A visualization method and system for power storage distribution and scheduling

Abstract

The present invention relates to the technical field of digital power load data processing, and in particular to a visualization method and system for power energy storage distribution and scheduling; power load data is acquired, function fitting is performed on the power load data, parameter difference is acquired according to parameters of power load functions of any two days, actual difference is acquired according to power load data of any two days, and power load similarity is acquired according to parameter difference and actual difference; power load data is clustered according to power load similarity, a central function of each clustering cluster is acquired, a separation degree between the latest power load data and the central function is acquired, a prediction function is acquired according to the separation degree, and power load prediction and power load visualization processing are performed according to the prediction function. The scheme of the present invention fully considers the changing trend of power load data every day, realizes more accurate power load prediction, and enables staff to understand power load data in time.

Description

Technical Field

[0001] The present invention relates to the technical field of digital power load data processing, and more specifically, to a visualization method and system for power storage distribution and scheduling. Background Art

[0002] Power load forecasting is an important technical means in the management of power energy storage distribution and dispatching systems. By accurately predicting the load of power energy storage distribution and dispatching systems, the power energy distribution and dispatching plan can be made more accurate, thereby better completing the visualization of power energy storage distribution and dispatching. Combined with the time series analysis method, historical power load data is analyzed to understand the regular characteristics of load such as seasonality, Sunday changes, holidays, etc., to provide a reference for future load forecasting, so as to predict the future load change trend.

[0003] The existing power load forecast can obtain the power load forecast value through analysis based on the historical power load data. When forecasting the power load, there are usually two forecasting methods:

[0004] Method 1: Overall rough prediction; perform function fitting based on the changing trend of historical power load data, obtain the future changing trend of power load data based on the fitting function, and when it is necessary to predict the power load of the next collection period, obtain the power load corresponding to the next collection period of the function as the predicted value.

[0005] Method 2: Prediction based on energy storage entropy of historical power load data; This prediction method corrects the power load data prediction model according to the historical power that meets the energy storage entropy of the power load data to perform power load prediction. Although it corrects the existing prediction method, it cannot accurately predict the periodic power load data for the periodic changes in the power load data. For example, the Chinese patent application document with publication number CN116307248A, entitled A Load Forecasting Method and System Considering Energy Storage Entropy Distribution, discloses a method for correcting power load prediction results based on energy storage entropy. This scheme corrects the prediction of power load according to the fitting function of the energy storage entropy of historical power load data.

[0006] At the same time, in the prediction process, the above two prediction methods often directly fit the function based on the historical power load data, and adjust the power load data prediction results on the fitting function in combination with the trend changes in some historical power load data, ignoring the periodicity in the daily changes of power load data, which will lead to inaccurate prediction of power load.

[0007] Therefore, in the process of power load forecasting, it is particularly important to obtain the periodic changes of power load in historical power load data and obtain the forecast results based on the change trend. Summary of the invention

[0008] The purpose of the present invention is to provide a visualization method and system for power storage distribution and scheduling, so as to solve the problem that in the process of existing power load data prediction by function fitting, the periodicity of daily changes in power load data is ignored, resulting in inaccurate prediction results. To this end, the present invention provides the following solution.

[0009] The present invention provides a visualization method for electric energy storage distribution and scheduling, comprising the following steps: performing function fitting on daily electric load data in historical time to obtain daily electric load functions, obtaining parameter differences of the electric load functions of any two days, obtaining actual differences of electric load data of any two days, and determining electric load similarity of any two days, wherein the electric load similarity is negatively correlated with the parameter differences and the actual differences;

[0010] According to the similarity of power load every two days in the historical time, the time periods in the historical time are clustered to obtain clusters, where the unit of the time period is day, and the central function of the cluster is determined, and the parameter of the central function is positively correlated with the parameter of the power load function corresponding to each time period in the cluster;

[0011] Perform function fitting based on the latest power load data collected in real time and a preset number of power load data collected before the latest power load data to determine the latest power load function, obtain the degree of separation between the latest power load function and the central function of each cluster, and the degree of separation is positively correlated with the distance between the parameters of the latest power load function and the corresponding parameters of the central function of the cluster; determine the central function with the smallest degree of separation as the prediction function of the latest power load data, and obtain the prediction result of the latest power load data according to the prediction function;

[0012] The forecast results are visualized to obtain the changes in power load in real time.

