A data fusion method and system based on distribution network cloud platform

Abstract

The present application relates to the field of electrical digital data processing technology, and specifically to a data fusion method and system based on a distribution network cloud platform, the method comprising: fitting all power data of each distribution network terminal within a sampling interval, determining the power mutation index of each distribution network terminal based on the difference between all power data of each distribution network terminal within the sampling interval and the corresponding fitting value; determining the power characteristic vector of each distribution network terminal based on the average distribution and discreteness of the power data of each distribution network terminal within the sampling interval, in combination with the power mutation index; determining the objective function of the algorithm based on the similarity between all power characteristic vectors in the distribution network, in combination with the fuzzy mean clustering algorithm, so as to fuse the power data. The present application is based on the degree of membership of the power characteristic vector to the cluster center, eliminates the influence of power mutation data, and improves the accuracy and reliability of power data fusion.

Description

Technical Field

[0001] The present application relates to the technical field of electrical digital data processing, and in particular to a data fusion method and system based on a distribution network cloud platform. Background Art

[0002] As a crucial component of the power transmission system, the distribution network impacts the reliability of the entire system. For large-scale power transmission systems, the data and information generated by the distribution network are vast and complex. By leveraging cloud platforms to implement functions such as distribution automation, the distribution network reduces energy consumption in local data processing centers and improves the efficiency of data processing.

[0003] Power data, as an important parameter reflecting the operating status of distribution networks, plays a key role in network dispatch optimization and other aspects. Distribution networks upload large amounts of power data to the distribution network cloud platform in real time, and through data fusion, improve subsequent data processing efficiency. Before data fusion, it is necessary to remove sudden changes in the power data to improve the accuracy of data fusion. Currently, when the industry uses the FCM algorithm to process distribution network power data, the utilization rate of the correlation characteristics of the distribution network power data is low, and the ability to identify sudden changes in power data is weak, making it difficult to effectively clean the power data, resulting in reduced reliability and accuracy of data fusion. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a data fusion method and system based on a distribution network cloud platform. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a data fusion method based on a distribution network cloud platform, the method comprising the following steps:

[0006] Obtain power data of each distribution network terminal in the distribution network within a sampling interval;

[0007] Fit all power data of each distribution network terminal within the sampling interval, and determine the power mutation index of each distribution network terminal based on the difference between all power data of each distribution network terminal within the sampling interval and the fitted result;

[0008] Determine the power characteristic vector of each distribution network terminal based on the average distribution and dispersion of the power data of each distribution network terminal within the sampling interval and the power mutation index;

[0009] According to the similarity between the power feature vectors of all distribution network terminals in the distribution network, the objective function is determined by combining the fuzzy mean clustering algorithm to fuse the power data.

[0010] Preferably, the method for determining the power mutation index of each distribution network terminal is:

[0011] Analyze the difference between each power data of each distribution network terminal in the sampling interval and the corresponding fitting result, and record it as the first difference of each power data of each distribution network terminal in the sampling interval;

[0012] Analyze the mean of the first differences of all power data of each distribution network terminal within the sampling interval, and record it as the first mean of each distribution network terminal;

[0013] The power mutation index of each distribution network terminal is a result of fusing the difference between the first difference and the first mean of all power data of each distribution network terminal within a sampling interval.

[0014] Preferably, the power characteristic vector of each distribution network terminal is a vector composed of the mean and variance of all power data of each distribution network terminal within a sampling interval, and the power mutation index.

[0015] Preferably, the objective function is determined by combining the fuzzy mean clustering algorithm, including:

[0016] Determining a power distribution frequency of each power characteristic vector based on a distribution of power characteristic vector moduli of all distribution network terminals in the distribution network, wherein one distribution network terminal corresponds to one power characteristic vector;

[0017] According to the power distribution frequency, the power feature vectors corresponding to the preset number of power distribution frequencies are selected from large to small as the initial cluster centers;

[0018] Based on the similarity between each power eigenvector and each initial cluster center, the initial membership of each power eigenvector to each initial cluster center is determined, and the membership of each power eigenvector to each cluster center at each iteration is obtained through iteration;

[0019] The objective function of the fuzzy mean clustering algorithm is determined based on the membership degree of each power eigenvector to each cluster center, the difference between each power eigenvector and each cluster center, and the power mutation index at each iteration.

