A method for equivalent modeling of low voltage area load and topology

Through the low-voltage table area load and topological equivalent modeling method based on the overall determination method, combined with the K-Means clustering algorithm and voltage similarity principle, the problems of inaccurate data and high cost in the existing technology are solved, and the rapid and Accurate load and topological equivalent modeling, suitable for complex line situations.

CN113935136BActive Publication Date: 2024-11-05YUNNAN POWER GRID CO LTD ELECTRIC POWER RES INST
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110734024.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-06-30
Publication Date
2024-11-05
Estimated Expiration
2041-06-30

AI Technical Summary

Technical Problem

The existing low-voltage platform load and topological equivalent modeling methods have problems such as inaccurate data, high cost and complexity, which are difficult to model. In particular, the traditional statistical comprehensive method is not suitable for time-varying load characteristics research, and the overall measurement method is A large number of monitoring equipment and long monitoring cycles are required.

Method used

The load and topological equivalent modeling method of low-voltage table area based on overall detection method is adopted to obtain line topology map information to perform load equivalent, and combined with the K-Means clustering algorithm and the principle of voltage similarity and distance similarity, simplify data acquisition and Compute to quickly complete topological equivalent modeling.

Benefits of technology

It realizes simple, fast, scientific and accurate load and topological equivalent modeling in low-voltage table areas, reducing data acquisition requirements and computing complexity, and is suitable for load modeling in case of complex line topology.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113935136B_ABST
    Figure CN113935136B_ABST
Patent Text Reader

Abstract

The present application shows a method for modeling the load and topology equivalence of a low-voltage substation, which is based on the overall measurement and identification method. The method steps are: obtaining the line topology map information of the low-voltage substation; performing load equivalence according to the line topology map information; after completing the load equivalence, obtaining the user load capacity, comparing the user load capacity with the topology library, determining the topology equivalence type, performing topology equivalence, and obtaining the low-voltage substation load and topology equivalent model. The present application shows a method for modeling the load and topology equivalence of a low-voltage substation, which is based on the overall measurement and identification method. The modeling is simple, the calculation is convenient, the data required to be collected is small, the line topology equivalence modeling work can be completed quickly, and the low-voltage substation load and topology are scientifically and accurately modeled equivalently.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of power system modeling, and in particular relates to a low-voltage area load and topology equivalent modeling method. Background Art

[0002] When planning, designing, dispatching, operating and analyzing power systems, load and topology equivalent modeling is required.

[0003] At present, there are mainly two methods for modeling the load and topology equivalent of low-voltage substations. The first is a load modeling method based on components, namely the statistical synthesis method. This method is based on statistical principles, regards the load as a collection of individual users, classifies electrical appliances and determines the average characteristics of various types of electrical appliances, and obtains a comprehensive load model based on the proportion of each type of electrical appliance. The basis of this method is to obtain the composition and capacity ratio of each typical electrical equipment through experiments and mathematical derivation, and finally integrate these data to obtain the static load model of the load point. When using this method, the average characteristic data of a single electrical equipment and the composition ratio data of electrical equipment in various types of loads are obtained through typical statistics. These two types of data are stable common data. At the same time, it is also necessary to obtain the composition ratio of various loads to reflect the particularity of the load. The second is a load modeling method based on measurement, namely the overall measurement and identification method. This method is based on system identification theory, takes the comprehensive load as a whole, obtains the voltage, frequency, current, active power, and reactive power data of the bus where the load is located through field tests and online capture of natural disturbances in the power system, and then identifies the structure and model parameters of the comprehensive load model based on these data.

[0004] The statistical synthesis method is a traditional modeling method. It does not require on-site measurements during modeling and is less costly. However, statistical data is time-consuming, labor-intensive and inaccurate, and cannot be performed frequently. It is not suitable for studying the time-varying nature of load characteristics. The overall identification method does not require understanding the complex internal structure of the load and is a feasible method for solving load group modeling. It is divided into online identification and offline identification. The online identification method can better reflect the changes in load characteristics over time, but it requires the installation of a large number of monitoring equipment at the load point, which is very expensive. In addition, since large disturbances do not occur frequently, the monitoring cycle is long. Summary of the invention

[0005] Based on the above problems, the present invention provides a low-voltage substation load and topology equivalent modeling method based on the overall measurement and identification method. The modeling is simple, the calculation is convenient, less data needs to be collected, and the line topology equivalent modeling work can be completed quickly.

