A hybrid energy storage frequency modulation control method and system for predicting frequency modulation commands

Abstract

The present invention relates to the technical field of power grid frequency modulation, and specifically to a hybrid energy storage frequency modulation control method and system for predicting frequency modulation instructions, comprising: obtaining an initial error value by using the difference between the predicted data of a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; obtaining several groups of sinusoidal adjustment sequences and exponential adjustment sequences, and obtaining sinusoidal error sequences and exponential error sequences by using the difference between the predicted data of the several groups of sinusoidal adjustment sequences and the rear frequency modulation instruction sequence respectively; selecting updated data in the front frequency modulation instruction sequence; correcting the updated data to obtain an updated frequency modulation instruction sequence; correcting the number of neurons in a GRU neural network to obtain the corrected number of neurons in the GRU neural network; and obtaining a final frequency modulation instruction sequence, and performing frequency modulation control of the hybrid energy storage using the final frequency modulation instruction sequence. The present invention improves the accuracy of frequency modulation control of the hybrid energy storage.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid frequency modulation, and in particular to a hybrid energy storage frequency modulation control method and system for predicting frequency modulation instructions. BACKGROUND

[0002] Energy storage frequency modulation instruction prediction refers to predicting changes in frequency modulation instructions when using energy storage systems to achieve frequency regulation (frequency modulation) tasks in a power system, thereby optimizing the scheduling and response of the energy storage system. Frequency regulation in a power system is very important, especially as the proportion of renewable energy (such as wind and solar energy) increases, frequency instability increases, and energy storage systems play an increasingly important role in this context.

[0003] Frequency modulation instructions refer to signals sent by power system dispatch centers or automation devices, telling energy storage systems how to adjust charging and discharging power to maintain the stability of the power grid frequency. Energy storage systems need to quickly respond to emergencies and failures according to these frequency modulation instructions to ensure frequency stability.

[0004] Traditional thermal power units require hybrid energy storage (ultra-capacitor and lithium battery) systems to assist in achieving certain requirements; therefore, when there is a difference between the output of the thermal power unit and the frequency modulation instruction, the hybrid energy storage system is needed to assist the thermal power unit to maintain system stability due to the slow response speed of the thermal power unit. Since hybrid energy storage includes both ultra-capacitors and lithium batteries, when using frequency modulation instructions to control hybrid energy storage, the transmission of frequency modulation instruction signals takes time, so the subsequent signal can be predicted by the frequency modulation instruction signal in the historical data to reduce the delay caused by transmission time. However, due to various noise disturbances in the transmission process of the frequency modulation instruction signal, there are errors in the frequency modulation instruction signal obtained from the historical data, so using the signal with errors to predict the subsequent signal will cause a large error, resulting in a decrease in the accuracy of hybrid energy storage frequency modulation control and affecting the stability of the system. SUMMARY

[0005] The present application provides a hybrid energy storage frequency modulation control method and system for predicting frequency modulation instructions to solve the problem of noise interference in frequency modulation instruction signals.

[0006] The purpose of the present application can be achieved by the following technical solutions:

[0007] The first aspect of the present application provides a hybrid energy storage frequency modulation control method for predicting frequency modulation instructions, comprising:

[0008] Obtaining a frequency modulation instruction signal;

[0009] The frequency modulation instruction signal is divided to obtain a frequency modulation instruction sequence; the frequency modulation instruction sequence is then divided into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; an initial error value is obtained by the difference between the predicted data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by a sine function to obtain several groups of sinusoidal adjustment sequences, and a sinusoidal error sequence is obtained by the difference between the predicted data of the several groups of sinusoidal adjustment sequences and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by an exponential function to obtain several groups of exponential adjustment sequences, and an exponential error sequence is obtained by the difference between the predicted data of the several groups of exponential adjustment sequences and the rear frequency modulation instruction sequence;

[0010] Selecting updated data in the previous frequency modulation instruction sequence based on the initial error value, the sine error sequence, and the exponential error sequence; correcting the updated data in the previous frequency modulation instruction sequence based on the sine error sequence and the exponential error sequence to obtain an updated frequency modulation instruction sequence; and correcting the number of neurons in the GRU neural network based on the sine error sequence to obtain a corrected number of neurons in the GRU neural network;

[0011] According to the corrected number of neurons, the updated frequency modulation instruction sequence and the previous frequency modulation instruction sequence are predicted respectively to obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence; then, according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, a sinusoidal error sequence and an exponential error sequence, a final frequency modulation instruction sequence is obtained, and the frequency modulation control of the hybrid energy storage is performed through the final frequency modulation instruction sequence.

