Regional autonomy-oriented power distribution network cluster division method

Abstract

The invention discloses a power distribution network cluster division method oriented to regional autonomy, and relates to the technical field of power distribution network cluster division, and the method comprises the steps of data acquisition, preprocessing, electrical distance calculation, similarity evaluation and cluster division. And abundant information sources are provided for subsequent analysis. During data preprocessing, abnormal values of electrical parameters are removed, mean values are taken for standardization, coordinates of geographic information are standardized, and data accuracy and uniformity are improved; the electrical distance calculation measures the node relation based on coordinates and the electrical coupling degree, and the similarity evaluation accurately calculates the similarity degree between the nodes through electrical parameter standardization characteristics. Cluster division starts from a single node, merging is carried out according to the fact that a similarity evaluation value is larger than a threshold value, merging rules are flexible, small clusters are merged into large clusters or targets are randomly selected on the same scale until no node pair meeting conditions exists, and efficient and accurate division is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network cluster division, and in particular to a distribution network cluster division method oriented to regional autonomy. Background Art

[0002] In modern power systems, efficient operation and management of distribution networks are crucial. With the development of social economy, electricity demand continues to grow, the scale of distribution networks is expanding, and the structure is becoming more complex. Traditional distribution network management methods face many challenges and cannot meet the growing demands for reliability, stability, and economy.

[0003] Accurate analysis and control of the distribution network depends on a large amount of accurate data information. The collected raw data often has various problems and needs to be preprocessed. Electrical parameter data may produce outliers due to equipment failure, interference and other factors. Through specific preprocessing methods, these outliers can be eliminated and standardized to make the data more usable and comparable. Geographic information data also needs to be standardized to unify the coordinate system for subsequent analysis and calculation. Calculating the electrical distance and similarity evaluation between nodes helps to gain a deeper understanding of the relationship between nodes in the distribution network. The electrical distance can be used to quantitatively analyze the node relationship in terms of the degree of electrical coupling, and the similarity evaluation measures the similarity between nodes from the perspective of electrical parameter data. Clustering is performed on this basis, and regional nodes with similar operating characteristics can be divided into the same cluster, which is conducive to the zoning management, fault diagnosis, load forecasting and other operations of the distribution network, thereby improving the operation and management efficiency of the distribution network, reducing operation and maintenance costs, and enhancing power supply reliability. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a distribution network cluster division method oriented to regional autonomy.

[0005] The distribution network cluster division method for regional autonomy includes the following steps:

[0006] Step 1: Data Collection

[0007] Collect data features of various attributes related to the distribution network;

[0008] Step 2: Preprocessing

[0009] Preprocess various data features related to the distribution network to obtain data features with standardized attributes;

[0010] Step 3: Calculate the electrical distance

[0011] According to the geographical location coordinates of the nodes in each area after preprocessing, the electrical distance between the nodes in the distribution network is calculated;

[0012] Step 4: Similarity Assessment

[0013] Based on the electrical distance value between nodes in each region and combined with the electrical parameter data, the similarity evaluation value between nodes in each region is calculated;

[0014] Step 5: Cluster division

[0015] The distribution network clusters are divided according to the similarity evaluation values ​​between the nodes in each area, and multiple target clusters are obtained based on the division results.

[0016] As a further solution of the present invention: the data collection method is:

[0017] First, the electrical parameter data on multiple regional nodes in the distribution network are collected, including voltage amplitude, current amplitude, active power, and reactive power;

[0018] Among them, the regional node is a node pre-set in the distribution network for data feature collection;

[0019] Among them, these data are obtained through smart meters and sensors installed at each node, and the collection frequency is set according to actual needs;

[0020] The electrical parameter data also includes: the resistance and reactance values ​​of the line between two adjacent regional nodes;

[0021] Among them, these data are stored in the power system database and can be obtained through the data interface;

[0022] At the same time, the geographic information data of the distribution network area is collected, and it is the geographic location coordinates of the nodes in each area;

[0023] Among them, geographic information data can be obtained through the geographic information system (GIS).