[0006] This application shows a method for modeling low-voltage load and topology equivalent, the method steps are:

[0007] Obtain low voltage area line topology information;

[0008] Perform load equivalence according to the line topology information;

[0009] After the load equivalence is completed, the user load capacity is obtained, the user load capacity is compared with the topology library, the topology equivalence type is determined, the topology equivalence is performed, and the low-voltage area load and topology equivalence model is obtained.

[0010] Optionally, the load equivalence method includes: using a K-Means clustering algorithm to perform load equivalence;

[0011] The method steps for load equivalence using the K-Means clustering algorithm are as follows:

[0012] According to the topology information, a load distribution distance value d is preset and the trunk line length L is obtained, the number of load division areas K=L / d, and the load area is divided using a K-Means clustering algorithm to obtain K load division areas;

[0013] Calculate the voltage mathematical expectations E(U) of the K load division areas respectively, and calculate the voltage standard deviation σ according to the voltage mathematical expectations E(U);

[0014] Load equivalence is performed according to the voltage standard deviation σ; the load equivalence method is: if the voltage standard deviation σ of any load division area satisfies the first value range, the area completes load equivalence; if the voltage standard deviation of any load division area does not satisfy the first value range, then the users in the area whose voltage deviation ΔU% satisfies the second value range are screened to determine whether the user can be summarized into an adjacent partition. If so, it is summarized into the adjacent partition and the area completes load equivalence. If not, the preset load distribution distance d is corrected according to the third value range, the K value is recalculated, and the calculation is performed according to the new K value, and it is iterated multiple times until all users complete load equivalence.

[0015] Optionally, the method steps of dividing the load area using the K-Means clustering algorithm are:

[0016] Obtain all user loads in the low-voltage area and establish a load sample set D, where D = {o i}; where oi (i = 1 to N) is the user load sample, and N is the number of connected users;

[0017] Randomly select K user load samples from the load sample set D as the initial center point m j (j=1~K); Calculate user load sample o i To the initial center point m j The distance l ij ;

[0018] The user load sample o i Divide to lij The clustering partition cluster C corresponding to the minimum value of j (j=1~K), and update the clustering cluster C j , where C j =C j ∪{o i}(j=1~K), all user load samples o i Clustering is performed to form K clustering clusters C j The K clusters form a sample set C = {C j}(j=1~K); Calculate each cluster partition cluster C j The square error SSE;

[0019] Divide each cluster into clusters C j Perform heuristic iterative calculations to divide the clusters into clusters C j Recalculate the new center point m from the sample points in j , re-cluster the clusters, and after M iterations, obtain the partition's minimized square error SSE, output the updated partition area sample set, complete the partition, and obtain K load partition areas.

[0020] Optionally, after completing load equivalence, the method for obtaining user load capacity is:

[0021] Calculate the user load capacity within the K load partition areas.

[0022] Optionally, the load equivalence method further includes: using a voltage similarity principle to perform load equivalence;

[0023] The method steps for load equivalence using the voltage similarity principle are as follows:

[0024] The average voltage value in the user area is obtained, and the user with the smallest voltage deviation is taken as the center of the circle and 45 meters as the radius. The users in the area with a deviation of less than or equal to 2% are grouped as the same node.

[0025] Optionally, the load equivalence method further includes: performing load equivalence using a principle of close distance;

[0026] The method steps for load equivalence using the principle of close distance are as follows:

[0027] Nodes with unknown voltage data whose distances to load nodes with known voltage data are within a first distance range are assimilated into the same node.

[0028] Optionally, the load equivalence method further includes: using a principle of close distance and a principle of similar voltage to perform load equivalence;

[0029] The method steps for load equivalence using the principle of close distance and similar voltage are as follows:

[0030] Assimilate the unknown voltage data nodes whose distances to the load nodes of known voltage data are within a first distance range as the same node;

[0031] The average voltage value in the user area is obtained, and the user with the smallest voltage deviation is taken as the center of the circle and 45 meters as the radius. The users in the area with a deviation of less than or equal to 2% are grouped as the same node.

