Low-voltage transformer area topology identification method, device and equipment and storage medium
Through the combination of dimensionality reduction and condensation hierarchical clustering algorithm, the problem of low topology recognition accuracy in the distribution network in the low-voltage platform area is solved, and efficient topology position recognition is achieved to adapt to the identification needs of complex distribution networks.
Patent Information
- Application Number
- CN202411053504.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-12-06
AI Technical Summary
In the low-voltage distribution network with increasing distributed power access, the accuracy of topology recognition is affected, and the synchronization of data acquisition time has a direct impact on the identification accuracy, resulting in topology recognition in complex distribution networks The requirements are getting higher and higher.
The dimensionality reduction algorithm is used to convert the high-dimensional voltage data set into a low-dimensional data set, and the voltage data is clustered using the condensed hierarchical clustering algorithm to identify the topological location of the target user, reduce identification errors and improve accuracy.
Through the combination of dimensionality reduction and clustering algorithms, the identification error is effectively reduced, the accuracy of topological recognition of low-voltage table areas is improved, and the phase and table box of the target user can be accurately determined, and its topological position in the distribution network can be determined.
Smart Images

Figure CN119093327A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of electrical control technology, and in particular to a low-voltage area topology identification method, a low-voltage area topology identification device, a computer device, and a computer-readable storage medium. Background Art
[0002] In terms of low-voltage substation topology identification, in the current relevant technologies, with the increase in the number of distributed power sources connected, active distribution networks have become the norm for distribution network applications, which will have a significant negative impact on the accuracy of low-voltage substation topology identification. At the same time, the synchronization of data collection time also has a direct impact on the accuracy of topology identification. As distribution networks become increasingly complex, the requirements for the accuracy of topology identification are also increasing, and the relevant requirements for topology identification solutions are also becoming higher and higher. Summary of the invention
[0003] The present application provides a low-voltage area topology identification method, a low-voltage area topology identification device, a computer device and a computer-readable storage medium.
[0004] The low voltage area topology identification method involved in the implementation mode of the present application includes the following steps:
[0005] According to the collected voltage data of the distribution network to be tested, a historical voltage data set of the substation is constructed;
[0006] According to the historical voltage data set of the substation, based on a preset dimensionality reduction algorithm, a plurality of response voltage data sets are determined, wherein the dimension of the response voltage data set is lower than that of the historical voltage data set of the substation;
[0007] According to the response voltage data set, based on a preset agglomerative hierarchical clustering algorithm, a topological position corresponding to the target user in the power distribution network to be tested is determined.
[0008] In this way, the present application can transform the high-dimensional original data set into a low-dimensional data set based on the preset dimensionality reduction algorithm, while retaining the effective information of the original data set, reducing the recognition errors caused by noise information and redundant information, and improving the accuracy of topology recognition. On this basis, the agglomerative hierarchical clustering algorithm is further used to cluster the voltage data after clustering, so as to identify the phase of the user, and further determine the meter box and branch line of the target user based on the phase, and finally determine the topological position of the target user in the distribution network to be tested, and complete the topology recognition process.
[0009] In some embodiments, the step of constructing a historical voltage data set of an area according to the collected voltage data of the distribution network to be tested includes:
[0010] At a target time interval, acquiring the collected voltage data;
[0011] The historical voltage data set of the substation is constructed according to the voltage value corresponding to each phase of the distribution network to be tested at each data collection time point in the collected voltage data.
[0012] In this way, the present application can also obtain an original data set for topology identification for the distribution network to be tested.
[0013] In some embodiments, the determining of multiple response voltage data sets based on the historical voltage data set of the substation and a preset dimensionality reduction algorithm includes:
[0014] Determine a single-phase user historical voltage data set according to the historical voltage data set of the substation area;
[0015] According to the single-phase user historical voltage data set and based on the dimension reduction algorithm, the response voltage data set is determined.
[0016] In this way, the present application can perform dimension reduction based on the single-phase user data set in the original data set using a preset dimension reduction algorithm.
