Hybrid energy storage frequency modulation instruction prediction method and system based on error feedback

Abstract

The invention discloses a hybrid energy storage frequency modulation instruction prediction method and system based on error feedback, and relates to the field of hybrid energy storage frequency modulation prediction optimization, and the method comprises the steps: collecting a frequency modulation instruction, obtaining first data, and carrying out the first screening and the second screening; based on the first data, the first screening result and the second screening result, a data abnormal value is found and repaired, and second data is obtained; decomposing the second data to obtain third data; inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; according to the method, through multi-level abnormal value detection and restoration, a decomposition strategy is optimized, the convolutional neural network is used for feature extraction and prediction, and a multi-level error feedback mechanism is adopted, so that the accuracy, stability, real-time performance and adaptability of frequency modulation instruction prediction are remarkably improved; the method can be adjusted and optimized according to different application scenes, and has important application value in the fields of power system frequency modulation, hybrid energy storage system control and the like.

Description

Technical Field

[0001] The present invention relates to the field of hybrid energy storage frequency modulation prediction and optimization, and in particular to a hybrid energy storage frequency modulation instruction prediction method and system based on error feedback. Background Art

[0002] In the power system, thermal power units are one of the main ways of generating electricity, and their frequency regulation capability is crucial to the stable operation of the power grid. The traditional frequency regulation method mainly relies on the regulation capability of the thermal power units themselves, but due to the slow response speed of thermal power units, it is difficult to meet the power grid's demand for fast frequency regulation. In order to improve the frequency regulation efficiency, hybrid energy storage systems (such as supercapacitors + lithium batteries) are introduced as auxiliary frequency regulation methods. The traditional hybrid energy storage frequency regulation method is to transmit the difference between the frequency regulation command and the actual output of the thermal power unit to the hybrid energy storage system, where the low-frequency part is borne by the lithium battery and the high-frequency part is borne by the supercapacitor.

[0003] It takes a certain amount of time for the frequency modulation command to be transmitted from the thermal power unit to the hybrid energy storage system. At the same time, there is also a delay in the response of the supercapacitor and lithium battery, resulting in poor overall frequency modulation effect, which in turn affects the revenue of the power plant. During the data collection stage, due to natural or human factors, the data may be distorted or missing, affecting the accuracy of the prediction. Traditional methods lack effective preprocessing of the original data, resulting in unreliable prediction results. When traditional data decomposition methods (such as VMD, CEEMD, etc.) process frequency modulation command signals, the decomposed subsequences may have modal aliasing, affecting the prediction accuracy. Traditional prediction methods usually lack feedback and correction mechanisms for prediction errors, resulting in deviations between the prediction results and actual needs, and unable to achieve dynamic optimization. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: how to improve the response speed and prediction accuracy of frequency regulation of hybrid energy storage-assisted thermal power units to enhance the stability of the power grid and the economic benefits of the power plant.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, an embodiment of the present invention provides a hybrid energy storage frequency modulation instruction prediction method based on error feedback, comprising: Collect frequency modulation instructions, obtain first data, and perform first screening and second screening; Based on the first data and the results of the first screening and the second screening, find and repair the data outliers to obtain the second data; Decomposing the second data to obtain third data; Decomposing the second data to obtain the third data includes: Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Perform five random decompositions, each generating K subsequences, and perform a first operation after the five random decompositions are completed. The first operation includes: calculating the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; finding the decomposition corresponding to the minimum error coefficient; Take the decomposition corresponding to the minimum error coefficient as the optimal decomposition of this round, and increase the value of p by one; Repeat the first decomposition. When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition. The directions of the five random decompositions in the next round of decomposition are based on the directions of the best decompositions in the previous round; When the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposition subsequence as the final decomposition result, and take the final decomposition result as the third data; Inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; The step of inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction includes: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result, where N is the total length of the subsequence.

[0007] As a preferred solution of the hybrid energy storage frequency modulation instruction prediction method based on error feedback, wherein: The step of finding and repairing data outliers based on the first data and the results of the first screening and the second screening to obtain the second data includes: If a point is both a suspicious point in the first screening and a suspicious point in the second screening, the point is identified as an outlier; after the outlier in the first data is repaired, the second data is obtained.

