Neutral line fault judgment method for low voltage substation based on voltage curve clustering

Abstract

The present invention discloses a method for judging a zero-line fault in a low-voltage substation based on voltage curve clustering. The method for judging a zero-line fault in a low-voltage substation based on voltage curve clustering of the present invention clusters the voltage curves of user tables in the low-voltage substation, and judges whether there is a zero-line fault in the low-voltage substation according to the clustering result. The method can solve the problems of low automation level, low judgment efficiency and high operation and maintenance cost of zero-line fault detection in traditional methods, and can provide a basis for the diagnosis and treatment of zero-line faults in low-voltage substations of power supply companies.

Description

Technical Field

[0001] The invention relates to a low-voltage station area zero-line fault judgment method based on voltage curve clustering, and belongs to the technical field of smart power and power grid. Background Art

[0002] In the three-phase four-wire power supply mode of low-voltage substations, a neutral line fault will cause an imbalance in the three-phase voltage of the substation, which may affect the power supply quality of the substation at the least, or even cause serious safety accidents at the worst. Therefore, both urban and rural power grids need to perform necessary maintenance work on the neutral line. Traditional neutral line fault judgment mainly relies on: (1) observing abnormal phenomena of power equipment in the substation; (2) on-site measurement through electrical testing equipment, which has a low degree of automation, low judgment efficiency, and high operation and maintenance costs.

[0003] Therefore, a low-voltage substation neutral line fault judgment method based on voltage curve clustering is needed to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to provide a method for determining a zero-line fault in a low-voltage transformer area based on voltage curve clustering, so as to solve the problems raised in the above-mentioned background technology.

[0005] A method for judging a low voltage transformer area neutral line fault based on voltage curve clustering comprises the following steps:

[0006] 1. Obtain the voltage curve x of the user meter in the low-voltage area within the period T c , where c∈[1,q], q is the number of voltage curves in the user table;

[0007] 2. The voltage curve x obtained from step 1 c Randomly select N voltage curves as cluster curve y d The initial value of , where d∈[1,N];

[0008] 3. Calculate the voltage curve x of the user meter c and N clustering curves y d The voltage difference between the two values ​​is the voltage curve x of the user table. c and N clustering curves y d The cluster curve with the smallest voltage difference is taken as the voltage curve x c The clusters to which they belong, and N groups of clusters are obtained;

[0009] 4. After clustering is completed, N groups of clusters are equivalently calculated as N clustering curves;

[0010] 5. If the clustering result to which the user table belongs is the same as the previous clustering result or the number of iterations exceeds L, the clustering iteration is terminated and the process goes to step 6; otherwise, steps 2 to 4 are repeated to cluster the voltage curve of the user table with the new N clustering curves again;

[0011] 6. Calculate the voltage difference between the N clustering curves after the final clustering;

[0012] 7. If the maximum voltage difference between the N cluster curves obtained in step 6 is greater than the fault judgment threshold dist max , it is determined that there is a neutral line fault in the substation, otherwise it is determined that there is no neutral line fault in the substation.

[0013] Furthermore, in step eight, dist max ≥6V.

[0014] Furthermore, in step 1, the period T is 1 day.

[0015] Furthermore, in step 2, N ≥ 4.

[0016] Furthermore, in step 2, L≤10.

[0017] Furthermore, in step 3, the voltage curve x of the user meter is calculated according to the following formula: c and clustering curve y d The voltage difference between:

[0018]

[0019] Where: dist cd is the user voltage curve x c and clustering curve y d The voltage difference between the two, n is the number of points on the voltage curve, i∈[1,n], x ci The voltage curve x of the user meter c The voltage value at the i-th point, y di is the clustering curve y d The voltage value at the i-th point.

[0020] Furthermore, in step 4, N groups of clusters are equivalently calculated as N cluster curves by the following formula:

[0021]

[0022] Where: Curve di is the clustering curve Curve d The value of the i-th point, p is the number of voltage curves in the d-group cluster, m∈[1,p], and Umi is the value of the i-th point of the m-th voltage curve.

[0023] Furthermore, in step 6, the voltage difference between the cluster curves is calculated according to the following formula:

[0024]

[0025] Where: distdk is the clustering curve y d With y k The voltage difference between them, n is the number of points of the clustering curve, j∈[1,n], y dj is the clustering curve y d The voltage value at point j, y kj is the clustering curve y k The voltage value at point j.

[0026] Beneficial effect: The low-voltage substation zero-line fault judgment method based on voltage curve clustering of the present invention clusters the voltage curves of the low-voltage substation user table, and judges whether there is a zero-line fault in the low-voltage substation according to the clustering results. It can solve the problems of low automation level, low judgment efficiency and high operation and maintenance cost of zero-line fault detection under traditional methods, and can provide a basis for the diagnosis and treatment of zero-line faults in low-voltage substations of power supply companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of the method for judging the zero line fault in the low voltage substation area based on voltage curve clustering of the present invention;

[0028] Figure 2 The user meter voltage curve of Example 1 of the present invention;

[0029] Figure 3 It is an equivalent curve of the 4 groups of user table curves after the final clustering in Example 1 of the present invention. DETAILED DESCRIPTION

[0030] 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.