[0013] The present invention analyzes the power load similarity of time periods of power load data in historical time, clusters all time periods according to the power load similarity of any two time periods, clusters time periods with higher similarity into one category according to the periodicity of power load data changes, and can make predictions based on the power load data of time periods with higher similarity. It can take corresponding considerations for power load data with different periodic changes, thereby improving the accuracy of prediction.

[0014] Optionally, the power load data includes voltage power load data, current power load data and power load data, the daily power load data includes values ​​at each moment, and the function fitting is polynomial function fitting.

[0015] Optionally, the parameter difference of the power load function of any two days satisfies the expression:

[0016] ;

[0017] in, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates the number of items of power load data, Indicates the number of parameters contained in the power load function, and Respectively represent Tianhedi Day's The first parameters.

[0018] The present invention analyzes the degree of difference in the parameters of the power load function for any two days, comprehensively considers the degree of difference in the various parameters of the two time periods, and obtains the parameter difference between the two time periods; a higher similarity can be determined for time periods with similar parameters of the power load function, thereby obtaining a more accurate clustering result.

[0019] Optionally, the actual difference of the power load data of any two days is specifically:

[0020] Get the Tianhedi Day's The absolute value of the difference between the values ​​of the power load data at each time, and the average value of all the absolute values ​​is recorded as the first Tianhedi Day's The average difference level of power load data , then Tianhedi The actual difference of the power load data of the day satisfies the expression:

[0021] ;

[0022] in, Indicates Tianhedi The actual difference of the power load data of the day, Indicates the number of items of power load data, Indicates the total number of times per day. and Respectively represent Tianhedi Day's The power load data is in The value of a moment.

[0023] The present invention considers the actual difference of power load data in two time periods on the basis of parameter difference, which can avoid classifying time periods with similar power load functions but large actual differences in power load data into the same category, thereby improving the accuracy of clustering.

[0024] Optionally, the power load similarity of any two days satisfies the expression:

[0025] ;

[0026] in, Indicates Tianhedi The similarity of power load on the day, represents a natural constant, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates Tianhedi The actual difference in power load data for the day.

[0027] Optionally, clustering the time periods within the historical time to obtain cluster clusters is specifically:

[0028] According to the similarity of power load every two days in the historical time, the time periods in the historical time are clustered by clustering algorithm. When judging the cluster to which each time period belongs, the average value of the power load similarity between each time period and all time periods in the cluster is recorded as the degree of belonging of each time period to the cluster, and each time period is classified into the cluster corresponding to the maximum degree of belonging.

[0029] The present invention clusters time periods according to the similarity of time periods in historical time, and groups time periods with high power load similarity into the same category. It can more accurately group time periods with similar changing trends of power load data into one category, so that a more accurate prediction function can be obtained after classifying the power load data collected in real time.

[0030] Optionally, the center function of the cluster is specifically:

[0031] According to the power load function corresponding to each power load data in each cluster, the central function of each power load data in the cluster is obtained. The first Item power load data, obtain the The first The parameters of the power load function of the power load data of the item are obtained, and the average value of the parameters of all the power load functions is obtained, which is recorded as the Clusters The parameters of the central function of the power load data are calculated according to The first cluster The parameters of the central function of the power load data are constructed The first The central function of each item of power load data is obtained by this method.

[0032] Optionally, the degree of separation satisfies the expression:

[0033] ;

[0034] in, Represents the latest power load function and The degree of separation between the central functions of the clusters, Indicates the number of items of power load data, Indicates The parameters of the latest power load function of the power load data and the The first The Euclidean distance of the parameters of the central function of the power load data.