[0017] In some embodiments, determining the response voltage data set based on the single-phase user historical voltage data set and the dimension reduction algorithm includes:
[0018] According to the single-phase user historical voltage data set, based on any two voltage data samples in the single-phase user historical voltage data set, the unilateral probability between the voltage data samples, the first probability distribution of the voltage data samples, the preset parameter perplexity, the second probability distribution of the voltage data samples corresponding to the sample mapping points after dimensionality reduction, and the information loss between the first probability distribution and the second probability distribution are optimized to determine the response voltage data set.
[0019] Thus, the present application provides a specific method for performing dimensionality reduction using a preset dimensionality reduction algorithm.
[0020] In some embodiments, determining the topological position corresponding to the target user in the distribution network to be tested based on the response voltage data set and a preset agglomerative hierarchical clustering algorithm includes:
[0021] According to the response voltage data set, based on any two response voltage data in the response voltage data set, optimizing the Euclidean distance between the response voltage data to determine the similarity between the response voltage data;
[0022] According to the Euclidean distance between any two of the response voltage data and the preset number of clusters, hierarchical clustering is performed on the response voltage data set to determine the single-phase user response voltage data of the same type;
[0023] According to the single-phase user response voltage data of the same type and the preset circuit configuration information, the topological position corresponding to the target user in the distribution network to be tested is determined.
[0024] Thus, the present application provides a method for performing clustering and topology identification for target users using a preset agglomerative hierarchical clustering algorithm.
[0025] In some embodiments, performing hierarchical clustering on the response voltage data set according to the Euclidean distance between any two of the response voltage data and a preset number of clusters to determine the same type of single-phase user response voltage data includes:
[0026] Based on the preset average chain method, according to the Euclidean distance, determine the cluster to which each single-phase user belongs and the distance between different clusters;
[0027] According to the distances between the different clusters and the preset number of clusters, a topological position corresponding to the target user in the power distribution network to be tested is determined.
[0028] Thus, the present application provides a method for performing cluster distance calculation using an agglomerative hierarchical clustering algorithm.
[0029] In some implementations, determining the topological position corresponding to the target user in the distribution network to be tested according to the distances between the different clusters and the preset number of clusters includes:
[0030] Determining a phase corresponding to a target user according to the distances between the different clusters and the preset number of clusters;
[0031] According to the phase corresponding to the target user, a topological position corresponding to the target user in the power distribution network to be tested is determined.
[0032] Thus, the present application provides a method for further determining the phase and topological position of the target user based on the cluster spacing.
[0033] The low voltage area topology identification device in the implementation mode of the present application includes:
[0034] The data collection module is used to construct a historical voltage data collection of the substation area based on the collected voltage data of the distribution network to be tested;
[0035] A data processing module, configured to determine a plurality of response voltage data sets according to the historical voltage data set of the substation area based on a preset dimensionality reduction algorithm, wherein the dimension of the response voltage data set is lower than that of the historical voltage data set of the substation area;
[0036] A topology positioning module is used to determine the topological position corresponding to the target user in the distribution network to be tested according to the response voltage data set and based on a preset agglomerative hierarchical clustering algorithm.
[0037] The computer device in the implementation mode of the present application includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the above method is implemented.
[0038] The computer-readable storage medium in the embodiments of the present application stores a computer program, and when the computer program is executed by one or more processors, the above method is implemented.
[0039] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:
[0041] Figure 1 This is a schematic diagram of an application scenario of the low-voltage area topology identification method in the implementation manner of this application;
[0042] Figure 2 This is a schematic diagram of an application scenario of the low-voltage area topology identification method in the implementation manner of this application;
[0043] Figure 3 This is a flow chart of a low voltage area topology identification method in an implementation manner of the present application;
[0044] Figure 4 This is a flow chart of a low voltage area topology identification method in an implementation manner of the present application;
[0045] Figure 5 This is a flow chart of a low voltage area topology identification method in an implementation manner of the present application;
[0046] Figure 6 This is a flow chart of a low voltage area topology identification method in an implementation manner of the present application;
[0047] Figure 7 This is a flow chart of a low voltage area topology identification method in an implementation manner of the present application;
[0048] Figure 8 It is a flow chart of the low voltage area topology identification method in the implementation mode of the present application. DETAILED DESCRIPTION
[0049] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of the present application, and cannot be understood as limiting the embodiments of the present application.