[0008] As a preferred solution of the hybrid energy storage frequency modulation instruction prediction method based on error feedback, wherein: The first screening is for size abnormalities: Assume the frequency modulation instruction is P t =[X1,X2,X3,...,X i ,...,X N ], the mean is J=avg(X1,X2,X3,...,X i ,...,X N ), where X1,X2,X3,...,X i ,...,X NAll are frequency modulation values; If a value of the frequency modulation instruction is greater than 3 times of J or less than 1 / 3 of J, it is listed as a suspicious point.

[0009] As a preferred solution of the hybrid energy storage frequency modulation instruction prediction method based on error feedback, wherein: The second screening is the screening of abnormal changes before and after: Assume that the three consecutive values ​​in the frequency modulation instruction are X i-1 , X i , X i+1 ; If sin(X i-1 -X i ) and sin(X i -X i+1 ) have different signs, and: , then X i Listed as suspicious points.

[0010] In a second aspect, an embodiment of the present invention provides a hybrid energy storage frequency modulation instruction prediction system based on error feedback, comprising: A screening module, used for collecting frequency modulation instructions, obtaining first data, and performing first screening and second screening; A repair module, used to find and repair data abnormal values ​​based on the first data and the results of the first screening and the second screening to obtain second data; A decomposition module, used for decomposing the second data to obtain third data; Decomposing the second data to obtain the third data includes: Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Perform five random decompositions, each generating K subsequences, and perform a first operation after the five random decompositions are completed. The first operation includes: calculating the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; finding the decomposition corresponding to the minimum error coefficient; Take the decomposition corresponding to the minimum error coefficient as the optimal decomposition of this round, and increase the value of p by one; Repeat the first decomposition. When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition. The directions of the five random decompositions in the next round of decomposition are based on the directions of the best decompositions in the previous round; When the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposition subsequence as the final decomposition result, and take the final decomposition result as the third data; A prediction module, used for inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; The step of inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction includes: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result, where N is the total length of the subsequence.

[0011] In a third aspect, an embodiment of the present invention provides a computing device, including: Memory and processor; The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the hybrid energy storage frequency modulation instruction prediction method based on error feedback as described in any embodiment of the present invention.

[0012] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the hybrid energy storage frequency modulation instruction prediction method based on error feedback.

[0013] Beneficial effects of the present invention: The present invention calculates the mean of the frequency modulation instruction sequence, identifies outliers, effectively filters extreme values, and avoids their negative impact on subsequent analysis and prediction; after the outliers are removed or repaired from the data, the error source of the prediction model is significantly reduced, and the reliability of the prediction results is improved; the random decomposition method is used to decompose the complex time series signal into multiple subsequences, each of which often represents a different frequency component or change mode, which is convenient for subsequent feature extraction and model prediction. This decomposition method can better capture the dynamic characteristics of the signal and improve the precision of signal processing; the convolutional neural network is used to predict the decomposed subsequences, which can more efficiently process complex frequency modulation instruction sequences and significantly improve the accuracy and robustness of the prediction. At the same time, the end-to-end training method of the convolutional neural network simplifies the design and optimization process of the model; through the multi-level error feedback mechanism, the decomposition and prediction process are gradually optimized, and the data processing and model prediction process can be dynamically adjusted and optimized, which significantly improves the prediction accuracy, especially in complex or noisy frequency modulation instruction data, showing higher stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0015] Figure 1 It is an overall flow chart of the hybrid energy storage frequency modulation instruction prediction method based on error feedback described in the present invention. DETAILED DESCRIPTION

[0016] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0017] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0018] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0019] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and provides a hybrid energy storage frequency modulation instruction prediction method based on error feedback, comprising: S1: Collect frequency modulation instructions, obtain first data, and perform first screening and second screening; S2: based on the first data and the results of the first screening and the second screening, find and repair the data outliers to obtain the second data; S3: Decompose the second data to obtain third data; S4: Input the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction.

[0020] It should be noted that through the above steps, a hybrid energy storage frequency modulation command prediction method based on error feedback is realized, data quality optimization is achieved, and the reliability of input data is improved; signal feature extraction and optimization decomposition are performed to enhance the effectiveness of signal features; CNN is used for efficient feature learning and prediction to significantly improve prediction accuracy; a multi-level error feedback mechanism is used to improve the stability and prediction performance of the model; rapid response and real-time prediction are achieved to meet the real-time control requirements of the power system; and a flexible system design is provided to adapt to a variety of application scenarios.