[0035] Optionally, the prediction result of obtaining the latest power load data according to the prediction function is specifically:

[0036] The cross-correlation function between the latest power load function and the prediction function is obtained, and the time delay corresponding to the maximum value of the cross-correlation function is recorded as the offset. According to the offset, each latest power load function and each prediction function are matched, and the function value of the latest power load data at the next moment in each prediction function is obtained as the prediction value of each power load data at the next moment.

[0037] The present invention obtains the offset between the latest power load function and the prediction function through the cross-correlation function, and obtains the corresponding correlation between the prediction function and the latest power load function at different offsets through the cross-correlation function, so as to more accurately obtain the offset of the two functions, and then more accurately obtain the corresponding relationship between the prediction function and the latest power load function.

[0038] In a second aspect, the present invention further provides a visualization system for power storage distribution and scheduling, comprising:

[0039] processor;

[0040] A memory stores computer instructions for a visualization method for electric power storage distribution and scheduling. When the computer instructions are executed by the processor, the system implements the above-mentioned visualization method for electric power storage distribution and scheduling.

[0041] The beneficial effects of the present invention are:

[0042] In the process of predicting the power load data through the daily periodic changes of the power load data, in order to solve the problem that the periodic change characteristics of the power load data are difficult to extract, the present invention analyzes the power load similarity of the time periods of the power load data in the historical time, clusters all the time periods according to the power load similarity of any two time periods, clusters the time periods with higher similarity into one category according to the periodicity of the power load data changes, and can make predictions based on the power load data of the time periods with higher similarity, and can take corresponding considerations for the power load data with different periodic changes, thereby improving the accuracy of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0044] Figure 1 A flowchart of the steps of a visualization method for power storage allocation and scheduling in this embodiment is schematically shown;

[0045] Figure 2 The structural block diagram of a visualization system for power storage distribution and scheduling in this embodiment is schematically shown. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0048] like Figure 1 As shown, a visualization method for power storage distribution scheduling in this embodiment includes the following steps:

[0049] S001: Collect the power load data in the historical period and segment the power load data.

[0050] The power load data collected during the power storage distribution and dispatching process can be used to analyze the power load changes, obtain the power load trend of the power storage distribution and dispatching system, and provide a practical basis for predicting subsequent power load data.

[0051] It should be noted that the power load in the power storage distribution and dispatching system is mainly reflected in its power load data such as voltage, current and power. Therefore, it is necessary to collect multiple power load data of the power storage distribution and dispatching system as a measure of the power load.

[0052] A power load data collection cycle is preset, wherein the specific value of the power load data collection cycle can be set according to actual application scenarios and requirements. The present invention sets the preset power load data collection cycle to 10 minutes.

[0053] Specifically, an electric load detection instrument is installed in the electric energy storage distribution and dispatching system, which can monitor and record the electric load data in real time. The electric load data within a period of time is collected according to the electric load data collection cycle. In this embodiment, all the electric load data within half a year is collected. The specific scope of the electric load data collection can be set according to the actual application scenario and needs. The electric load data to be collected includes voltage, current, and power data information. The day's collection No. The power load data is recorded as , divided into segments according to time, each day is recorded as a time period, and the number of power load data items contained in the power load data is recorded as , the total number of moments included in each day is recorded as .

[0054] At this point, all collected power load data of the power storage distribution and dispatching system are obtained.

[0055] S002: Obtain daily power load function based on daily power load data.

[0056] It should be noted that due to the different usage of the power storage distribution and dispatching system, there are obvious differences in the power load conditions of the power storage distribution and dispatching system during the day and at night, while the daily power load changes are roughly the same, so the daily power load can be fitted with a function. In this embodiment, the polynomial function has an obvious and smooth change cycle, which conforms to the change law of the power load data, so the power load data is fitted to a polynomial function. In other embodiments, implementers can choose the form of the fitting function according to actual application scenarios and needs.