[0020] Preferably, the power distribution frequency of each power eigenvector is the cumulative sum of the frequencies of occurrence of the moduli of all power eigenvectors in the neighborhood of each power eigenvector in the moduli of all power eigenvectors in the power distribution network.

[0021] Preferably, the initial membership degree of each power eigenvector to each initial cluster center is a normalized result of the difference between the modulus of each power eigenvector and the modulus of each initial cluster center.

[0022] Preferably, the expression of the membership degree of each power eigenvector to each cluster center during each iteration is: Where, Indicates the membership degree of the i-th power eigenvector to the r-th cluster center at the g-th iteration; Indicates the membership degree of the jth power eigenvector to the rth cluster center at the g-1th iteration; represents the i-th power eigenvector; represents the jth power eigenvector belonging to the rth cluster center at the g-1th iteration; Indicates the modulus value; represents the number of power eigenvectors belonging to the rth cluster center at the g-1th iteration; exp( ) represents the exponential function with a natural constant as the base; norm( ) represents the normalization function.

[0023] Preferably, the objective function is expressed as: Where, Represents the objective function at the g-th iteration; Represents a constant greater than 0; represents the power mutation index of the i-th power eigenvector; represents the membership index; represents the rth cluster center at the gth iteration; R represents the total number of cluster centers; M is the number of power characteristic vectors of all distribution network terminals in the distribution network.

[0024] Preferably, the fusing of power data includes:

[0025] Taking the objective function, all initial cluster centers and all power eigenvectors as inputs of a fuzzy mean clustering algorithm, and outputting each cluster;

[0026] Taking all the membership degrees in each cluster as input of the threshold segmentation algorithm to obtain the segmentation threshold;

[0027] All power data of each distribution network terminal within the sampling interval constitute the power sequence of each distribution network terminal;

[0028] The power sequences of distribution network terminals corresponding to all power feature vectors with membership greater than or equal to the segmentation threshold in each cluster are used as the input of the data fusion algorithm to obtain the power sequences corresponding to each cluster after fusion.

[0029] In a second aspect, an embodiment of the present application also provides a data fusion system based on a distribution network cloud platform, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0030] This application has at least the following beneficial effects:

[0031] This application determines the power mutation index of each distribution network terminal based on the difference between all power data of each distribution network terminal in the sampling interval and the corresponding fitting results. The power characteristic vector of each distribution network terminal is determined based on the average distribution and discreteness of the power data of each distribution network terminal in the sampling interval, combined with the power mutation index; the power sequence replaces the method of directly using power data as clustering samples to improve the discrimination between power sequences containing power mutation data and other sequences.

[0032] This application determines the initial membership of each eigenvector to each initial cluster center based on the similarity between the power eigenvector of each distribution network terminal and each initial cluster center, and combines the fuzzy mean clustering algorithm to obtain the membership of each power eigenvector to each cluster center at each iteration, thereby enhancing the correlation between the power eigenvector and the cluster center, thereby improving the redundancy between the power sequences in the cluster cluster and improving the reliability of subsequent data fusion.

[0033] This application calculates the objective function of the fuzzy mean clustering algorithm by using the membership degree and power mutation index to obtain clusters of power sequences. This enhances the ability to identify power feature vectors containing power mutation data, improves the data cleaning effect of power data, and thus improves the accuracy of data fusion. Finally, a data fusion algorithm is used to fuse the power sequences in each data cluster to obtain power data with high accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to 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 application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 This is a flowchart of the steps of a data fusion method based on a distribution network cloud platform;

[0036] Figure 2 Obtain a process diagram for the objective function;

[0037] Figure 3 Schematic diagram of data classification and data fusion process;

[0038] Figure 4 Schematic diagram of the power data fusion method for distribution network. DETAILED DESCRIPTION

[0039] In order to further illustrate the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a data fusion method and system based on a distribution network cloud platform proposed in this application, its specific implementation method, structure, features and effects. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0040] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0041] The following describes in detail a specific solution of a data fusion method and system based on a distribution network cloud platform provided by the present application with reference to the accompanying drawings.

[0042] See also Figure 1 , which shows a flowchart of a data fusion method based on a distribution network cloud platform provided by an embodiment of the present application, the method comprising the following steps:

[0043] Step S1: Obtain power data of each distribution network terminal in the distribution network.

