CN105116294A - Traveling wave polarity measure based power distribution network cable fault monitoring method - Google Patents

Abstract

The invention relates to a traveling wave polarity measure based power distribution network cable fault monitoring method. The method is applied to a neutral non-effectively grounding power grid with N feeders. The method is carried out through the following steps: 1) acquiring three-phase initial current traveling wave signals of the feeders; 2) acquiring zero modulus initial current traveling wave signals of the feeders; 3) acquiring a zero modulus initial current traveling wavewavelet modulus maximum m(j) of the feeders; 4) calculating traveling wave polarity measure D (j) of the feeders wherein D(j) is defined as description; 5) obtaining feeder choosing results according to the traveling wave polarity measure D(j)calculated the step 4); taking the feeder corresponding to the maximal value of {D(1), D(2)...D(j)...D(N)} as a fault feeder. The method utilizes a small current grounding fault feeder choosing principle of fault initial traveling waves to reduce interference, effectively finds out abnormal running state existing in small current grounding, and accurately determines a line with fault timely, and therefore, negative influence brought about by fault can be reduced. The method can also achieve high sensitivity.

Description

Based on the cable faults of distribution network monitoring method that polarity of traveling wave is estimated

Technical field

The present invention relates to the cable faults of distribution network monitoring method estimated based on polarity of traveling wave, belong to cable fault monitoring technical field.

Background technology

China's power distribution network extensively adopts neutral non-effective grounding mode, and this earthing mode has the high advantage of power supply reliability, but its singlephase earth fault incidence is the highest.When after generation singlephase earth fault, healthy phases voltage raises as original 1.732 times, under individual cases, earthing capacitance current may cause trouble spot electric arc to leap, instantaneous appearance, than the large 4-5 of phase voltage superpotential doubly, causes insulation breakdown, is expanded into or multipoint earthing short circuit further at 2; The electric arc of trouble spot also can cause total system superpotential, usually burns cable and even causes fire.Therefore, the safe reliability of the singlephase earth fault serious threat power distribution network of power distribution network is that Accident prevention expands, and wishes to select as early as possible faulty line and go forward side by side row relax in operation.But, because single-phase earthing is that the ground current of trouble spot is very little, and single-phase earth fault line selection and fault-location problem are not solved for a long time well by power supply winding and transmission line of electricity distributed capacitance formation short-circuit loop over the ground.

Failure line selection principle based on different principle cuts both ways.Although signal injection method, needs additionally to add equipment accurately and reliably, economy is too poor.When there is low current grounding, the extraction difficulty of Weak fault steady-state current is the subject matter utilizing fault steady-state component detection of ground faults.With utilize compared with steady-state signal, because transient signal amplitude is much larger than steady-state signal, utilizing transient signal to carry out fault detect has good sensitivity and selectivity, but the greatest problem of transient route selection is difficult to exclusive PCR, selects out of order feature band.

Chinese patent literature CN103217622A discloses a kind of distribution network fault line selection method based on multiterminal voltage traveling wave, belongs to Relay Protection Technology in Power System field.The steps include: the transformer station in power distribution network and each basic routing line branches end installed rows ripple harvester; Utilize transformer station's bus connection switch divide-shut brake to produce travelling wave signal, the mistiming of the initial time that on calculating basic routing line, the fault traveling wave of each branches end arrives and the initial time that the fault traveling wave at substation bus bar place arrives, obtain array reference time; The mistiming of the initial time that during calculating fault, on basic routing line, the fault traveling wave of each branches end arrives and the initial time that the fault traveling wave at substation bus bar place arrives, set up array fault-time; By reference time array and fault-time array carry out least square fitting, realize the use processing of row ripple time of arrival of each collection point record, find faulty line.But this patent does not consider that stake resistance is too high, the situation of fault initial phase angle, and fault simulation is too single, can not be advantageously applied to actual conditions.

Chinese patent literature CN102279346B low current neutral grounding system fault route selecting method, comprises the following steps: the sample data collected by protective device is divided into failure classes and non-faulting class according to space length, and calculates all kinds of center; When line outlet residual voltage is out-of-limit, record sample to be tested data; Calculate the distance at sample to be tested and failure classes center and non-faulting class center, judge whether circuit breaks down according to the distance of distance.The fault-line selecting method of this patent utilization sample data, proposes fault-line selecting method and utilizes space relative distance as criterion, broken tradition fault characteristic value and setting valve compared as Protection criteria.But this patent needs a large amount of sample datas to be used as foundation, is subject to the restriction of sample data accuracy, restricted too large.

In sum, to cause accurate measurements to go out earth fault by the impact of many factors very difficult for neutral non-effective grounding electrical network.Therefore, the fault-line selecting method that further research and utilization initial row ripple and fault measurement combine, has a good application prospect.