[0012] Furthermore, the frequency modulation instruction signal is divided to obtain a frequency modulation instruction sequence; the frequency modulation instruction sequence is then divided into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; and an initial error value is obtained by the difference between the prediction data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence, including:

[0013] At preset time intervals The frequency modulation command signal is divided to obtain a number of data corresponding to the divided frequency modulation command signal, and the data are grouped into a sequence in time order, which is recorded as a frequency modulation command sequence;

[0014] The first The data is composed of the first frequency modulation instruction sequence; the second frequency modulation instruction sequence The data is composed of, followed by a frequency modulation instruction sequence; among them, The percentage factor for the preset division;

[0015] Put the pre-FM instruction sequence into the GRU neural network for prediction to obtain a set of initial prediction sequences;

[0016] The error between the initial prediction sequence and the post-frequency modulation instruction sequence is calculated through MAPE and recorded as the initial error value.

[0017] Furthermore, the method of adjusting the pre-FM instruction sequence several times by using a sine function to obtain several groups of sine adjustment sequences, and obtaining a sine error sequence by the difference between the predicted data of the several groups of sine adjustment sequences and the post-FM instruction sequence, includes:

[0018]

[0019] Where, Indicates the first data in the pre-FM instruction sequence. Indicates the second data in the previous frequency modulation instruction sequence, Indicates the first data, represents the sine function, Indicates the number of all data in the previous FM instruction sequence, Represents the first set of sinusoidal adjustment sequences, Represents the second set of sinusoidal adjustment sequences, Indicates the Group sine adjustment sequence;

[0020] Several groups of sinusoidal adjustment sequences are sequentially put into the GRU neural network for prediction, and several groups of sinusoidal prediction sequences are obtained in turn; through MAPE, the errors between several groups of sinusoidal prediction sequences and post-frequency modulation instruction sequences are calculated respectively to obtain several error values, and these error values ​​are combined into a group of sequences, which are recorded as sinusoidal error sequences.

[0021] Furthermore, the method of adjusting the pre-FM instruction sequence several times by using an exponential function to obtain several groups of exponential adjustment sequences, and obtaining an exponential error sequence by using the differences between the predicted data of the several groups of exponential adjustment sequences and the post-FM instruction sequence, includes:

[0022]

[0023] Where, Indicates the first data in the pre-FM instruction sequence. Indicates the second data in the previous frequency modulation instruction sequence, Indicates the first data, represents a natural constant, Indicates the number of all data in the previous FM instruction sequence, represents the first set of index-adjusted series, represents the second set of index-adjusted series, Indicates the Group index adjustment series;

[0024] The several groups of index adjustment sequences are sequentially put into the GRU neural network for prediction, and several groups of index prediction sequences are sequentially obtained; the errors between the several groups of index prediction sequences and the subsequent frequency modulation instruction sequences are calculated respectively by MAPE, and several error values are obtained, which form a sequence, denoted as an index error sequence.

[0025] Further, the update data in the previous frequency modulation instruction sequence is selected according to the initial error value, the sinusoidal error sequence and the index error sequence; the update data in the previous frequency modulation instruction sequence is corrected according to the sinusoidal error sequence and the index error sequence to obtain an updated frequency modulation instruction sequence, comprising:

[0026] The first error value in the sinusoidal error sequence is denoted as the sinusoidal error value of the first data in the previous frequency modulation instruction sequence; the first error value in the index error sequence is denoted as the index error value of the first data in the previous frequency modulation instruction sequence.

[0027] Find the number of all data in the sinusoidal error sequence whose error value is greater than the initial error value W, denoted as the sinusoidal error number ; find the number of all data in the index error sequence whose error value is greater than the initial error value W, denoted as the index error number .

[0028] The data in the previous frequency modulation instruction sequence whose sinusoidal error value and index error value are both greater than the initial error value W are denoted as update data.

[0029]

[0030] In the formula, denotes the first data in the previous frequency modulation instruction sequence, denotes the sinusoidal error value of the first data in the previous frequency modulation instruction sequence, denotes the index error value of the first data in the previous frequency modulation instruction sequence, denotes the correction result value of the first data in the previous frequency modulation instruction sequence. denotes the data sequence number in the previous frequency modulation instruction sequence, denotes the absolute value symbol, denotes function for data normalization; denotes the factorial symbol, denotes the correction result value of the first data in the previous frequency modulation instruction sequence.

