A method and system for predicting frequency modulation instructions for optimizing molten salt heat storage systems

Abstract

The present invention relates to the technical field of power grid frequency regulation, and specifically to a method and system for optimizing frequency regulation instruction prediction for a molten salt heat storage system, comprising: clustering all sample data points to obtain a plurality of clusters; adjusting abnormal sample data points among all sample data points based on the degree of dispersion of all data in the plurality of clusters to obtain target data points; performing dimensionality reduction on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; obtaining a plurality of reference data points based on frequency regulation instruction signal data at a plurality of moments and various reference feature data; obtaining the corresponding errors of each decision tree under different sample sizes based on the plurality of reference data points; obtaining the optimal sample size of each decision tree based on the trend change of the corresponding errors under different sample sizes of each decision tree, and predicting instruction signal data. The present invention improves the accuracy of frequency regulation instruction prediction and reduces the amount of computation.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid frequency regulation, and in particular to a method and system for predicting frequency regulation instructions for optimizing a molten salt heat storage system. Background Art

[0002] The molten salt heat storage and peak-shaving frequency modulation system combines molten salt energy storage technology with the peak-shaving and frequency modulation capabilities of the power system, balancing fluctuations in energy supply and demand and improving grid stability. The molten salt heat storage and peak-shaving frequency modulation system is primarily used to store excess electricity and release the stored energy during peak demand periods, thereby balancing grid load fluctuations. Molten salt heat storage heats molten salt to store thermal energy; when needed, the molten salt can release the thermal energy to drive thermoelectric conversion equipment (such as steam turbines and combined heat and power generation) to generate electricity.

[0003] To ensure grid stability and reliability and address issues such as load and power generation fluctuations, unstable renewable energy generation, and frequency fluctuations, frequency modulation command signal prediction allows for pre-emptive adjustments to generator output and energy storage systems to avoid frequency anomalies, system overloads, and large-scale power outages. This improves grid stability, reduces operating costs, and enhances the grid's ability to respond to emergencies.

[0004] Existing methods for predicting frequency modulation command signals can use a random forest algorithm to construct a decision tree. However, in conventional prediction processes, the amount of data and the depth of each decision tree are manually set. When the amount of data per decision tree is small, the prediction results may be poor, reducing prediction accuracy. When the amount of data per decision tree is large, while prediction accuracy improves, the computational complexity also increases. Furthermore, the large amount of data per decision tree causes the model to memorize every detail of historical data, resulting in poor generalization when new data is presented. Furthermore, a deeper decision tree can lead to overfitting, while a smaller depth can lead to underfitting. By properly controlling the depth of the decision tree, a balance can be found between avoiding overfitting and underfitting. Summary of the Invention

[0005] The present invention provides a method and system for optimizing the prediction of frequency modulation instructions for a molten salt heat storage system, which is used to solve the problems of unstable prediction accuracy, increased calculation complexity, and overfitting or underfitting caused by improper setting of decision tree data volume and depth in existing methods.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A first aspect of the present invention is to provide a method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system, comprising:

[0008] Continuously obtain frequency modulation command signal data at several moments and various characteristic data that affect the stability of the power system; obtain several sample data points through the frequency modulation command signal data and various characteristic data at several moments;

[0009] All sample data points are clustered to obtain several clusters; through the degree of dispersion of all data in several clusters, abnormal sample data points are filtered out from all sample data points to obtain target sample points; all target sample points are interpolated to obtain target data points; all target data points are subjected to dimensionality reduction on the time series data of each dimension corresponding to the target data points to obtain reference features after dimensionality reduction; several reference data points are obtained through the frequency modulation command signal data at several moments and various reference feature data;

[0010] Through several reference data points, the corresponding errors of each decision tree under different sample sizes are obtained; through the trend changes of the corresponding errors of each decision tree under different sample sizes, the optimal sample size of each decision tree is obtained;

[0011] According to the optimal sample size of each decision tree, the command signal data is predicted using the random forest algorithm.

[0012] Furthermore, the frequency modulation command signal data and various characteristic data at several moments are used to obtain several sample data points, including:

[0013] By combining the data of all dimensions at each moment into a sample data point, a number of sample data points can be obtained; wherein all dimensions include all feature data and frequency modulation command signal data.

