CN104597378A - Fault line-selection method of power distribution network containing DG based on transient state non-power-frequency zero-sequence currents - Google Patents

Abstract

The invention relates to a fault line-selection method of power distribution network containing DG based on transient state non-power-frequency zero-sequence currents. A digital notching filter is utilized to obtain a fifth harmonic component of each line zero-sequence current. When the fifth harmonic currents have the same polarity, single-phase earth fault of a bus is judged to occur; otherwise, a fault of a certain outgoing line is judged to occur. An improved collection empirical mode decomposition algorithm is utilized to extract a non-power-frequency zero-sequence current component of each line. The energy weight coefficient of the non-power-frequency zero-sequence current of each line is calculated, and the line which has the largest energy weight coefficient is the fault line. According to proof of concrete calculation examples and a large number of simulation experiments, the fault line-selection method of the power distribution network containing the GD based on the transient state non-power-frequency zero-sequence currents can select the fault line accurately and reliably under the most unfavorable conditions such as small angle grounding fault, high resistance grounding fault and line end fault, is not influenced by the operation mode of the DG, does not need threshold setting according to human experiences, and improves the degree of automation of fault line selection.

Description

Based on the fault-line selecting method containing DG power distribution network of the non-power frequency zero-sequence current of transient state

Technical field

The present invention designs a kind of fault-line selecting method of power distribution network, particularly based on the fault-line selecting method containing DG new distribution net of the non-power frequency zero-sequence current of transient state.

Background technology

Along with user constantly increases the demand of electricity, the structure of power distribution network is just at upgrading, and cable line can improve the transmission capacity of circuit because of it, reduces the outlet scale of transformer station, take aboveground space few, the features such as failure rate is low are widely used in power distribution network.But, along with the increase of supplying power allocation capacity, when system generation single-phase earthing fault, respectively relatively capacitance current is much larger than before, in order to the size of limiting short-circuit current, the power distribution network containing wired-cable joint line adopts the neutral point resonance grounding method of operation more and more.And nowadays, along with the pollution of environment is more and more serious, China is just vigorously advocating the use of green energy resource; And day by day ripe along with distributed power source (Distributed Generator, DG) interconnection technology, a large amount of DG is directly or in being connected to the grid with the form of microgrid.So, the present invention say that new distribution net is the power distribution network adopting the resonance grounding method of operation containing DG and line-cable joint line, neutral point.

When not containing the conventional electrical distribution net neutral point employing resonance grounding of DG, under stable situation after there is singlephase earth fault, the phase place and the amplitude that perfect the zero-sequence current of circuit and faulty line do not have notable difference, make to utilize the selection method of steady state information amount to be limited to, in recent years, the research of transient information amount as route selection criterion is got more and more, and is studied much achievement.The characteristic quantity that route selection mainly utilizes has Sudden Changing Rate, polarity, amplitude and energy.The similarity perfecting zero-sequence current waveform between circuit in some research and utilization fault transient process is greater than this feature of similarity perfecting circuit and faulty line zero-sequence current waveform and realizes route selection; After some research and utilization fault occurs in first 1/4 cycle, the opposite polarity feature perfecting circuit and faulty line zero-sequence current realizes route selection; Some research and utilization perfects the capacity volume variance of the transient zero-sequence current between circuit and faulty line as route selection criterion; Recently, a large amount of comprehensive study is had to use energy feature, amplitude Characteristics and polar character to realize route selection.For a long time, although proposed a large amount of fault-line selecting methods, actual motion effect is unsatisfactory, and even to this day, many power stations are still with the method looking up the fault circuit manually drawing road.

For Novel distribution network, it is more complicated that the access of DG can make the structure of distribution become, it is made to become power network from single supply network, DG will change size, the flow direction of short-circuit current and distribute, utilize the reliability of the selection method of phase current and negative-sequence current feature to need to analyze separately, what the failure line selection problem of new distribution net was become is more difficult.The singlephase earth fault orientation problem of Novel distribution network is intelligent grid problem demanding prompt solution, and it has great importance for raising power supply reliability, minimizing loss of outage.

Summary of the invention

The technical problem to be solved in the present invention is the situation for the new distribution net generation singlephase earth fault containing DG, proposes a kind of fault-line selecting method containing DG power distribution network based on the non-power frequency zero-sequence current of transient state.

