CN101247040B - Permanent fault discrimination method for electric power line with shunt reactor - Google Patents

Abstract

The present invention relates to the field of the electric power system electric power transmission line relay protection, especially to a technique of the single-phase self-adapting reclosing with a parallel reactor electric power transmission line, and the invention discloses a method for judging the permanent fault of the electric power transmission line with a parallel reactor. The method adopts the parameter recognizing principal and constructs an identification equation which uses the circular parallel reactor and neutral point small reactor a parameter to be determined basing on the network equation when the instantaneous fault has disappeared. The solving to the inductor parameter is realized by the sound phase voltage, sound phase current, parallel reactor current and neutral point small reactor at this side. According to the least square parameter estimate redundant algorithm the inductor parameter is solved, and the instantaneous fault and permanent fault are discriminated using the comparison between the solved inductor parameter and the actual inductor parameter.

Description

A kind of permanent failure judgement method for power transmission line with shunt reactor

Technical field

The present invention relates to power system transmission line relaying protection field, relate in particular to the single-phase adaptive reclosing technology of band shunt reactor transmission line, particularly a kind of permanent failure judgement method for power transmission line with shunt reactor.

Background technology

Raising along with the transmission voltage grade, the increase of phase insulation distance, the ultra-high/extra-high voltage transmission line is its major failure form with single-phase transient fault, but there is certain blindness in the action of current single-pole reclosing, no matter be transient fault or permanent fault, all be behind definite-time, to overlap automatically, cause coinciding with permanent fault thus and bring the secondary pulse also more serious than normal short circuit to electric power system, the safety of electric equipments and the safe and stable operation of electric power system in this serious threat.For fear of the blindness of reclosing action, just must accomplish to determine that transmission line is transient fault or permanent fault before the reclosing, the action that determines reclosing by the judged result of nature of trouble whether, i.e. adaptive reclose.

The key problem of adaptive reclose is distinguished transient fault and permanent fault exactly, and correlative study has obtained certain achievement, and its achievement in research mainly is divided into two big classes at present: 1) based on fault phase recovery voltage characteristic; 2) based on the transient fault arc characteristic.The former has utilized the capacitor coupling voltage characteristic to distinguish transient fault and permanent fault, and its sensitivity and anti-transition resistance ability are influenced by line length; The latter mainly utilizes fault transient high-frequency signal feature to distinguish transient fault and permanent fault, and its discrimination precision and electric arc blow-out process, high-frequency signal obtains precision and malfunction is closely related, and it is unreliable to differentiate, and is difficult to practicability.

For the transmission line of band shunt reactor, because the compensating action of shunt reactor, it is lower to disconnect the phase capacitor coupling voltage, is not easy accurately to obtain the fault phase voltage, causes can't effectively using based on the method for discrimination of fault phase recovery voltage characteristic.Method of discrimination based on the transient fault arc characteristic will utilize the fault phase voltage, the influence that its discrimination precision and reliability directly are subjected to fault phase voltage amount to obtain precision and reliability.

In addition, transmission line for the band shunt reactor, utilizing shunt reactor side current transformer accurately to obtain the disconnection phase current does not have difficulties, but the low-frequency oscillation component during transient fault is big, make the method for discrimination that utilizes the power current component be difficult to determine fast nature of trouble, differentiation is unfavorable for the application of high-speed reclosure for up to milliseconds up to a hundred (ms).

Therefore, at the fault characteristic of band shunt reactor transmission line, need the quick method of discrimination of single-phase fault character that the research principle is simple, be easy to realize.

Summary of the invention

The object of the present invention is to provide a kind of permanent failure judgement method for power transmission line with shunt reactor, it can overcome existing transmission line single-phase fault character method of discrimination and be applied to defective with the shunt reactor transmission line.