[0014] Furthermore, all sample data points are clustered to obtain several clusters, including:

[0015] All sample data points are clustered using the K-means clustering algorithm to obtain several clusters; the number of clusters is obtained by the elbow method.

[0016] Furthermore, the method of filtering out abnormal sample data points from all sample data points by the degree of dispersion of all data in a plurality of clusters to obtain target sample points; and interpolating all target sample points to obtain target data points includes:

[0017] The outlier factor of each sample data point is obtained by calculating the LOF algorithm; the mean of the outlier factors of all sample data points in each cluster is recorded as the noise outlier factor of each cluster; all sample data points corresponding to the cluster with the largest noise outlier factor among all clusters are recorded as outlier sample data points; outlier sample data points are screened out from all sample data points, and the remaining sample data points are recorded as target sample points;

[0018] Interpolation is performed on all target sample points using a linear interpolation method to obtain interpolated data points; the interpolated data points and all target sample points are combined to form target data points.

[0019] Furthermore, the dimensionality reduction is performed on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; and several reference data points are obtained through the frequency modulation command signal data at several moments and various reference feature data, including:

[0020] According to the time series data of each dimension corresponding to all target data points, the PCA algorithm is used to reduce the dimension, and the features after dimensionality reduction are obtained and recorded as reference features;

[0021] By combining all reference characteristic data and frequency modulation command signal data at each moment into a reference data point, several reference data points can be obtained.

[0022] Furthermore, the error corresponding to each decision tree under different sample sizes is obtained by using a number of reference data points, including:

[0023] Sort several reference data points in chronological order to form a set of sequences, which are recorded as data point sequences; The reference data points form a sequence, which is recorded as the previous data point sequence; the latter data point sequence The reference data points of the posterior data point sequence are combined into a sequence, which is recorded as the posterior data point sequence; the frequency modulation command signal data corresponding to all reference data points in the posterior data point sequence are combined into a set of sequences, which is recorded as the posterior command signal data sequence; wherein, The percentage factor for the preset division;

[0024] Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision tree is used to predict the FM command signal data, and the predicted FM command signal data are organized into a set of sequences in chronological order, which is recorded as the first predicted command signal data sequence; the error between the first predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE, which is recorded as the first error. ; The first error The sample size of each decision tree is The corresponding error when

[0025] Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision tree is used to predict the FM command signal data, and the predicted FM command signal data are organized into a set of sequences in chronological order, which are recorded as the second predicted command signal data sequence; the error between the second predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE, which is recorded as the second error. ; The second error The sample size of each decision tree is The corresponding error when

[0026] And so on, the corresponding errors of each decision tree under different sample sizes are obtained in turn;

[0027] in, Indicates the number of preset decision trees, Express Round up; Indicates the number of all data points in the previous data point sequence.

[0028] Furthermore, obtaining the optimal sample size for each decision tree by analyzing the trend change of the error corresponding to different sample sizes of each decision tree includes:

[0029] A reference coordinate system is constructed with the sample size of each decision tree as the horizontal axis and the error corresponding to the sample size of each decision tree as the vertical axis; the different sample sizes and corresponding errors of each decision tree are mapped in the reference coordinate system to obtain several coordinate points; a curve fitting is performed on the several coordinate points using the least squares method to obtain an error fitting curve;

[0030] Get the inflection point where the second-order derivative of the function corresponding to the error fitting curve is equal to 0, and As the optimal sample size for each decision tree; Indicates the horizontal value of the inflection point, Express Round up.

[0031] A second aspect of the present invention is to provide a frequency modulation instruction prediction system for optimizing a molten salt thermal storage system, comprising:

[0032] Data acquisition module: used to continuously acquire frequency modulation command signal data at several moments and various characteristic data that affect the stability of the power system; the various characteristic data include humidity, temperature, wind speed and load; obtain a number of sample data points through the frequency modulation command signal data and various characteristic data at several moments;

[0033] Data processing module: used to cluster all sample data points to obtain several clusters; filter out abnormal sample data points from all sample data points through the degree of dispersion of all data in several clusters to obtain target sample points; interpolate all target sample points to obtain target data points; perform dimensionality reduction on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; obtain several reference data points through the frequency modulation command signal data and various reference feature data at several moments;

[0034] Error analysis module: used to obtain the corresponding error of each decision tree under different sample sizes through several reference data points; and obtain the optimal sample size of each decision tree through the trend change of the error corresponding to different sample sizes of each decision tree;

[0035] Prediction module: used to predict command signal data using the random forest algorithm based on the optimal sample size of each decision tree.