[0050] Exemplarily, the topological structure of the low-voltage distribution network in the current related technology refers to the power line connection structure from the low-voltage outlet side of the public transformer to the user's power consumption side. At present, there are basic wiring methods for low-voltage distribution network lines, such as radial, trunk and hybrid, among which the radial type is more common in urban low-voltage distribution networks. Under radial wiring, the 10kV voltage is stepped down to 0.4kV by the transformer, and the 0.4kV bus is divided into several 0.4kV branch feeders through several outlet cabinets. Different branch feeders distribute electrical energy to different power supply areas, and each branch feeder then delivers the electricity to the end user through the branch box. The radial wiring method has good stability and high reliability, and is often used in newly built communities.
[0051] Figure 1 A typical radial low-voltage distribution network topology is shown as an example. In the figure, the topology can be divided into four layers: transformer layer, branch line layer, meter box layer and user layer. Several meter boxes are connected in parallel behind each branch box, and there is no physical connection relationship between the meter boxes in the upstream and downstream. The line length between adjacent meter boxes is generally short and the voltage difference is small. W1~W4 represent user electricity meters. In order to make the three-phase load in the substation as balanced as possible, the single-phase electricity meters in the meter box are generally evenly connected to the three phases A, B, and C respectively. The trunk-type low-voltage distribution network structure is a 0.4kV trunk line led out by a transformer. The trunk line is divided into several branch lines through the distribution box busbar. Each branch line supplies power to several power load points (meter boxes) at the same time, and each power point is T-connected on the branch trunk line.
[0052] And for Figure 1 The structure shown in the figure can be modeled as follows: Figure 2 shown. Figure 2 U0 is the bus voltage on the low voltage side of the distribution transformer, U i is the voltage at node i, where i = 0, 1, 2, ..., n, P i , Q i Represent the active power and reactive power of node i respectively, P Zi , Q Zi Respectively represent the transmission active power and reactive power of branch i, R i , X i They represent the resistance and reactance of branch i respectively.
[0053] exist Figure 2Based on the above, for any node i of the line, its voltage U i The following relationship exists between the bus voltage U0 on the low-voltage side of the distribution transformer:
[0054]
[0055] Where P zk =P k +P loss-k ,Q Zk =Q k +Q loss-k
[0056] Where P loss-k , Q loss-k are respectively the active power loss and reactive power loss of the electricity transmission between the line segments of branch i. According to the above formula, the voltage of any node i is related to the bus voltage on the low-voltage side of the transformer of the phase line to which it belongs and the loads at all levels on the upstream side. The voltage curves of each node on the same phase line are similar, and the change trends are highly consistent. The closer the electrical distance, the smaller the voltage drop, and the higher the similarity of the voltages. Since there is a certain imbalance in the three phases of the substation, and the power load characteristics of each phase are basically different, the correlation between the node voltage curves of different phases is poor.
[0057] Then in Figure 1 and Figure 2 The low voltage section topology and equivalent model are shown on the basis of Figure 3 The low voltage area topology identification method in the implementation mode of the present application comprises the following steps:
[0058] 01: Construct a historical voltage data set of the substation area based on the collected voltage data of the distribution network to be tested;
[0059] 02: Based on the historical voltage data set of the substation area and the preset dimensionality reduction algorithm, multiple response voltage data sets are determined.
[0060] The dimension of the response voltage data set is lower than the historical voltage data set of the substation;
[0061] 03: According to the response voltage data set, based on the preset agglomerative hierarchical clustering algorithm, determine the topological position corresponding to the target user in the distribution network to be tested.
[0062] The low-voltage substation topology identification device in the implementation mode of the present application can implement the above-mentioned low-voltage substation topology identification method. Specifically, the low-voltage substation topology identification device includes a data collection module, a data processing module and a topology positioning module, wherein the data collection module is used to construct a substation historical voltage data set based on the collected voltage data of the distribution network to be tested, the data processing module is used to determine multiple response voltage data sets based on the historical voltage data set of the substation based on a preset dimensionality reduction algorithm, and the topology positioning module is used to determine the corresponding topological position of the target user in the distribution network to be tested based on the response voltage data set and a preset agglomerative hierarchical clustering algorithm.