[0031] ​​​​​​​Then, the corrected result values ​​of all updated data in the previous frequency modulation instruction sequence replace the original data values ​​to form a new sequence, which is recorded as the updated frequency modulation instruction sequence.

[0032] Furthermore, the method of correcting the number of neurons in the GRU neural network according to the sinusoidal error sequence to obtain the corrected number of neurons in the GRU neural network includes:

[0033]

[0034] Where, Indicates the sine error value of the first data in the previous frequency modulation instruction sequence, Indicates the sine error value of the second data in the previous frequency modulation instruction sequence, Indicates the first The sine error value of the data, Indicates the first The sine error value of the data, represents the maximum value function, Represents the revised number of neurons in the GRU neural network, Express Round to the nearest integer.

[0035] Furthermore, obtaining a final frequency modulation instruction sequence according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, the sinusoidal error sequence and the exponential error sequence includes:

[0036]

[0037] Where, Indicates the first The sine error value of the data, Indicates the first The exponential error value of the data, Indicates the number of all data in the previous FM instruction sequence, Indicates the first predicted frequency modulation sequence data, Indicates the first data, Indicates the first data, represents the activation function, represents the hyperbolic tangent function.

[0038] A second aspect of the present invention is to provide a hybrid energy storage frequency modulation control system for predicting frequency modulation instructions, comprising:

[0039] The data acquisition module is used for acquiring the frequency modulation instruction signal.

[0040] The error analysis module is used for dividing the frequency modulation instruction signal to obtain a frequency modulation instruction sequence; dividing the frequency modulation instruction sequence into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; obtaining an initial error value by predicting the difference between the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence; obtaining a sine error sequence by adjusting the front frequency modulation instruction sequence several times through a sine function, and predicting the difference between the several sine adjustment sequences and the rear frequency modulation instruction sequence; obtaining an exponential error sequence by adjusting the front frequency modulation instruction sequence several times through an exponential function, and predicting the difference between the several exponential adjustment sequences and the rear frequency modulation instruction sequence.

[0041] The correction adjustment module is used for selecting updated data in the front frequency modulation instruction sequence according to the initial error value, the sine error sequence and the exponential error sequence; correcting the updated data in the front frequency modulation instruction sequence according to the sine error sequence and the exponential error sequence to obtain an updated frequency modulation instruction sequence; and correcting the number of neurons in the GRU neural network according to the sine error sequence to obtain the number of neurons in the GRU neural network after correction.

[0042] The prediction control module is used for predicting the updated frequency modulation instruction sequence and the front frequency modulation instruction sequence according to the number of neurons after correction to obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence; and obtaining a final frequency modulation instruction sequence according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, the sine error sequence and the exponential error sequence, and performing frequency modulation control of hybrid energy storage through the final frequency modulation instruction sequence.

[0043] The third aspect of the present application provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to realize the hybrid energy storage frequency modulation control method of the predicted frequency modulation instruction.

[0044] The fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the hybrid energy storage frequency modulation control method of the predicted frequency modulation instruction.

[0045] Compared with the prior art, the beneficial effects of the present application are: obtaining the frequency modulation instruction sequence and the front and rear frequency modulation instruction sequences; obtaining the initial error value by predicting the difference between the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence, thereby improving the accuracy of the initial error analysis; obtaining a plurality of sets of sine adjustment sequences by adjusting the front frequency modulation instruction sequence a plurality of times through a sine function, and obtaining the sine error sequence by predicting the difference between the plurality of sets of sine adjustment sequences and the rear frequency modulation instruction sequence; obtaining a plurality of sets of exponential adjustment sequences by adjusting the front frequency modulation instruction sequence a plurality of times through an exponential function, and obtaining the exponential error sequence by predicting the difference between the plurality of sets of exponential adjustment sequences and the rear frequency modulation instruction sequence; adjusting through two functions and obtaining the error to improve the accuracy of error analysis under different adjustments; selecting the updated data in the front frequency modulation instruction sequence according to the initial error value, the sine error sequence and the exponential error sequence; correcting the updated data in the front frequency modulation instruction sequence according to the sine error sequence and the exponential error sequence to obtain the updated frequency modulation instruction sequence, thereby improving the accuracy of the correction of the frequency modulation instruction data; correcting the number of neurons in the GRU neural network according to the sine error sequence to obtain the number of neurons in the GRU neural network after correction, thereby improving the accuracy of the analysis of the number of neurons; predicting the updated frequency modulation instruction sequence and the front frequency modulation instruction sequence according to the number of neurons after correction to obtain the first predicted frequency modulation sequence and the second predicted frequency modulation sequence; and obtaining the final frequency modulation instruction sequence according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, the sine error sequence and the exponential error sequence, thereby performing frequency modulation control of the hybrid energy storage through the final frequency modulation instruction sequence, reducing the interference caused by noise, improving the accuracy of the frequency modulation control of the hybrid energy storage, and improving the stability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0047] Figure 1 A step flowchart of a hybrid energy storage frequency modulation control method for predicting frequency modulation instructions is provided for the present application.