[0024] As a further solution of the present invention: For electrical parameter data, the preprocessing method is as follows:

[0025] Step GS1, define vector

[0026] Select one data feature from multiple attributes and define it as a feature vector E;

[0027] The feature vector collected at multiple time nodes is obtained and recorded as Ee, where e=1, 2, ... t, and t represents the number of time nodes;

[0028] Step GS2: Vector calculation

[0029] Then the mean and standard deviation of all Ee were calculated and labeled as EP and EB respectively;

[0030] Step GS3: Vector analysis

[0031] Then each Ee is combined with EP and EB for outlier analysis:

[0032] If |Ee-EP|>k×EB, it means that the corresponding Ee is an abnormal value;

[0033] If |Ee-EP|≤k×EB, it means that the corresponding Ee is a normal value;

[0034] Wherein, k is a preset weight coefficient, which is obtained based on industry standard experience values ​​or through statistical analysis of a large amount of experimental data, wherein the value of k is 2.5;

[0035] Step GS4: Vector normalization

[0036] Then, all Ee with normal values ​​are extracted and their average value is calculated, and the average value is recorded as the standardized data feature of this attribute.

[0037] As a further solution of the present invention: For geographic information data, the preprocessing method is as follows:

[0038] Step KS1: Coordinate marking

[0039] Extract the geographic location coordinates of all regional nodes and mark the geographic location coordinates of each regional node as (x0 i ,y0 i ), i = 1, 2, ... n, n represents the number of regional nodes;

[0040] Step KS2: Extreme value selection

[0041] Then, among the geographical location coordinates of all regional nodes, the largest horizontal axis coordinate value and vertical axis coordinate value are obtained and marked as x respectively. max and max ; and obtain the smallest horizontal axis coordinate value and vertical axis coordinate value respectively, and mark them as x min and min ;

[0042] Step KS3: Coordinate standardization

[0043] Then select the geographical location coordinates of the regional nodes one by one (x0 i ,y0 i );

[0044] and through and Calculate the standardized geographic location coordinates (x i ,y i ).

[0045] As a further solution of the present invention: the electrical distance is calculated as follows:

[0046] Among all the regional nodes in the distribution network, the geographical coordinates of two regional nodes are selected and marked as (x j1 ,y j1 ) and (x j2 ,y j2 );

[0047] Wherein, j1 and j2 are two regional nodes in i, and j1≠j2;

[0048] Then through Calculate the electrical distance D between the j1th region node and the j2th region node j1，j2 .

[0049] As a further solution of the present invention: wherein the electrical distance is based on the degree of electrical coupling between nodes in each region.

[0050] As a further solution of the present invention: the similarity evaluation value is calculated as follows:

[0051] Step DS1: Data marking

[0052] The standardized data features corresponding to each attribute of the electrical parameter data in the j1th regional node are marked as E r，j1 , r = 1, 2, ... t, t represents the number of electrical parameter data corresponding to all attributes;

[0053] The standardized data features corresponding to each attribute of the electrical parameter data in the j2th regional node are marked as E r，j2 ,

[0054] Step DS2: Calculation of evaluation value

[0055] Then through:

[0056]

[0057] Calculate the similarity evaluation value S between the j1th regional node and the j2th regional node j1，j2 ;

[0058] In this way, the similarity evaluation values ​​between nodes in each region are calculated.

[0059] As a further solution of the present invention: the distribution network cluster division method is as follows:

[0060] First, define each regional node as a separate cluster;

[0061] According to the similarity evaluation values ​​between the regional nodes, two regional nodes whose similarity evaluation values ​​are greater than a preset similarity evaluation threshold are formed into regional node pairs, and then the clusters they belong to are merged;

[0062] Wherein, when the similarity evaluation value between other regional nodes and any regional node in the merged cluster is greater than the similarity evaluation threshold, the other regional nodes are merged into the merged cluster;

[0063] And so on, the merging operation is repeated until there are no more region node pairs that meet the merging conditions.

[0064] As a further solution of the present invention: the merging rule is: merge a smaller cluster into a larger cluster, and if two clusters are of the same size, randomly select one as the target cluster for merging.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] The present invention collects rich electrical parameter data on multiple regional nodes in the distribution network, including voltage amplitude, current amplitude, active power, reactive power, and resistance and reactance values ​​of the lines between two adjacent regional nodes, providing comprehensive electrical characteristic information for the analysis and optimization of the distribution network. At the same time, the geographic information data of the distribution network area is collected, including the geographical location coordinates of each regional node, which helps to comprehensively consider the impact of geographical factors and regional functions on the distribution network.

[0067] The present invention preprocesses the electrical parameter data, performs outlier analysis by calculating the mean value and standard deviation, and obtains the mean value of normal values ​​after removing the outliers as the standardized data feature, thereby improving the accuracy and reliability of the data; and standardizes the geographic information data so that the geographic location coordinates of nodes in different regions have a unified scale, which is convenient for subsequent calculation and analysis.