[0063] The computer device in the implementation mode of the present application can implement the above-mentioned low-voltage substation topology identification method. Specifically, the computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor is used to construct a substation historical voltage data set based on the collected voltage data of the distribution network to be tested, and is used to determine multiple response voltage data sets based on the historical voltage data set of the substation based on a preset dimensionality reduction algorithm, and is used to determine the topological position corresponding to the target user in the distribution network to be tested based on the response voltage data set and a preset agglomerative hierarchical clustering algorithm.
[0064] Specifically, on the basis of the above-mentioned radial low-voltage distribution network (corresponding to the distribution network to be tested), the distribution transformer monitoring terminal and the concentrator are used to collect the voltage data of the distribution network to be tested, and the historical voltage data set of the substation is constructed based on the collected data as a benchmark, and the voltage data of the single-phase user is randomly selected to perform the dimension reduction algorithm. Further based on the response voltage data set after dimension reduction, according to the preset agglomerative hierarchical clustering algorithm, the clusters are distinguished for each single-phase user, and each single-phase user with the same cluster has the same phase, so that the meter box to which each single-phase user belongs can be further determined, and then the corresponding topological position identification of all single-phase users in the distribution network to be tested is completed.
[0065] In this way, the present application can transform the high-dimensional original data set into a low-dimensional data set based on the preset dimensionality reduction algorithm, while retaining the effective information of the original data set, reducing the recognition errors caused by noise information and redundant information, and improving the accuracy of topology recognition. On this basis, the agglomerative hierarchical clustering algorithm is further used to cluster the voltage data after clustering, so as to identify the phase of the user, and further determine the meter box and branch line of the target user based on the phase, and finally determine the topological position of the target user in the distribution network to be tested, and complete the topology recognition process.
[0066] See also Figure 4 In some embodiments, step 01 includes:
[0067] 011: Get the collected voltage data at the target time interval;
[0068] 012: Construct a historical voltage data set of the substation area according to the voltage value corresponding to each phase of the distribution network to be tested at each data collection time point in the collected voltage data.
[0069] In some embodiments, the data collection module is also used to obtain collected voltage data at a target time interval, and to construct a historical voltage data set for the substation based on the voltage values corresponding to each phase of the distribution network to be tested at each data collection time point in the collected voltage data.
[0070] In some embodiments, the processor is also used to obtain collected voltage data at a target time interval, and to construct a historical voltage data set of the substation based on the voltage values corresponding to each phase of the distribution network to be tested at each data collection time point in the collected voltage data.
[0071] Specifically, based on the above implementation, in some examples, the distribution transformer monitoring terminal and the concentrator collect the voltage data of the above-mentioned power grid to be detected at intervals of 1 hour (corresponding to the target time interval), and upload it to the preset metering automation system, and construct the historical voltage data set U of the station area through calculation of the above-mentioned metering automation system. tr , as follows:
[0072]
[0073] The element u it It represents the voltage value of the i-th phase on the low-voltage side of the substation at time t, where T is the number of time points for collecting voltage data, where i = a, b or c, and t = 1, 2, …, T.
[0074] In this way, the present application can also obtain an original data set for topology identification for the distribution network to be tested.
[0075] See also Figure 5 In some embodiments, step 02 includes:
[0076] 021: Determine the single-phase user historical voltage data set based on the historical voltage data set of the substation area;
[0077] 022: According to the historical voltage data set of single-phase users, based on the dimension reduction algorithm, determine the response voltage data set.
[0078] In some embodiments, the data processing module is further used to determine a single-phase user historical voltage data set based on the historical voltage data set of the substation area, and to determine a response voltage data set based on the single-phase user historical voltage data set based on a dimensionality reduction algorithm.
