the prediction results, reasonable charging and discharging strategies can be formulated to avoid overcharging and discharging or frequent charging and discharging of energy storage devices, thereby extending their service life. Balanced equipment usage and reasonable allocation of usage frequency and load for different energy storage devices ensure a relatively even aging process, 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 obtain greater frequency regulation benefits while reducing power outage losses and equipment damage costs caused by grid frequency instability. Optimized investment costs: based on forecast results, the capacity and quantity of different energy storage devices can be rationally configured, avoiding over-investment and resource waste, and improving the return on investment.

[0003] Traditional prediction methods directly feed the original frequency-modulated (FM) sequence into a GRU network for prediction. The drawback of this approach is that the original sequence exhibits significant nonlinearity and irregularity, leading to inaccurate predictions and large errors. Therefore, preprocessing of the original FM sequence before prediction is necessary, such as... Figure 2 As shown, common methods such as VMD decomposition and CEEMD decomposition decompose the original frequency-modulated sequence into a series of low-frequency subsequences, which are then fed into a neural network for prediction. However, while combining VMD (Variational Mode Decomposition) and CEEMD (Complementary Ensemble Empirical Mode Decomposition) for the decomposition and prediction of the original frequency-modulated sequence can improve the accuracy of non-stationary signal processing, it also has the following drawbacks:

[0004] 1. High computational complexity. The decomposition process is time-consuming: both VMD and CEEMD require multiple iterations (CEEMD requires multiple additions of noise and decomposition). Combining the two leads to a significant increase in computational load, especially for long-term data where real-time performance is poor. Parameter tuning is complex: VMD requires preset modal number (K) and penalty factor (α), while CEEMD requires setting noise amplitude and ensemble count. Inappropriate parameter selection can lead to over-decomposition or under-decomposition.

[0005] 2. Modal redundancy and over-decomposition risks. Sensitive to the number of modes: If the number of modes (K) in VMD is set too large, it may decompose into noisy modes with no physical meaning; if it is too small, it will fail to capture key features. Information fragmentation: CEEMD may generate redundant modes due to residual noise or over-decomposition, resulting in too many subsequences and increasing the complexity of subsequent prediction models.

[0006] 3. Mode aliasing and frequency confusion. Residual mode aliasing: Although CEEMD reduces mode aliasing compared to traditional EMD, it does not completely eliminate it, especially in areas where the original FM sequence abruptly changes or in high-frequency noise regions, overlap may still occur. VMD frequency boundary problem: VMD assumes that the modes are narrowband signals, but the actual original FM sequence may contain broadband components, resulting in blurred frequency boundaries between modes after decomposition.

[0007] 4. Difficulty in integrating prediction models. Multimodal modeling complexity: Each subsequence after decomposition needs to be modeled individually or predicted jointly, multiplying the number of models. Integration strategies (such as linear superposition or neural network fusion) may introduce errors. Temporal synchronization issues: The decomposed subsequences need to maintain time alignment. If phase shifts occur during decomposition (such as the endpoint effect of CEEMD), the prediction results may be distorted.

[0008] 5. Information Loss and Reconstruction Errors. Decomposition-Reconstruction Distortion: After VMD and CEEMD decomposition, the sub-modes need to be reconstructed into the original signal. Decomposition errors or improper parameters can cause the reconstructed signal to deviate from the true value, affecting the quality of the prediction input. High-Frequency Noise Amplification: Random fluctuations in the original FM sequence may be decomposed into high-frequency modes. If these are directly ignored or simply processed, key transient information may be lost.

[0009] 6. Insufficient handling of sudden changes and non-stationarity. Lagging response to extreme events: The corresponding modes may be dispersed across multiple subsequences, making it difficult for the prediction model to capture sudden patterns in a timely manner. Residual non-stationarity: If the decomposition fails to completely separate stationary and non-stationary components, the residuals still contain complex fluctuations, reducing prediction accuracy. Summary of the Invention

[0010] To address the shortcomings of existing technologies and improve prediction accuracy while overcoming the drawbacks of VMD / CEEMD, this invention proposes a distance optimization method for processing the original frequency modulation sequence.

[0011] A distance-optimized method for predicting frequency regulation commands in energy storage, characterized by comprising:

[0012] S101, Obtain the original frequency modulation command sequence;

[0013] S102. Calculate the nonlinearity value of each signal in the original frequency modulation command sequence and form a nonlinearity value sequence.