[0031] Example 1

[0032] See also Figure 1 As shown, the low-voltage station area zero-line fault judgment method based on voltage curve clustering of the present invention comprises the following steps:

[0033] Step 1: Obtain the voltage curve x of the low-voltage area user meter within the calculation period c , c∈[1,q], q is the number of voltage curves of the user meter in this area, see Appendix Figure 2 ;

[0034] Step 2: Randomly select 4 curves as clustering curve y d The initial value of , d∈[1,4];

[0035] Step 3: Calculate the voltage difference between the user meter voltage curve and the four cluster curves respectively. The curve voltage difference calculation formula is as follows:

[0036]

[0037] Where: dist cd is the user voltage curve x c and clustering curve y d The voltage difference between them, n is the number of points on the curve, i∈[1,n], x ci The voltage curve x for the user meter c The voltage value at the i-th point, y di is the clustering curve y d The voltage value at the i-th point;

[0038] Step 4: Set the user meter voltage curve x c The cluster curve with the smallest voltage difference among the four cluster curves is taken as the cluster to which the user meter voltage curve belongs, and four clusters are obtained;

[0039] Step 5: After this round of clustering is completed, the four groups of clusters are equivalently calculated as four clustering curves;

[0040]

[0041] Where: Curve di is the clustering curve Curve d The value of the i-th point, p is the number of curves in the d-group cluster, m∈[1,p], Umi is the value of the i-th point of the m-th curve;

[0042] Step 6: If the clustering result to which the user table belongs is the same as the previous clustering result or the number of iterations exceeds 10, the clustering iteration process ends; otherwise, repeat the process in steps 4 and 5 to cluster the user table voltage curve with the new 4 clustering curves again;

[0043] Step 7: After 6 rounds of iterations, the final 4 clustering curves are obtained (see attached Figure 3 ), calculate the voltage difference between the four clustering curves after the final clustering;

[0044] The voltage difference between cluster curves is calculated according to the following formula:

[0045]

[0046] Where: dist dk is the clustering curve y d With y k The voltage difference between them, n is the number of points of the clustering curve, j∈[1,n], y dj is the clustering curve y dThe voltage value at point j, y kj is the clustering curve y k The voltage value at point j.

[0047] Step 8: According to step 7, the maximum voltage difference of the final four cluster curves is 21.15 (unit: V), which is greater than the judgment threshold 6 (unit: V), and it is determined that there is a neutral line fault in the substation.

Claims

1. A method for judging neutral line fault in low voltage substation based on voltage curve clustering, characterized in that: The following steps are involved:

1. Obtain the voltage curve x of the user meter in the low-voltage area within the period T c , where c∈[1,q], q is the number of voltage curves in the user table; 2. The voltage curve x obtained from step 1 c Randomly select N voltage curves as cluster curve y d The initial value of , where d∈[1,N]; 3. Calculate the voltage curve x of the user meter c and N clustering curves y d The voltage difference between the two values ​​is the voltage curve x of the user table. c and N clustering curves y d The cluster curve with the smallest voltage difference is taken as the voltage curve x c The clusters to which it belongs, and N groups of clusters are obtained; In step 3, the voltage curve x of the user meter is calculated according to the following formula: c and clustering curve y d The voltage difference between: Where: dist cd is the user voltage curve x c and clustering curve y d The voltage difference between the two, n is the number of points on the voltage curve, i∈[1,n], x ci The voltage curve x of the user meter c The voltage value at the i-th point, y di is the clustering curve y d The voltage value at the i-th point; 4. After clustering is completed, N groups of clusters are equivalently calculated as N clustering curves; In step 4, N groups of clusters are equivalently calculated as N cluster curves using the following formula: Where: Curve di is the clustering curve Curve d The value of the i-th point, p is the number of voltage curves in the d-group cluster, m∈[1,p], Umi is the value of the i-th point of the m-th voltage curve; 5. If the clustering result to which the user table belongs is the same as the previous clustering result or the number of iterations exceeds L, the clustering iteration is terminated and the process goes to step 6; otherwise, steps 2 to 4 are repeated to cluster the voltage curve of the user table with the new N clustering curves again; 6. Calculate the voltage difference between the N clustering curves after the final clustering; In step 6, the voltage difference between the cluster curves is calculated according to the following formula: Where: dist dk is the clustering curve y d With y k The voltage difference between them, n is the number of points of the clustering curve, j∈[1,n], y dj is the clustering curve y d The voltage value at point j, y kj is the clustering curve y k The voltage value at point j; 7. If the maximum voltage difference between the N cluster curves obtained in step 6 is greater than the fault judgment threshold dist max , it is determined that there is a neutral line fault in the substation, otherwise it is determined that there is no neutral line fault in the substation.

2. The method for determining a low voltage station neutral line fault based on voltage curve clustering according to claim 1, characterized in that: Step 8 dist max ≥6V.

3. The method for determining a low voltage station neutral line fault based on voltage curve clustering according to claim 1, characterized in that: In step 1, the period T is 1 day.

4. The method for determining a low voltage station zero line fault based on voltage curve clustering according to claim 1, characterized in that: In step 2, N ≥ 4.

5. The method for determining a low voltage station neutral line fault based on voltage curve clustering according to claim 1, characterized in that: In step 5, L≤10.

Images