[0036] The third aspect of the present invention is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the computer program, the method for predicting frequency modulation instructions for optimizing a molten salt heat storage system is implemented.

[0037] A fourth aspect of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for predicting frequency modulation instructions for optimizing a molten salt heat storage system is implemented.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows: clustering all sample data points to obtain several clusters; filtering out abnormal sample data points from all sample data points through the degree of discreteness of all data in several clusters to obtain target sample points; interpolating all target sample points to obtain target data points and reduce the influence of noise data; performing dimensionality reduction on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; obtaining several reference data points through the frequency modulation instruction signal data and various reference feature data at several moments, and reducing the amount of calculation through dimensionality reduction; obtaining the corresponding errors of each decision tree under different sample sizes through several reference data points; obtaining the optimal sample size of each decision tree through the trend change of the errors corresponding to different sample sizes of each decision tree; predicting the instruction signal data through the random forest algorithm based on the optimal sample size of each decision tree, thereby improving the accuracy of frequency modulation instruction prediction and reducing the amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 The present invention provides a step flow diagram of a method for optimizing the frequency modulation instruction prediction of a molten salt heat storage system;

[0041] Figure 2 The present invention provides a module flow diagram for optimizing the frequency modulation instruction prediction system of a molten salt heat storage system. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0043] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products or devices.

[0044] In response to the problems existing in the background technology, a method and system for predicting the frequency modulation instructions for optimizing the molten salt heat storage system was studied and designed, which has important practical significance.

[0045] like Figure 1 As shown, the first aspect of the present invention is to provide a method for predicting frequency modulation instructions for optimizing a molten salt heat storage system, comprising the following steps:

[0046] Step S001: continuously collecting frequency modulation command signal data and various characteristic data affecting the stability of the power system at several moments.

[0047] It should be noted that when there is a deviation in the balance between power supply and demand (such as load changes, power generation fluctuations, etc.), the power system's dispatch center or automation system generates frequency regulation instructions based on the system's frequency fluctuations, load demand, and power generation conditions. These frequency regulation instructions maintain frequency stability, respond to emergencies, and reduce the possibility of failures.

[0048] It should be further explained that, since the power system is easily disturbed by the environment, weather and other conditions during normal power generation, the power system has poor stability when disturbed. When the power system is disturbed, in order to maintain the stability of the system, the molten salt heat storage peak and frequency regulation system will perform frequency and peak regulation to maintain the stability of the power system through frequency and peak regulation. Therefore, the frequency in the command signal data corresponding to different environmental conditions is different. Therefore, the frequency in the command signal data can be determined through environmental and weather data. Therefore, it is necessary to collect all types of data related to the frequency in the command signal data.

[0049] Specifically, at a preset time interval To get the preset time before the current moment The frequency modulation command signal data and various characteristic data of the power system dispatching center at all times in the hour; among them, the various characteristic data include humidity, temperature, wind speed and load, etc.

[0050] By combining the data of all dimensions at each moment into a sample data point, a number of sample data points can be obtained; wherein all dimensions include all feature data and frequency modulation command signal data.

[0051] Among them, in this embodiment, the preset time length , wherein, in this embodiment, the preset time There is no specific limitation, and the implementer can determine it according to the specific situation. seconds, where the preset time interval There is no specific limitation and implementers can decide based on specific circumstances.

[0052] At this point, several sample data points are obtained.

[0053] Step S002: Cluster all sample data points to obtain several clusters; filter out abnormal sample data points from all sample data points through the degree of discreteness of all data in several clusters to obtain target sample points; interpolate all target sample points to obtain target data points; perform dimensionality reduction on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; obtain several reference data points through the frequency modulation command signal data at several moments and various reference feature data.