[0048] Figure 2 A module flowchart of a hybrid energy storage frequency modulation control system for predicting frequency modulation instructions is provided for the present application. DETAILED DESCRIPTION

[0049] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative effort should belong to the protection scope of the present application.

[0050] It should be noted that the terms "first", "second" and the like in the description and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in other than the order illustrated or described herein. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a list of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0051] In view of the problems in the background art, a hybrid energy storage frequency modulation control method and system for predicting frequency modulation instructions are designed, which has important practical significance.

[0052] As shown in Figure 1 The first aspect of the present application provides a hybrid energy storage frequency modulation control method for predicting frequency modulation instructions, comprising the following steps:

[0053] Step S001: Collecting frequency modulation instruction signals.

[0054] It should be noted that when the power supply and demand balance deviates (for example, load changes, power fluctuations, etc.), the dispatching center or automatic system of the power system generates frequency modulation instructions according to the frequency fluctuation of the system, load demand and power generation, and maintains frequency stability through the frequency modulation instructions, which can respond to emergencies and reduce the possibility of failure.

[0055] Specifically, the frequency modulation instruction signals of the power system dispatching center in the preset time length hours before the current time are obtained, wherein in the present embodiment, the preset time length is not limited specifically, and the implementer can determine it according to the specific circumstances. is not limited specifically, and the implementer can determine it according to the specific circumstances.

[0056] At this point, the frequency modulation instruction signals are obtained.

[0057] Step S002: dividing the frequency modulation instruction signal to obtain a frequency modulation instruction sequence, a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; obtaining an initial error value by predicting the difference between the data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence; obtaining a sine error sequence by predicting the difference between the data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence through a plurality of sine adjustment sequences obtained by adjusting the front frequency modulation instruction sequence through a sine function for a plurality of times; and obtaining an exponential error sequence by predicting the difference between the data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence through a plurality of exponential adjustment sequences obtained by adjusting the front frequency modulation instruction sequence through an exponential function for a plurality of times.

[0058] It should be noted that, since the distribution of data in the frequency modulation instruction signal can reflect the change characteristics of the signal data, and the change fluctuation difference of the signal data can reflect the error characteristics of the data, the frequency modulation instruction signal is divided into data points to analyze the data error situation.

[0059] Specifically, the frequency modulation instruction signal is divided at a preset time interval to obtain a plurality of data corresponding to the frequency modulation instruction signal after division, and the plurality of data is arranged in time sequence to form a sequence, which is denoted as a frequency modulation instruction sequence. In this embodiment, the preset time interval is 1 second, and the preset time interval is not specifically limited and can be determined by the implementer according to the specific situation.

[0060] Thus, the frequency modulation instruction sequence is obtained.

[0061] It should be noted that, in order to analyze the error of the data affected by noise in the collected frequency modulation instruction signal when the data is predicted, the frequency modulation instruction sequence is divided into two parts, and the difference between the predicted data of the former part and the data of the latter part after division is analyzed.

[0062] Specifically, the data in the front of the frequency modulation instruction sequence is arranged to form a front frequency modulation instruction sequence, and the data in the rear of the frequency modulation instruction sequence is arranged to form a rear frequency modulation instruction sequence. In this embodiment, the preset division percentage factor is 50%.

[0063] In this embodiment, the preset division percentage factor is 50%, and the preset division percentage factor is not specifically limited and can be determined by the implementer according to the specific situation.

[0064] Put the front frequency modulation instruction sequence into the GRU (Gated Recurrent Unit, Gated Recurrent Unit) neural network for prediction to obtain an initial prediction sequence; wherein the number of data obtained by prediction is the same as the number of data in the rear frequency modulation instruction sequence.

[0065] Calculate the error between the initial prediction sequence and the rear frequency modulation instruction sequence by MAPE (Mean Absolute Percentage Error, Mean Absolute Percentage Error), denoted as the initial error value W; wherein MAPE is a known technology, which will not be described in detail here.