For achieving the above object, technical scheme of the present invention is: based on the fault-line selecting method containing DG new distribution net of the non-power frequency zero-sequence current of transient state, it is characterized in that: comprise the following steps: based on the fault-line selecting method containing DG power distribution network of the non-power frequency zero-sequence current of transient state, it is characterized in that: comprise the following steps: step S01: when DG power distribution network generation singlephase earth fault, extract each outlet L _jat the zero-sequence current i in fault initial time front 1/2 cycle, rear 5/2 cycle _0j, to each zero-sequence current i _0jcarry out filtering, obtain 5 order harmonic components i of each zero-sequence current _05j, wherein j is the numbering of circuit, and j is natural number;

Step S02: differentiate each outlet L _j5 subharmonic current i _05jpolar relationship, if 5 subharmonic current i of all outlets _05jpolarity identical, then judge that bus there occurs singlephase earth fault, otherwise, continue perform step S03; Step S03: utilize the set empirical mode decomposition algorithm improved successively to each zero-sequence current i _0jdecompose, extract each transient state non-power frequency zero-sequence current i of all outlets _fjk, and ask energy and the E of the non-power frequency zero-sequence current of the transient state of every bar circuit _j, and then obtain the energy weight Coefficient m of the non-power frequency zero-sequence current of each circuit _j; Step S04: differentiate each energy weight Coefficient m _jmagnitude relationship, the circuit with ceiling capacity weight coefficient is faulty line.

In an embodiment of the present invention, described digital notch filter comprises the following steps: step S011: utilize digital notch filter by 5 subharmonic current i in each circuit _05jfiltering, obtains filtered signal x ' (t); Step S012: original signal x (t) is poor with filtered signal x ' (t), just obtains 5 order harmonic components i _05j.

In an embodiment of the present invention, 5 subharmonic current i of each outlet of described differentiation _05jpolar relationship comprise following concrete steps: step S021: calculate correlation coefficient ρ _xy, its computing formula is: n is total sampling number of coherent signal x (n), y (n); ρ _xyfor signal x (n) and the related coefficient of y (n); Step S022: obtaining correlation matrix is: $M=\left[\begin{array}{c}{\mathrm{\&rho;}}_{11}{\mathrm{\&rho;}}_{12}...{\mathrm{\&rho;}}_{1n}\\ {\mathrm{\&rho;}}_{21}{\mathrm{\&rho;}}_{22}...{\mathrm{\&rho;}}_{2n}\\ .\\ .\\ .\\ {\mathrm{\&rho;}}_{n1}{\mathrm{\&rho;}}_{n2}...{\mathrm{\&rho;}}_{\mathrm{nn}}\end{array}\right],$ Shape correction is carried out to correlation matrix M: the elements in a main diagonal first deleting M, if ρ _ij>0, then by ρ _ijbe set to 1, if ρ _ij<0, then by ρ _ijbe set to 0; Obtain the correlation matrix after preliminary shaping if in the element of certain a line be 0 entirely, then can judge that the polarity of circuit 5 subharmonic current corresponding to this row is negative, and this line polarity information is set to 0, otherwise judge that the polarity of circuit 5 subharmonic corresponding to this row is as just, and this line polarity information is set to 1; Finally obtain the polarity discriminating row vector α only having 0,1 element, wherein in α, i-th element represents the polarity of i-th circuit.

In an embodiment of the present invention, described set empirical mode decomposition algorithm comprises the following steps: the likelihood probability ρ (D) asking for the first two IMF obtained through set empirical mode decomposition, if its ρ (D) is higher, then merge two IMF, then calculate the ρ (D) of the signal after merging and the 3rd IMF; Otherwise, calculate the ρ (D) of second IMF and the 3rd IMF; Down carry out successively, until ask and the ρ of last IMF (D); The computing formula of ρ (D) is: $\left\{\begin{array}{c}\mathrm{\&lambda;}=(\sqrt{\frac{N_1{N}_2}{N_1+N_2}}+0.12+\frac{0.11}{\sqrt{\frac{N_1{N}_2}{N_1+N_2}}})D\\ D=\underset{-\&infin;\&le;x\&le;\&infin;}{\max }|f\left(i\right)-r\left(i\right)|\\ \mathrm{\&rho;}\left(D\right)=Q_{\mathrm{ks}}\left(\mathrm{\&lambda;}\right)=2\underset{j=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}{(-1)}^{j-1}{e}^{-2j^2{\mathrm{\&lambda;}}^2}\end{array}\right.,$ In formula: N ₁, N ₂be two time series x, the counting of y; F (i), r (i) are respectively the Cumulative Distribution Function of x (i), y (i), f (i)=N ₁(i)/N ₁, r (i)=N ₂(i)/N ₂, N ₁(i), N ₂i () is respectively total number of the point less than x (i), y (i); ρ (D) is the likelihood probability of two time series x (i), y (i).