[0068] The present invention calculates the electrical distance between nodes in the distribution network according to the geographic location coordinates of the nodes in each area after preprocessing, and measures the relationship between the nodes by the degree of electrical coupling, thus providing a scientific and reasonable basis for similarity evaluation and cluster division.

[0069] The present invention calculates the standardized data features of each attribute in the electrical parameter data to obtain the similarity evaluation value between the regional nodes, which can accurately reflect the similarity of different nodes in electrical characteristics.

[0070] The present invention divides the distribution network into clusters according to the similarity evaluation value, merges the regional nodes with higher similarity into one cluster, and the merging rules are flexible, which can adapt to distribution networks of different sizes. At the same time, the merging operation is repeated until the conditions are met, ensuring the accuracy and stability of the cluster division.

[0071] In summary, the present invention provides strong support for the optimized management and operation of the distribution network through comprehensive data collection, accurate data preprocessing, scientific electrical distance calculation, accurate similarity evaluation and flexible and efficient cluster division. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is a schematic diagram of the process of the present invention;

[0073] Figure 2 The process diagram of pretreatment in the method of the present invention is shown in FIG. Figure 1 ;

[0074] Figure 3 The process diagram of pretreatment in the method of the present invention is shown in FIG. Figure 2 . DETAILED DESCRIPTION

[0075] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0076] Embodiment 1

[0077] See also Figure 1 As shown, the present invention is a method for dividing distribution network clusters for regional autonomy, comprising the following steps:

[0078] Step 1: Data Collection

[0079] Collect data features of various attributes related to the distribution network;

[0080] Specifically:

[0081] First, the electrical parameter data on multiple regional nodes in the distribution network are collected, including voltage amplitude, current amplitude, active power, and reactive power;

[0082] Among them, the regional node is a node pre-set in the distribution network for data feature collection;

[0083] In this embodiment, these data are acquired through devices such as smart meters and sensors installed at each node. The acquisition frequency is set according to actual needs. In this embodiment, the acquisition frequency is once per minute;

[0084] The electrical parameter data also includes: the resistance and reactance values ​​of the line between two adjacent regional nodes;

[0085] In this embodiment, these data are stored in the power system database and can be obtained through the data interface;

[0086] At the same time, the geographic information data of the distribution network area is collected, and it is the geographic location coordinates of the nodes in each area;

[0087] In this embodiment, the geographic information data may be obtained through a geographic information system (GIS);

[0088] Step 2: Calculate the electrical distance

[0089] Calculate the electrical distance between nodes in the distribution network based on the geographical location coordinates of nodes in each region;

[0090] Among them, the electrical distance is based on the electrical coupling tightness between nodes in each region;

[0091] The specific method is as follows:

[0092] Among all the regional nodes in the distribution network, the geographical coordinates of two regional nodes are selected and marked as (x j1 ,y j1 ) and (x j2 ,y j2 );

[0093] Wherein, j1 and j2 are two regional nodes in i, and j1≠j2;

[0094] Then through Calculate the electrical distance D between the j1th region node and the j2th region node j1，j2 ;

[0095] Step 3: Similarity Assessment

[0096] Based on the electrical distance value between nodes in each region and combined with the electrical parameter data, the similarity evaluation value between nodes in each region is calculated;

[0097] The specific method is as follows:

[0098] The standardized data features corresponding to each attribute of the electrical parameter data in the j1th regional node are marked as E r，j1 , r = 1, 2, ... t, t represents the number of electrical parameter data corresponding to all attributes;

[0099] The standardized data features corresponding to each attribute of the electrical parameter data in the j2th regional node are marked as E r，j2 ;

[0100] Then through:

[0101]

[0102] Calculate the similarity evaluation value S between the j1th regional node and the j2th regional node j1，j2 ;

[0103] By analogy, the similarity evaluation values ​​between nodes in each region are calculated;

[0104] Step 4: Cluster Division

[0105] The distribution network clusters are divided according to the similarity evaluation values ​​between nodes in each area;

[0106] The specific method is as follows:

[0107] First, define each regional node as a separate cluster;

[0108] According to the similarity evaluation values ​​between the regional nodes, two regional nodes whose similarity evaluation values ​​are greater than a preset similarity evaluation threshold are formed into regional node pairs, and then the clusters they belong to are merged;

[0109] Wherein, when the similarity evaluation value between other regional nodes and any regional node in the merged cluster is greater than the similarity evaluation threshold, the other regional nodes are merged into the merged cluster;