[0079] In some embodiments, the processor is further configured to determine a single-phase user historical voltage data set based on the area historical voltage data set, and to determine a response voltage data set based on the single-phase user historical voltage data set and a dimensionality reduction algorithm.
[0080] Specifically, on the basis of the above implementation mode, according to the above historical voltage data set of the substation area, a single-phase user historical voltage data set U can also be formed based on the historical voltage data of each single-phase user, as follows:
[0081]
[0082] Among them, the element u nt It represents the voltage value of the nth single-phase user in the substation at time t, and the number of single-phase users in the substation is N. The row vector U of U n The column vector U represents the voltage measurement value of a single user n during the acquisition period, t represents the voltage measurement value of all single-phase users belonging to the same phase in the substation at time t. Then, we can use the multiple column vectors U t Based on this, dimensionality reduction is performed based on the preset t-SNE algorithm to obtain the response voltage data set. This can reduce the recognition errors caused by noise and redundant information while retaining the effective information of the original data set, thereby improving the accuracy of topological recognition.
[0083] In this way, the present application can perform dimension reduction based on the single-phase user data set in the original data set using a preset dimension reduction algorithm.
[0084] In some embodiments, step 022 includes:
[0085] According to the historical voltage data set of single-phase users, based on any two voltage data samples in the historical voltage data set of single-phase users, the unilateral probability between the voltage data samples, the first probability distribution of the voltage data samples, the preset parameter perplexity, the second probability distribution of the voltage data samples corresponding to the sample mapping points after the dimension is reduced, and the information loss between the first probability distribution and the second probability distribution are optimized to determine the response voltage data set.
[0086] In some embodiments, the data processing module is also used to determine the response voltage data set based on any two voltage data samples in the single-phase user historical voltage data set by optimizing the unilateral probability between the voltage data samples, the first probability distribution of the voltage data samples, the preset parameter perplexity, the second probability distribution of the voltage data samples corresponding to the sample mapping points after dimensionality reduction, and the information loss between the first probability distribution and the second probability distribution.
[0087] In some embodiments, the processor is also used to determine a response voltage data set based on a single-phase user historical voltage data set and based on any two voltage data samples in the single-phase user historical voltage data set, optimize the unilateral probability between the voltage data samples, the first probability distribution of the voltage data samples, the preset parameter perplexity, the second probability distribution of the voltage data samples corresponding to the sample mapping points after dimensionality reduction, and the information loss between the first probability distribution and the second probability distribution.
[0088] Specifically, based on the above implementation, in some examples, t-SNE is used to reduce the dimension of voltage data based on the single-phase user historical voltage data set U. The t-SNE algorithm represents the similarity between voltage samples by converting the Euclidean distance into conditional probability.
[0089] For example, given a voltage data set U of a single-phase user at the same time t =[U1,U2,…,U N ], for any two voltage samples U i , U j , use p j|i Represents the unilateral probability between the above sample points:
[0090]
[0091] Among them, p j|i Indicates voltage data point U j Appears at voltage data point U i The conditional probability of voltage data point U j And voltage data point U i The closer the distance, the j|i The larger the value, that is, the voltage data point U j And voltage data point U i The higher the similarity, the more likely it is that the phases of the two are the same.
[0092] On this basis, in order to avoid the problem of asymmetric conditional probability, for example, the voltage data point U can be calculated in the following way: j And voltage data point U i The probability distribution between ij (corresponding to the first probability distribution):
[0093]
[0094] In addition, σ i The voltage sample U j The standard deviation of the Gaussian distribution centered at . i , can be obtained by given parameter perplexity Perp(P i) is obtained, and the definition of parameter confusion is as follows:
[0095]
[0096] P i =[p i1 ,p i2 ,…,p ij ,…,p iN ]
[0097] Among them, H(P i ) is the information entropy. i is the voltage data point U i The similarity measurement matrix between other data points in the single-phase user historical voltage data set U can be solved by binary search to obtain the optimal standard deviation σ i , so that the similarity measurement matrix P of user i in the station area i Satisfy the preset parameter perplexity.