[0066] It should be noted that in order to analyze the degree of interference of the data in the frequency modulation instruction sequence by noise, each data in the front frequency modulation instruction sequence is adjusted, and the adjusted data is used to predict the data, and then the adjusted data is used to analyze the adjustment of each data according to the predicted data and the rear frequency modulation instruction sequence.

[0067] Specifically, the data in the front frequency modulation instruction sequence is adjusted by a sine function to obtain a plurality of sine adjustment sequences; the plurality of sine adjustment sequences are specifically represented as:

[0068]

[0069] In the formula, represents the first data in the front frequency modulation instruction sequence, represents the second data in the front frequency modulation instruction sequence, represents the data in the front frequency modulation instruction sequence, represents the sine function, represents the number of all data in the front frequency modulation instruction sequence, represents the first sine adjustment sequence, represents the second sine adjustment sequence, represents the sine adjustment sequence.

[0070] Put the plurality of sine adjustment sequences into the GRU neural network for prediction to obtain a plurality of sine prediction sequences; calculate the error between the plurality of sine prediction sequences and the rear frequency modulation instruction sequence by MAPE, respectively, to obtain a plurality of error values, and form a sequence with the plurality of error values, denoted as a sine error sequence.

[0071] At this point, the sine error sequence is obtained.

[0072] Adjust the data in the front frequency modulation instruction sequence by an exponential function to obtain a plurality of exponential adjustment sequences; the plurality of exponential adjustment sequences are specifically represented as:

[0073]

[0074] Where, Indicates the first data in the pre-FM instruction sequence. Indicates the second data in the previous frequency modulation instruction sequence, Indicates the first data, represents a natural constant, Indicates the number of all data in the previous FM instruction sequence, represents the first set of index-adjusted series, represents the second set of index-adjusted series, Indicates the Group index adjustment series.

[0075] Several groups of exponential adjustment sequences are sequentially put into the GRU neural network for prediction, and several groups of exponential prediction sequences are obtained in turn; through MAPE, the errors between several groups of exponential prediction sequences and post-frequency modulation instruction sequences are calculated respectively to obtain several error values, and the several error values ​​are combined into a group of sequences, which are recorded as exponential error sequences.

[0076] At this point, the exponential error sequence is obtained.

[0077] Step S003: Select the updated data in the previous frequency modulation instruction sequence according to the initial error value, the sine error sequence and the exponential error sequence; correct the updated data in the previous frequency modulation instruction sequence according to the sine error sequence and the exponential error sequence to obtain an updated frequency modulation instruction sequence; correct the number of neurons in the GRU neural network according to the sine error sequence to obtain the corrected number of neurons in the GRU neural network.

[0078] It should be noted that both the sinusoidal error sequence and the exponential error sequence are obtained by adjusting each data in the previous frequency modulation instruction sequence and then performing prediction and error calculation to obtain the adjustment result of each data. The larger the error result, the less ideal the adjustment of the data at the same position. Therefore, correction is performed by analyzing the error.

[0079] Specifically, the first The error value is recorded as the first error in the previous frequency modulation instruction sequence. The sine error value of the data; the first The error value is recorded as the first error in the previous frequency modulation instruction sequence. The exponential error value of the data.

[0080] Find the number of all data in the sine error sequence whose error value is greater than the initial error value W, and record it as the number of sine errors find all the data whose error value is greater than the initial error value W in the exponential error sequence, and record the number as the exponential error number .

[0081] record the data whose sine error value and exponential error value are both greater than the initial error value W as the update data.

[0082] According to the sine error sequence and the exponential error sequence, the sine error number and the exponential error number, all the update data in the previous frequency modulation instruction sequence are corrected; the specific formula is:

[0083]

[0084] In the formula, indicates the sine error number, indicates the exponential error number, indicates the first data in the previous frequency modulation instruction sequence, indicates the second data in the previous frequency modulation instruction sequence, indicates the sine error value of the first data in the previous frequency modulation instruction sequence, indicates the exponential error value of the first data in the previous frequency modulation instruction sequence, indicates the data sequence number in the previous frequency modulation instruction sequence, indicates the absolute value symbol, indicates the function, which is used for data normalization; indicates the factorial symbol, indicates the correction result value of the first data in the previous frequency modulation instruction sequence. At this point, the correction result values of all the update data in the previous frequency modulation instruction sequence are obtained. Then, the correction result values of all the update data in the previous frequency modulation instruction sequence are replaced by the original data values to form a new sequence, which is recorded as the updated frequency modulation instruction sequence.

[0085] At this point, the updated frequency modulation instruction sequence is obtained.