[0110] And so on, the merging operation is repeated until there are no more regional node pairs that meet the merging conditions;

[0111] The merging rule in this embodiment is: merge the smaller cluster into the larger cluster, and if the two clusters are of the same size, randomly select one as the target cluster for merging;

[0112] This embodiment comprehensively collects electrical parameter data and geographic information data of multiple regional nodes in the distribution network through the data collection step, providing a rich data basis for subsequent analysis and calculation, and facilitating more accurate distribution network cluster division; the electrical distance calculation is based on the geographical location coordinates of the regional nodes, and the electrical distance between nodes is measured by the degree of electrical coupling, which provides an important basis for similarity evaluation and can more objectively reflect the actual relationship between nodes; the similarity evaluation combines the electrical distance value and the electrical parameter data to calculate the similarity evaluation value between each regional node, making the similarity judgment between nodes more scientific and reasonable; cluster division is performed based on the similarity evaluation value, and regional nodes with high similarity are grouped into clusters, realizing the reasonable division of the distribution network, which is conducive to the realization of regional autonomy and improving the operation efficiency and management level of the distribution network.

[0113] Embodiment 2

[0114] See also Figure 2 and Figure 3 As shown, as the second embodiment of the present invention, when the present application is implemented, compared with the first embodiment, the technical solution of this embodiment is different from that of the first embodiment only in that:

[0115] This embodiment also includes the following pre-processing steps:

[0116] Preprocess various data features related to the distribution network to obtain data features with standardized attributes;

[0117] For electrical parameter data, the preprocessing method is as follows:

[0118] Select a data feature from multiple attributes and define it as a feature vector E;

[0119] The feature vector collected at multiple time nodes is obtained and recorded as Ee, where e=1, 2, ... t, and t represents the number of time nodes;

[0120] Then the mean and standard deviation of all Ee were calculated and labeled as EP and EB respectively;

[0121] Then each Ee is combined with EP and EB for outlier analysis:

[0122] If |Ee-EP|>k×EB, it means that the corresponding Ee is an abnormal value;

[0123] If |Ee-EP|≤k×EB, it means that the corresponding Ee is a normal value;

[0124] Wherein, k is a preset weight coefficient, and in this embodiment, the value of k is 2.5;

[0125] Then, all Ee with normal values ​​are extracted and their average value is calculated, and then the average value is recorded as the standardized data feature of this attribute;

[0126] For geographic information data, the preprocessing method is as follows:

[0127] Extract the geographic location coordinates of all regional nodes and mark the geographic location coordinates of each regional node as (x0 i ,y0 i ), i = 1, 2, ... n, n represents the number of regional nodes;

[0128] Then, among the geographical location coordinates of all regional nodes, the largest horizontal axis coordinate value and vertical axis coordinate value are obtained and marked as x respectively.max and max ; and obtain the smallest horizontal axis coordinate value and vertical axis coordinate value respectively, and mark them as x min and min ;

[0129] Then select the geographical location coordinates of the regional nodes one by one (x0 i ,y0 i );

[0130] and through and Calculate the standardized geographic location coordinates (x i ,y i );

[0131] This embodiment adds a preprocessing step to standardize various data features related to the distribution network, so that data with different attributes are comparable, thereby improving the accuracy of subsequent analysis and calculation; for the preprocessing of electrical parameter data, outliers are removed through outlier analysis, and the average value of normal values ​​is taken as the standardized data feature, thereby effectively improving the quality of electrical parameter data; for the preprocessing of geographic information data, the geographic location coordinates are standardized, so that the geographic location information of nodes in different regions can be compared at the same scale, thereby facilitating subsequent calculations and analysis.

[0132] Embodiment 3

[0133] As the third embodiment of the present invention, when the present application is specifically implemented, compared with the first and second embodiments, the technical solution of this embodiment is to combine the solutions of the first and second embodiments mentioned above for implementation.

[0134] This embodiment combines the solutions of Embodiment 1 and Embodiment 2, which can not only comprehensively collect and analyze the data characteristics of the distribution network, perform accurate electrical distance calculation, similarity evaluation and cluster division, but also improve data quality and comparability through preprocessing steps, further improving the scientificity and effectiveness of distribution network cluster division; it provides a more complete technical means for realizing distribution network management for regional autonomy, and helps to improve the operating stability, reliability and management efficiency of the distribution network.