[0021] Example 2, reference Figure 1 , which is an embodiment of the present invention, provides a hybrid energy storage frequency modulation instruction prediction method based on error feedback based on the previous embodiment, including: In the embodiment of the present application, in the above step S1, the frequency modulation instruction is collected to obtain the first data, and the first screening and the second screening include: It should be noted that the collected scheduling instructions are frequency modulation values ​​at intervals of 1 second, which are subsequently processed in the form of a data sequence, for example, [1.1, 2.0, 3]. The first data is the filtered data sequence.

[0022] Specifically, the screening includes the first screening: Screening for size abnormalities: Assume the frequency modulation instruction is P t =[X1,X2,X3,...,X i ,...,X N ], the mean is J=avg(X1,X2,X3,...,X i ,...,X N ), where X1,X2,X3,...,X i ,...,X N All are frequency modulation values; If a value of the frequency modulation instruction is greater than 3 times of J or less than 1 / 3 of J, it is listed as a suspicious point.

[0023] The screening also includes a second screening: screening for abnormal changes before and after: Assume that the three consecutive values ​​in the frequency modulation instruction are X i-1 , X i , X i+1 ; If sin(X i-1 -X i ) and sin(X i -X i+1 ) have different signs, and: , then X i Listed as suspicious points.

[0024] In another possible implementation, the first screening and the second screening may also be Z-score detection: the Z score of each data point is calculated, and points that deviate from the mean by more than a certain standard deviation are identified as outliers.

[0025] Rolling average filtering: Use a rolling window to calculate the average and mark points that deviate from the average as anomalies.

[0026] Window size: Select an appropriate window size based on data characteristics.

[0027] Boxplot method: Use quartiles to identify outliers, for example, values ​​outside Q1-1.5IQR and Q3+1.5IQR.

[0028] In the embodiment of the present application, in the above step S2, based on the first data and the results of the first screening and the second screening, the data outliers are found and repaired, and the second data obtained includes: If a point is a suspicious point in both the first screening and the second screening, the point is identified as an outlier.

[0029] Let X i is an outlier, for X i The correction is expressed as: .

[0030] After the outliers in the first data are repaired, the second data is obtained.

[0031] In another possible implementation, polynomial interpolation may be used to estimate outliers. Machine learning models such as K-nearest neighbors and decision trees may also be used to predict outliers based on surrounding data points.

[0032] In the embodiment of the present application, the second data is decomposed in the above step S3 to obtain the third data including: Suppose the sequence obtained after data processing is P t =[X1',X2',X3',...,X i ',...,X N ']; Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Five random decompositions are performed, each time generating K subsequences. After the five random decompositions are completed, the first operation is performed to obtain the optimal decomposition of this round, and the value of p is increased by one.

[0033] When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition; Each time the next round of five random decompositions is performed, the directions of the five random decompositions in the next round are based on the directions of the optimal decompositions in the previous round.

[0034] Specifically, the initial value of the iteration round p is set to 0, and the initial decomposition layer number K is set. In this embodiment, the preferred value of the iteration round p is 5, K is preferably 6, and the layer number setting value is preferably 12; it should be noted that the constraint condition of random decomposition is that the sum of all subsequences decomposed each time is P t '.

[0035] Perform the first decomposition: Perform five random decompositions, each time dividing P t ' is divided into K subsequences, represented as: IMF1, IMF i ,...,IMF5, where i is the order of decomposition; that is, there are K subsequences in IMF1, IMF i ,...,IMF5 has K subsequences respectively.

[0036] The first operation is performed on the subsequence after the five random decompositions: calculate the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; find the decomposition corresponding to the minimum error coefficient; let W j Minimum; record the subsequence IMF generated by this decomposition j and error coefficient W j ; It should be noted that the calculation of the error coefficient mainly adopts MAPE, and the calculation formula is: , exemplarily, MAPE is calculated for each of the K subsequences generated by each decomposition, and the K MAPE values ​​of each decomposition are summarized into an error coefficient (which can be an average value, maximum value, minimum value, etc.).

[0037] It should also be noted that the directions of the five random decompositions in the next step are based on the directions of the optimal decomposition in the previous step.