[0032] The beneficial effects of this application are:

[0033] A method for modeling low-voltage substation load and topology equivalence shown in the present application is based on the overall measurement and identification method, and the method steps include: obtaining the line topology map information of the low-voltage substation; performing load equivalence according to the line topology map information; after completing the load equivalence, obtaining the user load capacity, comparing the user load capacity with the topology library, determining the topology equivalence type, performing topology equivalence, and obtaining the low-voltage substation load and topology equivalent model. A method for modeling low-voltage substation load and topology equivalence shown in the present application is based on the overall measurement and identification method, with simple modeling, convenient calculation, less data required to be collected, and the line topology equivalence modeling work can be completed quickly, and the low-voltage substation load and topology can be modeled equivalently in a scientific and accurate manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solution of the application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 A schematic diagram of the steps of a low voltage area load and topology equivalent modeling method shown in this application;

[0036] Figure 2 This is a step diagram of the method for load equivalence using the K-Means clustering algorithm shown in this application;

[0037] Figure 3 This is a flow chart of a method for load equivalence using a K-Means clustering algorithm as shown in this application;

[0038] Figure 4 This is the line topology information of an embodiment shown in this application;

[0039] Figure 5 This is a schematic diagram of the K-Means algorithm principle of an embodiment of the present application;

[0040] Figure 6This is a load equivalent schematic diagram of an embodiment of the present application;

[0041] Figure 7 This is a schematic diagram of load distribution of an embodiment of the present application;

[0042] Figure 8 For this application Figure 7 The load equivalent schematic diagram of the embodiment shown;

[0043] Fig. 9 A schematic diagram of load distribution of another embodiment of the present application;

[0044] Fig.10 For this application Fig. 9 A schematic diagram of load equivalence of the embodiment;

[0045] Fig.11 This is a schematic diagram of load distribution of an embodiment of the present application;

[0046] Fig.12 This is a schematic diagram of a load equivalent model with a deviation of less than 2% according to the voltage similarity principle according to an embodiment of the present application;

[0047] Fig.13 This is a schematic diagram of load distribution of another embodiment of the present application;

[0048] Fig.14 A schematic diagram of a load equivalent model with a deviation greater than 2% according to the voltage similarity principle according to another embodiment of the present application;

[0049] Fig.15 This is a schematic diagram of load distribution of an embodiment of the present application;

[0050] Fig.16 This is a schematic diagram of a load equivalent model with a small distance difference according to the principle of close distance according to an embodiment of the present application;

[0051] Fig.17 This is a schematic diagram of load distribution of another embodiment of the present application;

[0052] Fig.18 This is another embodiment of the present application, which is a schematic diagram of a load equivalent model when the distance difference is large according to the principle of close distance;

[0053] Fig.19 This is a schematic diagram of a load equivalent model according to the voltage similarity principle and the distance proximity principle according to an embodiment of the present application;

[0054] Fig. 20 This is a model of uniform load distribution along the line according to an embodiment of the present application;

[0055] Fig.21This is a load distribution model along the line that is an embodiment of the present application;

[0056] Fig. 22 This is a load reduction distribution model along the line according to an embodiment of the present application;

[0057] Fig.23 This is a load increment sawtooth distribution model of an embodiment shown in the present application;

[0058] Fig.24 This is a load reduction sawtooth distribution model of an embodiment shown in the present application;

[0059] Fig.25 This is a load trapezoidal distribution model of an embodiment shown in this application;

[0060] Fig.26 This is a schematic diagram of load and topology equivalent modeling according to an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solutions and advantages of the exemplary embodiments of the present application clearer, the technical solutions in the exemplary embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the exemplary embodiments of the present application. Obviously, the described exemplary embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0062] In order to facilitate the understanding of this application, the professional terms involved in this application are first introduced:

[0063] The overall measurement method is based on the data measured on site. The structure and model parameters of the model are determined according to the measured fault data, so that the model response fits the measured load data well. It is also required to pass the model verification to ensure that the constructed model has good extrapolation and interpolation capabilities within the simulation calculation requirements, so that the model can both highlight the essence and simplify the description of the load behavior. Determining the model structure of the measured load characteristic data and identifying the model parameters are two important issues in the overall measurement method load modeling. The overall measurement method has two data sources: data collected under human interference and data collected under natural disturbances. Therefore, from the perspective of parameter identification, it can be divided into online identification and offline identification. The online identification method can better reflect the changes in load characteristics over time, but it requires the installation of a large number of monitoring equipment at the load point, which is very expensive, and because large disturbances do not occur frequently, the monitoring cycle is long. The overall measurement method does not need to understand the complex structure of the load, and is a feasible method for solving the modeling of load groups composed of thousands of electrical equipment.