[0098] Furthermore, suppose the voltage data set after dimension reduction by the t-SNE algorithm is V = [V1, V2,,…, V N ], which is a mapping from high-dimensional space U to low-dimensional space V, that is, the voltage data point U in high-dimensional space j And voltage data point U i Points V corresponding to the data in the low-order space j and V i . Using t distribution in low-dimensional space to calculate the probability distribution q between and ij (corresponding to the second probability distribution):
[0099]
[0100] For the probability distribution p in high-dimensional space ij And the probability distribution q in low-dimensional space ij , the information loss between the two probability distributions can be analyzed by calculating the KL divergence, that is, the retention of voltage data features after the high-dimensional space U is mapped to the low-dimensional space V. In order to keep the distance distribution relationship between data points before and after mapping consistent as much as possible, the final optimization goal of the t-SNE algorithm is to minimize the KL divergence, that is:
[0101]
[0102] Where P and Q are both N×N matrices, whose elements are p ij and q ijThat is, the smaller the optimization object C is, the more consistent the probability distribution of the user voltage data set before and after dimensionality reduction is, that is, the voltage data set after dimensionality reduction can better replace the high-dimensional data set for substation topology identification. Since minimizing KL divergence is a non-convex optimization, it can be solved iteratively by the stochastic gradient descent method. The gradient of the KL divergence of the t-SNE optimization objective function is calculated as follows:
[0103]
[0104] On this basis, the update method for solving the response voltage data set V is as follows:
[0105]
[0106] Where s is the number of iterations of the data model executing the t-SNE algorithm, η is the learning rate of the above data model, and α(t) is the power factor of the above data model.
[0107] Thus, the present application provides a specific method for performing dimensionality reduction using a preset dimensionality reduction algorithm.
[0108] See also Figure 6 In some embodiments, step 03 includes:
[0109] 031: According to the response voltage data set, based on any two response voltage data in the response voltage data set, optimizing the Euclidean distance between the response voltage data to determine the similarity between the response voltage data;
[0110] 032: Performing hierarchical clustering on the response voltage data set according to the Euclidean distance between any two response voltage data and the preset number of clusters to determine the single-phase user response voltage data of the same type;
[0111] 033: According to the response voltage data of similar single-phase users and the preset circuit configuration information, the topological position corresponding to the target user in the distribution network to be tested is determined.
[0112] In certain embodiments, the topology positioning module is also used to optimize the Euclidean distance between the response voltage data based on the response voltage data set and based on any two response voltage data in the response voltage data set to determine the similarity between the response voltage data, and to perform hierarchical clustering on the response voltage data set based on the Euclidean distance between any two response voltage data and a preset number of clustering clusters to determine similar single-phase user response voltage data, and to determine the corresponding topological position of the target user in the distribution network to be tested based on similar single-phase user response voltage data and preset circuit configuration information.
[0113] In certain embodiments, the processor is also used to optimize the Euclidean distance between the response voltage data based on the response voltage data set and based on any two response voltage data in the response voltage data set to determine the similarity between the response voltage data, and to perform hierarchical clustering on the response voltage data set based on the Euclidean distance between any two response voltage data and a preset number of clustering clusters to determine similar single-phase user response voltage data, and to determine the corresponding topological position of the target user in the distribution network to be tested based on the similar single-phase user response voltage data and preset circuit configuration information.
[0114] Specifically, when the response voltage data set V has been determined, Euclidean distance calculation is performed based on any pair of data in the response voltage data set V to determine the similarity between any two voltage data in the response voltage data set V, and hierarchical clustering is performed on the data in the response voltage data set V based on the above-mentioned similarity and a preset number of clustering clusters, so as to classify the data in the response voltage data set V, and the single-phase users corresponding to the data belonging to the same category have the same phase.
[0115] Finally, according to the voltage data corresponding to the single-phase users with the same phase, based on the meter information corresponding to the above single-phase users in the distribution network to be tested, the meter box to which each single-phase user belongs is identified, thereby realizing the topological position identification process of the target user in the distribution network to be tested.