Principle of the present invention is based on parameter recognition thought.For linear Electric Network, the structure of network is depended in the response of network, component parameters and excitation.If the structure of known network and excitation can solve network element parameters R, L, C by its response, be network parameter identification.The present invention is by setting up the parameter recognition equation based on transient fault (disappearance back, fault point) network, be parameter to be asked with transmission line shunt reactor and the little reactance inductance parameters of neutral point, according to the transient fault of relatively distinguishing transmission line and the permanent fault of corresponding inductance parameters of finding the solution and actual parameter, thereby whether the decision transmission line needs reclosing.This method is not subjected to the influence of transition resistance, abort situation and load current substantially.

Based on above-mentioned parameter identification thought, technical scheme of the present invention is achieved in that a kind of permanent failure judgement method for power transmission line with shunt reactor, based on transmission line MN, and its a two ends band shunt reactor or an end band shunt reactor; Suppose N side band shunt reactor, this shunt reactor has the little reactance of neutral point, it is characterized in that, may further comprise the steps:

Step 1 is gathered transmission line N side three-phase voltage u _Na, u _Nb, u _Nc, three-phase current i _Al, i _Bl, i _Cl, and shunt reactor three-phase current i _Nax, i _Nbx, i _NcxWith the little reactor current i of neutral point _N0, the single-phase fault of determining transmission line mutually with perfect mutually;

Step 2 will perfect phase voltage, perfect phase current, parallel three phase reactor current and little reactor current of neutral point and system parameters, substitution parameter recognition equation A ₁L _N1+ A ₂L ₀₁=B, its corresponding discrete parameter recognition equation is A _1kL _N1+ A _2kL ₀₁=B _k, parameter to be asked is L _N1, L ₀₁, wherein: L _N1, L ₀₁Be respectively the little reactor inductance value of N side shunt reactor and neutral point thereof, A _1k, A _2kBe the coefficient value of k sampling instant correspondence, k is the counting in sampling period, k=1, and 2 ..., n, n are the data window length of the required sampled data of least-squares parameter estimation redundant arithmetic, make up matrix equation group AX=B, wherein

$A=\left[\begin{array}{cc}{A}_{11}& A_{21}\\ A_{12}& A_{22}\\ \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\\ A_{1n}& A_{2n}\end{array}\right],$ $X=\left[\begin{array}{c}{L}_{n1}\\ L_{01}\end{array}\right],$ $B=\left[\begin{array}{c}{B}_1\\ {\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">B</figure-callout>}}_2\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ B_n\end{array}\right]$

Can try to achieve X=[A _{2 * n}] ⁺[B], i.e. Dui Ying inductance parameters L _N1And L ₀₁, pass data window in turn, in like manner, can ask for corresponding inductance parameters sequence L _N1(i) and L ₀₁(i), i is the inductance parameters sequence number;

Step 3 is calculated corresponding inductance parameter error estimated value $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|,$ $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|,$ Wherein m is the inductance parameters sequence length, L _N1(i) and L ₀₁(i) inductance parameters sequence, L _N1rAnd L _01rBe respectively the little reactor inductance value of circuit N side shunt reactor and neutral point thereof;

Step 4 will $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|,$ $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|$ Respectively with setting value K ₁L _N1r, K ₂L _01rCompare, wherein K ₁And K ₂Be the nargin coefficient, if $\left\{\begin{array}{c}\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|<K_1{L}_{n1r}\\ \frac{1}{n}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|<K_2{L}_{01r}\end{array}\right.$ When the both set up, line fault was judged to be transient fault; Otherwise line fault is judged to be permanent fault.

Characteristics of the present invention also are:

The described data window of passing in turn, this data window passing time is 20～40ms, the length n of the required sampled data of described least-squares estimation redundant arithmetic is 40～80.

Described inductance parameters sequence length m value gets 10～20.

Described nargin COEFFICIENT K ₁Get 0.1～0.2, the nargin COEFFICIENT K ₂Get 0.1～0.2.

Compared with prior art, the present invention has outstanding advantage:

1, the transient fault model with the recovery voltage stage is that reference model is set up the parameter recognition network equation, has avoided owing to the uncertain problem that is difficult to accurately set up the permanent fault model in fault point; And only keep the little reactor inductance parameter of shunt reactor and neutral point as identification parameter, reduced the order of finding the solution of parameter recognition equation, improved the precision of parameter identification.