[0057] It should be further explained that in order to ensure the accuracy of subsequent difference calculations, the fitting function form of the power load data needs to remain consistent, that is, the number of parameters included in the power load function is the same and the function terms of each power load function are consistent; for the polynomial function in this embodiment, the degree of the polynomial function should be the same.

[0058] Specifically, obtain the All the days The power load data is fitted into a polynomial function, and the number of parameters contained in the polynomial function is recorded as ; Day's The fitting function of the power load data is recorded as the power load function , the independent variable of the function is The function value is the moment of power load data within a day. The size of the power load data.

[0059] At this point, the power load function of the power load data for each time period is obtained.

[0060] S003: Obtain the power load similarity of two days according to the difference between the power load data of any two days and the difference degree of the parameters of the power load function.

[0061] It should be noted that in the process of clustering the power load functions, it is necessary to measure the difference between the two power load functions. The difference between the power load functions is mainly reflected in the difference in the parameters of the power load functions. Therefore, the parameter difference between the power load functions of any two days can be obtained based on the difference between the parameters of the power load functions.

[0062] Specifically, obtain the The parameters of the power load function of the day are The power load function of the power load data The parameters are recorded as According to Tianhedi The parameters of the power load function of the day are used to obtain the parameter difference of the power load function:

[0063] ;

[0064] in, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates the number of items of power load data, Indicates the number of parameters contained in the power load function, and Respectively represent Tianhedi Day's The first parameters.

[0065] It should be noted that the Partially indicates Tianhedi Day's The degree of difference in parameters of the power load function of the corresponding power load data can be used to evaluate the degree of difference in parameters of two power load functions. The greater the overall difference in parameters, the greater the difference between the two power load functions.

[0066] It should be further explained that for the parameter difference of the power load function of any two days, it is necessary to comprehensively consider the difference of all power load data. and The average value of the parameter difference of the power load function of all power load data of the day is taken as the and The parameter difference of power load during the day.

[0067] Specifically, the parameter difference between the power load functions of any two days is obtained.

[0068] It should be noted that the parameter difference measures the degree of difference in the change of power load based on the parameter difference of the power load function on any two days. In addition to the parameter difference, the fluctuation degree of the power load data difference on any two days should also be considered. The greater the fluctuation degree, the greater the difference in the power load data on the two days.

[0069] Specifically, obtain the Tianhedi Day's The absolute value of the difference between the values ​​of the power load data at each time, and the average value of all the absolute values ​​is recorded as the first Tianhedi Day's The average difference level of power load data , then Tianhedi The actual difference of the power load data of the day satisfies the expression:

[0070] ;

[0071] in, Indicates Tianhedi The actual difference of the power load data of the day, Indicates the number of items of power load data, Indicates the number of moments contained in each time period. and Respectively represent Tianhedi Day's The power load data is in The value of a moment.

[0072] What needs to be explained about the actual difference formula is that Partially indicates Tianhedi Day's The absolute value of the difference between the absolute value of each difference and the average value of the absolute value of all differences indicates that the Tianhedi Day's The greater the difference in the power load data, the greater the The average value of the difference corresponding to all the time periods of the two days is taken as the first The fluctuation degree of each item of power load data is taken into consideration, and the fluctuation degree of all items of power load data is taken into consideration to obtain the Tianhedi The actual difference in power load data for the day.

[0073] At this point, the actual difference in power load data for any two days is obtained.

[0074] It should be noted that the changes in daily power load data during the actual operation of the power storage distribution and dispatching system often have certain regularities. For example, the changes in power load data on two days of the week with the same weather conditions are similar or even the same within one day. Therefore, the daily power load function can be classified, and the power load functions with high power load similarity can be classified into the same category. In the subsequent operation of the power storage distribution and dispatching system, the corresponding power load function can be used to obtain the cluster to which it belongs, and the power load data can be predicted based on all the power load data in the cluster to which it belongs. In order to classify the power load data, it is necessary to obtain the similarity of the power load data of any two days.