[0054] It should be noted that in order to determine the prediction accuracy of each decision tree constructed and the number of layers of the corresponding decision tree, the characteristic data that affects the frequency modulation command signal data is analyzed. However, the redundancy between some characteristic data is very high. Constructing a decision tree with a higher number of layers will only increase the amount of calculation, and will not have much effect on improving the performance of frequency modulation command signal data prediction. Therefore, the number of layers of the decision tree can be optimized by reducing the data of some characteristic dimensions through dimensionality reduction. In order to determine the prediction accuracy of each decision tree constructed and the number of leaf nodes of the corresponding decision tree, the prediction analysis error is performed by iteratively analyzing the number of leaf nodes of each decision tree in turn to analyze and determine the optimal number of leaf nodes of the decision tree.

[0055] It should be further explained that due to the interference of noise during the data collection process, there is a certain difference between the collected data and the real data, so denoising is necessary. If a sample data point among all the collected sample data points is far away from the other sample data points, it means that the sample data point is interfered by noise. Therefore, denoising is performed by distributing all the sample data points.

[0056] Specifically, all sample data points are clustered using the K-means clustering algorithm (K-means clustering) to obtain a number of clusters; the number of clusters is obtained by the elbow method; the K-means clustering algorithm and the elbow method are both well-known technologies and will not be described in detail here.

[0057] The abnormal outlier factor of each sample data point is calculated by the LOF (Local Outlier Factor) algorithm; the mean of the abnormal outlier factors of all sample data points in each cluster is recorded as the noise anomaly factor of each cluster; all sample data points corresponding to the cluster with the largest noise anomaly factor in all clusters are recorded as abnormal sample data points; abnormal sample data points are screened out from all sample data points, and the remaining sample data points are recorded as target sample points; wherein, the LOF algorithm is a well-known technology and will not be described in detail here.

[0058] Interpolation is performed on all target sample points using linear interpolation to obtain interpolated data points; the interpolated data points and all target sample points are combined to form target data points. The linear interpolation method is a well-known technique and will not be described in detail here.

[0059] The time series data of each dimension corresponding to all target data points is subjected to dimensionality reduction using the PCA (Principal Component Analysis) algorithm to obtain features after dimensionality reduction, which are recorded as reference features. The PCA algorithm is a well-known technology and will not be described in detail here.

[0060] By combining all reference characteristic data and frequency modulation command signal data at each moment into a reference data point, several reference data points can be obtained.

[0061] So far, several reference data points have been obtained through the above method.

[0062] Step S003: Obtain the error corresponding to each decision tree under different sample sizes through several reference data points; obtain the optimal sample size for each decision tree through the trend change of the error corresponding to different sample sizes of each decision tree.

[0063] It should be noted that in order to reduce the number of data points in each decision tree and thus reduce the amount of calculation, the optimal number of data points is obtained by iteratively analyzing the number of data points in each decision tree against the prediction error.

[0064] Specifically, several reference data points are sorted in chronological order to form a set of sequences, which are recorded as data point sequences. The reference data points form a sequence, which is recorded as the previous data point sequence; the latter data point sequence The reference data points of the AFTER data point sequence are combined into a sequence, which is recorded as the AFTER data point sequence; the FM command signal data corresponding to all reference data points in the AFTER data point sequence are combined into a set of sequences, which is recorded as the AFTER command signal data sequence. Divide the preset into percentage factors.

[0065] In this embodiment, the division percentage factor is preset , wherein in this embodiment the preset division percentage factor There is no specific limitation and implementers can decide based on specific circumstances.

[0066] Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision trees is used to predict the FM command signal data. The predicted FM command signal data are organized into a set of sequences in chronological order, which are recorded as the first predicted command signal data sequence. The error between the first predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE (Mean Absolute Percentage Error), which is recorded as the first error. ; The first error The sample size of each decision tree is The corresponding error when

[0067] Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision tree is used to predict the FM command signal data, and the predicted FM command signal data are organized into a set of sequences in chronological order, which are recorded as the second predicted command signal data sequence; the error between the second predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE, which is recorded as the second error. ; The second error The sample size of each decision tree is The corresponding error when

[0068] And so on, the corresponding errors of each decision tree under different sample sizes are obtained in turn;

[0069] in, Indicates the number of preset decision trees; wherein, in this embodiment, the number of preset decision trees , in this embodiment, the number of decision trees is preset There is no specific limitation, and the implementer can decide according to the specific situation. Express Round up; Indicates the number of all data points in the previous data point sequence.