[0064] K-Means algorithm is the most commonly used clustering algorithm. The input of this algorithm is a sample set (or point set). Through this algorithm, samples with similar characteristics can be clustered into one category.

[0065] The specific implementation process of the algorithm is as follows: calculate the distance from each point to all center points, select the center point with the closest distance, and then assign the point to the cluster represented by this center point. After one iteration, recalculate the center point for each cluster, and each point re-searches for the center point closest to itself. Repeat multiple iterations until a certain termination condition is met. The termination condition is any of the following conditions: (1) no (or a minimum number of) objects are reallocated to different clusters; (2) no (or a minimum number of) cluster centers change again; (3) the sum of squared errors is minimized.

[0066] See also Figure 1 , Figure 1 A schematic diagram of a method for modeling low-voltage load and topology equivalent is shown, and the method steps are:

[0067] S1: Obtain the line topology information of the low voltage area;

[0068] S2: performing load equivalence according to the line topology information;

[0069] In a feasible embodiment, the load equivalence method includes: using a K-Means clustering algorithm to perform load equivalence;

[0070] See also Figure 2 and Figure 3 , Figure 2 The figure shows the steps of the method for load equivalence using the K-Means clustering algorithm. Figure 3 A flow chart of a method for load equivalence using a K-Means clustering algorithm is shown;

[0071] See also Figure 4 , Figure 4 The line topology information in a feasible embodiment is shown. Taking the 10KV Dazhai line Fengjia public substation as an example, the capacity of the distribution transformer in this substation is 50KVA. The low-voltage substation line topology information obtained includes: the length of the main line, the number of main line towers, the distance between towers, the number of access users, the location of access users and the voltage data of access users. In this embodiment, the length of the main line is 376 meters, the number of main line towers is 9, the distance between towers is 47 meters, the number of access users is 677, and the voltage data of access users can be found in Table 1.

[0072] Table 1 Voltage data of connected users

[0073]

[0074] The method steps for load equivalence using the K-Means clustering algorithm are as follows:

[0075] S211: Preset a load distribution distance value d according to the topology information and obtain the trunk line length L, the number of load division areas K=L / d, and use the K-Means clustering algorithm to divide the load area to obtain K load division areas;

[0076] The preset load distribution distance value d is set according to the average distance between towers.

[0077] See also Figure 5 , Figure 5 A schematic diagram of the K-Means algorithm principle of a feasible embodiment is shown, and the steps of this embodiment are as follows:

[0078] S101: Set the clustering sample targets to be clustered into 3 categories;

[0079] S102: Selecting three center points for clustering according to the clustering sample target;

[0080] S103: Calculate the distance from each point to all center points, select the center point with the closest distance, and then classify the point into the cluster represented by this center point. After one iteration, determine whether the clustering of the sample points is the same as before this clustering. If so, the algorithm terminates, if not, proceed to the next step.

[0081] S104: Update the center point and determine whether the clustering situation of the sample points before clustering is the same. If the center point is the same, the algorithm terminates, otherwise proceed to the next step.

[0082] S105: Determine whether the clustering situation of the sample points is the same as that before the current clustering. If the clustering situation is the same, the algorithm terminates.

[0083] S106: Obtain algorithm results.

[0084] The method of dividing the load area using the K-Means clustering algorithm is:

[0085] S2101: Obtain all user loads in the low-voltage area and establish a load sample set D, where D = {o i}; where o i (i=1~n) is the user load sample, n is the number of connected users;

[0086] S2102: Randomly select K user load samples from the load sample set D as the initial center point m j (j=1~K);

[0087] S2103: Calculate user load sample o i To the initial center point m j The distance l ij , the l ij The calculation formula is:

[0088]

[0089] The user load sample o i Divide to l ij The clustering partition cluster C corresponding to the minimum value of j (j=1~K), and update the clustering cluster C j , where C j =C j ∪{o i}(j=1~K), all user load samples o i Clustering is performed to form K clustering clusters C j The K clusters form a sample set C = {C j}(j=1~K); Calculate each cluster partition cluster C j The square error SSE;

[0090] The calculation formula of the SSE is:

[0091]

[0092] S2104: Divide each cluster into clusters C j Perform heuristic iterative calculations to divide the clusters into clusters C j Recalculate the new center point from the sample points in Re-clustering is performed, and after M iterations, the partitioning minimizes the square error SSE, outputs the updated partition area sample set, completes the partitioning, and obtains K load partition areas.