[0116] Furthermore, the above method of determining the Euclidean distance can refer to the following example:
[0117] Assume that the voltage data feature vectors of any two single-phase users after being processed by the above t-SNE dimension reduction algorithm are V i and V j , where V i =[v i1 ,v i2 ,…,v id ],V j =[v j1 ,v j2 ,…,v jd ], the Euclidean distance between the above two is:
[0118]
[0119] Among them, d is the dimension of the voltage data set after dimensionality reduction, and N is the number of single-phase users in the substation.
[0120] Thus, the present application provides a method for performing clustering and topology identification for target users using a preset agglomerative hierarchical clustering algorithm.
[0121] See also Figure 7In some embodiments, step 032 includes:
[0122] 0321: Based on the preset average chain method and Euclidean distance, determine the cluster to which each single-phase user belongs and the distances between different clusters;
[0123] 0322: According to the distances between different clusters and the preset number of clusters, determine the topological position of the target user in the distribution network to be tested.
[0124] In some embodiments, the topological positioning module is also used to determine the cluster to which each single-phase user belongs and the distance between different clusters based on the preset average chain method and according to the Euclidean distance, and is used to determine the topological position corresponding to the target user in the distribution network to be tested based on the distance between different clusters and the preset number of clusters.
[0125] In some embodiments, the processor is also used to determine the cluster to which each single-phase user belongs and the distance between different clusters based on the preset average chain method and according to the Euclidean distance, and is used to determine the topological position corresponding to the target user in the distribution network to be tested based on the distance between different clusters and the preset number of clusters.
[0126] Specifically, on the basis of the above implementation, when the Euclidean distance between the voltage data corresponding to any two single-phase users is determined, the cluster to which each single-phase user belongs is determined according to the preset number of clusters and the degree of clustering of the Euclidean distance between the voltage data. Exemplarily, the number of clusters can be preset to 3.
[0127] Then, for the distance between any two clusters, the preset average chain method can be used to solve it:
[0128]
[0129] Where ‖xz‖ represents the distance between voltage data point x and voltage data point z, |C i | represents the number of preset clusters, |C j | represents the number of voltage data points.
[0130] Based on the above calculation method, the cluster to which each single-phase user belongs can be determined by calculating the distance between clusters, so that single-phase users with the same phase can be classified into the same cluster, so as to further determine the topological position of the single-phase user in the distribution network to be tested according to the cluster of the single-phase user.
[0131] Thus, the present application provides a method for performing cluster distance calculation using an agglomerative hierarchical clustering algorithm.
[0132] See also Figure 8 In some embodiments, step 033 includes:
[0133] 0331: Determine the phase corresponding to the target user according to the distances between different clusters and the preset number of clusters;
[0134] 0332: According to the phase corresponding to the target user, determine the topological position corresponding to the target user in the distribution network to be tested.
[0135] In some embodiments, the topology positioning module is also used to determine the phase corresponding to the target user based on the distances between different clusters and the preset number of clusters, and to determine the topological position corresponding to the target user in the distribution network to be tested based on the phase corresponding to the target user.
[0136] In some embodiments, the processor is further used to determine the phase corresponding to the target user based on the distances between different clusters and a preset number of clusters, and to determine the topological position corresponding to the target user in the distribution network to be tested based on the phase corresponding to the target user.
[0137] Specifically, on the basis of the above implementation mode, further, when each single-phase user is classified into different clusters based on the phase as the distinguishing condition, within each of the above clusters, based on the above-mentioned average chain method, taking the meter box distribution information in the distribution network to be tested as a reference, the number of corresponding meter boxes is re-used as the preset number of clusters, and hierarchical clustering is performed again based on the average chain method in the above implementation mode to determine the meter box to which each single-phase user belongs.
[0138] In addition, while determining the meter box to which each single-phase user belongs, it is also possible to determine, based on the size of the average voltage, whether the access position of each single-phase user on the line in the distribution network to be tested is upstream or downstream of the line, thereby completing the identification of the topological position of each single-phase user in the distribution network to be tested.
[0139] Thus, the present application provides a method for further determining the phase and topological position of the target user based on the cluster spacing.
[0140] The computer-readable storage medium in the embodiments of the present application stores a computer program, and when the computer program is executed by one or more processors, the above method is implemented.