Summary of the invention

For the deficiencies in the prior art, the invention provides the cable faults of distribution network monitoring method estimated based on polarity of traveling wave;

Terminological interpretation

1, fault measurement, refers to a kind of arithmetic number of the close degree that can characterize the feature of respective circuit and faulty line under certain criterion benchmark.In many circuits of same electric pressure electrical network, if the fault measurement of a certain bar circuit is maximum, then can show that this circuit be the possibility of faulty line is maximum.See Jia Qingquan, Yang Qixun, Yang Yihan. the many criterions of one-phase earthing failure in electric distribution network based on fault measurement concept and evidence theory merge [J]. Proceedings of the CSEE, 2003,12:9-14.

2, polarity of traveling wave is estimated, and refers to a kind of arithmetic number that can characterize the close degree of the feature of respective circuit and faulty line under zero mould initial current polarity of traveling wave character references.In many circuits of same electric pressure electrical network, if the fault measurement of a certain bar circuit is maximum, then can show that this circuit be the possibility of faulty line is maximum.

Technical scheme of the present invention is:

Based on the cable faults of distribution network monitoring method that polarity of traveling wave is estimated, be applied to the neutral non-effective grounding electrical network of total N bar feeder line, concrete steps comprise:

(1) the three-phase initial current travelling wave signal of each bar feeder line is gathered: the initial current travelling wave signal being gathered three-phase by each bar feeder line of current transformer to power distribution network;

(2) obtain zero mould initial current travelling wave signal of each bar feeder line: by phase-model transformation, the initial current travelling wave signal of three-phase step (1) obtained changes zero mould initial current travelling wave signal into;

(3) capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line is obtained: wavelet transformation is carried out to zero mould initial current travelling wave signal of each bar feeder line, obtain capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line, j ∈ [1, N];

(4) polarity of traveling wave calculating each bar feeder line estimates D (j), and described polarity of traveling wave estimates the defined formula of D (j) such as formula shown in (I):

$D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\×\frac{\underset{l=1,l\overline{)=}j}{\overset{N}{\Σ}}sgn\[m\left(j\right)\×m\left(l\right)\]}{N-1}---\left(I\right)$

In formula (I), D (j) represents that the polarity of traveling wave of jth bar feeder line is estimated, and namely represents the possibility of fault generation jth bar feeder line; The span of D (j) is [0,1]; M (j) represents the amplitude of the capable ripple Wavelet Modulus Maxima of jth bar feeder line zero mould initial current, M (j) > 0, i.e. M (j)=| m (j) |; m (l) represents the amplitude of the l article of capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, M (l) > 0, i.e. M (l)=| m (l) |, l ∈ [1, N] and l ≠ j; Function sgn [M (j)-M (l)] is used for the size of the amplitude difference reflecting certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1,0 or-1, as M (j) > M (l), sgn [M (j)-M (l)]=1; As M (j)=M (l), sgn [M (j)-M (l)]=0; As M (j) < M (l), sgn [M (j)-M (l)]=-1;

(5) polarity of traveling wave of each bar feeder line calculated according to step (4) is estimated D (j) and exports the result of low-current ground fault line selection: get { D (1), D (2) ... D (j) ... D (N) } in feeder line corresponding to maximal value be faulty line.

Working foundation principle of the present invention is:

In power distribution network, the tie point of bus and fault feeder is wave impedance point of discontinuity.When fault initial row ripple arrives this, there is refraction and reflection in fault initial row ripple, and refraction row ripple enters bus, and reflected traveling wave returns fault feeder.The wave process of refraction row ripple in bus terminates very soon, continue to be advanced in other feeder line and go, refraction row ripple reflects at other feeder terminal, before this reflected traveling wave is advanced to bus, there is incident wave and reflection wave in fault feeder position near bus, other circuit position near bus only exists refraction wave.This process is exactly so-called primary wave process, as shown in Figure 1.

The key distinction of resonant earthed system and isolated neutral system is the equivalent wave impedance of neutral point, the refraction of fault initial row ripple on bus and reflection, and power distribution network adopts the resonance grounding method of operation, i.e. neutral by arc extinction coil grounding.The equivalent wave impedance of neutral point is equivalent to the equivalent wave impedance of arc suppression coil.Because initial row wave frequency is very high, the wave impedance of substation transformer is directly proportional to frequency, and therefore, row ripple does not propagate into power supply by substation transformer substantially.

The present invention is directed to zero mould initial current travelling wave analysis, do not consider the decay on the line of the capable ripple of initial current, think that the incident row ripple that is advanced to bus is identical with the row ripple advanced to bus from trouble spot.

Circuit Fault on Secondary Transformer zero mould wave impedance is called the zero mould equivalence wave impedance of neutral point to bus, with Z ' with connecting of neutral point wave impedance _eqrepresent.