[0093] In a feasible embodiment, the data of this embodiment are as shown in the above embodiment. The number of load division areas is set to 7, and 7 users are randomly selected as the initial cluster centers. Then, the distance between each user and each seed cluster center is calculated, and each user is assigned to the cluster center closest to it. Each time a sample is assigned, the cluster center will be recalculated based on the existing objects until the algorithm terminates.

[0094] See also Figure 6 , Figure 6 FIG. 2 shows a load equivalent schematic diagram of this embodiment.

[0095] S212: Calculate the voltage mathematical expectations E(U) of the K load division areas respectively, and calculate the voltage standard deviation σ based on the voltage mathematical expectations E(U); the voltage standard deviation σ is used to calculate the discrete degree of voltage data in the load division area and the difference between each user, and then evaluate the voltage situation in the area to determine whether the voltage similarity principle can be applied.

[0096] The voltage mathematical expectation E(U) calculation formula is:

[0097]

[0098] Among them U i (i=1~n) is the voltage of the connected users, and n is the number of connected users;

[0099] The voltage standard deviation σ is calculated as follows:

[0100]

[0101] In a feasible embodiment, the data of this embodiment are as shown in the above embodiment, and the voltage mathematical expectation E(U) of the seven load division areas of this embodiment and the voltage standard deviation σ data calculated according to the voltage mathematical expectation E(U) are shown in the following table:

[0102] Table 2 Voltage mathematical expectation E(U) and voltage standard deviation σ

[0103]

[0104] S213: Perform load equivalence according to the voltage standard deviation σ; the load equivalence method is: if the voltage standard deviation σ of any load division area satisfies a first value range, the load equivalence of the area is completed; the first value range is: σ≤4.4;

[0105] See also Figure 7 , Figure 7 A load distribution diagram of a feasible embodiment is shown. In this embodiment, the voltage standard deviation σ≤4.4, then the area can be directly equivalent. The equivalent result can be found in Figure 8 ;

[0106] If the voltage standard deviation of any load division area does not meet the first value range, and its voltage standard deviation σ>4.4, then select users in the area whose voltage deviation ΔU% meets the second value range, and the second value range is ΔU%>2%;

[0107] Determine whether the user can be classified into an adjacent partition. If so, classify it into the adjacent partition, and the load equivalence of the area is completed;

[0108] See also Fig. 9 , Fig. 9 This is another load distribution diagram of a feasible embodiment. In this embodiment, the voltage standard deviation σ>4.4, and the voltage deviation of each user in the area is calculated respectively. The user voltage deviations are:

[0109] ...

[0111]

[0112]

[0113] If ΔU1%>2%, the equivalent result can be found in Fig.10 ; U1 is then adapted to the adjacent area. If ΔU1% meets the standard of the adjacent area, U1 can be equivalent to this area;

[0114] If not, the preset load distribution distance d is corrected according to the third value range, wherein the third value range is: -5%; the K value is recalculated, and the calculation is performed according to the new K value, and it is iterated multiple times until all users complete the load equivalence;

[0115] In a feasible embodiment, the data of the embodiment are as shown in the above embodiment, and the equivalent voltage conditions of each user in each area are shown in the following table:

[0116] Table 3 The first area

[0117] Serial number user Voltage Voltage deviation 1 Gao*Shun 217.117 1.44% 2 Ceramics 221.53 0.56% 3 Wang*Lei 219.457 0.38% 4 Luo*Ju 221.774 0.67% 5 Tao*Shou 222.891 1.18% 6 Li*Ping 223.001 1.23%

[0118] Table 4 Second Region

[0119] Serial number user Voltage Voltage deviation 1 Ruan*Ju 221.819 0.51% 2 Wang*qing 223.465 1.25% 3 Salary* 223.457 1.25% 4 Salary* key 222.217 0.69% 5 Salary* 223.84 1.42% 6 Feng*chang 225.041 1.97% 7 Pottery bag 224.114 1.55% 8 Tao*Qing 224.505 1.72%