2, only utilize single-end electrical quantity, the electric parameters of offside is obtained indirectly according to known quantity, not influenced by passage; And do not utilize to disconnect the phase voltage amount, it is too small and cause the error problem measured to have avoided disconnecting the phase voltage amplitude.

3, adopt the parameter recognition principle, be not subjected to direct current, the humorous influence that involves the low-frequency oscillation component basically, do not have filtering low frequency free component.Therefore, with respect to the method for discrimination based on the power frequency amount, this method does not need the long data window filtering, can determine fast to help the realization of high-speed reclosure by nature of trouble.

4, this method of discrimination has good applicability, is applicable to the differentiation of an end, the single-phase fault character of two ends band shunt reactor circuit under diverse location, different transition resistance situation.

Description of drawings

Fig. 1 is the protective device hardware configuration schematic diagram of part of path MN.

Fig. 2 is the π type equivalent circuit diagram behind the band shunt reactor circuit transient fault arc extinction of two ends.

Fig. 3 is the π type equivalent circuit diagram behind the end band shunt reactor circuit transient fault arc extinction.

Fig. 4 is that single-phase fault character of the present invention is differentiated flow chart.

Fig. 5 is an end band shunt reactor transmission line figure, and line parameter circuit value is: R ₁=0.027 Ω/km, R ₀=0.1957 Ω/km, L ₁=0.9651mH/km, L ₀=2.2110mH/km, C ₁=0.0136 μ F/km, C ₀=0.0092 μ F/km.The shunt reactor inductance L _N1=6.4203H, the little reactor inductance L of neutral point ₀₁=1.7507H.

Fig. 6 is both-end band shunt reactor transmission line figure, and line parameter circuit value is: R ₁=0.0195 Ω/km, R ₀=0.1675 Ω/km, L ₁=0.9134mH/km, L ₀=2.719mH/km, C ₁=0.014 μ F/km, C ₀=0.00834 μ F/km.The shunt reactor inductance L _N1=5.3494H, the little reactor inductance L of neutral point ₀₁=1.3815H.

Embodiment

With reference to Fig. 1, MN two ends, protective wire highway section have shunt reactor, and for realizing protective device hardware block diagram of the present invention, it is made of data acquisition system, microcomputer main system (DSP), output system three parts in the N side frame of broken lines.The data acquisition system of N side protective device; gather transmission line A; B; C three-phase voltage amount; the magnitude of current; gather shunt reactor A; B; the magnitude of current of C three-phase electricity flow and the little reactor of neutral point thereof; wherein transmission line voltage is obtained by line side voltage transformer TV; electric current is obtained by line side current transformer TA1; the shunt reactor three-phase current is obtained by current transformer TA2; the electric current of the little reactor of neutral point is obtained by current transformer TA3; all measuring-signals pass through low pass filter; sampling/retainer; A/D converter is imported microcomputer main system (DSP) then.Microcomputer main system (DSP) goes out inductance parameters to be identified by corresponding algorithm computation, judges that according to inductance parameters result of calculation the single-phase fault that circuit takes place is transient fault or permanent fault then, produces judged result.This judged result is admitted to the photoisolator of output system, isolates the rear drive relay through photoelectricity and produces the forceful electric power output signal, and the forceful electric power output signal is finished the reclosing action of respective circuit breakers through outlet.This microcomputer main system (DSP) also has the human-computer dialogue function, can by the user realize easily setting value setting, monitoring running state, discriminating element throw move back, upper machine communication and printing function.

Below in conjunction with accompanying drawing parameter recognition principle of the present invention is elaborated.

When considering circuit generation permanent fault, because the uncertainty of fault point, set up accurately that equivalent model has difficulties, the transient fault model after the present invention disappears with the fault point is that identification model is analyzed.