[0038] For example, assuming that the error of W3 is the lowest, the subsequence IMF3 after this decomposition corresponding to W3 is expressed as: IMF 31 =[X 11 ,X 12 ,...,X 1N ],IMF 32 =[X 21 ,X 22 ,...,X 2N ],...,IMF 3K =[X K1 ,X K2 ,...,X KN], set it as set B, then the random direction of the five random decompositions generated in the next round is specified. For example, let the subsequence generated by any random decomposition in the next round of five random decompositions be IMF t , expressed as: IMF t1 =[X 11 ,X 12 ,...,X 1N ],IMF t2 =[X 21 ,X 22 ,...,X 2N ],...,IMF tK =[X K1 , X K2 ,...,X KN ], set it to set A; All corresponding sets A and B need to satisfy: ,in, refers to the Nth element in the Kth subsequence of set A, Refers to the Nth element in the Kth subsequence in set B; further, when the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposed subsequence as the final decomposition result, and take the final decomposition result as the third data.

[0039] In the embodiment of the present application, in the above step S4, the third data is input into the convolutional neural network, and the prediction result of the frequency modulation instruction is obtained, including: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result.

[0040] It should be noted that the convolutional neural network used in the present invention has shown excellent performance in the field of signal processing. Its core advantage is that CNN can automatically extract features from data without human intervention. CNN is usually composed of a multi-layer structure, mainly including convolutional layers, pooling layers and fully connected layers. As a key part of CNN, the convolutional layer extracts local features from the data through moving convolution kernels; the pooling layer is used to reduce the spatial dimension of the features, thereby simplifying the model; at the end of the network, the fully connected layer maps the features extracted by the previous layers to the final output to achieve the classification effect. During the training process, CNN continuously updates the weights and biases of the network by back propagation.

[0041] Embodiment 3, the above is a schematic scheme of the hybrid energy storage frequency modulation instruction prediction method based on error feedback of this embodiment. It should be noted that the technical scheme of the hybrid energy storage frequency modulation instruction prediction system based on error feedback and the technical scheme of the hybrid energy storage frequency modulation instruction prediction method based on error feedback belong to the same concept, and the details of the technical scheme of the hybrid energy storage frequency modulation instruction prediction system based on error feedback in this embodiment that are not described in detail can all be referred to the description of the technical scheme of the hybrid energy storage frequency modulation instruction prediction method based on error feedback.

[0042] This embodiment also provides a system for predicting a hybrid energy storage frequency modulation instruction based on error feedback, including: A screening module, used for collecting frequency modulation instructions, obtaining first data, and performing first screening and second screening; A repair module, used to find and repair data abnormal values ​​based on the first data and the results of the first screening and the second screening to obtain second data; A decomposition module, used for decomposing the second data to obtain third data; Decomposing the second data to obtain the third data includes: Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Perform five random decompositions, each generating K subsequences, and perform a first operation after the five random decompositions are completed. The first operation includes: calculating the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; finding the decomposition corresponding to the minimum error coefficient; Take the decomposition corresponding to the minimum error coefficient as the optimal decomposition of this round, and increase the value of p by one; Repeat the first decomposition. When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition. The directions of the five random decompositions in the next round of decomposition are based on the directions of the best decompositions in the previous round; When the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposition subsequence as the final decomposition result, and take the final decomposition result as the third data; A prediction module, used for inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; The step of inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction includes: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result, where N is the total length of the subsequence.

[0043] This embodiment further provides a computing device, which is applicable to the hybrid energy storage frequency modulation instruction prediction method based on error feedback, and includes: Memory and processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the hybrid energy storage frequency modulation instruction prediction method based on error feedback as proposed in the above embodiment.

[0044] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the hybrid energy storage frequency modulation instruction prediction method based on error feedback proposed in the above embodiment is implemented.

[0045] The storage medium proposed in this embodiment and the hybrid energy storage frequency modulation instruction prediction method based on error feedback proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0046] Example 4, referring to Tables 1 and 2, is an embodiment of the present invention, which provides a hybrid energy storage frequency modulation instruction prediction method based on error feedback. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0047] The method of the present invention and the prediction method of VMD-CNN are used to predict the frequency modulation sequence, and the results are as follows: Table 1 Comparison of FM sequence prediction results , Table 2 Four evaluation indicators , Where N represents the sample size, and Represent the actual value and predicted value at time n respectively.