[0135] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0136] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A distribution network cluster division method for regional autonomy, characterized in that: The following steps are involved: Step 1: Collect data features related to the distribution network, including electrical parameter data and geographic location coordinates of multiple attributes at multiple regional nodes in the distribution network; Step 2: Preprocess various data features related to the distribution network to obtain data features with standardized attributes; Step 3: Calculate the electrical distance between nodes in the distribution network based on the geographical location coordinates of nodes in each region after preprocessing; Step 4: Based on the electrical distance value between the nodes in each region and in combination with the electrical parameter data, the similarity evaluation value between the nodes in each region is calculated; Step 5: Divide the distribution network clusters according to the similarity evaluation values ​​between the nodes in each area, and obtain multiple target clusters based on the division results.

2. The method for dividing distribution network clusters for regional autonomy according to claim 1, characterized in that: For electrical parameter data, the preprocessing method is as follows: Select one data feature from multiple attributes and define it as a feature vector E; The feature vector collected at multiple time nodes is obtained and recorded as Ee, where e=1, 2, ... t, and t represents the number of time nodes; Then the mean and standard deviation of all Ee were calculated and labeled as EP and EB respectively; Then, each Ee is combined with EP and EB for outlier analysis. According to the analysis results, it is determined whether Ee is an outlier: Then, all Ee with normal values ​​are extracted and their average value is calculated, and the average value is recorded as the standardized data feature of this attribute.

3. The method for dividing distribution network clusters for regional autonomy according to claim 2, characterized in that: The outlier analysis is as follows: If |Ee-EP|>k×EB, it means that the corresponding Ee is an abnormal value; If |Ee-EP|≤k×EB, it means that the corresponding Ee is a normal value; Among them, k is a preset weight coefficient.

4. The method for dividing distribution network clusters for regional autonomy according to claim 2, characterized in that: For geographic information data, the preprocessing method is as follows: Extract the geographic location coordinates of all regional nodes and mark the geographic location coordinates of each regional node as (x0 i ,y0 i ), i = 1, 2, ... n, n represents the number of regional nodes; Then, among the geographical location coordinates of all regional nodes, the largest horizontal axis coordinate value and vertical axis coordinate value are obtained and marked as x respectively. max and max ; And get the smallest horizontal axis coordinate value and vertical axis coordinate value respectively, and mark them as x min and min ; Then select the geographical location coordinates of the regional nodes one by one (x0 i ,y0 i ); and through and Calculate the standardized geographic location coordinates (x i ,y i ).

5. The method for dividing distribution network clusters for regional autonomy according to claim 1, characterized in that: The electrical distance is calculated as follows: Among all the regional nodes in the distribution network, the geographical coordinates of two regional nodes are selected and marked as (x j1 ,y j1 ) and (x j2 ,y j2 ); Wherein, j1 and j2 are two regional nodes in i, and j1≠j2; Then, the electrical distance D between the j1th regional node and the j2th regional node is calculated using the Euclidean distance formula. j1，j2 .

6. The method for dividing distribution network clusters for regional autonomy according to claim 5, characterized in that: in, The electrical distance is the degree of electrical coupling between nodes in each area.

7. The method for dividing distribution network clusters for regional autonomy according to claim 5, characterized in that: The similarity rating is calculated as follows: The standardized data features corresponding to each attribute of the electrical parameter data in the j1th regional node are marked as E r，j1 , r = 1, 2, ... t, t represents the number of electrical parameter data corresponding to all attributes; The standardized data features corresponding to each attribute of the electrical parameter data in the j2th regional node are marked as E r，j2 , Then it is combined with the electrical distances in the corresponding adjacent area nodes and the similarity evaluation value formula is used to calculate the similarity. Calculate the similarity evaluation value S between the j1th regional node and the j2th regional node j1，j2 ; In this way, the similarity evaluation values ​​between nodes in each region are calculated.

8. The method for dividing distribution network clusters for regional autonomy according to claim 7, characterized in that: The similarity evaluation value formula is as follows:

9. The method for dividing distribution network clusters for regional autonomy according to claim 1, characterized in that: The distribution network clusters are divided as follows: First, define each regional node as a separate cluster; According to the similarity evaluation values ​​between the regional nodes, two regional nodes whose similarity evaluation values ​​are greater than a preset similarity evaluation threshold are formed into regional node pairs, and then the clusters they belong to are merged; And so on, the merging operation is repeated until there are no more region node pairs that meet the merging conditions.

10. The method for dividing distribution network clusters for regional autonomy according to claim 9, characterized in that: in, When the similarity evaluation value between other regional nodes and any regional node in the merged cluster is greater than the similarity evaluation threshold, the other regional nodes are merged into the merged cluster.

Images