Circuit with two ends band shunt reactor is that example describes earlier.Suppose part of path MN, this side (N side) is provided with protective device, and A is the fault phase mutually, and the transient fault л type equivalent electric circuit behind the arc extinction of fault point as shown in Figure 2.Wherein: C ₀Be line mutual-ground capacitor, C _mBe the alternate electric capacity of circuit, L _N1, L ₀₁Be respectively the little reactor inductance value of N side shunt reactor and neutral point thereof, L _N2, L ₀₂Be respectively the little reactor inductance value of M side shunt reactor and neutral point, Z _sBe circuit self-impedance, i _Bl, i _ClFor perfecting phase B, C, circuit N side protects measuring point place electric current, i mutually _b, i _cBe the electric current that flows through in perfecting mutually, i _Nax, i _Nbx, i _NcxBe respectively N side A, B, the C reactor current that is in parallel, i _Max, i _Mbx, i _McxBe respectively M side A, B, the C reactor current that is in parallel, i _N0Be the little reactive current of neutral point of N side shunt reactor, i _M0Be the little reactive current of neutral point of M side shunt reactor, i ₁, i ₂Being respectively the N side perfects and fault three-phase current mutually, i ₃, i ₄Being respectively the M side perfects and fault three-phase current mutually, i ₅, i ₆Be respectively fault phase corresponding N side, M side to earth-current, u _Ma, u _Mb, u _Mc, u _Na, u _Nb, u _NcBe respectively circuit M side, N side three phase terminals voltage, u _M0, u _N0Be respectively circuit M side, N side shunt reactor neutral point voltage.

Can obtain following relation by Fig. 2

i ₁+i ₂+i ₃+i ₄-i ₅-i ₆-i _nax-i _max＝0 (1)

In the formula: ${\mathrm{<figure-callout\; id="1"\; label="i"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">i</figure-callout>}}_1=\frac{C_m}{2}\frac{d(u_{\mathrm{nb}}-u_{\mathrm{na}})}{\mathrm{dt}},$ ${\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">i</figure-callout>}}_2=\frac{C_m}{2}\frac{d(u_{\mathrm{nc}}-u_{\mathrm{na}})}{\mathrm{dt}},$ $i_3=\frac{C_m}{2}\frac{d(u_{\mathrm{mb}}-u_{\mathrm{ma}})}{\mathrm{dt}},$ $i_4=\frac{C_m}{2}\frac{d(u_{\mathrm{mc}}-u_{\mathrm{ma}})}{\mathrm{dt}},$

$i_5=\frac{C_0}{2}\frac{{\mathrm{du}}_{\mathrm{na}}}{\mathrm{dt}},$ $i_6=\frac{C_0}{2}\frac{{\mathrm{du}}_{\mathrm{ma}}}{\mathrm{dt}},$ $i_{\max }=\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n}\&Integral;(u_{\mathrm{ma}}-u_{m0})\mathrm{dt}.$

This side (N side) electric parameters i _Nax, i _Nbx, i _Ncx, i _N0, i _Bl, i _Cl, u _Nb, u _NcCan directly measure by protective device and obtain u _NaCan be by calculating: $u_{\mathrm{na}}={\mathrm{<figure-callout\; id="1"\; label="L\; n"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{\mathrm{nax}}}{\mathrm{dt}}+L_{01}\frac{{\mathrm{di}}_{n0}}{\mathrm{dt}}---\left(2\right)$

In addition, the electric parameters relation according to the M reactor can get:

$\left\{\begin{array}{c}{i}_{m0}=i_{\max }+i_{\mathrm{mbx}}+i_{\mathrm{mcx}}\\ u_{\mathrm{ma}}-u_{m0}={\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{\max }}{\mathrm{dt}}\\ u_{\mathrm{mb}}-u_{m0}={\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{\mathrm{mbx}}}{\mathrm{dt}}\\ u_{\mathrm{mc}}-u_{m0}={\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{\mathrm{mcx}}}{\mathrm{dt}}\\ u_{m0}=L_{02}\frac{{\mathrm{di}}_{m0}}{\mathrm{dt}}\end{array}\right.---\left(3\right)$