The wave impedance setting each feeder line is all equal, then make Z _l10=Z _l20=...=Z _lN0=Z.Fault feeder produces fault initial row ripple i _f0, refraction and reflection occur when being advanced to bus along fault feeder, and the wave impedance of incident wave is the wave impedance Z of each feeder line, the wave impedance Z of refraction wave _zfor N-1 bar perfects feeder line wave impedance and neutral point equivalence wave impedance Z ' _eqparallel connection, as shown in electric current catadioptric coefficient formula (II):

$Z_Z=\frac{1}{\frac{1}{Z_{eq}^{\&prime;}}+\frac{N-1}{Z}}<\frac{1}{0+\frac{N-1}{Z}}=\frac{Z}{N-1}---\left(II\right)$

Released by formula (II), the refraction wave i of fault feeder _zZand the reflection wave i of fault feeder _fbe respectively:

$\left\{\begin{array}{c}{i}_{ZZ}=\frac{2Z}{Z+Z_Z}{i}_{F0}\\ i_f=\frac{Z-Z_Z}{Z+Z_Z}{i}_{F0}\end{array}\right.---\left(III\right)$

Every bar perfects the refraction wave i that feeder line flows through _zfor:

$i_Z=\frac{Z_Z}{Z}{i}_{ZZ}<\frac{1}{N-1}{i}_{ZZ}---\left(IV\right)$

The velocity of wave of row ripple is very fast, and on bus, the current transformer mounting points of each feeder line is extremely short from the distance of bus, therefore, from the capable ripple i of faulty line zero mould initial current that current transformer records _nfor the superposition of incident wave and reflection wave, and perfect the capable ripple i of line zero mould initial current _jbe only refraction wave, i _jand i _namplitude expression such as formula shown in (V):

$\left\{\begin{array}{c}{i}_N=i_{F0}+i_f=\frac{2Z}{Z+Z_Z}{i}_{F0}\\ i_J=i_Z=\frac{2Z_Z}{Z+Z_Z}{i}_{F0}\end{array}\right.---\left(V\right)$

Uniform provisions according to electric system: it is positive dirction that electric current flows to circuit from bus.The capable ripple i of faulty line zero mould initial current _nwith perfect the capable ripple i of line zero mould initial current _jfor:

$\left\{\begin{array}{c}{i}_N=i_{F0}+i_f=-\frac{2Z}{Z+Z_Z}{i}_{F0}\\ i_J=i_Z=\frac{2Z_Z}{Z+Z_Z}{i}_{F0}\end{array}\right.---\left(VI\right)$

In the moment of current traveling wave through inductance, inductance is equivalent to open circuit to current traveling wave, and negative total reflection occurs current traveling wave.Therefore, be easy analysis, when the capable ripple of zero mould initial current is advanced to arc suppression coil, the capable ripple of zero mould initial current is considered as not being refracted to arc suppression coil, neutral point equivalence wave impedance Z _eqfor infinity, formula (VI) changes formula (VII) into:

$\left\{\begin{array}{c}{i}_N=-\frac{2(N-1)}{N}{i}_{F0}\\ i_J=\frac{2}{N}{i}_{F0}\end{array}\right.---\left(VII\right)$

When neutral point ungrounded electric network generation singlephase earth fault, can be obtained by formula (VII):

(1) the zero mould initial current polarity of traveling wave perfecting feeder line is all identical;

(2) the capable wave amplitude of zero mould initial current of fault feeder is maximum, and it equals (N-1) with the ratio of the capable wave amplitude of zero mould initial current perfecting circuit;

(3) the capable ripple of zero mould initial current of fault feeder is contrary with the zero mould initial current polarity of traveling wave perfecting feeder line.

Adopt Wavelet Modulus Maxima to represent the capable ripple of zero mould initial current of each bar feeder line, the feature of the capable ripple Wavelet Modulus Maxima of zero mould fault initial current is not by the impact that neutral operation method changes, and principle is reliably clear and definite, and concrete rule is as follows:

For total N bar feeder line system with non-effectively earthed neutral, when there is line side singlephase earth fault:

(1) the zero mould initial current capable ripple Wavelet Modulus Maxima polarity perfecting feeder line is all identical;

(2) fault feeder is contrary with the zero mould initial current capable ripple Wavelet Modulus Maxima polarity perfecting feeder line;

With perfect compared with feeder line, the amplitude maximum of the capable ripple Wavelet Modulus Maxima of fault feeder zero mould initial current, its with perfect circuit ratio equal (N-1), namely the amplitude of the zero mould initial current capable ripple Wavelet Modulus Maxima of fault feeder equals the amplitude sum that other (N-1) bars perfect the capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current.