[0120] Table 5 The third region

[0121] Serial number user Voltage Voltage deviation 1 Luo*Qing 224.036 1.65% 2 Salary* 224.539 1.88% 3 Feng*Mei 223.233 1.29% 4 Wang*Qi 224.621 1.92% 5 Feng*Hua 223.001 1.18% 6 Wang*Ming 221.819 0.64% 7 Liu*shun 223.465 1.39% 8 Zhao*xue 223.457 1.39% 9 Sun*Lei 222.217 0.82%

[0122] Table 6 The fourth region

[0123]

[0124]

[0125] Table 7 Fifth Region

[0126] Serial number user Voltage Voltage deviation 1 Zhao*shun 222.217 1.10% 2 Ceramics 223.84 1.84% 3 Wang*Lei 224.141 1.97% 4 Luo*Ju 224.114 1.96% 5 Tao*Shou 224.105 1.96% 6 Tao*Ping 224.036 1.93% 7 Wang*xiu 224.039 1.93% 8 Wang*qing 223.233 1.56% 9 Wu*Xian 223.621 1.74% 10 Wu*Jian 224.005 1.91% 11 Wu*Xue 222.891 1.41% 12 Feng*Lan 223.001 1.46%

[0127] Table 8 Region 6

[0128] Serial number user Voltage Voltage deviation 1 pottery* 221.819 1.15% 2 pottery* 223.465 1.90% 3 Luo*Qing 223.457 1.90% 4 Salary* 222.891 1.64% 5 Salary* key 223.008 1.69% 6 Wang*Qi 223.036 1.70% 7 Feng*Hua 223.539 1.93% 8 Wang*Bo 223.233 1.79% 9 Tao*Li 223.114 1.74%

[0129] Table 9 Region 7

[0130] Serial number user Voltage Voltage deviation 1 Nguyen*Chi 217.117 0.68% 2 Wang*Han 219.457 0.39% 3 Salary* 221.774 1.45% 4 Salary*Celebration 222.891 1.96% 5 Stable salary 222.901 1.97% 6 Wang*tu 221.819 1.47%

[0131] It can be seen from the data in the table that the regional division according to the K-Means clustering algorithm is reasonable. In the first to seventh regions, the voltage standard deviation is within the first value range, and the user deviation in the region is within the second value range.

[0132] In a feasible embodiment, the load equivalence method further includes: performing load equivalence using a voltage similarity principle;

[0133] The method steps for load equivalence using the voltage similarity principle are as follows:

[0134] The average voltage value in the user area is obtained, and the user with the smallest voltage deviation is taken as the center of the circle and 45 meters as the radius. The users in the area with a deviation of less than or equal to 2% are grouped as the same node.

[0135] The calculation formula of the voltage average value is:

[0136]

[0137] Said is the average voltage, U1~U n It is the user's load voltage.

[0138] The calculation formula of the voltage deviation is:

[0139]

[0140] The ΔU n % is the voltage deviation, the U n It is the user's load voltage.

[0141] See also Fig.11 , Fig.11 This is a schematic diagram of load distribution in a feasible embodiment; the load voltage information is:

[0142] U1=220V、U2=221V、U3=119V,According to the calculation formula of the average voltage:

[0143]

[0144] The voltage average

[0145] The voltage deviations of U1, U2 and U3 are:

[0146]

[0147]

[0148]

[0149] In a feasible embodiment, the first voltage deviation range is 2%. Since 0.45%<2% and the distances between the three points are all less than 90 meters, U1, U2, and U3 can be classified as the same node. Fig.12 , Fig.12 A schematic diagram of the load equivalent model of this embodiment is shown.

[0150] See also Fig.13 , Fig.13 This is a schematic diagram of load distribution in another feasible embodiment, wherein the load voltage information is:

[0151] U4=220V、U5=221V、U6=230V,According to the calculation formula of the average voltage:

[0152]

[0153] The voltage average

[0154] The voltage deviations of U4, U5 and U6 are:

[0155]

[0156]

[0157]

[0158] Since 2.8%>2%, U5 and U6 can be classified as the same node, and U4 as another node. Fig.14 , Fig.14 A schematic diagram of the load equivalent model of this embodiment is shown.

[0159] In a feasible embodiment, the load equivalence method further includes: performing load equivalence using a principle of close distance;

[0160] The method steps for load equivalence using the principle of close distance are as follows:

[0161] Nodes with unknown voltage data whose distances to load nodes with known voltage data are within a first distance range are assimilated into the same node.