[0048] It can be seen that the four evaluation indicators can be detected to be decreased from FM sequence 1, indicating that the method proposed by the present invention can well improve the prediction accuracy. FM sequence 2 illustrates the same problem.

[0049] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A hybrid energy storage frequency modulation instruction prediction method based on error feedback, characterized in that: include: Collect frequency modulation instructions, obtain first data, and perform first screening and second screening; Based on the first data and the results of the first screening and the second screening, find and repair the data outliers to obtain the second data; Decomposing the second data to obtain third data; Decomposing the second data to obtain the third data includes: Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Perform five random decompositions, each generating K subsequences, and perform a first operation after the five random decompositions are completed. The first operation includes: calculating the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; finding the decomposition corresponding to the minimum error coefficient; Take the decomposition corresponding to the minimum error coefficient as the optimal decomposition of this round, and increase the value of p by one; Repeat the first decomposition. When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition. The directions of the five random decompositions in the next round of decomposition are based on the directions of the best decompositions in the previous round; When the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposition subsequence as the final decomposition result, and take the final decomposition result as the third data; Inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; The step of inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction includes: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result, where N is the total length of the subsequence.

2. The hybrid energy storage frequency modulation instruction prediction method based on error feedback according to claim 1 is characterized in that: The step of finding and repairing data outliers based on the first data and the results of the first screening and the second screening to obtain the second data includes: If a point is both a suspicious point in the first screening and a suspicious point in the second screening, the point is identified as an outlier; after the outlier in the first data is repaired, the second data is obtained.

3. The hybrid energy storage frequency modulation instruction prediction method based on error feedback as claimed in claim 2 is characterized in that: The first screening is for size abnormalities: Assume the frequency modulation instruction is P t =[X1,X2,X3,...,X i ,...,X N ], the mean is J=avg(X1,X2,X3,...,X i ,...,X N ), where X1,X2,X3,...,X i ,...,X N All are frequency modulation values; If a value of the frequency modulation instruction is greater than 3 times of J or less than 1 / 3 of J, it is listed as a suspicious point.

4. The hybrid energy storage frequency modulation instruction prediction method based on error feedback according to claim 3 is characterized in that: The second screening is the screening of abnormal changes before and after: Assume that the three consecutive values ​​in the frequency modulation instruction are X i-1 , X i , X i+1 ; If sin(X i-1 -X i ) and sin(X i -X i+1 ) have different signs, and: , then X i Listed as suspicious points.

5. A hybrid energy storage frequency modulation instruction prediction system based on error feedback, using any method as claimed in claim 1 to claim 4, characterized in that: include: A screening module, used for collecting frequency modulation instructions, obtaining first data, and performing first screening and second screening; A repair module, used to find and repair data abnormal values ​​based on the first data and the results of the first screening and the second screening, so as to obtain second data; A decomposition module, used for decomposing the second data to obtain third data; Decomposing the second data to obtain the third data includes: Set the initial value of the iteration round p to 0, set the initial decomposition layer number K, and perform the first decomposition: Perform five random decompositions, each generating K subsequences, and perform a first operation after the five random decompositions are completed. The first operation includes: calculating the error coefficient of each decomposition, expressed as [W1, W2, W3, W4, W5]; finding the decomposition corresponding to the minimum error coefficient; Take the decomposition corresponding to the minimum error coefficient as the optimal decomposition of this round, and increase the value of p by one; Repeat the first decomposition. When the number of iterations p reaches the set value, if K has not reached the set number of layers, increase the value of K by one and continue to repeat the first decomposition. The directions of the five random decompositions in the next round of decomposition are based on the directions of the best decompositions in the previous round; When the number of iterations p reaches the set value and K reaches the set value of the number of layers, find the corresponding decomposition with the lowest decomposition error in the corresponding round, take the decomposition subsequence as the final decomposition result, and take the final decomposition result as the third data; A prediction module, used for inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction; The step of inputting the third data into the convolutional neural network to obtain a prediction result of the frequency modulation instruction includes: The numerical value with a length of 0.9N before each subsequence in the third data is input into the convolutional neural network, the value of the corresponding unknown group is output, and then the value of the unknown group predicted by each subsequence is added to obtain the prediction result, where N is the total length of the subsequence.

6. A computing device comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method described in any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the steps of the method according to any one of claims 1 to 4.

Images