Can push away by formula (3): u _M0=λ (u _Ma+ u _Mb+ u _Mc) (4)

In the formula: λ=L ₀₂/ (L _N2+ 3L ₀₂)

In the formula (4), electric parameters u _Ma, u _Mb, u _McCan pass through this side (N side) known electric tolerance i _Nax, i _Nbx, i _Ncx, i _Bl, i _Cl, u _Nb, u _NcObtain by finding the solution following equation group indirect calculation:

$u_{\mathrm{ma}}=u_{\mathrm{na}}+R_m(i_b+i_c)+L_m\frac{d(i_b+i_c)}{\mathrm{dt}}$

$u_{\mathrm{mb}}+u_{\mathrm{mc}}=u_{\mathrm{nb}}+u_{\mathrm{nc}}+(R_s+R_m)(i_b+i_c)+(L_s+L_m)\frac{d(i_b+i_c)}{\mathrm{dt}}$

$i_b+i_c=i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}}+\frac{C_0+C_m}{2}\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}}---\left(5\right)$

In the formula (5), R _sBe circuit self-resistance, R _mBe circuit mutual resistance, L _sBe circuit self-inductance, L _mBe the circuit mutual inductance.Asking for u _Mb, u _Mc, i _b, i _cThe time, consider that the electric current and voltage of fault phase is sound relatively very little mutually, therefore can ignore fault phase voltage electric current to perfecting the influence of phase.

Formula (1) is organized into shown in the following equation (6):

p ₁+p ₂-G＝0 (6)

Wherein:

${\mathrm{<figure-callout\; id="1"\; label="p"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">p</figure-callout>}}_1=-[(1-\mathrm{\&lambda;})/{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n.i_{\mathrm{nax}}+({2C}_m+C_0)\frac{d^2{i}_{\mathrm{nax}}}{{\mathrm{dt}}^2}]{\mathrm{<figure-callout\; id="1"\; label="L\; n"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">L</figure-callout>}}_n,$ ${\mathrm{<figure-callout\; id="2"\; label="p"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">p</figure-callout>}}_2=-[(1-\mathrm{\&lambda;})/{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n.i_{n0}+({2C}_m+C_0)\frac{d^2{i}_{n0}}{{\mathrm{dt}}^2}]L_{01},$

${\mathrm{<figure-callout\; id="1"\; label="g"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">g</figure-callout>}}_1=-\frac{(1-2\mathrm{\&lambda;})R_m-\mathrm{\&lambda;}{R}_{\mathrm{<figure-callout\; id="2"\; label="s\; L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">s</figure-callout>}}}{L_n}\&Integral;(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})\mathrm{dt},$ ${\mathrm{<figure-callout\; id="2"\; label="g"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">g</figure-callout>}}_2=-\frac{(1-2\mathrm{\&lambda;})R_m-\mathrm{\&lambda;}{R}_s}{L_{n2}}\frac{C_0+C_m}{2}(u_{\mathrm{nb}}+u_{\mathrm{nc}}),$

$g_3=-\frac{(1-2\mathrm{\&lambda;})L_m-\mathrm{\&lambda;}{L}_s}{L_{n2}}(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}}),$ $g_4=-\frac{(1-2\mathrm{\&lambda;})L_m-\mathrm{\&lambda;}{L}_s}{L_{n2}}\frac{C_0+C_m}{2}\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}},$

$g_5=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{d(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{\mathrm{dt}},$ $g_6=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^2(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$

$g_7=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{d^2(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$ $g_8=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^3(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{dt}}^3}$

$G=i_{\mathrm{nax}}-\mathrm{\&lambda;}/{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\&Integral;(u_{\mathrm{nb}}+u_{\mathrm{nc}})\mathrm{dt}-C_m\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}}-\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{g}_i$

Formula (6) is about model parameter C _n, C ₀, L _N1, L ₀₁, L _N2, L ₀₂, R _s, R _m, L _s, L _mFunction can be expressed as f (C _m, C ₀, L _N1, L ₀₁, L _N2, L ₀₂, R _s, R _m, L _s, L _m)=0.Can obtain above-mentioned model parameter by formula (6), but formula (6) is a nonlinear equation, least-squares algorithm is found the solution all parameters and is realized comparatively difficulty, therefore, only keeps L _N1, L ₀₁As model parameter to be asked, all the other parameters C _m, C ₀, L _N2, L ₀₂, R _s, R _m, L _s, L _mUtilize known system parameters.