For the singlephase earth fault of system with non-effectively earthed neutral generation bus bar side, the polarity of the zero mould initial current capable ripple Wavelet Modulus Maxima of each feeder line is identical.

Preferred according to the present invention, formula (I) definition is estimated definition one and polarity of traveling wave by polarity of traveling wave and is estimated definition two and combine, and polarity of traveling wave is estimated definition one and used D ₁j () represents, polarity of traveling wave estimates definition dual-purpose D ₂j () represents; The derivation of formula (I) is:

According to the relation perfecting circuit and faulty line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, perfect circuit contrary with faulty line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, and it is all identical to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, constructs the polarity of traveling wave relevant with zero mould initial current capable ripple Wavelet Modulus Maxima and estimate definition one D ₁(j), that is:

$D_1\left(j\right)=\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}\mathrm{sgn}\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(VIII\right)$

In formula (VIII), sgn () is sign function, is defined as formula (Ⅸ):

$\mathrm{sgn}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(IX\right)$

From formula (Ⅸ), described sign function sgn [m (j) × m (l)], be used to the polar relationship reflecting certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1 or-1;

If m (j) × m (l) > 0, namely by after the polarity that compares two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, show that capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of jth bar feeder line is identical with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of kth bar feeder line, then sgn [m (j) × m (l)]=1;

If m (j) × m (l) < 0, namely by after the polarity of the capable ripple Wavelet Modulus Maxima of zero mould initial current that compares two feeder lines, show that capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of jth bar feeder line is contrary with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of kth bar feeder line, then sgn [m (j) × m (l)]=-1;

If m (j) × m (l)=0, namely the capable ripple Wavelet Modulus Maxima of zero mould initial current of jth article or l article of feeder line has one to be 0, then sgn [m (j) × m (l)]=0; For wavelet transformation, the capable ripple Wavelet Modulus Maxima of zero mould initial current be 0 situation do not exist; Therefore the value of sgn [m (j) × m (l)] can not be 0;

When jth bar feeder line is faulty line, capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current is contrary with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of other all feeder line, l ∈ [1, N] and l ≠ j, for value from 1 successively to N l article of feeder line m (l) for, the cumulative sum of sgn [m (j) × m (l)] equals-(N-1), D ₁(j)=-1;

When jth bar feeder line is for perfecting circuit, to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity identical with other for capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current, for value from 1 successively to N l article of feeder line m (l) for, the cumulative sum of sgn [m (j) × m (l)] is greater than-(N-1), D ₁(j) >-1;

Therefore, polarity of traveling wave estimates definition one D ₁j () meets the requirement of fault measurement universal, obviously distinguished faulty line and the result perfecting circuit.

Polarity of traveling wave estimates being constructed as follows of definition two:

In a practical situation, due to the reason such as measuring error, interference, make the amplitude of the zero mould initial current capable ripple Wavelet Modulus Maxima of each bar feeder line very little.From magnitude relation, the amplitude perfecting the zero mould initial current capable ripple Wavelet Modulus Maxima of circuit is minimum, and close to 0, now, then carrying out computing with more respective polarity to the zero mould initial current capable ripple Wavelet Modulus Maxima of each bar feeder line, is nonsensical.

Polarity of traveling wave estimates definition two D ₂(j), also known as amplitude coefficient, that is:

$D_2\left(j\right)=\frac{M\left(j\right)}{Sum}---\left(X\right)$

The definition that the polarity of traveling wave of jth bar circuit estimates D (j) estimates definition one D by polarity of traveling wave ₁j () and polarity of traveling wave estimate definition two D ₂j () combines, for:

$D\left(j\right)=\frac{1}{2}-D_1\left(j\right)\&times;D_2\left(j\right)---\left(XI\right)$

That is: $D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\&times;\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}\mathrm{sgn}\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(XIII\right)$

For perfecting circuit, polarity of traveling wave estimates definition two D ₂j the span of () is 0 < D ₂(j) < 0.5, and D ₂j the numerical value of () is very little, close to 0, when disturbing larger, the polarity of traveling wave perfecting circuit estimates definition one D ₁j () may have error, add amplitude coefficient D ₂j (), substantially reduces the error that polarity of traveling wave estimates D (j).For faulty line, amplitude coefficient D ₂(j)=0.5, D (j)=1.Therefore, definition two D is estimated by polarity of traveling wave ₂j (), makes D ₁j defect that () is disturbed impact improves, and makes polarity of traveling wave estimate the antijamming capability of D (j) large.

Polarity of traveling wave is estimated D (j) and can be distinguished faulty line significantly and perfect circuit.Ideally, the polarity of traveling wave of faulty line estimates D (j)=1, and the polarity of traveling wave perfecting circuit estimates D (j)=0.