[0086] Then, the correction result values of all the update data in the previous frequency modulation instruction sequence are replaced by the original data values to form a new sequence, which is recorded as the updated frequency modulation instruction sequence.

[0087] At this point, the updated frequency modulation instruction sequence is obtained.

[0088] According to the sine error sequence, the number of neurons in the GRU neural network is corrected, and the specific correction process is represented by the formula:

[0089]

[0090] In the formula, indicates the sine error value of the first data in the previous frequency modulation instruction sequence, indicates the sine error value of the second data in the previous frequency modulation instruction sequence, ​Indicates the first The sine error value of the data, Indicates the first The sine error value of the data, represents the maximum value function, Represents the revised number of neurons in the GRU neural network, Express Round to the nearest integer.

[0091] Step S004: According to the corrected number of neurons, the updated frequency modulation instruction sequence and the previous frequency modulation instruction sequence are predicted respectively to obtain a final frequency modulation instruction sequence, and the frequency modulation control of the hybrid energy storage is performed through the final frequency modulation instruction sequence.

[0092] According to the revised number of neurons in the GRU neural network, the updated frequency modulation instruction sequence and the previous frequency modulation instruction sequence are predicted respectively to obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence.

[0093] Then, according to the first predicted frequency modulation sequence, the second predicted frequency modulation sequence, the sine error sequence and the exponential error sequence, the final frequency modulation instruction sequence is obtained; specifically, it is expressed as follows:

[0094]

[0095] Where, Indicates the first The sine error value of the data, Indicates the first The exponential error value of the data, Indicates the number of all data in the previous FM instruction sequence, Indicates the first predicted frequency modulation sequence data, Indicates the first data, Indicates the first data, represents the activation function, represents the hyperbolic tangent function.

[0096] At this point, the final frequency modulation instruction sequence is obtained.

[0097] The frequency modulation control of the hybrid energy storage is carried out through the final frequency modulation instruction sequence.

[0098] like Figure 2 As shown, the second aspect of the present invention is to provide a hybrid energy storage frequency modulation control system for predicting frequency modulation instructions, including the following modules:

[0099] The data acquisition module 101 is configured to acquire the frequency modulation instruction signal.

[0100] The error analysis module 102 is configured to divide the frequency modulation instruction signal to obtain a frequency modulation instruction sequence, cut the frequency modulation instruction sequence into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence, obtain an initial error value by predicting the difference between the data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence, obtain a sine error sequence by adjusting the front frequency modulation instruction sequence for several times through a sine function and predicting the difference between the data of the several sine adjustment sequences and the rear frequency modulation instruction sequence, and obtain an exponential error sequence by adjusting the front frequency modulation instruction sequence for several times through an exponential function and predicting the difference between the data of the several exponential adjustment sequences and the rear frequency modulation instruction sequence.

[0101] The correction adjustment module 103 is configured to select updated data in the front frequency modulation instruction sequence according to the initial error value, the sine error sequence and the exponential error sequence, correct the updated data in the front frequency modulation instruction sequence according to the sine error sequence and the exponential error sequence to obtain an updated frequency modulation instruction sequence, and correct the number of neurons in the GRU neural network according to the sine error sequence to obtain the number of neurons in the GRU neural network after correction.

[0102] The prediction control module 104 is configured to predict the updated frequency modulation instruction sequence and the front frequency modulation instruction sequence according to the number of neurons after correction to obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence, and obtain a final frequency modulation instruction sequence according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, the sine error sequence and the exponential error sequence, and perform frequency modulation control of hybrid energy storage through the final frequency modulation instruction sequence.

[0103] The third aspect of the present application provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements a hybrid energy storage frequency modulation control method for predicting a frequency modulation instruction when executing the computer program.

[0104] The fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the computer program implements a hybrid energy storage frequency modulation control method for predicting a frequency modulation instruction when executed by a processor.

[0105] Those skilled in the art will appreciate that embodiments of the application can be devised for a method, a system, or a computer program product. Accordingly, the present application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage etc.) embodying computer readable program code.

[0106] The present application is described in reference to the flowchart and / or block diagram of the method, system, and computer program product according to embodiments of the application. It will be understood that each block of the flowchart and / or block diagram, and combinations of blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processing system, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for carrying out each of the one or more functions specified in the flowchart and / or block diagram block or blocks.

[0107] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for carrying out each of the one or more functions specified in the flowchart and / or block diagram block or blocks.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified in the flowchart and / or block diagram block or blocks. Figure 1 means for carrying out each of the one or more functions specified in the flowchart and / or block diagram block or blocks.