Can obtain the parameter recognition equation (7) of following form, for fear of integral operation, formula (7) is the abbreviation arrangement result of formula (6) behind first derivation:

A ₁L _n1+A ₂L ₀₁＝B (7)

Wherein:

$A_1=-(1-\mathrm{\&lambda;})/{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{\mathrm{nax}}}{\mathrm{dt}}-({2C}_m+C_0)\frac{d^3{i}_{\mathrm{nax}}}{{\mathrm{dt}}^3},$ $A_2=-(1-\mathrm{\&lambda;})/{\mathrm{<figure-callout\; id="2"\; label="L\; n"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">L</figure-callout>}}_n\frac{{\mathrm{di}}_{n0}}{\mathrm{dt}}-({2C}_m+C_0)\frac{d^3{i}_{n0}}{{\mathrm{dt}}^3},$

${\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">B</figure-callout>}}_1=-\frac{(1-2\mathrm{\&lambda;}{)R}_m-\mathrm{\&lambda;}{R}_s}{L_{n2}}(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}}),$ ${\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">B</figure-callout>}}_2=-\frac{(1-2\mathrm{\&lambda;})R_m-\mathrm{\&lambda;}{R}_s}{L_{n2}}\frac{C_0+C_m}{2}\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}},$

$B_3=-\frac{(1-2\mathrm{\&lambda;})L_m-\mathrm{\&lambda;}{L}_s}{L_{n2}}\frac{d(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{\mathrm{dt}},$ $B_4=-\frac{(1-2\mathrm{\&lambda;})L_m-\mathrm{\&lambda;}{L}_s}{L_{n2}}\frac{C_0+C_m}{2}\frac{d^2(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$

$B_5=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{d^2(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$ $B_6=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^3(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{dt}}^3},$

$B_7=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{d^3(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{{\mathrm{dt}}^3},$ $B_8=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^4(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{dt}}^4}$

$B=\frac{{\mathrm{di}}_{\mathrm{nax}}}{\mathrm{dt}}-\mathrm{\&lambda;}/L_{n2}(u_{\mathrm{nb}}+u_{\mathrm{nc}})-C_m\frac{d^2(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{dt}}^2}-\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{B}_i$

I in the equation (7) _Nax, i _Nbx, i _Ncx, i _Hl, i _Cl, u _Nb, u _NcInstantaneous value at each moment k can be obtained by the protective device of N side, and the differential of corresponding electric parameters can be realized by difference equation.

For the transmission line of an end band shunt reactor, suppose this side (N side) band shunt reactor, offside (M side) is not with shunt reactor, A be mutually fault phase (behind the arc extinction of fault point) transient fault л type equivalent model as shown in Figure 3.

Can obtain following relation by Fig. 3

i ₁+i ₂+i ₃+i ₄-i ₅-i ₆-i _nax＝0 (8)

In the formula: $i_1=\frac{C_m}{2}\frac{d(u_{\mathrm{nb}}-u_{\mathrm{na}})}{\mathrm{dt}},$ ${\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">i</figure-callout>}}_2=\frac{C_m}{2}\frac{d(u_{\mathrm{nc}}-u_{\mathrm{na}})}{\mathrm{dt}},$ $i_3=\frac{C_m}{2}\frac{d(u_{\mathrm{mb}}-u_{\mathrm{ma}})}{\mathrm{dt}},$ $i_4=\frac{C_m}{2}\frac{d(u_{\mathrm{mc}}-u_{\mathrm{ma}})}{\mathrm{dt}}.$