Beneficial effect of the present invention is:

1, the present invention is on the basis of labor singlephase earth fault initial row wave characteristic, by the capable ripple Wavelet Modulus Maxima of zero mould initial current, defines the concept that polarity of traveling wave is estimated, proposes the cable faults of distribution network monitoring method estimated based on polarity of traveling wave.

2, the present invention utilizes the low-current ground fault line selection principle of fault initial row ripple to reduce interference, effectively detect the irregular operating state existed in small current neutral grounding mode, and accurately and timely determine guilty culprit circuit, reduce the harmful effect that fault is brought, there is higher sensitivity.

Accompanying drawing explanation

Fig. 1 the present invention is based on the cable faults of distribution network monitoring method workflow diagram that polarity of traveling wave estimates;

Fig. 2 is that polarity of traveling wave of the present invention estimates D (j) derivation schematic diagram;

Fig. 3 is the catadioptric schematic diagram of initial row ripple at bus place;

The ATP realistic model schematic diagram that Fig. 4 is system with non-effectively earthed neutral described in embodiment 1;

Fig. 5 is six feeder line zero mould initial current traveling-wave waveform schematic diagram in embodiment 1;

In Fig. 5, horizontal ordinate refers to the time, and unit is ms, and ordinate refers to electric current, and unit is A, I ₁₀, I ₂₀, I ₃₀, I ₄₀, I ₅₀, I ₆₀represent the zero mould initial current traveling-wave waveform of feeder line 1-6 successively;

The capable ripple schematic diagram of zero mould initial current of feeder line 1 when Fig. 6 is fault initial phase angle 90 ° in embodiment 1;

Fig. 7 is the wavelet conversion coefficient schematic diagram of feeder line 1 in embodiment 1;

The capable ripple schematic diagram of zero mould initial current of feeder line 2 when Fig. 8 is fault initial phase angle 90 ° in embodiment 1;

Fig. 9 is the wavelet conversion coefficient schematic diagram of feeder line 2 in embodiment 1;

The capable ripple schematic diagram of zero mould initial current of feeder line 3 when Figure 10 is fault initial phase angle 90 ° in embodiment 1;

Figure 11 is the wavelet conversion coefficient schematic diagram of feeder line 3 in embodiment 1.

The capable ripple schematic diagram of zero mould initial current of feeder line 4 when Figure 12 is fault initial phase angle 90 ° in embodiment 1;

Figure 13 is the wavelet conversion coefficient schematic diagram of feeder line 4 in embodiment 1.

The capable ripple schematic diagram of zero mould initial current of feeder line 5 when Figure 14 is fault initial phase angle 90 ° in embodiment 1;

Figure 15 is the wavelet conversion coefficient schematic diagram of feeder line 5 in embodiment 1.

The capable ripple schematic diagram of zero mould initial current of feeder line 6 when Figure 16 is fault initial phase angle 90 ° in embodiment 1;

Figure 17 is the wavelet conversion coefficient schematic diagram of feeder line 6 in embodiment 1.

Embodiment

Below in conjunction with Figure of description and embodiment, the present invention is further qualified, but is not limited thereto.

Embodiment 1

Set up the ATP realistic model of system with non-effectively earthed neutral, system with non-effectively earthed neutral comprises: the transformer station of 110kV/10.5kV, 6 feeder lines, the grounding transformer of 0.4kV/10kV, K switch, arc suppression coil K _p, 6 feeder lines adopt distributed parameter model, and namely in Fig. 4, the grounding transformer neutral point of label 1,2,3,4,5,6,0.4kV/10kV connects arc suppression coil K by K switch _p, being isolated neutral system when K switch is opened, is resonant earthed system when K switch closes.As shown in Figure 4.

The feeder line parameter of system with non-effectively earthed neutral is as follows: feeder line positive sequence impedance Z ₁=(0.17+j0.38) Ω/km, feeder line zero sequence impedance Z ₀=(0.23+j1.72) Ω/km; Positive sequence is admittance b over the ground ₁=(j3.045) μ s/km, zero sequence is admittance b over the ground ₀=(j1.884) μ s/km, under 10kV electric pressure, the approximate range of feeder line transmission distance is 6-20km.Article 6, the length of feeder line is respectively: 6km, 9km, 12km, 15km, 18km, 20km;

The grounding transformer parameter of the 0.4kV/10kV of system with non-effectively earthed neutral is as follows: rated capacity: 40000kVA, open circuit loss: 35.65kW.Former and deputy limit single-phase centering point coil resistance: 0.4 Ω, 0.006 Ω; The inductance of former and deputy limit single-phase centering point coil: 12.2 Ω, 0.183 Ω; Exciting current: 0.675A, magnetic flux: 202.2WB, magnetic circuit resistance: 400k Ω.