[0014] S103. The signal in the original frequency modulation command sequence is corrected by using the nonlinearity value sequence to obtain the corrected frequency modulation command sequence.

[0015] S104. The modified frequency modulation command sequence is predicted using a neural network to obtain the final prediction result.

[0016] Furthermore,

[0017] In step S101, the original frequency modulation command signal sequence is Pt=[X1,X2,X3,…,X…]. i ,…,X N The original frequency modulation command signal sequence Pt is a function of time t. The numerical points in the original frequency modulation command signal sequence Pt are projected onto a plane coordinate system, with the vertical axis representing the magnitude of the value and the horizontal axis representing the number of samples.

[0018] Furthermore,

[0019] Step S102 includes,

[0020] Step 1: Establish a signal value X in the sequence i Connect the N-1 signal values ​​in the sequence to the N-1 line segments, and calculate the length d of each line segment to obtain the length value sequence [d]. i1 ,d i2 ,d i3 ,…,d i(N-1) ];

[0021] Step 2: Calculate the signal value X based on the length value sequence i The regular distance r between the signal values ​​and the rest of the sequence

[0022] Obtain the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ];

[0023] Step 3: Using signal value X i Using the line segment connecting the first-ranked signal value among the remaining N-1 signal values ​​as the baseline segment, calculate the angle θ between the N-1 connecting line segments and the baseline segment to obtain the angle value sequence.

[0024] [θ i1 ,θ i2 ,θ i3 ,…,θ i(N-1) ], where θ i1 =0;

[0025] Step 4: Calculate the signal value X i The degree of nonlinearity D in the sequence i Its formula is,

[0026] D i =[θ i1 r i1 +θ i2 r i2 +θ i3 r i3 +…+θ i(N-1) r i(N-1) ];

[0027] The final generated sequence of nonlinearity values ​​is [D1, D2, D3, ..., D]. i ,…,D N ].

[0028] Furthermore,

[0029] When the signal value X i When it is X1;

[0030] First calculate X1 to [X2, X3, X4, ..., X N The sequence of length values ​​of each connection [d] 11 ,d 12 ,d 13 ,…,d 1(N-1) Then calculate the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ],

[0031] Connect X1 and X2 as the baseline line segment, and calculate the sequence of angles [θ] between each connecting line and the baseline line segment. 11 ,θ 12 ,θ 13 ,…,θ 1(N-1) ], θ 12 The angle is the angle between the line connecting X1 and X3 and the line connecting X1 and X2, θ. 13 The angle is the angle between the line connecting X1 and X4 and the line connecting X1 and X2, and other angles are deduced similarly; the degree of nonlinearity of X1 in the sequence, D1, is defined as D1 = [θ]. 11 r 11 +θ 12 r 12 +θ 13 r13 +…+θ 1(N-1) r 1(N-1) ].

[0032] Furthermore,

[0033] When the signal value X i For X m At that time, 1 <m<N;

[0034] First calculate X m to [X1,X2,…,X m-1 ,X m+1 ,…,X N The sequence of length values ​​of each connection [d] m1 ,d m2 ,d m3 ,…,d m(N-1) Then calculate the regular distance sequence [r] m1 ,r m2 ,…,r m(N-1) ],

[0035] Connect X m Using X1 as the baseline segment, calculate the sequence of angles [θ] between each connecting line and the baseline segment. m1 ,θ m2 ,θ m3 ,…,θ m(N-1) ], where θ m1 For X m The angle between the line connecting X1 and the baseline segment is 0.

[0036] X m The degree of nonlinearity D in the sequence m Defined as D m =[θ m1 r m1 +θ m2 r m2 +θ m3 r m3 +…+θ m(N-1) r m(N-1) ].

[0037] Furthermore,

[0038] When the signal value X i For X N hour;

[0039] First calculate X N to [X1,X2,X3,…,X N-1 The sequence of length values ​​of each connection [d] N1 ,d N2 ,d N3 ,…,d N(N-1)Then calculate the regular distance sequence [r] N1 ,r N2 ,r N3 ,…,r N(N-1) ],

[0040] Connect X N Using X1 as the baseline segment, calculate the sequence of angles [θ] between each connecting line and the baseline segment. N1 ,θ N2 ,θ N3 ,…,θ N(N-1) ], where θ N1 For X N The angle between the line connecting X1 and the baseline segment is 0.