In an embodiment of the present invention, in step S03, the energy weight Coefficient m of the non-power frequency zero-sequence current of each circuit _jask for and comprise the following steps: step S031: ask for each bar circuit L _jtransient state non-power frequency zero-sequence current i _fjkenergy and E _j: $E_j=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}\left({\&Integral;}_{-\&infin;}^{+\&infin;}{c}_{\mathrm{jk}}^2\left(t\right)\mathrm{dt}\right)+{\&Integral;}_{-\&infin;}^{+\&infin;}{r_j}^2\left(t\right)\mathrm{dt},j=\mathrm{1,2},...,s;$ In formula: j is the numbering of circuit; c _jkt () is jth bar circuit L _jkth non-power frequency IMF component; N is jth bar circuit L _jthe quantity of non-power frequency mode; R (t) is the residual components of jth bar circuit; S is the total number of circuit; Step S032: the energy weight coefficient asking the non-power frequency zero-sequence current of each circuit: m _j=E _j/ (E ₁+ E ₂+ ... + E _s), j=1,2 ..., s.

Compared with prior art the present invention has following beneficial effect:

1, the present invention utilize the set empirical mode decomposition algorithm of improvement can decompose signal adaptive according to signal unique characteristics, without the need to selecting basis function, and adopt K-S method of inspection effectively can process the modal overlap problem occurred in set empirical mode decomposition algorithm;

2, the phase information of digital notch filter energy Obtaining Accurate each circuit zero-sequence current 5 order harmonic components of the present invention's utilization;

3, fault-line selecting method that the present invention carries can select faulty line accurately and reliably in the worst situations such as little corner-of-delta grounding fault, high resistance earthing fault, arc-earth faults, and not by the impact of DG volume change, without the need to artificial experience selected threshold.

Accompanying drawing explanation

Fig. 1 is the route selection process flow diagram of the embodiment of the present invention.

Fig. 2 be in the embodiment of the present invention adopt the frequency response characteristic of second order digital notch filter.

Fig. 3 is containing DG new distribution network simulation illustraton of model in the embodiment of the present invention.

Fig. 4 be in the embodiment of the present invention containing DG1 and DG2, only containing DG2, containing in DG1 (also not containing its feeder line) and DG2 tri-kinds of situations, the oscillogram of its head end zero-sequence current when singlephase earth fault occurs circuit 2.

Fig. 5 is when in the embodiment of the present invention, DG1 and DG2 accesses electrical network simultaneously, the zero-sequence current oscillogram of faulty line 2 and non-fault line 1 head end.

Fig. 6 is the oscillogram of each circuit zero-sequence current 5 order harmonic components in the embodiment of the present invention.

Fig. 7 is the needle-like figure of related coefficient between two of each 5 subharmonic currents in the embodiment of the present invention.

Fig. 8 is the energy weight coefficient bar chart of non-power frequency zero-sequence current in the embodiment of the present invention.

Embodiment

Below in conjunction with the drawings and specific embodiments, the present invention will be further described.

During new distribution net generation singlephase earth fault, the zero-sequence current that the access of DG can not change faulty line head end perfects the zero-sequence current of circuit and this feature of electric current sum of arc suppression coil.When the present invention fully utilizes arbitrary outlet fault, the energy Ratios of the transient state non-power frequency zero-sequence current of faulty line perfects the large of circuit; During bus-bar fault, the feature that all outlet zero-sequence current 5 order harmonic components polarity is identical, proposes a kind of fault-line selecting method containing DG new distribution net based on the non-power frequency zero-sequence current of transient state.