The load parameter of system with non-effectively earthed neutral is as follows: the positive sequence impedance only considering load, and negative phase-sequence, the zero mould impedance of load are set to infinity, and the equivalent load impedance of each outlet is unified adopts Z _l=(360+j30) Ω.

The parameter of the arc suppression coil of system with non-effectively earthed neutral: in resonant earthed system, compensativity is defined as 8%, L _p=6.126 Η, its resistance in series R _p=192 Ω.The material calculation of simulation software ATP is: 1 μ s, namely 1 × 10 ^-6s.

Based on the cable faults of distribution network monitoring method that polarity of traveling wave is estimated, be applied to the ATP realistic model of system with non-effectively earthed neutral, fault condition is respectively: (1) fault initial phase angle, is respectively 30 °, 60 ° and 90 °; (2) transition resistance, is respectively 10 Ω, 100 Ω and 500 Ω; (3) position of failure point, be respectively distance bus 4km, 5km, 6km place, concrete steps comprise:

(1) the three-phase initial current travelling wave signal of each bar feeder line is gathered: the initial current travelling wave signal being gathered three-phase by each bar feeder line of current transformer to power distribution network;

(2) obtain zero mould initial current travelling wave signal of each bar feeder line: by phase-model transformation, the initial current travelling wave signal of three-phase step (1) obtained changes zero mould initial current travelling wave signal into; Article six, feeder line zero mould initial current traveling-wave waveform schematic diagram as shown in Figure 5; During fault initial phase angle 90 °, the capable wavelength-division of zero mould initial current of feeder line 1, feeder line 2, feeder line 3, feeder line 4, feeder line 5, feeder line 6 is not as shown in Fig. 6, Fig. 8, Figure 10, Figure 12, Figure 14, Figure 16;

(3) capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line is obtained: carry out wavelet transformation to zero mould initial current travelling wave signal of each bar feeder line, the wavelet conversion coefficient schematic diagram of feeder line 1, feeder line 2, feeder line 3, feeder line 4, feeder line 5, feeder line 6 is respectively as shown in Fig. 7, Fig. 9, Figure 11, Figure 13, Figure 15, Figure 17; Obtain capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line, j ∈ [1, N]; When fault initial phase angle is respectively 30 °, 60 ° and 90 °, the zero mould initial current capable ripple Wavelet Modulus Maxima of each bar feeder line is as shown in table 1:

Table 1

(4) polarity of traveling wave calculating each bar feeder line estimates D (j), and described polarity of traveling wave estimates the defined formula of D (j) such as formula shown in (I):

$D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\&times;\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}sgn\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(I\right)$

In formula (I), D (j) represents that the polarity of traveling wave of jth bar feeder line is estimated, and namely represents the possibility of fault generation jth bar feeder line; The span of D (j) is [0,1]; M (j) represents the amplitude of the capable ripple Wavelet Modulus Maxima of jth bar feeder line zero mould initial current, M (j) > 0, i.e. M (j)=| m (j) |; m (l) represents the amplitude of the l article of capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, M (l) > 0, i.e. M (l)=| m (l) |, l ∈ [1, N] and l ≠ j; Function sgn [M (j)-M (l)] is used for the size of the amplitude difference reflecting certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1,0 or-1, as M (j) > M (l), sgn [M (j)-M (l)]=1; As M (j)=M (l), sgn [M (j)-M (l)]=0; As M (j) < M (l), sgn [M (j)-M (l)]=-1; When fault initial phase angle is respectively 30 °, 60 ° and 90 °, the polarity of traveling wave of each bar feeder line is estimated as shown in table 2:

Table 2

(5) polarity of traveling wave of each article of feeder line calculated according to step (4) is estimated D (j) and exports the result of low-current ground fault line selection: the 1st article of feeder line is faulty line; Other feeder line is for perfecting circuit.

The simulation results show validity of the method.

Formula (I) definition is estimated definition one and polarity of traveling wave by polarity of traveling wave and is estimated definition two and combine, and polarity of traveling wave is estimated definition one and used D ₁j () represents, polarity of traveling wave estimates definition dual-purpose D ₂j () represents; The derivation of formula (I) is:

According to the relation perfecting circuit and faulty line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, perfect circuit contrary with faulty line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, and it is all identical to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, constructs the polarity of traveling wave relevant with zero mould initial current capable ripple Wavelet Modulus Maxima and estimate definition one D ₁(j), that is:

$D_1\left(j\right)=\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}\mathrm{sgn}\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(VIII\right)$

In formula (VIII), sgn () is sign function, is defined as formula (Ⅸ):

$\mathrm{sgn}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(IX\right)$

From formula (Ⅸ), described sign function sgn [m (j) × m (l)], be used to the polar relationship reflecting certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1 or-1;

If m (j) × m (l) > 0, namely by after the polarity that compares two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, show that capable ripple Wavelet Modulus Maxima m (j) of jth bar feeder line zero mould initial current is identical with kth bar feeder line zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity, then sgn [m (j) × m (l)]=1;