[0075] Specifically, according to Tianhedi The parameter difference of the power load function of the day and the actual difference of the power load data are obtained. Tianhedi Similarity of power load in time periods of the day:

[0076] ;

[0077] in, Indicates Tianhedi The similarity of power load on the day, represents a natural constant, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates Tianhedi The actual difference in power load data for the day.

[0078] It should be noted that the Partially indicates Tianhedi The result of normalizing the parameter difference of the power load function of the day and the actual difference of the power load data is obtained by comprehensively considering and normalizing the parameter difference of the power load function and the actual difference of the power load data in two time periods, and then subtracting the normalized result from one can be used to represent the similarity between the power load data in different time periods.

[0079] Specifically, the power load similarity between any two time periods is obtained.

[0080] S004: Classify all time periods according to the similarity of power loads between any two time periods to obtain a final classification result.

[0081] It should be noted that the greater the power load similarity, the more likely the corresponding two time periods are of the same type, so the time periods can be clustered according to the power load similarity between each two time periods.

[0082] Preset the number of clusters for a time period , where the number of time period clusters The specific value of can be set according to the actual application scenario and requirements. The preset number is required to be a positive integer greater than 1. The present invention sets the preset number to 6.

[0083] Specifically, randomly select from all time periods The time periods are taken as the initial cluster centers, and each initial cluster center corresponds to an initial cluster.

[0084] It should be noted that, to cluster time periods, time periods with high similarity need to be clustered into the same category, so time periods can be clustered directly according to the similarity of power load.

[0085] Specifically, each initial cluster center constitutes its own cluster. All time periods outside the initial cluster center are traversed, and the average value of the power load similarity between each time period and all time periods in each cluster is obtained, which is recorded as the degree of belonging of the time period and the cluster. Each time period is classified into the cluster corresponding to the maximum degree of belonging, and the original time period in the cluster forms a new cluster.

[0086] It should be noted that in the process of calculating the degree of belonging between time periods and clusters and updating clusters, since the time periods included in the clusters change, the similarities between each time period and the clusters will also change accordingly. Therefore, the clusters need to be iteratively updated to improve the accuracy of cluster classification.

[0087] Specifically, all time periods are traversed to obtain their degree of belonging to all time periods in each cluster. If the maximum degree of belonging between each cluster and the time period does not correspond to the cluster where the time period is located, the time period is reclassified into the cluster corresponding to the maximum degree of belonging. This iterative traversal is performed until the clusters of all time periods remain unchanged.

[0088] It should be noted that the prediction of power load data requires prediction of each power load data separately, and the similarity between the power load data and all the power load data in the cluster is relatively large. The changing trend of the power load data can be estimated based on the power load function of all time periods in the cluster. Therefore, the central function is obtained based on the power load function corresponding to all time periods in the same cluster to represent the changing trend of the power load data.

[0089] Specifically, the central function of the power load data in each cluster is obtained according to the power load function corresponding to the power load data in the cluster. The first Item power load data, obtain the parameters of all items of power load function, obtain the average value of the corresponding parameters of each item of power load function, and record it as The first The central function is constructed according to the parameters of the central function. , the independent variable of the central function is the collection time of the power load data in a time period, and the function value is the value of the power load data. In this way, the central function of each power load data of all clusters is obtained.

[0090] At this point, the central function of each power load data in all clusters is obtained.

[0091] S005: Obtain the latest power load data, determine the latest power load function by performing function fitting based on a preset number of power load data collected before the latest power load data, classify the latest power load function to obtain the belonging cluster, predict the power load data based on the central function of the belonging cluster, and visualize the prediction results.

[0092] It should be noted that, for the latest power load data acquired in real time, function fitting is performed based on the power load data collected in the most recent period of time, and the power load data is predicted based on the fitting result.

