A method for identifying branch topology of low voltage substations based on density peak clustering

Abstract

The present invention discloses a method for identifying branch topology of low-voltage substations based on density peak clustering. Based on the impedance values ​​between nodes, the method for identifying substation branch topology is realized through the density peak clustering method. The present invention first installs smart meters on users in the substation, and extracts the impedance values ​​of different measurement points based on the current advanced measurement system; then, considering that the physical space of users in the same branch is similar and the line distance with other users is close, the impedance space distance between users is defined; finally, based on the impedance space distance between each user, users are clustered based on density peak clustering to obtain the relationship between branches and meter boxes. The spatial topology is identified by the impedance value without installing new devices, providing a feasibility for realizing the automatic topology identification system.

Description

Technical Field

[0001] The invention relates to a low-voltage station area branch topology identification method based on density peak clustering, belonging to the technical field of low-voltage power distribution. Background Art

[0002] Low-voltage topology relationship recognition technology is the foundation and research hotspot for the current refinement of distribution network management and loss reduction and energy conservation. It is one of the core technologies and key technologies for the construction of intelligent distribution substations. The low-voltage distribution substation topology automatic recognition system is also one of the key systems of intelligent distribution substations. In the construction of smart grids, low-voltage topology relationship recognition technology can realize lean management such as accurate calculation of low-voltage distribution lines, three-phase imbalance control and fault interval judgment. The low-voltage distribution substation topology automatic recognition system can realize automatic identification and monitoring of low-voltage distribution lines, improving the safety, reliability and economy of the distribution network.

[0003] At present, the identification of low-voltage distribution topology mainly relies on manual on-site verification, which requires a lot of manpower and is chaotic. The emergence of an automatic identification system for low-voltage distribution area topology is urgent.

[0004] Therefore, a new low-voltage substation branch topology identification method is needed to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide a low-voltage station branch topology identification method based on density peak clustering to solve the problems raised in the above background technology.

[0006] A low voltage station area branch topology identification method based on density peak clustering includes the following steps:

[0007] 1. Measure the current data and voltage data of each user node in the topology area to be identified;

[0008] 2. Calculate the impedance matrix value between different user nodes;

[0009] 3. Cluster the user nodes based on the density peak clustering method to obtain the relationship between branches and table boxes.

[0010] Furthermore, the impedance values ​​between different users are calculated in step 2, including the following steps:

[0011] Step 2.1: The voltage data and current data of user node s and user node j are expressed by the following formula:

[0012] V s = {v s1 ,v s2 ,…,v sq}, I l1 ={i s1,i s2 ,…,i sq}

[0013] V j = {v j1 ,v j2 ,…,v jq}, I l2 ={i j1 ,i j2 ,…,i jq}

[0014] Where V s Represents the voltage data of user node s, I l1 Represents the current data of user node s, V j Represents the voltage data of user node j, I l2 represents the current data of user node j, q represents the time series length;

[0015] Step 2.2: Calculate the mutual impedance R between user node s and user node j s,j , mutual impedance R s,j represents the calculated mutual impedance value of the current change of user node s based on the measured value of user node j, where R s,j =R j,s ;

[0016] The mutual impedance R is calculated according to the following formula s,j :

[0017]

[0018] Where V s (t) and V s (t1) represents the voltage data of user node s at time t and time t1, respectively, I l1 (t) and I l1 (t1) represents the current data of user node s at time t and time t1, V j (t) and V j (t1) represents the voltage data of user node j at time t and time t1, respectively, I l2 (t) and I l2 (t1) represents the current data of user node j at time t and time t1, V T (t) and V T (t1) represents the transformer voltage data at time t and time t1, respectively. T (t) and I T (t1) represents the transformer current data at time t and time t1, R l1 , R l2 and R Tdenote the line resistance at user node s, user node j and transformer respectively;

[0019] At time t1, the current I l1 changes, and the current I of user node j l2 If no changes have occurred, then:

[0020]

[0021] In the formula, ΔI T =ΔI l1 , ΔU=ΔU T -ΔU j , ΔU T is the voltage difference between the transformer at time t1 and time t, ΔU j is the voltage difference between user node j at time t1 and time t, ΔI T is the current difference between the transformer at time t1 and time t, ΔI l1 is the current difference between time t1 and time t of user node s;

[0022] Step 2.3: Calculate the mutual impedance between other users according to step 2.2 and obtain the impedance matrix:

[0023]

[0024] In the formula, R k,p represents the mutual impedance between user node k and user node p.