[0044] As the scale of power systems expands, the amount of data generated by distribution networks is becoming increasingly large. Cloud platforms offer powerful data processing and storage capabilities. To improve the reliability of distribution network power supply, the distribution network cloud platform acquires real-time power data uploaded by each distribution network terminal. N power data points are collected during the sampling interval of each distribution network terminal, where the total number of distribution network terminals is L.

[0045] It should be noted that the value of the sampling interval, the value of the number N of collected power data, and the value of the total number L are all set manually. In this embodiment, the value of the sampling interval is 1s, the value of the number N of collected power data is 50, and the value of the total number L is 1000. The implementer can set them according to the specific situation, and this embodiment does not impose any special restrictions.

[0046] Step S2: Fit the power data of each distribution network terminal in the sampling interval to obtain the fitting curve of each distribution network terminal.

[0047] The power mutation index of each distribution network terminal is determined based on the difference between the power data of each distribution network terminal and the corresponding fitting value on the fitting curve.

[0048] As the number of terminal devices in the distribution network continues to grow, the distribution network cloud platform will receive an increasing amount of power data from these devices. Due to the increasing number of terminal devices and the network's inherent topology, the power series captured by the distribution network cloud platform are highly redundant. Leveraging this redundancy to fuse the power series captured by the distribution network cloud platform can effectively improve data quality and processing efficiency.

[0049] Distribution network terminal devices are widely distributed, and some are deployed in harsh environments and have complex working conditions. This can easily lead to deviations between some of the power data measured and uploaded to the distribution network cloud platform and the actual power data. This can generate power mutation data within the data sampling interval, reducing the accuracy and reliability of data fusion. In order to distinguish between normal power data and power mutation data, the power characteristic vector of each distribution network terminal is determined, specifically:

[0050] (1) The occurrence of power mutation data is highly random and short-lived, making it difficult to identify through instantaneous analysis. At the same time, the magnitude of power mutation data differs significantly from the overall trend of power data.

[0051] Therefore, all power data of each distribution network terminal within the sampling interval are fitted to obtain the fitting curve of each distribution network terminal.

[0052] According to the difference between all power data of each distribution network terminal in the sampling interval and the corresponding fitting value, the power mutation index of each distribution network terminal is determined to judge whether there is power mutation data in the power data of each distribution network. Specifically:

[0053] Analyze the difference between each power data of each distribution network terminal in the sampling interval and the corresponding fitting value in the fitting curve, and record it as the first difference of each power data of each distribution network terminal in the sampling interval;

[0054] Analyze the mean of the first differences of all power data of each distribution network terminal within the sampling interval, and record it as the first mean of each distribution network terminal;

[0055] The power mutation index of each distribution network terminal is a result of fusing the first difference of all power data of each distribution network terminal within a sampling interval with the difference of the first mean.

[0056] It should be noted that there are many ways to measure differences. In this embodiment, the difference between the actual power data and the fitting value is measured by taking the absolute value of the difference between each power data and the corresponding fitting value. The implementer can also use ratios and other methods to measure the difference according to the specific situation. This embodiment does not impose any special restrictions.

[0057] It should be understood that fusion refers to combining two or more indicators through addition or multiplication to obtain a comprehensive indicator, thereby more comprehensively and accurately evaluating a phenomenon or problem. This fusion method is not limited to simple arithmetic operations and can also include more complex statistical models and analysis methods. Implementers can choose according to their specific circumstances and this embodiment does not impose any special restrictions.

[0058] Preferably, as an embodiment of the present application, the power mutation index of each distribution network terminal is the result of accumulating the difference between the first difference of all power data of each distribution network terminal in the sampling interval and the first mean value.

[0059] Preferably, as another embodiment of the present application, the power mutation index of each distribution network terminal is an exponential function value with a natural constant as the base and the result of accumulating the difference between the first difference of the power data of each distribution network terminal in the sampling interval and the first mean as the independent variable.

[0060] The greater the difference between the power data and the corresponding fitting value in the fitting curve, the greater the possibility that the power data is power mutation data; conversely, the smaller the difference between the power data and the corresponding fitting value in the power fitting line, the smaller the possibility of power data mutation.

[0061] (2) Further, the mean and variance of all power data of each distribution network terminal within the sampling interval are analyzed. For the convenience of expression, they are recorded as the average power and power variance of each distribution network terminal respectively.