[0109] Finally, it should be noted that the above-mentioned embodiments are merely intended for describing the technical solutions of the present application, but not for limiting it. Although the present application is described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modification or equivalent replacement without departing from the spirit and scope of the present application should be covered within the protection scope of the present application.

Claims

1. A hybrid energy storage frequency modulation control method for predicting frequency modulation instructions, characterized in that: include: Obtain frequency modulation command signal; Dividing the frequency modulation command signal to obtain a frequency modulation command sequence; The frequency modulation instruction sequence is then divided into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; an initial error value is obtained by the difference between the predicted data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by a sine function to obtain several groups of sinusoidal adjustment sequences, and a sinusoidal error sequence is obtained by the difference between the predicted data of the several groups of sinusoidal adjustment sequences and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by an exponential function to obtain several groups of exponential adjustment sequences, and an exponential error sequence is obtained by the difference between the predicted data of the several groups of exponential adjustment sequences and the rear frequency modulation instruction sequence; Selecting updated data in the previous frequency modulation instruction sequence based on the initial error value, the sine error sequence, and the exponential error sequence; correcting the updated data in the previous frequency modulation instruction sequence based on the sine error sequence and the exponential error sequence to obtain an updated frequency modulation instruction sequence; and correcting the number of neurons in the GRU neural network based on the sine error sequence to obtain a corrected number of neurons in the GRU neural network; According to the corrected number of neurons, the updated frequency modulation instruction sequence and the previous frequency modulation instruction sequence are predicted respectively to obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence; then, according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, a sinusoidal error sequence and an exponential error sequence, a final frequency modulation instruction sequence is obtained, and the frequency modulation control of the hybrid energy storage is performed through the final frequency modulation instruction sequence.

2. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The frequency modulation instruction signal is divided to obtain a frequency modulation instruction sequence; the frequency modulation instruction sequence is then divided into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; The initial error value is obtained by the difference between the predicted data of the pre-FM instruction sequence and the post-FM instruction sequence, including: At preset time intervals The frequency modulation command signal is divided to obtain a number of data corresponding to the divided frequency modulation command signal, and the data are grouped into a sequence in time order, which is recorded as a frequency modulation command sequence; The first The data is composed of the first frequency modulation instruction sequence; the second frequency modulation instruction sequence The data is composed of, followed by a frequency modulation instruction sequence; among them, The percentage factor for the preset division; Put the pre-FM instruction sequence into the GRU neural network for prediction to obtain a set of initial prediction sequences; The error between the initial prediction sequence and the post-frequency modulation instruction sequence is calculated through MAPE and recorded as the initial error value.

3. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The method of adjusting the pre-FM instruction sequence several times by using a sine function to obtain several groups of sine adjustment sequences, and obtaining a sine error sequence by using the difference between the predicted data of the several groups of sine adjustment sequences and the post-FM instruction sequence, includes: Where, Indicates the first data in the pre-FM instruction sequence. Indicates the second data in the previous frequency modulation instruction sequence, Indicates the first data, represents the sine function, Indicates the number of all data in the previous FM instruction sequence, Represents the first set of sinusoidal adjustment sequences, Represents the second set of sinusoidal adjustment sequences, Indicates the Group sine adjustment sequence; Several groups of sinusoidal adjustment sequences are sequentially put into the GRU neural network for prediction, and several groups of sinusoidal prediction sequences are obtained in turn; through MAPE, the errors between several groups of sinusoidal prediction sequences and post-frequency modulation instruction sequences are calculated respectively to obtain several error values, and these error values ​​are combined into a group of sequences, which are recorded as sinusoidal error sequences.

4. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The method of adjusting the pre-FM instruction sequence several times by using an exponential function to obtain several groups of exponential adjustment sequences, and obtaining an exponential error sequence by using the differences between the predicted data of the several groups of exponential adjustment sequences and the post-FM instruction sequence, includes: Where, Indicates the first data in the pre-FM instruction sequence. Indicates the second data in the previous frequency modulation instruction sequence, Indicates the first data, represents a natural constant, Indicates the number of all data in the previous FM instruction sequence, represents the first set of index-adjusted series, represents the second set of index-adjusted series, Indicates the Group index adjustment series; Several groups of exponential adjustment sequences are sequentially put into the GRU neural network for prediction, and several groups of exponential prediction sequences are obtained in turn; through MAPE, the errors between several groups of exponential prediction sequences and post-frequency modulation instruction sequences are calculated respectively to obtain several error values, and the several error values ​​are combined into a group of sequences, which are recorded as exponential error sequences.

5. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The updating data in the pre-FM instruction sequence is selected according to the initial error value, the sinusoidal error sequence and the exponential error sequence; According to the sinusoidal error sequence and the exponential error sequence, the update data in the previous frequency modulation instruction sequence is corrected to obtain an updated frequency modulation instruction sequence, including: The first The error value is recorded as the first error value in the previous frequency modulation instruction sequence. The sine error value of the data; the first The error value is recorded as the first error in the previous frequency modulation instruction sequence. The exponential error value of each data; Find the number of all data in the sine error sequence whose error value is greater than the initial error value W, and record it as the number of sine errors ; Find the number of all data in the exponential error sequence whose error value is greater than the initial error value W, and record it as the number of exponential errors ; The data whose sine error value and exponential error value of each data in the previous frequency modulation instruction sequence are both greater than the initial error value W is recorded as the updated data; Where, Indicates the first data, Indicates the first The sine error value of the data, Indicates the first The exponential error value of the data, Indicates the data sequence number in the previous frequency modulation instruction sequence, Indicates the absolute value symbol, express Function, used for data normalization; represents the factorial symbol, Indicates the first The corrected result value of each data; Then, the corrected result values ​​of all updated data in the previous frequency modulation instruction sequence replace the original data values ​​to form a new sequence, which is recorded as the updated frequency modulation instruction sequence.

6. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The method of correcting the number of neurons in the GRU neural network according to the sinusoidal error sequence to obtain the corrected number of neurons in the GRU neural network includes: Where, Indicates the sine error value of the first data in the previous frequency modulation instruction sequence, Indicates the sine error value of the second data in the previous frequency modulation instruction sequence, Indicates the first The sine error value of the data, Indicates the first The sine error value of the data, represents the maximum value function, Represents the revised number of neurons in the GRU neural network, Express Round to the nearest integer.

7. The hybrid energy storage frequency modulation control method for predicting frequency modulation instructions according to claim 1, characterized in that: The method of obtaining a final frequency modulation instruction sequence according to the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, the sinusoidal error sequence and the exponential error sequence comprises: Where, Indicates the first The sine error value of the data, Indicates the first The exponential error value of the data, Indicates the number of all data in the previous FM instruction sequence, Indicates the first predicted frequency modulation sequence data, Indicates the first data, Indicates the first data, represents the activation function, represents the hyperbolic tangent function.

8. A hybrid energy storage frequency modulation control system for predicting frequency modulation instructions, characterized in that: include: Data acquisition module: used to obtain frequency modulation command signals; Error analysis module: used to divide the FM command signal and obtain the FM command sequence; The frequency modulation instruction sequence is then divided into a front frequency modulation instruction sequence and a rear frequency modulation instruction sequence; an initial error value is obtained by the difference between the predicted data of the front frequency modulation instruction sequence and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by a sine function to obtain several groups of sinusoidal adjustment sequences, and a sinusoidal error sequence is obtained by the difference between the predicted data of the several groups of sinusoidal adjustment sequences and the rear frequency modulation instruction sequence; the front frequency modulation instruction sequence is adjusted several times by an exponential function to obtain several groups of exponential adjustment sequences, and an exponential error sequence is obtained by the difference between the predicted data of the several groups of exponential adjustment sequences and the rear frequency modulation instruction sequence; Correction and adjustment module: used to select the update data in the previous frequency modulation instruction sequence according to the initial error value, the sine error sequence and the exponential error sequence; correct the update data in the previous frequency modulation instruction sequence according to the sine error sequence and the exponential error sequence to obtain the updated frequency modulation instruction sequence; correct the number of neurons in the GRU neural network according to the sine error sequence to obtain the corrected number of neurons in the GRU neural network; Prediction control module: used to predict the updated frequency modulation instruction sequence and the previous frequency modulation instruction sequence according to the corrected number of neurons, and obtain a first predicted frequency modulation sequence and a second predicted frequency modulation sequence; then obtain a final frequency modulation instruction sequence based on the first predicted frequency modulation sequence and the second predicted frequency modulation sequence, a sinusoidal error sequence and an exponential error sequence, and perform frequency modulation control of the hybrid energy storage through the final frequency modulation instruction sequence.

9. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the hybrid energy storage frequency modulation control method for predicting frequency modulation instructions as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the hybrid energy storage frequency modulation control method of predicting frequency modulation instructions according to any one of claims 1 to 7 is implemented.

Images