[0041] X N The degree of nonlinearity D in the sequence N Defined as D N =[θ N1 r N1 +θ N2 r N2 +θ N3 r N3 +…+θ N(N-1) r N(N-1) ].

[0042] Furthermore,

[0043] Step S103 includes,

[0044] For the original frequency modulation command signal sequence [X1,X2,X3,…,X…] i ,…,X N Based on the nonlinearity value sequence [D1, D2, D3, ..., D] i ,…,D N The original sequence is corrected to [X1] 、 X2 、 X3 、 ,…,X i 、 ,…,X N 、 ];

[0045] For each value X in the original sequence i The corrected formula is as follows:

[0046] .

[0047] Furthermore,

[0048] Step S104 includes,

[0049] The corrected sequence [X1] 、X2 、 X3 、 ,…,X i 、 ,…,X N 、 The data is fed into a GRU network for prediction, and the final prediction result is obtained. The final prediction result is then applied to the power plant grid for frequency regulation response.

[0050] A distance-optimized energy storage frequency regulation command prediction system is also provided, characterized in that it includes,

[0051] The instruction acquisition module acquires the original frequency modulation instruction sequence;

[0052] The data processing module calculates the degree of nonlinearity of each signal in the original frequency modulation command sequence and forms a sequence of nonlinearity values.

[0053] The signal correction module uses a nonlinearity value sequence to correct the signal in the original frequency modulation command sequence, resulting in a corrected frequency modulation command sequence.

[0054] The prediction processing module uses a neural network to predict the corrected frequency modulation command sequence to obtain the final prediction result.

[0055] Furthermore,

[0056] The data processing module further includes:

[0057] Establish line segments connecting a certain signal value in the sequence to the other signal values ​​in the sequence, and calculate the length of each line segment to obtain the length value sequence;

[0058] Calculate the regular distance between the signal value and the other signal values ​​in the sequence based on the length value sequence to obtain the regular distance sequence;

[0059] Using the line segment connecting the signal value to the first-ranked signal value among the remaining signal values ​​as the baseline line segment, calculate the angle between each connecting line segment and the baseline line segment to obtain the angle value sequence.

[0060] The degree of nonlinearity of the signal value in the sequence is calculated based on the regular distance sequence and the included angle value sequence.

[0061] The final result is a sequence of nonlinearity values.

[0062] The beneficial effects of this invention are:

[0063] Reducing the nonlinearity of the original sequence greatly reduces the difficulty of prediction and improves the accuracy of frequency regulation command prediction, thereby improving the accuracy of power plant response and frequency regulation benefits. Attached Figure Description

[0064] Figure 1This is a diagram illustrating the method steps.

[0065] Figure 2 This is a diagram illustrating a traditional algorithm.

[0066] Figure 3 This is a schematic diagram of the coordinate system.

[0067] Figure 4 This is a schematic diagram of a data processing algorithm. Detailed Implementation

[0068] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0069] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0070] It should be understood that the terminology used in this application specification is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this application 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.

[0071] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0072] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0073] This disclosure provides a distance-optimized method for predicting energy storage frequency regulation commands, such as... Figure 1 As shown, it includes:

[0074] S101, Obtain the original frequency modulation command sequence;

[0075] S102. Calculate the nonlinearity value of each signal in the original frequency modulation command sequence and form a nonlinearity value sequence.

[0076] S103. The signal in the original frequency modulation command sequence is corrected by using the nonlinearity value sequence to obtain the corrected frequency modulation command sequence.

[0077] S104. The modified frequency modulation command sequence is predicted using a neural network to obtain the final prediction result.

[0078] The frequency regulation command prediction method based on supercapacitor energy storage provided in this disclosure uses novel supercapacitors and lithium batteries as energy storage devices and regulates the frequency of the power grid in the power plant when fluctuations occur. It monitors the state of the power plant's power grid in real time, and when fluctuations occur in the power supply frequency of the power plant's power grid, it generates a corresponding frequency regulation command signal. After acquiring this original signal, it implements frequency regulation of the power plant's power grid by predicting future signals.

[0079] Step S101, obtaining the original frequency modulation command signal sequence, includes:

[0080] The original frequency modulation command signal sequence is Pt=[X1,X2,X3,…,X…]. i ,…,X N The original frequency modulation command signal sequence Pt is a function of time t.