[0025] Furthermore, in step 3, clustering users based on the density peak clustering method to obtain the relationship between branches and table boxes includes the following steps:

[0026] Step 3.1: Get the user's mutual impedance dataset Ω in the past week:

[0027] Ω={(R 1,1 … R 1,n ),…,(R m,1 … R m,n )}

[0028] In the formula, m is the total number of data, and n is the number of users;

[0029] Step 3.2: Calculate the distance between two data points according to the following formula:

[0030]

[0031] In the formula, sim(R s , R j ) represents the mutual impedance similarity between user node s and user node j;

[0032] The local density ρ of the data point is calculated according to the following formula i :

[0033]

[0034]

[0035] Where D b is the cutoff distance, ρ i Indicates the number of points whose distance to the data point is less than the cutoff distance;

[0036] Step 3.3: Determine the density threshold ρ N and distance threshold D0, when the local density of data points ρ i Greater than the density threshold ρ N And the distance between the data points is greater than the distance threshold D0, then these data points are a cluster. When the local density ρ of the data points i Less than the density threshold ρ N And the distance between data points is less than the distance threshold D0, then these points are noise or boundary points.

[0037] Furthermore, in step 3.2, D b is 3.5.

[0038] Furthermore, the distance threshold D0 in step 3.3 is 3.5.

[0039] Furthermore, the density threshold ρ in step 3.3 N is the data point with the maximum data change rate.

[0040] Principle of the invention: First, smart meters are installed on users in the substation area, and the impedance values ​​of different measurement points are extracted based on the current advanced measurement system; then, considering that the physical space of users in the same branch is similar and the line distance with other users is close, the impedance space distance between users is defined; finally, based on the impedance space distance between each user, users are clustered based on density peak clustering to obtain the relationship between branches and meter boxes.

[0041] Beneficial effect: The low-voltage substation branch topology identification method based on density peak clustering of the present invention calculates the impedance value between each node based on an advanced measurement system, identifies the spatial topology through the impedance value, and realizes the low-voltage substation branch topology identification without installing new devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of the low voltage station area branch topology identification method based on density peak clustering of the present invention;

[0043] Figure 2 This is a general area topology map. DETAILED DESCRIPTION

[0044] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent forms of modifications to the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0045] See also Figure 1 and Figure 2 The low-voltage station area branch topology identification method based on density peak clustering of the present invention comprises the following steps:

[0046] 1. Measure the current data and voltage data of each user node in the topology area to be identified;

[0047] 2. Calculate the impedance matrix value between different user nodes;

[0048] 3. Cluster the user nodes based on the density peak clustering method to obtain the relationship between branches and table boxes.

[0049] Preferably, calculating the impedance values ​​between different users in step 2 includes the following steps:

[0050] Step 2.1: The voltage data and current data of user node s and user node j are expressed by the following formula:

[0051] V s = {v s1 ,v s2 ,…,v sq}, I l1 ={i s1 ,i s2 ,…,i sq}

[0052] v j = {v j1 ,v j2 ,…,v jq}, I l2 ={i j1 ,i j2 ,…,i jq}

[0053] Where V s Represents the voltage data of user node s, I l1 Represents the current data of user node s, V j Represents the voltage data of user node j, I l2 represents the current data of user node j, q represents the time series length;

[0054] Step 2.2: Calculate the mutual impedance R between user node s and user node j s,j , mutual impedance R s,j represents the calculated mutual impedance value of the current change of user node s based on the measured value of user node j, where R s,j =R j,s ;

[0055] The mutual impedance R is calculated according to the following formula s,j :

[0056]

[0057] Where V s (t) and V s (t1) represents the voltage data of user node s at time t and time t1, respectively, I l1 (t) and I l1 (t1) represents the current data of user node s at time t and time t1, V j (t) and V j (t1) represents the voltage data of user node j at time t and time t1, respectively, I l2 (t) and I l2 (t1) represents the current data of user node j at time t and time t1, V T (t) and V T (t1) represents the transformer voltage data at time t and time t1, respectively. T (t) and I T (t1) represents the transformer current data at time t and time t1, R l1 , R l2 and R T denote the line resistance at user node s, user node j and transformer respectively;

[0058] At time t1, the current I l1 changes, and the current I of user node j l2 If no changes have occurred, then:

[0059]

[0060] In the formula, ΔI T =ΔI l1 , ΔU=ΔU T -ΔU j , ΔU T is the voltage difference between the transformer at time t1 and time t, ΔU j is the voltage difference between user node j at time t1 and time t, ΔI T is the current difference between the transformer at time t1 and time t, ΔI l1is the current difference between time t1 and time t of user node s;

[0061] Step 2.3: Calculate the mutual impedance between other users according to step 2.2 and obtain the impedance matrix:

[0062]

[0063] In the formula, R k,p represents the mutual impedance between user node k and user node p.

[0064] Preferably, in step 3, clustering users based on a density peak clustering method to obtain the relationship between branches and meter boxes includes the following steps:

[0065] Step 3.1: Get the user's mutual impedance dataset Ω in the past week:

[0066] Ω={(R 1,1 … R 1,n ),…,(R m,1 … R m,n )}

[0067] In the formula, m is the total number of data, and n is the number of users;

[0068] Step 3.2: Calculate the distance between two data points according to the following formula:

[0069]

[0070] In the formula, sim(R s , R j ) represents the mutual impedance similarity between user node s and user node j;

[0071] The local density ρ of the data point is calculated according to the following formula i :

[0072]

[0073]

[0074] Where D b is the cutoff distance, ρ i Indicates the number of points whose distance to the data point is less than the cutoff distance; in the present invention, D b is 3.5.