$i_5=\frac{C_0}{2}\frac{{\mathrm{du}}_{\mathrm{na}}}{\mathrm{dt}},$ $i_6=\frac{C_0}{2}\frac{{\mathrm{du}}_{\mathrm{ma}}}{\mathrm{dt}}.$

Formula (8) put in order:

$i_{\mathrm{nax}}=\frac{C_m}{2}[\frac{d(u_{\mathrm{mb}}+u_{\mathrm{mc}})}{\mathrm{dt}}+\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}}]-(C_m+\frac{C_0}{2})[\frac{{\mathrm{du}}_{\mathrm{ma}}}{\mathrm{dt}}+\frac{{\mathrm{du}}_{\mathrm{na}}}{\mathrm{dt}}]---\left(9\right)$

In the formula (9), remove u _Nb, u _NcOutside known, electric parameters u _Mb, u _McCan be by N side electric parameters by formula (5) indirect calculation, u _NaCan hold electric parameters by formula (2) indirect calculation by N.

With two ends band shunt reactor circuit analysis in like manner, formula (9) is put in order shape suc as formula the parameter recognition equation A of (7) ₁L _N1+ A ₂L ₀₁=B, its coefficient and two ends band shunt reactor circuit are slightly different, and each coefficient is followed successively by:

$A_1=-({2C}_m+C_0)\frac{d^2{\mathrm{<figure-callout\; id="2"\; label="i\; nax\; dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">i</figure-callout>}}_{\mathrm{nax}}}{{\mathrm{dt}}^{}},A_2=-({2C}_m+C_0)\frac{d^2{i}_{n0}}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$ ${\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="S2008100175744E00031.png"\; state="\{\{state\}\}">B</figure-callout>}}_1=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{d(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{\mathrm{dt}},$

${\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">B</figure-callout>}}_2=\frac{R_s{C}_m-R_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^2(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2},$ $B_3=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{d^2(i_{\mathrm{bl}}+i_{\mathrm{cl}}+i_{\mathrm{nbx}}+i_{\mathrm{ncx}})}{{\mathrm{<figure-callout\; id="2"\; label="dt"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">dt</figure-callout>}}^2}$

$B_4=\frac{L_s{C}_m-L_m(C_0+C_m)}{2}\frac{C_0+C_m}{2}\frac{d^3(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{{\mathrm{dt}}^3},$ $B=i_{\mathrm{nax}}-C_m\frac{d(u_{\mathrm{nb}}+u_{\mathrm{nc}})}{\mathrm{dt}}-B_1-B_2-B_3-B_4.$

I wherein _Nax, i _Nbx, i _Ncx, i _Bl, i _Cl, u _Nb, u _NcInstantaneous value at each moment k can be obtained by the protective device of N side, and the differential of corresponding electric parameters can be realized by difference equation.

In theory, for the parameter recognition equation of shape suc as formula (7), two independent equations just can be found the solution inductance parameters L _N1, L ₀₁, but because the influence of the sum of errors sampled data error that computation model is simplified, and there be the low-frequency oscillation component of amplitude during transient fault near the power frequency amount, cause the parametric solution the possibility of result big ups and downs to occur.In order to improve the reliability of fault distinguishing, the present invention adopts the data redundancy mode to carry out the least square method parameter Estimation, promptly draws an estimates of parameters with a period of time sampled value, passes data window then in turn, asks for the sequence of identified parameters.In a data window, according to the discrete form of parameter recognition equation

A _1kL _n1+A _2kL ₀₁＝B _k(k＝1，2，...，n) (10)

Wherein: A _1k, A _2kIt is the coefficient value of k sampling instant correspondence.