[0093] A number of recent power load data is preset, wherein the specific value of the number of recent power load data can be set according to actual application scenarios and requirements. The present invention sets the preset number to 30.

[0094] Specifically, the real-time collected power load data is recorded as the latest power load data, and several power load data closest to the collection time of the latest power load data are obtained and recorded as the latest range power load data, wherein the latest range power load data is composed of all power load data whose difference from the collection time of the latest power load data is less than the number of the latest power load data. Function fitting is performed based on the latest range power load data to obtain the latest power load function of the latest power load data.

[0095] It should be noted that the latest power load function can describe the changes in the latest power load data in the recent period of time, and then classify the latest power load data according to the power load function of the latest power load data, and predict the latest power load data according to the cluster to which it belongs.

[0096] Specifically, the The latest power load data parameters and The first The Euclidean distance of the parameters of the central function of the power load data is recorded as The latest power load data is The separation degree of the power load data of each cluster is recorded as:

[0097] ;

[0098] in, Represents the latest power load function and The degree of separation between the central functions of the clusters, Indicates the number of items of power load data, Indicates Parameters and specifications of power load data The first The Euclidean distance of the parameters of the central function of the power load data.

[0099] What needs to be explained about the separation degree function is that, for any item of power load data, the larger the Euclidean distance between the parameters of its latest power load function and the parameters of the central function of the power load data of each cluster, the larger the gap between the latest power load function and the cluster, and they cannot be classified into the same cluster. The cluster that is most similar to the latest power load function can be judged according to the degree of separation, so as to obtain the cluster to which the latest power load data belongs.

[0100] Specifically, the separation degree between the latest power load function and the central functions of all clusters is obtained, and the central function corresponding to the minimum separation degree is obtained. The cluster corresponding to the central function is recorded as the subordinate class of the latest power load data. And the central functions of the subordinate classes of the latest power load data are used as the prediction functions of the latest power load data of the corresponding items.

[0101] It should be noted that the prediction function can accurately describe the changing trend of the latest power load data. However, the prediction function and the latest power load function cannot directly correspond to each other. It is necessary to calculate the offset between the two and correspond the latest power load function and the prediction function according to the offset.

[0102] Specifically, the cross-correlation function between the latest power load function and the prediction function is obtained, the time delay corresponding to the maximum value of the cross-correlation function is recorded as the offset, each latest power load function and each prediction function are matched according to the offset, and the function value of the latest power load data at the next moment in each prediction function is obtained as the predicted value of each power load data at the next moment.

[0103] It should be noted that the predicted value of power load data can help power operation and maintenance personnel to more clearly obtain the operating status of power storage distribution and scheduling, so as to warn and respond in advance to possible operational failures that may occur during the change of power load data.

[0104] The predicted values ​​of power load data are visualized so that operation and maintenance personnel can conduct more accurate risk assessment and status understanding of power load data.

[0105] The present invention also provides a visualization system for power storage distribution and scheduling. Figure 2 As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a visualization method for power storage distribution and scheduling according to the first aspect of the present invention is implemented.

[0106] The system also includes other components well known to those skilled in the art, such as a communication interface, whose configuration and functions are known in the art and thus will not be described in detail here.

[0107] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as RRAM (Resistive Random Access Memory), DRAM (Dynamic Random Access Memory), SRAM (Static Random-Access Memory), EDRAM (Enhanced Dynamic Random Access Memory), HBM (High-Bandwidth Memory), HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0108] In the description of this specification, “first” and “second” are only used to distinguish terms and do not have any limiting meaning.