The process flow diagram of method provided by the invention, see Fig. 1, comprises the following steps: step S01: when DG power distribution network generation singlephase earth fault, extract each outlet L _jat the zero-sequence current i in fault initial time front 1/2 cycle, rear 5/2 cycle _0j, utilize digital notch filter to each zero-sequence current i _0jcarry out filtering, obtain 5 order harmonic components i of each zero-sequence current _05j, wherein j is the numbering of circuit, and j is natural number;

Step S02: differentiate each outlet L _j5 subharmonic current i _05jpolar relationship, if 5 subharmonic current i of all outlets _05jpolarity identical, then judge that bus there occurs singlephase earth fault, otherwise, continue perform step S03;

Step S03: utilize the set empirical mode decomposition algorithm improved successively to each zero-sequence current i _0jdecompose, extract each transient state non-power frequency zero-sequence current i of all outlets _fjk, and ask energy and the E of the non-power frequency zero-sequence current of the transient state of every bar circuit _j, and then obtain the energy weight Coefficient m of the non-power frequency zero-sequence current of each circuit _j;

Step S04: differentiate each energy weight Coefficient m _jmagnitude relationship, the circuit with ceiling capacity weight coefficient is faulty line.

Further in step S01, utilize digital notch filter to each zero-sequence current i _0jcarry out filtering, comprise the following steps: step S011: utilize digital notch filter by 5 subharmonic current i in each circuit _05jfiltering, obtains filtered signal x ' (t); Step S012: original signal x (t) is poor with filtered signal x ' (t), just obtains 5 order harmonic components i _05j.In the specific embodiment of the invention, the frequency response characteristic of second order digital notch filter is see Fig. 2.

To each outlet 5 subharmonic current i _05jpolarity discriminating can differentiate as follows: if 5 subharmonic current polarity of two circuits are contrary, then its related coefficient is negative, and close-1; If 5 subharmonic current polarity of two circuits are identical, then its related coefficient is just, and close to 1.Therefore, the present invention obtains by the related coefficient between two of circuit 5 subharmonic current the 5 subharmonic current i that its polarity information differentiates each outlet _05jpolar relationship.

Concrete steps are as follows:

Step S021: calculate correlation coefficient ρ _xy, its computing formula is: n is total sampling number of coherent signal x (n), y (n); ρ _xyfor signal x (n) and the related coefficient of y (n);

Step S022: obtaining correlation matrix is: $M=\left[\begin{array}{c}{\mathrm{\&rho;}}_{11}{\mathrm{\&rho;}}_{12}...{\mathrm{\&rho;}}_{1n}\\ {\mathrm{\&rho;}}_{21}{\mathrm{\&rho;}}_{22}...{\mathrm{\&rho;}}_{2n}\\ .\\ .\\ .\\ {\mathrm{\&rho;}}_{n1}{\mathrm{\&rho;}}_{n2}...{\mathrm{\&rho;}}_{\mathrm{nn}}\end{array}\right],$ Shape correction is carried out to correlation matrix M: the elements in a main diagonal first deleting M, if ρ _ij>0, then by ρ _ijbe set to 1, if ρ _ij<0, then by ρ _ijbe set to 0; Obtain the correlation matrix after preliminary shaping if in the element of certain a line be 0 entirely, then can judge that the polarity of circuit 5 subharmonic current corresponding to this row is negative, and this line polarity information is set to 0, otherwise judge that the polarity of circuit 5 subharmonic corresponding to this row is as just, and this line polarity information is set to 1; Finally obtain the polarity discriminating row vector α only having 0,1 element, wherein in α, i-th element represents the polarity of i-th circuit.

Further in step S03, the modal overlap problem as follows in process set empirical mode decomposition algorithm:

Adopt the modal overlap problem in K-S method of inspection process set empirical mode decomposition algorithm, K-S method is for analyzing the similarity of two groups of signals, and its principle is:

$\left\{\begin{array}{c}\mathrm{\&lambda;}=(\sqrt{\frac{N_1{N}_2}{N_1+N_2}}+0.12+\frac{0.11}{\sqrt{\frac{N_1{N}_2}{N_1+N_2}}})D\\ D=\underset{-\&infin;\&le;x\&le;\&infin;}{\max }|f\left(i\right)-r\left(i\right)|\\ \mathrm{\&rho;}\left(D\right)=Q_{\mathrm{ks}}\left(\mathrm{\&lambda;}\right)=2\underset{j=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}{(-1)}^{j-1}{e}^{-2j^2{\mathrm{\&lambda;}}^2}\end{array}\right.,$

In formula: N ₁, N ₂be two time series x, the counting of y; F (i), r (i) are respectively the Cumulative Distribution Function of x (i), y (i), f (i)=N ₁(i)/N ₁, r (i)=N ₂(i)/N ₂, N ₁(i), N ₂i () is respectively total number of the point less than x (i), y (i); ρ (D) is the likelihood probability of two time series x (i), y (i).