[0162] See also Fig.15 , Fig.15 This is a schematic diagram of load distribution in a feasible embodiment, and the load voltage information is: U'1 = 219V;

[0163] U'2=220V,U'4=220V,U'6=223V,wherein the voltage data of U'3 and U'5 are unknown; according to the principle of close distance, within a certain distance, the voltage deviation will not be too large. Therefore, in the absence of voltage data, the load with unknown voltage data can be classified into one category with other loads within a certain distance. Fig.16 , Fig.16 A schematic diagram of the load equivalent model of this embodiment is shown.

[0164] See also Fig.17 , Fig.17FIG. 1 is a schematic diagram of load distribution in another feasible embodiment, wherein U'7 and U'8 are relatively close in space, while U'9 is relatively far away in space. Therefore, U'7 and U'8 are classified as one type of node, and U'9 is classified as a separate type. Fig.18 , Fig.19 A schematic diagram of the load equivalent model of this embodiment is shown.

[0165] In a feasible embodiment, the load equivalence method further includes: using the principle of close distance and the principle of similar voltage to perform load equivalence;

[0166] The method steps for load equivalence using the principle of close distance and similar voltage are as follows:

[0167] Assimilate the unknown voltage data nodes whose distances to the load nodes of known voltage data are within a first distance range as the same node;

[0168] The average voltage value in the user area is obtained, and the user with the smallest voltage deviation is taken as the center of the circle and 45 meters as the radius. The users in the area with a deviation of less than or equal to 2% are grouped as the same node.

[0169] See also Fig.19 , Fig.19 The figure is a schematic diagram of a load equivalent model using the principle of close distance and the principle of similar voltage in a feasible embodiment.

[0170] S3: When the load equivalence is completed, the user load capacity is obtained, the load capacity is compared with the topology library, the topology equivalence type is determined, the topology equivalence is performed, and the low-voltage area load and topology equivalence model is obtained.

[0171] After completing the load equivalence, the load distribution along the line is established to establish a topological equivalent model. Figure 20-Figure 25 , six topological equivalent models are shown respectively, and a topological library is constructed, which includes: Fig. 20 The load is evenly distributed along the line. Fig.21 The load distribution model along the line is shown. Fig. 22 The load reduction distribution model along the line is shown. Fig.23 The load increasing sawtooth distribution model shown, Fig.24 The load reduction sawtooth distribution model shown and Fig.25 The load trapezoidal distribution model shown;

[0172] Among them, in the model of load distribution along the line, other lines at the head end that do not need to be studied are merged into the head end load, defined as S1, and the line impedance between S1 and the transformer is negligible. The load on the line that needs to be studied is defined as S2, and S2 is processed according to the load uniform, increasing or decreasing distribution, increasing sawtooth distribution, decreasing sawtooth distribution and trapezoidal distribution. For the convenience of theoretical calculation, the load distribution is processed with an arithmetic distribution according to the tolerance d. In the low-voltage distribution network, the earth-to-ground susceptance effect is not obvious. The ground branch in the line and transformer equivalent model is ignored in the calculation. The transformer is equivalent to a series resistance and reactance, and the specific value is obtained by the given transformer parameters.

[0173] In a feasible embodiment, the data of this embodiment are as shown in the above embodiment. Since the main line length is 376 meters, there are 9 main line towers, and the distance between towers is 47 meters, the load distribution is processed by arithmetic distribution according to d=47. After considering the load distribution, the topological equivalent schematic diagram is shown in FIG. Fig.26 .

[0174] The present application shows a method for equivalent modeling of low-voltage substation load and topology, which is based on the K-Means algorithm and combines the line topology structure and load voltage data to perform equivalent modeling of the low-voltage substation load and topology. The proposed load equivalent method based on the K-Means algorithm effectively solves the problem of difficulty in load modeling when the line topology is complex. The present application shows a method for equivalent modeling of low-voltage substation load and topology, which is simple in modeling, convenient in calculation, requires less data to be collected, and can quickly complete the line topology equivalent modeling work.

[0175] The present application is described in detail above in conjunction with specific embodiments and exemplary examples, but these descriptions cannot be understood as limiting the present application. Those skilled in the art understand that, without departing from the spirit and scope of the present application, a variety of equivalent substitutions, modifications or improvements can be made to the technical solution of the present application and its implementation methods, all of which fall within the scope of the present application, and the scope of protection of the present application shall be subject to the attached claims.