[0081] Project each value point in the original frequency modulation command signal sequence Pt onto a plane coordinate system, with the vertical axis representing the magnitude of the value and the horizontal axis representing the number of samples.

[0082] like Figure 3 As shown in the coordinate system, the horizontal axis represents the number of samples, and the vertical axis represents the signal values ​​X1, X2, X3, ..., X. i ,…,X N .

[0083] In another embodiment disclosed in this application, such as Figure 4 As shown,

[0084] For the original frequency modulation command signal sequence, the stronger the nonlinearity, the more difficult it is to predict. An algorithm is used to quantify the strength of the irregularity of each signal value point in the sequence. For a single value point, the corresponding dimensionless value is obtained by calculation to represent the nonlinearity of that value point. The dimensionless value is the nonlinearity value.

[0085] S102 above involves calculating the degree of nonlinearity for each signal in the original frequency modulation command sequence and forming a sequence of nonlinearity values; including:

[0086] Step 1: Establish a signal value X in the sequence i Connect the N-1 signal values ​​in the sequence to the N-1 line segments, and calculate the length d of each line segment to obtain the length value sequence [d]. i1 ,d i2 ,d i3 ,…,d i(N-1) ].

[0087] Step 2: Calculate the signal value Xi The regular distance r between the signal values ​​and the rest of the sequence.

[0088] Obtain the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ].

[0089] Step 2: Using signal value X i The line segment connecting the N-1 signal values ​​to the first-ranked signal value among the remaining N-1 signal values ​​is used as the baseline segment. The angle θ between the N-1 connecting line segments and the baseline segment is calculated. This yields a sequence of angle values.

[0090] [θ i1 ,θ i2 ,θ i3 ,…,θ i(N-1) ], obviously where θ i1 It is 0.

[0091] Step 3: Calculate the signal value X i The degree of nonlinearity D in the sequence i Its formula is,

[0092] D i =[θ i1 r i1 +θ i2 r i2 +θ i3 r i3 +…+θ i(N-1) r i(N-1) ].

[0093] like Figure 4 As shown,

[0094] 1. When the signal value X i When it is X1;

[0095] First calculate X1 to [X2, X3, X4, ..., X N The sequence of length values ​​of each connection [d] 11 ,d 12 ,d 13 ,…,d 1(N-1) Then calculate the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ],

[0096] Connect X1 and X2 as the baseline line segment, and calculate the sequence of angles between each connecting line and the baseline line segment.

[0097] [θ 11 ,θ 12 ,θ 13 ,…,θ1(N-1) ], θ 12 The angle is the angle between the line connecting X1 and X3 and the line connecting X1 and X2, θ. 13 The angle is the angle between the line connecting X1 and X4 and the line connecting X1 and X2. Other angles follow the same principle.

[0098] Where θ 11 Let θ be the angle between the line connecting X1 and X2 and the baseline segment. Clearly, the angle value is θ. 11 =0;

[0099] Among them, [θ 11 ,θ 12 ,θ 13 ,…,θ 1(N-1) ] represents X1 and [X2,X3,X4,…,X N The magnitude of the trend change.

[0100] [r i1 ,r i2 ,…,r i(N-1) ] represents X1 and [X2,X3,X4,…,X N The magnitude of the trend change and the closeness of the linear relationship.

[0101] Therefore, the degree of nonlinearity D1 of X1 in the sequence is defined as

[0102] D1=[θ 11 r 11 +θ 12 r 12 +θ 13 r 13 +…+θ 1(N-1) r 1(N-1) ].

[0103] II. When the signal value X i For X m At that time, 1 <m<N;

[0104] First calculate X m to [X1,X2,…,X m-1 ,X m+1 ,…,X N The sequence of length values ​​of each connection [d] m1 ,d m2 ,d m3 ,…,d m(N-1) Then calculate the regular distance sequence [r] m1 ,r m2 ,…,r m(N-1) ],

[0105] Connect X mUsing X1 as the baseline segment, calculate the sequence of angles between each connecting line and the baseline segment.

[0106] [θ m1 ,θ m2 ,θ m3 ,…,θ m(N-1) ], where θ m1 For X m The angle between the line connecting X1 and the baseline segment is 0.