[0062] The average power and power variance of each distribution network terminal reflect the average size and distribution of power data at the distribution network terminals. The closer the average power and power variance of two distribution network terminals are, the smaller the characteristic difference between their power series, the higher the redundancy during data fusion, and the more reliable the corresponding data fusion results.

[0063] (3) Furthermore, the average power, power variance and power mutation index of each distribution network terminal are combined to form the power feature vector of each distribution network terminal, which reflects the change characteristics of the power data of the distribution network terminal within the sampling interval.

[0064] It should be noted that one distribution network terminal corresponds to one power characteristic vector, so each power characteristic vector represents the power characteristic vector of each distribution network terminal.

[0065] Step S3: Determine the objective function of the algorithm based on the similarity between all power feature vectors in the distribution network and the fuzzy mean clustering algorithm.

[0066] When power mutation data is generated at the distribution network terminal, the power mutation index increases significantly. In addition, power mutation data will also affect the size of the average power and power variance, and improve the degree of separation from each cluster center during subsequent clustering. Therefore, in order to enhance the distinction between power mutation data and normal power data, the similarity between the power feature vectors of all distribution network terminals in the distribution network is analyzed, and the objective function of the algorithm is determined by combining the fuzzy mean clustering algorithm to facilitate the screening of normal power data and the elimination of power mutation data. Specifically:

[0067] (1) Due to the strong correlation and redundancy between the power data of different distribution networks, there is a good similarity between the power feature vectors of the distribution network terminals. Therefore, according to the distribution characteristics of the modulus value of the power feature vector, the initial cluster center is selected, specifically:

[0068] W neighborhoods are divided with each power eigenvector as the center. The power distribution frequency of each power eigenvector is the cumulative sum of the frequencies of the moduli of all power eigenvectors in the neighborhood of the power eigenvector of each distribution network terminal in the moduli of all power eigenvectors in the distribution network.

[0069] It should be noted that the value of the neighborhood W is set manually. In this embodiment, the value of the neighborhood W is 20. The implementer can also set it according to the specific situation. This embodiment does not impose any special restrictions.

[0070] According to the power distribution frequency, the power characteristic vectors corresponding to the preset number of power distribution frequencies are selected from large to small as the initial clustering centers, where the value of the preset number is set artificially. In this embodiment, the value of the preset number is 100, and the implementer can also set it according to the specific situation. This embodiment does not impose any special restrictions.

[0071] (2) Further, based on the similarity between each power eigenvector and each initial cluster center, the initial membership of each power eigenvector to each initial cluster center is determined. The initial membership of each power eigenvector to each initial cluster center is the normalized result of the difference between the modulus value of each power eigenvector and the modulus value of each initial cluster center.

[0072] It should be noted that there are many ways to measure differences. In this embodiment, the absolute value of the difference between the modulus of each power feature vector and the modulus of each initial cluster center is taken to measure the difference between the two. Implementers can also use other methods to measure differences such as ratios, and this embodiment does not impose any special restrictions.

[0073] (3) When performing power data fusion of the distribution network, it is necessary to analyze the correlation of the power sequences uploaded by multiple terminals. Therefore, based on the initial membership of each eigenvector to each initial cluster center, the membership of each power eigenvector to each cluster center at each iteration is obtained through iteration to improve the accuracy of clustering. The expression of the membership of each power eigenvector to each cluster center at each iteration is: Where, Indicates the membership degree of the i-th power eigenvector to the r-th cluster center at the g-th iteration; Indicates the membership degree of the jth power eigenvector to the rth cluster center at the g-1th iteration; represents the i-th power eigenvector; represents the jth power eigenvector belonging to the rth cluster center at the g-1th iteration; Indicates the modulus value; represents the number of power eigenvectors belonging to the rth cluster center at the g-1th iteration; exp( ) represents the exponential function with a natural constant as the base; norm( ) represents the normalization function.

[0074] The smaller the modulus of the difference between the power eigenvector and the power eigenvector belonging to the cluster center in the previous iteration, the stronger the correlation between the two. Furthermore, the greater the membership of the power eigenvector belonging to the cluster center in the previous iteration, the greater the influence of the power eigenvector's membership, and the larger the calculated new membership. By expanding the analysis range of the power eigenvector to calculate the membership, the correlation between the power eigenvector and the cluster center is enhanced, thereby increasing the redundancy between power sequences in the cluster and improving the reliability of subsequent data fusion.