[0141] In the description of this specification, the descriptions with reference to the terms "certain embodiments", "in an example", "exemplarily", etc., mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.
[0142] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.
[0143] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.
Claims
1. A method for identifying low voltage area topology, characterized in that: The method comprises: According to the collected voltage data of the distribution network to be tested, a historical voltage data set of the substation is constructed; According to the historical voltage data set of the substation, based on a preset dimensionality reduction algorithm, a plurality of response voltage data sets are determined, wherein the dimension of the response voltage data set is lower than that of the historical voltage data set of the substation; According to the response voltage data set, based on a preset agglomerative hierarchical clustering algorithm, a topological position corresponding to the target user in the power distribution network to be tested is determined.
2. The method according to claim 1, characterized in that The step of constructing a historical voltage data set of the substation area according to the collected voltage data of the distribution network to be tested includes: At a target time interval, acquiring the collected voltage data; The historical voltage data set of the substation is constructed according to the voltage value corresponding to each phase of the distribution network to be tested at each data collection time point in the collected voltage data.
3. The method according to claim 1, characterized in that The step of determining a plurality of response voltage data sets based on the historical voltage data set of the substation area and a preset dimensionality reduction algorithm includes: Determine a single-phase user historical voltage data set according to the historical voltage data set of the substation area; According to the single-phase user historical voltage data set and based on the dimension reduction algorithm, the response voltage data set is determined.
4. The method according to claim 3, characterized in that: The step of determining the response voltage data set based on the single-phase user historical voltage data set and the dimension reduction algorithm includes: According to the single-phase user historical voltage data set, based on any two voltage data samples in the single-phase user historical voltage data set, the unilateral probability between the voltage data samples, the first probability distribution of the voltage data samples, the preset parameter perplexity, the second probability distribution of the voltage data samples corresponding to the sample mapping points after dimensionality reduction, and the information loss between the first probability distribution and the second probability distribution are optimized to determine the response voltage data set.
5. The method according to claim 1, characterized in that The step of determining the topological position of the target user in the distribution network to be tested based on the response voltage data set and a preset agglomerative hierarchical clustering algorithm includes: According to the response voltage data set, based on any two response voltage data in the response voltage data set, optimizing the Euclidean distance between the response voltage data to determine the similarity between the response voltage data; According to the Euclidean distance between any two of the response voltage data and the preset number of clusters, hierarchical clustering is performed on the response voltage data set to determine the single-phase user response voltage data of the same type; According to the single-phase user response voltage data of the same type and the preset circuit configuration information, the topological position corresponding to the target user in the distribution network to be tested is determined.
6. The method according to claim 5, characterized in that The step of performing hierarchical clustering on the response voltage data set according to the Euclidean distance between any two of the response voltage data and a preset number of clusters to determine the single-phase user response voltage data of the same type includes: Based on the preset average chain method, according to the Euclidean distance, determine the cluster to which each single-phase user belongs and the distance between different clusters; According to the distances between the different clusters and the preset number of clusters, a topological position corresponding to the target user in the power distribution network to be tested is determined.
7. The method according to claim 6, characterized in that Determining the topological position corresponding to the target user in the distribution network to be tested according to the distances between the different clusters and the preset number of clusters includes: Determining a phase corresponding to a target user according to the distances between the different clusters and the preset number of clusters; According to the phase corresponding to the target user, a topological position corresponding to the target user in the power distribution network to be tested is determined.
8. A low voltage area topology identification device, characterized in that: The device comprises: The data collection module is used to construct a historical voltage data collection of the substation area based on the collected voltage data of the distribution network to be tested; A data processing module, configured to determine a plurality of response voltage data sets according to the historical voltage data set of the substation area based on a preset dimensionality reduction algorithm, wherein the dimension of the response voltage data set is lower than that of the historical voltage data set of the substation area; A topology positioning module is used to determine the topological position corresponding to the target user in the distribution network to be tested according to the response voltage data set and based on a preset agglomerative hierarchical clustering algorithm.
9. A computer device, characterized in that: The computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the method according to any one of claims 1 to 7 is implemented.