[0107] X m The degree of nonlinearity D in the sequence m Defined as

[0108] D m =[θ m1 r m1 +θ m2 r m2 +θ m3 r m3 +…+θ m(N-1) r m(N-1) ].

[0109] III. When the signal value X i For X N hour;

[0110] First calculate X N to [X1,X2,X3,…,X N-1 The sequence of length values ​​of each connection [d] N1 ,d N2 ,d N3 ,…,d N(N-1) Then calculate the regular distance sequence [r] N1 ,r N2 ,r N3 ,…,r N(N-1) ],

[0111] Connect X N Using X1 as the baseline segment, calculate the sequence of angles between each connecting line and the baseline segment.

[0112] [θ N1 ,θ N2 ,θ N3 ,…,θ N(N-1) ], where θ N1 For X N The angle between the line connecting X1 and the baseline segment is 0.

[0113] X N The degree of nonlinearity D in the sequence N Defined as

[0114] D N =[θN1 r N1 +θ N2 r N2 +θ N3 r N3 +…+θ N(N-1) r N(N-1) ].

[0115] Following the above method, the final nonlinearity value sequence [D1,D2,D3,…,D] is generated. i ,…,D N ].

[0116] The above-mentioned S103, using a nonlinearity level value sequence to correct the original frequency modulation command sequence signal, yields a corrected frequency modulation command sequence; including:

[0117] For the original sequence [X1,X2,X3,…,X…] i ,…,X N Based on the nonlinearity value sequence [D1, D2, D3, ..., D] i ,…,D N The original sequence is corrected to [X1] 、 X2 、 X3 、 ,…,X i 、 ,…,X N 、 ]

[0118] For each value X in the original sequence i The corrected formula is as follows:

[0119] .

[0120] The above-mentioned S104 involves using a neural network to predict the modified frequency modulation command sequence to obtain the final prediction result; including:

[0121] The corrected sequence [X1] 、 X2 、 X3 、 ,…,X i 、 ,…,X N 、 The data is fed into a GRU network for prediction, and the final prediction result is obtained.

[0122] The final prediction results are then applied to the power plant grid for frequency regulation response.

[0123] With the above Figure 1 Correspondingly, this disclosure also provides a distance optimization method-based energy storage frequency regulation command prediction system, including:

[0124] The instruction acquisition module acquires the original frequency modulation instruction sequence;

[0125] The data processing module calculates the degree of nonlinearity of each signal in the original frequency modulation command sequence and forms a sequence of nonlinearity values.

[0126] The signal correction module uses a nonlinearity value sequence to correct the signal in the original frequency modulation command sequence, resulting in a corrected frequency modulation command sequence.

[0127] The prediction processing module uses a neural network to predict the corrected frequency modulation command sequence to obtain the final prediction result.

[0128] The data processing module further includes:

[0129] Establish line segments connecting a certain signal value in the sequence to the other signal values ​​in the sequence, and calculate the length of each line segment to obtain the length value sequence;

[0130] Calculate the regular distance between the signal value and the other signal values ​​in the sequence based on the length value sequence to obtain the regular distance sequence;

[0131] Using the line segment connecting the signal value to the first-ranked signal value among the remaining signal values ​​as the baseline line segment, calculate the angle between each connecting line segment and the baseline line segment to obtain the angle value sequence.

[0132] The degree of nonlinearity of the signal value in the sequence is calculated based on the regular distance sequence and the included angle value sequence.

[0133] The final result is a sequence of nonlinearity values.

[0134] To further verify the advantages of the present invention, the present invention uses both the method of the present invention and the GRUS prediction method to predict the frequency modulation sequence, and the results are as follows.

[0135] The frequency modulation commands for the experiment were acquired from a power plant in Gansu Province. Frequency modulation sequence 1 was collected from 4:00 AM to 6:00 PM on a certain day in January 2021, with one data point collected every second. Frequency modulation sequence 2 was collected from 12:00 AM to 7:00 PM on a certain day in March 2022, with one data point collected every second.

[0136] The comparison results are shown in the table below:

[0137]

[0138] The four evaluation indicators are shown in the table below:

[0139]

[0140] N represents the sample size. and These represent the actual value and the predicted value at time n, respectively.

[0141] The experimental results show that all four evaluation indicators decreased, indicating that the proposed model can significantly improve prediction accuracy.

[0142] This disclosure provides a computer device, including a processor, a memory, and a bus. 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. When the machine-readable instructions are executed by the processor, they perform the steps provided in any embodiment of this disclosure.