[0075] (4) Furthermore, since the power mutation data deviates greatly from the cluster center, in order to enhance the degree of distinction of the power feature vector containing the power mutation data and avoid the reduction of data fusion accuracy caused by clustering overfitting, the objective function is calculated. The expression of the objective function is:

[0076]

[0077] Where, Represents the objective function at the g-th iteration; Indicates a constant greater than 0. In order to prevent the denominator from being zero, in this embodiment The value is 1; represents the power mutation index of the i-th power eigenvector; Indicates the membership degree of the i-th power eigenvector to the r-th cluster center at the g-th iteration; represents the membership index; represents the i-th power eigenvector; represents the rth cluster center at the gth iteration; R represents the total number of cluster centers; M is the number of power characteristic vectors of all distribution network terminals in the distribution network.

[0078] Among them, the membership index It is a key parameter in the FCM algorithm, which controls the fuzziness of the clustering results and the membership index. Too large may cause the algorithm to become unstable when iteratively updating the membership index. Too small may cause the algorithm to converge prematurely, so the membership index The value of is usually 2.

[0079] The larger the power mutation index of the power eigenvector, the greater the probability that the corresponding distribution network terminal will generate power mutation data, the smaller the weight it occupies in the calculation of the objective function, and the smaller the impact on the clustering of the power eigenvector. At the same time, the smaller its membership degree, the greater the difference from the other power eigenvectors, thereby enhancing the ability to recognize power eigenvectors containing power mutation data, improving the data cleaning effect of power data and the accuracy of data fusion.

[0080] Preferably, as an embodiment of the present application, the target function acquisition process diagram is as follows: Figure 2 shown.

[0081] Step S4: Using a data fusion algorithm to fuse the power sequences in each data cluster, obtaining a fused power sequence, and completing the fusion of the power data of the distribution network.

[0082] (1) The objective function, all initial cluster centers and all power eigenvectors are used as inputs of the FCM algorithm, and the clusters are output.

[0083] (2) All power data of each distribution network terminal within the sampling interval constitute the power sequence of each distribution network terminal.

[0084] (3) The membership degree of all power feature vectors belonging to the cluster center is used as the input of the threshold segmentation algorithm to obtain the segmentation threshold.

[0085] It should be noted that there are many commonly used threshold segmentation algorithms. In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold. The implementer may also use other threshold segmentation algorithms according to specific circumstances. This embodiment does not impose any special restrictions.

[0086] (4) The power sequences of the distribution network terminals corresponding to the power feature vectors in each cluster whose membership is less than the segmentation threshold are eliminated, and the power sequences of the distribution network terminals corresponding to the power feature vectors in each cluster whose membership is greater than or equal to the segmentation threshold are used as the input of the data fusion algorithm to obtain the power sequences corresponding to each cluster after fusion.

[0087] It should be noted that there are many commonly used data fusion algorithms. In this embodiment, the Kalman filter algorithm is used to perform data fusion on the power sequence. The implementer may also use other data fusion methods, and this embodiment does not impose any special restrictions.

[0088] At this point, based on the membership degree of each power feature vector belonging to each cluster center, combined with the fuzzy mean clustering algorithm and data fusion algorithm, the fusion of power data in the distribution network is completed.

[0089] Preferably, as an embodiment of the present application, the data classification and data fusion process diagram is as follows: Figure 3 shown.

[0090] Preferably, as an embodiment of the present application, the schematic diagram of the power distribution network data fusion method is as follows: Figure 4 shown.

[0091] Based on the same inventive concept as the above method, an embodiment of the present application also provides a data fusion system based on a distribution network cloud platform, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned data fusion methods based on the distribution network cloud platform are implemented.

[0092] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0093] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0094] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A data fusion method based on a distribution network cloud platform, characterized in that: The method comprises the following steps: Obtain power data of each distribution network terminal in the distribution network within a sampling interval; Fit all power data of each distribution network terminal within the sampling interval, and determine the power mutation index of each distribution network terminal based on the difference between all power data of each distribution network terminal within the sampling interval and the fitted result; Determine the power characteristic vector of each distribution network terminal based on the average distribution and dispersion of the power data of each distribution network terminal within the sampling interval and the power mutation index; According to the similarity between the power feature vectors of all distribution network terminals in the distribution network, the objective function is determined by combining the fuzzy mean clustering algorithm to fuse the power data.