In the specific embodiment of the invention, first the likelihood probability ρ (D) of the first two IMF obtained through set empirical mode decomposition is asked for, if its ρ (D) is higher, then merge two IMF, then calculate the ρ (D) of the signal after merging and the 3rd IMF; Otherwise, calculate the ρ (D) of second IMF and the 3rd IMF.Down carry out successively, until ask and the ρ of last IMF (D).After above-mentioned steps process, the modal overlap problem that the signal of same frequency is broken down in different modalities can effectively be solved.

The energy weight Coefficient m of the non-power frequency zero-sequence current of each circuit _jacquiring method as follows:

First each bar circuit L is asked for _jtransient state non-power frequency zero-sequence current i _fjkenergy and E _j:

$E_j=\underset{k=1}{\overset{n}{\mathrm{\&Sigma;}}}\left({\&Integral;}_{-\&infin;}^{+\&infin;}{c}_{\mathrm{jk}}^2\left(t\right)\mathrm{dt}\right)+{\&Integral;}_{-\&infin;}^{+\&infin;}{r_j}^2\left(t\right)\mathrm{dt},j=\mathrm{1,2},...,s;$ In formula: j is the numbering of circuit; c _jkt () is jth bar circuit L _jkth non-power frequency IMF component; N is jth bar circuit L _jthe quantity of non-power frequency mode; R (t) is the residual components of jth bar circuit; S is the total number of circuit;

Ask the energy weight coefficient of the non-power frequency zero-sequence current of each circuit again:

m _j＝E _j/(E ₁+E ₂+…+E _s),j＝1,2,…,s。

In the specific embodiment of the invention, utilize MATLAB/SimPowersystem simulation software foundation new distribution network as shown in Figure 3, distribution contains 6 feeder lines and 2 distributed power sources DG1, DG2, wherein, feeder line is made up of pole line, cable, built on stilts-cable hybrid line three kinds of circuits, the parameter of overhead transmission line: positive sequence resistance is 0.17 Ω/km, positive sequence inductance is 1.21mH/km, positive sequence electric capacity is 0.0097uF/km, zero sequence resistance is 0.23 Ω/km, zero sequence inductance is 5.48mH/km, and zero sequence electric capacity is 0.006uF/km; The parameter of cable line: positive sequence resistance is 0.265 Ω/km, positive sequence inductance is 0.255mH/km, and positive sequence electric capacity is 0.17uF/km, and zero sequence resistance is 2.54 Ω/km, and zero sequence inductance is 1.019mH/km, and zero sequence electric capacity is 0.153uF/km.The capacity of DG1 is 3MVA, is equivalent to rotary-type DG, and it is grid-connected on circuit 6 by the cable feeder line of 6km; The capacity of DG2 is 2MW, and an active power of output, is equivalent to inverse type DG, and its direct grid-connected is on circuit 2.

Now suppose apart from the position of bus 5km, Single Phase Metal earth fault to occur on circuit 2, in simulation process, the compensativity cancelling arc coil is 10%, Fig. 4 illustrate containing DG1 and DG2, only containing DG2, containing in DG1 (also not containing its feeder line) and DG2 tri-kinds of situations, the Changing Pattern of circuit 2 head end zero-sequence current, known, when disregarding the feeder line of DG, the access of DG can not make the zero-sequence current of circuit 2 head end change.Fig. 5 illustrates DG1 and DG2 when accessing electrical network simultaneously, and faulty line 2 and the zero-sequence current oscillogram of non-fault line 1 head end, can find out, the non-power frequency zero-sequence current of transient state of faulty line much larger than non-fault line, the optional circuit that is out of order accordingly.

Utilize the present invention can carry out route selection to different singlephase earth fault, list part typical fault situation in a particular embodiment.