[0070] It should be noted that when the sample size of each decision tree is small, there is a large deviation in its prediction results. Therefore, the prediction performance is often improved by increasing the sample size of each decision tree. However, in the process of increasing the sample size of each decision tree, its prediction performance is gradually improved. When it is improved to a certain extent, increasing the number of decision trees may not improve the prediction effect or the improvement speed is slow. Therefore, the most appropriate sample size is obtained by analyzing the error change curve during the increase of the sample size of each decision tree.

[0071] Specifically, a reference coordinate system is constructed with the sample size of each decision tree as the horizontal axis and the error corresponding to the sample size of each decision tree as the vertical axis. The different sample sizes and corresponding errors of each decision tree are mapped into the reference coordinate system to obtain a number of coordinate points. A curve fitting is performed on these coordinate points using the least squares method to obtain an error fitting curve. The least squares method is a well-known technique and will not be described in detail here.

[0072] Get the inflection point where the second-order derivative of the function corresponding to the error fitting curve is equal to 0, and As the optimal sample size for each decision tree; Indicates the horizontal value of the inflection point, Express Round up.

[0073] So far, the optimal sample size for each decision tree is obtained through the above method.

[0074] Step S004: According to the optimal sample size of each decision tree, the command signal data is predicted using the random forest algorithm.

[0075] According to the optimal sample size of each decision tree, the command signal data is predicted by the random forest algorithm to improve the accuracy of the prediction and reduce the amount of calculation in the prediction process.

[0076] like Figure 2 As shown, the second aspect of the present invention is to provide a frequency modulation instruction prediction system for optimizing a molten salt heat storage system, comprising the following modules:

[0077] Data acquisition module 101: used to continuously acquire frequency modulation command signal data and various characteristic data affecting the stability of the power system at a certain time; the various characteristic data include humidity, temperature, wind speed and load; and obtain a number of sample data points through the frequency modulation command signal data and various characteristic data at a certain time;

[0078] Data processing module 102: used to cluster all sample data points to obtain several clusters; filter out abnormal sample data points from all sample data points based on the degree of dispersion of all data in several clusters to obtain target sample points; interpolate all target sample points to obtain target data points; perform dimensionality reduction on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; obtain several reference data points based on the frequency modulation command signal data at several moments and various reference feature data;

[0079] Error analysis module 103: used to obtain the corresponding errors of each decision tree under different sample sizes through a number of reference data points; and to obtain the optimal sample size of each decision tree through the trend change of the corresponding errors under different sample sizes of each decision tree;

[0080] Prediction module 104: used to predict the instruction signal data using a random forest algorithm according to the optimal sample size of each decision tree.

[0081] The third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements a method for predicting frequency modulation instructions for optimizing a molten salt heat storage system.

[0082] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, a method for predicting frequency modulation instructions for optimizing a molten salt heat storage system is implemented.

[0083] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, optical storage, etc.) containing computer-usable program code.