2. The data fusion method based on the distribution network cloud platform according to claim 1, characterized in that: The method for determining the power mutation index of each distribution network terminal is as follows: Analyze the difference between each power data of each distribution network terminal in the sampling interval and the corresponding fitting result, and record it as the first difference of each power data of each distribution network terminal in the sampling interval; Analyze the mean of the first differences of all power data of each distribution network terminal within the sampling interval, and record it as the first mean of each distribution network terminal; The power mutation index of each distribution network terminal is a result of fusing the difference between the first difference and the first mean of all power data of each distribution network terminal within a sampling interval.

3. The data fusion method based on the distribution network cloud platform according to claim 1, characterized in that: The power characteristic vector of each distribution network terminal is a vector composed of the mean and variance of all power data of each distribution network terminal within a sampling interval, and the power mutation index.

4. The data fusion method based on the distribution network cloud platform according to claim 1, characterized in that: The objective function is determined by combining the fuzzy mean clustering algorithm, including: Determining a power distribution frequency of each power characteristic vector based on a distribution of power characteristic vector moduli of all distribution network terminals in the distribution network, wherein one distribution network terminal corresponds to one power characteristic vector; According to the power distribution frequency, the power feature vectors corresponding to the preset number of power distribution frequencies are selected from large to small as the initial cluster centers; Based on the similarity between each power eigenvector and each initial cluster center, the initial membership of each power eigenvector to each initial cluster center is determined, and the membership of each power eigenvector to each cluster center at each iteration is obtained through iteration; The objective function of the fuzzy mean clustering algorithm is determined based on the membership degree of each power eigenvector to each cluster center, the difference between each power eigenvector and each cluster center, and the power mutation index at each iteration.

5. The data fusion method based on the distribution network cloud platform according to claim 4, characterized in that: The power distribution frequency of each power eigenvector is the cumulative sum of the frequencies of occurrence of the moduli of all power eigenvectors in the neighborhood of each power eigenvector in the moduli of all power eigenvectors in the power distribution network.

6. The data fusion method based on the distribution network cloud platform according to claim 4, characterized in that: The initial membership degree of each power eigenvector to each initial cluster center is a normalized result of the difference between the modulus value of each power eigenvector and the modulus value of each initial cluster center.

7. The data fusion method based on the distribution network cloud platform according to claim 4, characterized in that: The expression of the membership degree of each power eigenvector to each cluster center during each iteration is: Where, Indicates the membership degree of the i-th power eigenvector to the r-th cluster center at the g-th iteration; Indicates the membership degree of the jth power eigenvector to the rth cluster center at the g-1th iteration; represents the i-th power eigenvector; represents the jth power eigenvector belonging to the rth cluster center at the g-1th iteration; Indicates the modulus value; represents the number of power eigenvectors belonging to the rth cluster center at the g-1th iteration; exp( ) represents the exponential function with a natural constant as the base; norm( ) represents the normalization function.

8. The data fusion method based on the distribution network cloud platform according to claim 7, characterized in that: The expression of the objective function is: Where, Represents the objective function at the g-th iteration; Represents a constant greater than 0; represents the power mutation index of the i-th power eigenvector; represents the membership index; represents the rth cluster center at the gth iteration; R represents the total number of cluster centers; M is the number of power characteristic vectors of all distribution network terminals in the distribution network.

9. The data fusion method based on the distribution network cloud platform according to claim 8, characterized in that: The power data is integrated, including: Taking the objective function, all initial cluster centers and all power eigenvectors as inputs of a fuzzy mean clustering algorithm, and outputting each cluster; Taking all the membership degrees in each cluster as input of the threshold segmentation algorithm to obtain the segmentation threshold; All power data of each distribution network terminal within the sampling interval constitute the power sequence of each distribution network terminal; The power sequences of distribution network terminals corresponding to all power feature vectors with membership greater than or equal to the segmentation threshold in each cluster are used as the input of the data fusion algorithm to obtain the power sequences corresponding to each cluster after fusion.

10. A data fusion system based on a distribution network cloud platform, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the data fusion method based on the distribution network cloud platform as described in any one of claims 1 to 9 are implemented.

Images