In simulation process, getting sample frequency is 100kHz; In the set empirical mode decomposition improved, the variance of added white noise is 0.02, and cycle index is 50; The compensativity of arc suppression coil is 8%, i.e. the L=0.3885H of arc suppression coil, R=3.662 Ω.When phase voltage crosses peak value, Single Phase Metal earth fault is there is in circuit 6 apart from the position of bus 5km, as shown in Figure 6, the related coefficient between two of each 5 subharmonic currents as shown in Figure 7, tries to achieve polarity discriminating row vector α=[1 now to the oscillogram of each circuit zero-sequence current 5 order harmonic components obtained through digital notch filter, 1,1,1,1,0], can judge that bus does not break down by polarity discriminating row vector α.And the energy weight coefficient of each circuit as shown in Figure 8, the energy weight coefficient of circuit 6 is maximum as can be seen from Figure 8, and can prepare to learn, fault occurs on circuit 6.

Typical fault route selection result:

In following table, L _ibe i-th circuit, X is the distance of trouble spot apart from bus, and R is stake resistance, and p is the compensativity of arc suppression coil, and θ is fault angle, and R1 is the capacity of DG1, and R2 is the capacity of DG2, and α is 5 subharmonic polarity discriminating row vectors, and m is energy weight coefficient.

(1) when earth point position is different

When earthing position is different, the zero sequence impedance of system will change, and then the amplitude of each circuit zero-sequence current and phase place also will change.When the compensativity of arc suppression coil is 8%, stake resistance is 50 Ω, and fault angle is time, route selection result when fault occurs in diverse location is as shown in table 1.

The asynchronous route selection result in table 1 earth point position

(2) when the compensativity of arc suppression coil is different

Under different compensativity, there is Single Phase Metal earth fault in the position apart from bus 8km in circuit 2, fault angle is route selection result is as shown in table 2.

The asynchronous route selection result of table 2 compensativity

(3) when stake resistance is different

When high resistance ground, each circuit zero-sequence current will reduce, and fault transient feature will be affected.When compensativity is 10%, circuit 5 is in the position apart from bus 5km through different resistance eutral grounding, and fault angle is route selection result is as shown in table 3.

Route selection result during table 3 different faults resistance

(4) when fault initial phase angle is different

Generally speaking, when there is little corner-of-delta grounding fault, faulty line is not easily selected.Under different fault angle, when compensativity is 10%, there is earth fault apart from bus 10km place in circuit 3, fault resstance is 100 Ω, and route selection result is as shown in table 4.

Route selection result during table 4 different faults angle

(5) when the DG method of operation changes

Because DG capacity is different, the electric current externally provided is also different, and especially when there is singlephase earth fault, the capacity of DG is different, and the fault current distribution of circuit can change, and when taking into account DG feeder line, the zero-sequence current of part circuit also will change.When the capacity of DG1 and DG2 changes, compensativity is 10%, and circuit 2, in the position apart from bus 10km, metallic earthing fault occurs, and fault angle is route selection result is as shown in table 5.

Route selection result during table 5 DG volume change

(6) arc-earth faults

When the compensativity of arc suppression coil is 8%, fault angle is time, there is intermittent arc-earth fault apart from bus 10km place in circuit 3.Characterize the extinguishing of electric arc with the folding of switch and restrike, blow-out and arcing total degree 3, the arcing moment is 0.065s, 0.085s, 0.105s, and the blow-out time is 0.075s, 0.095s.Route selection result is as shown in table 6.

There is route selection result during arc fault in table 6

(7) when Conductor transposition is not good

The overhead transmission line transposition of power distribution network is ill-conditioned, causes the three-phase ground capacitance of pole line uneven, the neutral point voltage of distribution is offset, produces out-of-balance current.The unit resistance of pole line L3 is made to be [0.1803,0.048,0.048; 0.048,0.1803,0.048; 0.048,0.048,0.1803] * Ω/km, unit inductance is [2.3157,1.1361,0.9975; 1.1361,2.3157,1.1361; 0.9975,1.1361,2.3175] * 10^ (-3) H/km, specific capacitance is [7.8571 ,-1.8039 ,-0.8857;-1.8039,8.1715 ,-1.8039;-0.8857 ,-1.8039,7.7851] * 10^ (-9) * F/km.When compensativity is 8%, there is metallic earthing fault in the position of distance bus 15km and bus in circuit 3, fault angle is all respectively route selection result is as shown in table 7

Route selection result when table 7 Conductor transposition is not good enough

Be more than preferred embodiment of the present invention, all changes done according to technical solution of the present invention, when the function produced does not exceed the scope of technical solution of the present invention, all belong to protection scope of the present invention.