[0075] Step 3.3: Determine the density threshold ρ N and distance threshold D0, when the local density of data points ρ i Greater than the density threshold ρ N And the distance between the data points is greater than the distance threshold D0, then these data points are a cluster. When the local density ρ of the data pointsi Less than the density threshold ρ N And the distance between data points is less than the distance threshold D0, then these points are noise or boundary points. Specifically, in the present invention, the distance threshold D0 is 3.5. The density threshold ρ N is the data point with the maximum data change rate.

[0076] Principle of the invention: First, smart meters are installed on users in the substation area, and the impedance values ​​of different measurement points are extracted based on the current advanced measurement system; then, considering that the physical space of users in the same branch is similar and the line distance with other users is close, the impedance space distance between users is defined; finally, based on the impedance space distance between each user, users are clustered based on density peak clustering to obtain the relationship between branches and meter boxes.

[0077] The low-voltage substation branch topology identification method based on density peak clustering of the present invention calculates the impedance value between each node based on an advanced measurement system, identifies the spatial topology through the impedance value, and realizes the low-voltage substation branch topology identification without installing new devices.

Claims

1. A low voltage station area branch topology identification method based on density peak clustering, characterized in that: The following steps are involved:

1. Measure the current data and voltage data of each user node in the topology area to be identified; 2. Calculate the impedance matrix value between different user nodes; The impedance values ​​between different users are calculated in step 2, including the following steps: Step 2.1: The voltage data and current data of user node s and user node j are expressed by the following formula: V s ={v s1 ,v s2 ,…,v sq },I l1 ={i s1 ,i s2 ,…,i sq } V j ={v j1 ,v j2 ,…,v jq },I l2 ={i j1 ,i j2 ,…,i jq } Where V s Represents the voltage data of user node s, I l1 Represents the current data of user node s, V j Represents the voltage data of user node j, I l2 represents the current data of user node j, q represents the time series length; Step 2.2: Calculate the mutual impedance R between user node s and user node j s,j , mutual impedance R s,j represents the calculated mutual impedance value of the current change of user node s based on the measured value of user node j, where R s,j =R j,s ; The mutual impedance R is calculated according to the following formula s,j : Where V s (t) and V s (t1) represents the voltage data of user node s at time t and time t1, I l1 (t) and I l1 (t1) represents the current data of user node s at time t and time t1, V j (t) and V j (t1) represents the voltage data of user node j at time t and time t1, respectively, I l2 (t) and I l2 (t1) represents the current data of user node j at time t and time t1, V T (t) and V T (t1) represents the transformer voltage data at time t and time t1, respectively. T (y) and I T (t1) represents the transformer current data at time t and time t1, R l1 , R l2 and R T denote the line resistance at user node s, user node j and transformer respectively; At time t1, the current I l1 changes, and the current I of user node j l2 If no changes have occurred, then: In the formula, ΔI T =ΔI l1 , ΔU=ΔU T -ΔU j , ΔU T is the voltage difference between the transformer at time t1 and time t, ΔU j is the voltage difference between user node j at time t1 and time t, ΔI T is the current difference between the transformer at time t1 and time t, ΔI l1 is the current difference between time t1 and time t of user node s; Step 2.3: Calculate the mutual impedance between other users according to step 2.2 and obtain the impedance matrix: In the formula, R k,p represents the mutual impedance between user node k and user node p; 3. Cluster the user nodes based on the density peak clustering method to obtain the relationship between branches and table boxes.

2. The low voltage station area branch topology identification method based on density peak clustering according to claim 1, characterized in that: In step 3, users are clustered based on the density peak clustering method to obtain the relationship between branches and meter boxes, including the following steps: Step 3.1: Get the user's mutual impedance dataset Ω in the past week: Ω={(R 1,1 …R 1,n ),…,(R m,1 …R m,n )} In the formula, m is the total number of data, and n is the number of users; Step 3.2: Calculate the distance between two data points according to the following formula: In the formula, sim(R s , R j ) represents the mutual impedance similarity between user node s and user node j, and n represents the number of users; The local density ρ of the data point is calculated according to the following formula i : Where D b is the cutoff distance, ρ i Indicates the number of points whose distance to the data point is less than the cutoff distance; Step 3.3: Determine the density threshold ρ N and distance threshold D0, when the local density of data points ρ i Greater than the density threshold ρ N And the distance between the data points is greater than the distance threshold D0, then these data points are a cluster. When the local density ρ of the data points i Less than the density threshold ρ N And the distance between data points is less than the distance threshold D0, then these points are noise or boundary points.

3. The low voltage station area branch topology identification method based on density peak clustering according to claim 2, characterized in that: Step 3.2 D b is 3.

5.

4. The low voltage station area branch topology identification method based on density peak clustering according to claim 2, characterized in that: The distance threshold D0 in step 3.3 is 3.

5.

5. The low voltage station area branch topology identification method based on density peak clustering according to claim 2, characterized in that: Density threshold ρ in step 3.3 N is the data point with the maximum data change rate.

Images