[0143] The computer device provided in this application includes a processor, a memory, and a bus. The memory, also known as internal memory, stores execution instructions and includes main memory and external memory. The main memory temporarily stores data processed by the processor, as well as data exchanged with external storage devices such as hard disks. The processor exchanges data with external storage devices through main memory. When the electronic device is running, the processor and memory communicate via the bus, enabling the processor to execute the following instructions:

[0144] Obtain the original frequency modulation command sequence;

[0145] For each signal in the original frequency modulation command sequence, calculate the degree of nonlinearity in the sequence and form a sequence of nonlinearity values;

[0146] The signal in the original frequency modulation command sequence is corrected by using a sequence of nonlinearity values ​​to obtain a corrected frequency modulation command sequence;

[0147] The modified frequency modulation command sequence is predicted using a neural network to obtain the final prediction result.

[0148] This disclosure provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps provided in any embodiment of this disclosure. The storage medium may be volatile or non-volatile computer-readable storage.

[0149] Through the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of this disclosure can be implemented in hardware or by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions of the embodiments of this 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 disc, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.

[0150] 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 drawings are not necessarily essential for implementing this disclosure.

[0151] Those skilled in the art will understand that the modules in the apparatus of the embodiments can be distributed in the apparatus of the embodiments as described in the embodiments, or they can be located in one or more devices different from this embodiment with corresponding changes. The modules of the above embodiments can be combined into one module, or they can be further divided into multiple sub-modules.

[0152] The sequence numbers of the embodiments disclosed above are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0153] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.

[0154] Finally, it should be noted that the above description is merely an explanation of the present invention and is not intended to limit the invention. Although the present invention has been described in detail, those skilled in the art can still modify the technical solutions described above or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A distance-optimized method for predicting frequency regulation commands in energy storage, characterized in that, include: S101, Obtain the original frequency modulation command sequence; S102. Calculate the nonlinearity value of each signal in the original frequency modulation command sequence and form a nonlinearity value sequence. S103. The signal in the original frequency modulation command sequence is corrected by using the nonlinearity value sequence to obtain the corrected frequency modulation command sequence. S104. The modified frequency modulation command sequence is predicted using a neural network to obtain the final prediction result; In step S101, the original frequency modulation command sequence is Pt=[X1,X2,X3,…,X…]. i ,…,X N The original frequency modulation command sequence Pt is a function of time t. The numerical points in the original frequency modulation command sequence Pt are projected onto a plane coordinate system, with the vertical axis representing the magnitude of the value and the horizontal axis representing the number of samples. Step S102 includes, Step 1: Establish a signal value X in the sequence i Connect the N-1 signal values ​​in the sequence to the N-1 line segments, and calculate the length d of each line segment to obtain the length value sequence [d]. i1 ,d i2 ,d i3 ,…,d i(N-1) ]; Step 2: Calculate the signal value X based on the length value sequence i The regular distance r between the signal values ​​and the rest of the sequence Obtain the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ]; Step 3: Using signal value X i Using the line segment connecting the first-ranked signal value among the remaining N-1 signal values ​​as the baseline segment, calculate the angle θ between the N-1 connecting line segments and the baseline segment to obtain the angle value sequence. [θ i1 , θ i2 , θ i3 , …, θ i(N-1) , where θ i1 is 0; Step 4: Calculate the signal value X i The degree of nonlinearity D in the sequence i Its formula is, D i =[θ i1 r i1 +θ i2 r i2 +θ i3 r i3 +…+θ i(N-1) r i(N-1) ]; The final generated sequence of nonlinearity values ​​is [D1, D2, D3, ..., D]. i ,…,D N ].

2. The energy storage frequency regulation command prediction method based on the distance optimization method according to claim 1, characterized in that, When the signal value X i When it is X1; First calculate X1 to [X2, X3, X4, ..., X N The sequence of length values ​​of each connection [d] 11 ,d 12 ,d 13 ,…,d 1(N-1) Then calculate the regular distance sequence [r] i1 ,r i2 ,…,r i(N-1) ], Connect X1 and X2 as the baseline line segment, and calculate the sequence of angles [θ] between each connecting line and the baseline line segment. 11 ,θ 12 ,θ 13 ,…,θ 1(N-1) ], θ 12 The angle is the angle between the line connecting X1 and X3 and the line connecting X1 and X2, θ. 13 The angle is the angle between the line connecting X1 and X4 and the line connecting X1 and X2, and other angles are deduced similarly; the degree of nonlinearity of X1 in the sequence, D1, is defined as D1 = [θ]. 11 r 11 +θ 12 r 12 +θ 13 r 13 +…+θ 1(N-1) r 1(N-1) ].