If m (j) × m (l) < 0, namely by after the polarity that compares two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, show that capable ripple Wavelet Modulus Maxima m (j) of jth bar feeder line zero mould initial current is contrary with kth bar feeder line zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity, then sgn [m (j) × m (l)]=-1;

If m (j) × m (l)=0, namely the capable ripple Wavelet Modulus Maxima of zero mould initial current of jth article or l article of feeder line has one to be 0, then sgn [m (j) × m (l)]=0; For wavelet transformation, the capable ripple Wavelet Modulus Maxima of zero mould initial current be 0 situation do not exist; Therefore the value of sgn [m (j) × m (l)] can not be 0;

When jth bar feeder line is faulty line, capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current is contrary with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of other all circuit, l ∈ [1, N] and l ≠ j, for value from 1 successively to N l article of feeder line m (l) for, the cumulative sum of sgn [m (j) × m (l)] equals-(N-1), D ₁(j)=-1;

When jth bar feeder line is for perfecting circuit, to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity identical with other for capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current, for value from 1 successively to N l article of circuit m (l) for, the cumulative sum of sgn [m (j) × m (l)] is greater than-(N-1), D ₁(j) >-1;

Therefore, polarity of traveling wave estimates definition one D ₁j () meets the requirement of fault measurement universal, obviously distinguished fault feeder and the result perfecting feeder line.

Polarity of traveling wave estimates being constructed as follows of definition two:

In a practical situation, due to the reason such as measuring error, interference, make the amplitude of the zero mould initial current capable ripple Wavelet Modulus Maxima of each bar feeder line very little.From magnitude relation, the amplitude perfecting the zero mould initial current capable ripple Wavelet Modulus Maxima of feeder line is minimum, and close to 0, now, then carrying out computing with more respective polarity to the zero mould initial current capable ripple Wavelet Modulus Maxima of each bar feeder line, is nonsensical.

Polarity of traveling wave estimates definition two D ₂(j), also known as amplitude coefficient, that is:

$D_2\left(j\right)=\frac{M\left(j\right)}{Sum}---\left(X\right)$

The definition that the polarity of traveling wave of jth bar circuit estimates D (j) estimates definition one D by polarity of traveling wave ₁j () and polarity of traveling wave estimate definition two D ₂j () combines, for:

$D\left(j\right)=\frac{1}{2}-D_1\left(j\right)\&times;D_2\left(j\right)---\left(XI\right)$

That is: $D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\&times;\frac{\underset{l=1,l\overline{)=}j}{\overset{N}{\&Sigma;}}sgn\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(XII\right)$

For perfecting circuit, polarity of traveling wave estimates definition two D ₂j the span of () is 0 < D ₂(j) < 0.5, and D ₂j the numerical value of () is very little, close to 0, when disturbing larger, the polarity of traveling wave perfecting circuit estimates definition one D ₁j () may have error, add amplitude coefficient D ₂j (), substantially reduces the error that polarity of traveling wave estimates D (j).For faulty line, amplitude coefficient D ₂(j)=0.5, D (j)=1.Therefore, definition two D is estimated by polarity of traveling wave ₂j (), makes D ₁j defect that () is disturbed impact improves, and makes polarity of traveling wave estimate the antijamming capability of D (j) large.

Polarity of traveling wave is estimated D (j) and can be distinguished faulty line significantly and perfect circuit.Ideally, the polarity of traveling wave of faulty line estimates D (j)=1, and the polarity of traveling wave perfecting circuit estimates D (j)=0.

Claims (2)

1., based on the cable faults of distribution network monitoring method that polarity of traveling wave is estimated, it is characterized in that, concrete steps comprise:

(1) the three-phase initial current travelling wave signal of each bar feeder line is gathered: the initial current travelling wave signal being gathered three-phase by each bar feeder line of current transformer to power distribution network;

(2) obtain zero mould initial current travelling wave signal of each bar feeder line: by phase-model transformation, the initial current travelling wave signal of three-phase step (1) obtained changes zero mould initial current travelling wave signal into;

(3) capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line is obtained: wavelet transformation is carried out to zero mould initial current travelling wave signal of each bar feeder line, obtain capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of each bar feeder line, j ∈ [1, N];

(4) polarity of traveling wave calculating each bar feeder line estimates D (j), and described polarity of traveling wave estimates the defined formula of D (j) such as formula shown in (I):

$D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\&times;\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}sgn\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(I\right)$