Claims

1. A method for modeling low voltage area load and topology equivalent, based on the overall measurement method, characterized in that: The method comprises: Obtain low voltage area line topology information; Load equivalence is performed according to the line topology information, and the load equivalence method includes: using the K-Means clustering algorithm to perform load equivalence, and the use of the K-Means clustering algorithm to perform load equivalence includes: presetting the load distribution distance value d according to the topology information and obtaining the trunk line length L, the number of load division areas K=L / d, using the K-Means clustering algorithm to divide the load area to obtain K load division areas; calculating the voltage mathematical expectations E(U) of the K load division areas respectively, and calculating the voltage standard deviation σ according to the voltage mathematical expectation E(U); performing load equivalence according to the voltage standard deviation σ; the load equivalence method is: if the voltage standard deviation σ of any load division area meets the first value range, then the area completes load equivalence; if the voltage standard deviation of any load division area does not meet the first value range, then the load division area is screened. For users whose voltage deviation ΔU% in the area meets the second value range, it is determined whether the user can be classified into an adjacent partition. If so, the user is classified into the adjacent partition and the area completes load equivalence. If not, the preset load distribution distance d is corrected according to the third value range, the K value is recalculated, and the calculation is performed according to the new K value, and it is iterated multiple times until all users complete load equivalence. The load equivalence method also includes: using the principle of similar distance and the principle of similar voltage for load equivalence; the method of using the principle of similar distance and the principle of similar voltage for load equivalence includes: assimilating unknown voltage data nodes whose distances to load nodes with known voltage data are within the first distance range into the same node; obtaining the average voltage value of the user area, taking the user with the smallest voltage deviation as the center of the circle and the first value as the radius, and assimilating users in the area whose deviations are less than or equal to the first threshold into the same node; After the load equivalence is completed, the user load capacity is obtained, the user load capacity is compared with the topology library, the topology equivalence type is determined, the topology equivalence is performed, and the low-voltage area load and topology equivalence model is obtained.

2. A low voltage area load and topology equivalent modeling method according to claim 1, characterized in that: The first value includes 45 meters; The first threshold comprises 2%.

3. A low voltage area load and topology equivalent modeling method according to claim 1, characterized in that: The method of dividing the load area by using the K-Means clustering algorithm includes: Obtain all user loads in the low-voltage area and establish a load sample set D, where D = {o i }; where o i (i=1~N) is the user load sample, N is the number of connected users; Randomly select K user load samples from the load sample set D as the initial center point m j (j=1~K); Calculate user load sample o i To the initial center point m j The distance l ij ; The user load sample o i Divide to l ij The clustering partition cluster C corresponding to the minimum value of j (j=1~K), and update the clustering cluster C j , where C j =C j ∪{o i }(j=1~K), all user load samples o i Clustering is performed to form K clustering clusters C j The K clusters form a sample set C = {C j }(j=1~K); Calculate each cluster partition cluster C j The square error SSE; Divide each cluster into clusters C j Perform heuristic iterative calculations to divide the clusters into clusters C j Recalculate the new center point m from the sample points in j , re-cluster the clusters, and after M iterations, obtain the partition's minimized square error SSE, output the updated partition area sample set, complete the partition, and obtain K load partition areas.

4. A method for modeling low voltage area load and topology equivalent according to claim 3, characterized in that: After the load equivalence is completed, the method to obtain the user load capacity is: Calculate the user load capacity within the K load partition areas.

5. A low voltage area load and topology equivalent modeling method according to claim 1, characterized in that: The load equivalence method also includes: using the voltage similarity principle to perform load equivalence; The method for load equivalence using the voltage similarity principle includes: The average voltage value in the user area is obtained, and the user with the smallest voltage deviation is taken as the center of the circle and 45 meters as the radius. The users in the area with a deviation of less than or equal to 2% are grouped as the same node.

6. A method for modeling low voltage area load and topology equivalent according to claim 1, characterized in that: The load equivalence method also includes: using the principle of close distance to perform load equivalence; The method of load equivalence using the principle of close distance includes: Nodes with unknown voltage data whose distances to load nodes with known voltage data are within a first distance range are assimilated into the same node.