[0084] The present invention is described with reference to flowcharts and / or block diagrams of methods, systems, and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0085] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system, characterized in that: include: Continuously obtain frequency modulation command signal data and various characteristic data that affect the stability of the power system at several moments; Obtain a number of sample data points through the frequency modulation command signal data and various characteristic data at a number of moments; All sample data points are clustered to obtain several clusters; through the degree of dispersion of all data in several clusters, abnormal sample data points are filtered out from all sample data points to obtain target sample points; all target sample points are interpolated to obtain target data points; all target data points are subjected to dimensionality reduction on the time series data of each dimension corresponding to the target data points to obtain reference features after dimensionality reduction; several reference data points are obtained through the frequency modulation command signal data at several moments and various reference feature data; Through several reference data points, the corresponding errors of each tree under different sample sizes are obtained, including sorting several reference data points in chronological order to form a set of sequences, recorded as data point sequences; The reference data points form a sequence, which is recorded as the previous data point sequence; the latter data point sequence The reference data points of the posterior data point sequence are combined into a sequence, which is recorded as the posterior data point sequence; the frequency modulation command signal data corresponding to all reference data points in the posterior data point sequence are combined into a set of sequences, which is recorded as the posterior command signal data sequence; wherein, The percentage factor for the preset division; Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision tree is used to predict the FM command signal data, and the predicted FM command signal data are organized into a set of sequences in chronological order, which is recorded as the first predicted command signal data sequence; the error between the first predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE, which is recorded as the first error. ; The first error As the sample size of each tree is The corresponding error when Select preset parameters with replacement from previous data point series data points to form a data set, and then select Second, composition data sets; Data set construction A decision tree, through The random forest constructed by the decision tree is used to predict the FM command signal data, and the predicted FM command signal data are organized into a set of sequences in chronological order, which are recorded as the second predicted command signal data sequence; the error between the second predicted command signal data sequence and the subsequent command signal data sequence is calculated by MAPE, which is recorded as the second error. ; The second error As the sample size of each tree is The corresponding error when And so on, the corresponding errors of each tree under different sample sizes are obtained in turn; in, Indicates the number of preset decision trees, Express Round up; Represents the number of all data points in the previous data point sequence; The optimal sample size for each tree is obtained by analyzing the trend of the error corresponding to different sample sizes of each tree, including constructing a reference coordinate system with the sample size of each tree as the horizontal axis and the error corresponding to the sample size of each tree as the vertical axis; mapping the different sample sizes of each tree and the corresponding errors in the reference coordinate system to obtain a number of coordinate points; and performing curve fitting on the several coordinate points using the least squares method to obtain an error fitting curve. Get the inflection point where the second derivative of the function corresponding to the error fitting curve is equal to 0, and as the optimal sample size for each tree; where, Indicates the horizontal value of the inflection point, Express Round up; According to the optimal sample size of each tree, the command signal data is predicted using the random forest algorithm.

2. A method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system according to claim 1, characterized in that: The method obtains a plurality of sample data points by using the frequency modulation command signal data and various characteristic data at a plurality of moments, including: By combining the data of all dimensions at each moment into a sample data point, a number of sample data points can be obtained; wherein all dimensions include all feature data and frequency modulation command signal data.

3. A method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system according to claim 1, characterized in that: The clustering of all sample data points is performed to obtain several clusters, including: All sample data points are clustered using the K-means clustering algorithm to obtain several clusters; the number of clusters is obtained by the elbow method.

4. A method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system according to claim 1, characterized in that: The method of filtering out abnormal sample data points from all sample data points by determining the discreteness of all data in a plurality of clusters to obtain target sample points; and interpolating all target sample points to obtain target data points includes: The outlier factor of each sample data point is obtained by calculating the LOF algorithm; the mean of the outlier factors of all sample data points in each cluster is recorded as the noise outlier factor of each cluster; all sample data points corresponding to the cluster with the largest noise outlier factor among all clusters are recorded as outlier sample data points; outlier sample data points are screened out from all sample data points, and the remaining sample data points are recorded as target sample points; Interpolation is performed on all target sample points using a linear interpolation method to obtain interpolated data points; the interpolated data points and all target sample points are combined to form target data points.

5. The method for predicting frequency modulation instructions for optimizing a molten salt thermal storage system according to claim 1, characterized in that: The dimensionality reduction is performed on the time series data of each dimension corresponding to all target data points to obtain reference features after dimensionality reduction; Through the frequency modulation command signal data and various reference characteristic data at several moments, several reference data points are obtained, including: According to the time series data of each dimension corresponding to all target data points, the PCA algorithm is used to reduce the dimension, and the features after dimensionality reduction are obtained and recorded as reference features; By combining all reference characteristic data and frequency modulation command signal data at each moment into a reference data point, several reference data points can be obtained.

6. A frequency modulation instruction prediction system for optimizing a molten salt heat storage system, characterized in that: It includes a data acquisition module, a data processing module, an error analysis module and a prediction module. When the data acquisition module, the data processing module, the error analysis module and the prediction module are executed, a method for predicting frequency modulation instructions for optimizing a molten salt heat storage system as described in claim 1 is implemented.

7. An electronic device, characterized in that: It 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, it implements a method for predicting frequency modulation instructions for optimizing a molten salt heat storage system as described in any one of claims 1 to 5.

8. 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, it implements the method for predicting frequency modulation instructions for optimizing a molten salt heat storage system according to any one of claims 1 to 5.

Images