In formula (I), D (j) represents that the polarity of traveling wave of jth bar feeder line is estimated, and namely represents the possibility of fault generation jth bar feeder line; The span of D (j) is [0,1]; M (j) represents the amplitude of the capable ripple Wavelet Modulus Maxima of jth bar feeder line zero mould initial current, M (j) > 0, i.e. M (j)=| m (j) |; m (l) represents the amplitude of the l article of capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, M (l) > 0, i.e. M (l)=| m (l) |, l ∈ [1, N] and l ≠ j; Function sgn [M (j)-M (l)] is used for the size of the amplitude difference reflecting certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1,0 or-1, as M (j) > M (l), sgn [M (j)-M (l)]=1; As M (j)=M (l), sgn [M (j)-M (l)]=0; As M (j) < M (l), sgn [M (j)-M (l)]=-1;

(5) polarity of traveling wave of each bar feeder line calculated according to step (4) is estimated D (j) and exports the result of low-current ground fault line selection: get { D (1), D (2) ... D (j) ... D (N) } in feeder line corresponding to maximal value be fault feeder.

2. the cable faults of distribution network monitoring method estimated based on polarity of traveling wave according to claim 1, it is characterized in that, formula (I) definition is estimated definition one and polarity of traveling wave by polarity of traveling wave and is estimated definition two and combine, and polarity of traveling wave is estimated definition one and used D ₁j () represents, polarity of traveling wave estimates definition dual-purpose D ₂j () represents; The derivation of formula (I) is:

According to the relation perfecting circuit and the capable ripple Wavelet Modulus Maxima of faulty line zero mould initial current, perfect circuit contrary with faulty line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, and it is all identical to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity, constructs the polarity of traveling wave relevant with zero mould initial current capable ripple Wavelet Modulus Maxima and estimate definition one D ₁(j), that is:

$D_1\left(j\right)=\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}\mathrm{sgn}\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(VIII\right)$

In formula (VIII), sgn () is sign function, is defined as formula (Ⅸ):

$\mathrm{sgn}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(IX\right)$

From formula (Ⅸ), described sign function sgn [m (j) × m (l)], be used for reflecting the polar relationship of certain two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, value is 1 or-1;

If m (j) × m (l) > 0, namely by after the polarity that compares two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, show that capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of jth bar feeder line is identical with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of kth bar feeder line, then sgn [m (j) × m (l)]=1;

If m (j) × m (l) < 0, namely by after the polarity that compares two capable ripple Wavelet Modulus Maxima of feeder line zero mould initial current, show that capable ripple Wavelet Modulus Maxima m (j) of zero mould initial current of jth bar feeder line is contrary with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of kth bar feeder line, then sgn [m (j) × m (l)]=-1;

If m (j) × m (l)=0, namely the capable ripple Wavelet Modulus Maxima of zero mould initial current of jth article or l article of feeder line has one to be 0, then sgn [m (j) × m (l)]=0; For wavelet transformation, the capable ripple Wavelet Modulus Maxima of zero mould initial current be 0 situation do not exist; Then judge that the situation of sgn [m (j) × m (l)]=0 does not exist;

When jth bar feeder line is faulty line, capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current is contrary with zero mould initial current capable ripple Wavelet Modulus Maxima m (l) polarity of other all feeder line, l ∈ [1, N] and l ≠ j, for value from 1 successively to N l article of feeder line m (l) for, the cumulative sum of sgn [m (j) × m (l)] equals-(N-1), D ₁(j)=-1;

When jth bar feeder line is for perfecting circuit, to perfect line zero mould initial current capable ripple Wavelet Modulus Maxima polarity identical with other for capable ripple Wavelet Modulus Maxima m (j) of its zero mould initial current, for value from 1 successively to N l article of circuit m (l) for, the cumulative sum of sgn [m (j) × m (l)] is greater than-(N-1), D ₁(j) >-1;

Polarity of traveling wave estimates being constructed as follows of definition two:

Polarity of traveling wave estimates definition two D ₂(j), also known as amplitude coefficient, that is:

$D_2\left(j\right)=\frac{M\left(j\right)}{Sum}---\left(X\right)$

The definition that the polarity of traveling wave of jth bar circuit estimates D (j) estimates definition one D by polarity of traveling wave ₁j () and polarity of traveling wave estimate definition two D ₂j () combines, for:

$D\left(j\right)=\frac{1}{2}-D_1\left(j\right)\&times;D_2\left(j\right)---\left(XI\right)$

That is: $D\left(j\right)=\frac{1}{2}-\frac{M\left(j\right)}{Sum}\&times;\frac{\underset{l=1,l\&NotEqual;j}{\overset{N}{\&Sigma;}}\mathrm{sgn}\&lsqb;m\left(j\right)\&times;m\left(l\right)\&rsqb;}{N-1}---\left(XII\right)$

Ideally, the polarity of traveling wave of fault feeder estimates D (j)=1, and the polarity of traveling wave perfecting feeder line estimates D (j)=0.