[0109] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A visualization method for power storage distribution and scheduling, characterized in that: The following steps are involved: Perform function fitting on the daily power load data in the historical time to obtain the daily power load function, obtain the parameter difference of the power load function on any two days, obtain the actual difference of the power load data on any two days, and determine the power load similarity on any two days, wherein the power load similarity is negatively correlated with the parameter difference and the actual difference; Clustering the time periods in the historical time according to the similarity of the power loads every two days in the historical time to obtain clustering clusters, and determining the central function of the clustering cluster, wherein the parameters of the central function are positively correlated with the parameters of the power load function corresponding to each time period in the clustering cluster; Perform function fitting based on the latest power load data collected in real time and a preset number of power load data collected before the latest power load data to determine the latest power load function, and obtain the degree of separation between the latest power load function and the central function of each cluster, wherein the degree of separation is positively correlated with the distance between the parameters of the latest power load function and the corresponding parameters of the central function of the cluster; Determine the central function with the smallest separation degree as the prediction function of the latest power load data, and obtain the prediction result of the latest power load data according to the prediction function; The forecast results are visualized to obtain the changes in power load in real time.

2. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The power load data include voltage power load data, current power load data and power load data. The daily power load data includes values ​​at each moment. The function fitting is a polynomial function fitting.

3. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The parameter difference of the power load function of any two days satisfies the expression: ； in, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates the number of items of power load data, Indicates the number of parameters contained in the power load function, and Respectively represent Tianhedi Day's The first parameters.

4. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The actual difference between the power load data of any two days is specifically: Get the Tianhedi Day's The absolute value of the difference between the values ​​of the power load data at each time, and the average value of all the absolute values ​​is recorded as the first Tianhedi Day's The average difference level of power load data , then Tianhedi The actual difference of the power load data of the day satisfies the expression: ； in, Indicates Tianhedi The actual difference of the power load data of the day, Indicates the number of items of power load data, Indicates the total number of times per day. and Respectively represent Tianhedi Day's The power load data is in The value of a moment.

5. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The similarity of power loads on any two days satisfies the expression: ； in, Indicates Tianhedi The similarity of power load on the day, represents a natural constant, Indicates Tianhedi The parameter difference of the power load function of the day, Indicates Tianhedi The actual difference in power load data for the day.

6. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The clustering of time periods within the historical time to obtain cluster clusters is specifically as follows: According to the similarity of power load every two days in the historical time, the time periods in the historical time are clustered by clustering algorithm. When judging the cluster to which each time period belongs, the average value of the power load similarity between each time period and all time periods in the cluster is recorded as the degree of belonging of each time period to the cluster, and each time period is classified into the cluster corresponding to the maximum degree of belonging.

7. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The central function of the cluster is specifically: According to the power load function corresponding to each power load data in each cluster, the central function of each power load data in the cluster is obtained. The first Item power load data, obtain the The first The parameters of the power load function of the power load data of the item are obtained, and the average value of the parameters of all the power load functions is obtained, which is recorded as the Clusters The parameters of the central function of the power load data are calculated according to The first cluster The parameters of the central function of the power load data are constructed The first cluster The central function of each item of power load data is obtained by this method.

8. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The degree of separation satisfies the expression: ； in, Represents the latest power load function and The degree of separation between the central functions of the clusters, Indicates the number of items of power load data, Indicates The parameters of the latest power load function of the power load data and the The first cluster The Euclidean distance of the parameters of the central function of the power load data.

9. A visualization method for power storage distribution and scheduling according to claim 1, characterized in that: The prediction result of obtaining the latest power load data according to the prediction function is specifically: The cross-correlation function between the latest power load function and the prediction function is obtained, and the time delay corresponding to the maximum value of the cross-correlation function is recorded as the offset. According to the offset, each latest power load function and each prediction function are matched, and the function value of the latest power load data at the next moment in each prediction function is obtained as the prediction value of each power load data at the next moment.

10. A visualization system for power storage distribution and scheduling, characterized in that: include: processor; A memory storing computer instructions for a visualization method for electric power storage distribution and scheduling, wherein when the computer instructions are executed by the processor, the system executes a visualization method for electric power storage distribution and scheduling according to any one of claims 1 to 9.

Images