3. The energy storage frequency regulation command prediction method based on the distance optimization method according to claim 1, characterized in that, When the signal value X i For X m At that time, 1 <m<N; First calculate X m to [X1,X2,…,X m-1 ,X m+1 ,…,X N The sequence of length values ​​of each connection [d] m1 ,d m2 ,d m3 ,…,d m(N-1) Then calculate the regular distance sequence [r] m1 ,r m2 ,…,r m(N-1) ], Connect X m Using X1 as the baseline segment, calculate the sequence of angles [θ] between each connecting line and the baseline segment. m1 ,θ m2 ,θ m3 ,…,θ m(N-1) ], where θ m1 For X m The angle between the line connecting X1 and the baseline segment is 0. X m The degree of nonlinearity D in the sequence m Defined as D m =[θ m1 r m1 +θ m2 r m2 +θ m3 r m3 +…+θ m(N-1) r m(N-1) ].

4. The energy storage frequency regulation command prediction method based on the distance optimization method according to claim 1, characterized in that, When the signal value X i For X N hour; First calculate X N to [X1,X2,X3,…,X N-1 The sequence of length values ​​of each connection [d] N1 ,d N2 ,d N3 ,…,d N(N-1) Then calculate the regular distance sequence [r] N1 ,r N2 ,r N3 ,…,r N(N-1) ], Connect X N Using X1 as the baseline segment, calculate the sequence of angles [θ] between each connecting line and the baseline segment. N1 ,θ N2 ,θ N3 ,…,θ N(N-1) ], where θ N1 For X N The angle between the line connecting X1 and the baseline segment is 0. X N The degree of nonlinearity D in the sequence N Defined as D N =[θ N1 r N1 +θ N2 r N2 +θ N3 r N3 +…+θ N(N-1) r N(N-1) ].

5. The energy storage frequency regulation command prediction method based on the distance optimization method according to claim 1, characterized in that, Step S103 includes, For the original frequency modulation command signal sequence [X1,X2,X3,…,X…] i ,…,X N Based on the nonlinearity value sequence [D1, D2, D3, ..., D] i ,…,D N The original sequence is corrected to [X1] 、 X2 、 X3 、 ,…,X i 、 ,…,X N 、 ]; For each value X in the original sequence i The corrected formula is as follows: 。 6. The energy storage frequency regulation command prediction method based on the distance optimization method according to claim 5, characterized in that, Step S104 includes, The corrected sequence [X1] 、 X2 、 X3 、 ,…,X i 、 ,…,X N 、 The data is fed into a GRU network for prediction, and the final prediction result is obtained. The final prediction result is then applied to the power plant grid for frequency regulation response.

7. A prediction system employing the distance optimization method of claim 1 for predicting energy storage frequency regulation commands, characterized in that, include, The instruction acquisition module acquires the original frequency modulation instruction sequence; The data processing module calculates the degree of nonlinearity of each signal in the original frequency modulation command sequence and forms a sequence of nonlinearity values. The signal correction module uses a nonlinearity value sequence to correct the signal in the original frequency modulation command sequence, resulting in a corrected frequency modulation command sequence. The prediction processing module uses a neural network to predict the corrected frequency modulation command sequence to obtain the final prediction result.

8. The prediction system according to claim 7, characterized in that, The data processing module further includes: Establish line segments connecting a certain signal value in the sequence to the other signal values ​​in the sequence, and calculate the length of each line segment to obtain the length value sequence; Calculate the regular distance between the signal value and the other signal values ​​in the sequence based on the length value sequence to obtain the regular distance sequence; Using the line segment connecting the signal value to the first-ranked signal value among the remaining signal values ​​as the baseline line segment, calculate the angle between each connecting line segment and the baseline line segment to obtain the angle value sequence. The degree of nonlinearity of the signal value in the sequence is calculated based on the regular distance sequence and the included angle value sequence. The final result is a sequence of nonlinearity values.

Images