Claims (5)

1., based on the fault-line selecting method containing DG power distribution network of the non-power frequency zero-sequence current of transient state, it is characterized in that: comprise the following steps:

Step S01: when DG power distribution network generation singlephase earth fault, extract each outlet at the zero-sequence current in fault initial time front 1/2 cycle, rear 5/2 cycle , to each zero-sequence current carry out filtering, obtain 5 order harmonic components of each zero-sequence current , wherein jfor the numbering of circuit, jfor natural number;

Step S02: differentiate each outlet 5 subharmonic currents polar relationship, if 5 subharmonic currents of all outlets polarity identical, then judge that bus there occurs singlephase earth fault, otherwise, continue perform step S03;

Step S03: utilize the set empirical mode decomposition algorithm improved successively to each zero-sequence current decompose, extract the non-power frequency zero-sequence current of each transient state of all outlets , and ask the non-power frequency zero-sequence current of the transient state of every bar circuit energy and , and then obtain the energy weight coefficient of the non-power frequency zero-sequence current of each circuit ;

Step S04: differentiate each energy weight coefficient magnitude relationship, the circuit with ceiling capacity weight coefficient is faulty line.

2. the fault-line selecting method containing DG power distribution network based on the non-power frequency zero-sequence current of transient state according to claim 1, is characterized in that: adopt digital notch filter to each zero-sequence current carry out filtering, comprise the following steps:

Step S011: utilize digital notch filter by 5 subharmonic currents in each circuit filtering, obtains filtered signal ; Step S012: by original signal with filtered signal differ from, just obtain 5 order harmonic components .

3. the fault-line selecting method containing DG power distribution network based on the non-power frequency zero-sequence current of transient state according to claim 1, is characterized in that: 5 subharmonic currents of each outlet of described differentiation polar relationship comprise following concrete steps:

Step S021: calculate related coefficient , its computing formula is: , for coherent signal , total sampling number; for signal with related coefficient;

Step S022: obtaining correlation matrix is: , to correlation matrix mcarry out Shape correction: first delete mthe elements in a main diagonal, if , then will be set to 1, if , then will be set to 0; Obtain the correlation matrix after preliminary shaping if, in the element of certain a line be 0 entirely, then can judge that the polarity of circuit 5 subharmonic current corresponding to this row is negative, and this line polarity information is set to 0, otherwise judge that the polarity of circuit 5 subharmonic corresponding to this row is as just, and this line polarity information is set to 1; Finally obtain the polarity discriminating row vector only having 0,1 element , wherein in iindividual element represents ithe polarity of bar circuit.

4. the fault-line selecting method containing DG power distribution network based on the non-power frequency zero-sequence current of transient state according to claim 1, is characterized in that: described set empirical mode decomposition algorithm comprises the following steps: the likelihood probability asking for the first two IMF obtained through set empirical mode decomposition if, its higher, then merge two IMF, then calculate the signal after merging and the 3rd IMF ; Otherwise, calculate second IMF and the 3rd IMF's ; Down carry out successively, until ask with last IMF's till; computing formula be: ,

In formula: , be two time serieses , count; , be respectively , cumulative Distribution Function, , , , be respectively ratio , total number of little point; be two time serieses , likelihood probability.

5. the fault-line selecting method containing DG power distribution network based on the non-power frequency zero-sequence current of transient state according to claim 1, is characterized in that: in step S03, the energy weight coefficient of the non-power frequency zero-sequence current of each circuit ask for and comprise the following steps:

Step S031: ask for each bar circuit the non-power frequency zero-sequence current of transient state energy and :

；

In formula: jfor the numbering of circuit; be jbar circuit ? kindividual non-power frequency IMF component; nbe jbar circuit the quantity of non-power frequency mode; be jthe residual components of bar circuit; for the total number of circuit;

Step S032: the energy weight coefficient asking the non-power frequency zero-sequence current of each circuit: .