Can get: $A=\left[\begin{array}{cc}{A}_{11}& A_{21}\\ A_{12}& A_{22}\\ \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\\ A_{1n}& A_{2n}\end{array}\right],$ $X=\left[\begin{array}{c}{L}_{n1}\\ L_{01}\end{array}\right],$ $B=\left[\begin{array}{c}{B}_1\\ {\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="S2008100175744E00021.png"\; state="\{\{state\}\}">B</figure-callout>}}_2\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ B_n\end{array}\right]$

Formula (10) can be expressed as AX=B, can try to achieve X=[A _{2 * n}] ⁺[B] can obtain corresponding inductance parameters L _N1And L ₀₁Pass data window in turn, in like manner, can ask for corresponding inductance parameters sequence L _N1(i) and L ₀₁(i), i is the inductance parameters sequence number.

In theory, under the transient fault situation, the true fault model is consistent with identification model, and the inductance parameters of finding the solution is consistent with true inductance parameters, should not change in time and changes; And under the permanent fault situation, true fault model and identification model inconsistent, inductance parameters of finding the solution and true inductance obvious difference.Therefore, can realize the differentiation of permanent fault according to shunt reactor and the little reactor inductance parameter recognition of neutral point result.Corresponding criterion is as follows:

$\left\{\begin{array}{c}\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|<K_1{L}_{n1r}\\ \frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|<K_2{L}_{01r}\end{array}\right.---\left(11\right)$

In the formula (11), m is the result of calculation sequence length, L _N1(i) and L ₀₁(i) least-squares calculation inductance sequence, L _N1rAnd L _01rBe the actual value of this side of circuit (N side) parallel reactance, the little reactance of neutral point, K ₁And K ₂Be the nargin coefficient, according to fault simulation analysis result under a large amount of different situations, nargin COEFFICIENT K ₁And K ₂Get 0.1～0.2 and can satisfy nature of trouble differentiation requirement.When formula (11) is set up, can be judged to transient fault, protective device sends the reclosing order; Otherwise, just being judged to permanent fault, reclosing is failure to actuate.

The present invention is based on the parameter recognition principle, sample rate is too low can to increase the error of calculation, and sample rate can meet the demands in 2000Hz～10000Hz scope, finishes primary parameter and estimates the desirable 20～40ms of desired data window.

With reference to Fig. 4, decision method of the present invention is at first gathered transmission line three-phase voltage u _Na, u _Nb, u _Nc, three-phase current i _Al, i _Bl, i _Cl, and shunt reactor three-phase current i _Nax, i _Nbx, i _NcxWith the little reactor current i of neutral point _N0, the single-phase fault of determining transmission line mutually with perfect mutually, and coupling system parameter is asked for parameter recognition equation A ₁L _N1+ A ₂L ₀₁The coefficient A of=B ₁, A ₂, B, setting up corresponding discrete parameter recognition equation is A _1kL _N1+ A _2kL ₀₁=B _k(k=1,2 ..., n), utilize the least-squares estimation redundant arithmetic to calculate corresponding inductance parameters sequence L _N1(i) and L ₀₁(i), i is the inductance parameters sequence number; With corresponding inductance estimates of parameters $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|,$ $\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|,$ Respectively with setting value K ₁L _N1r, K ₂L _01rCompare, if $\left\{\begin{array}{c}\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|<K_1{L}_{n1r}\\ \frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|<K_2{L}_{01r}\end{array}\right.$ When the both set up, line fault was judged to be single-phase transient fault, and protective device sends the reclosing order; Otherwise line fault is judged to be permanent fault, and reclosing is failure to actuate.

Utilize ATP (electromagnetic transient simulation software) that the Typical Route model is carried out emulation under the different faults situation respectively below, the present invention is done further confirmation.The Typical Route model is referring to Fig. 5, Fig. 6.

Suppose that Typical Route model two ends power supply phase angle difference θ is 45 °, earth fault takes place in A mutually, and circuit breaker is in the 130ms tripping, back 300ms takes place and begins to calculate in fault, sample frequency is 2000HZ, and the used data window length of least square method is 20ms, promptly finds the solution inductance parameters L _N1And L ₀₁Utilize 40 continuous sampling point data.In the corresponding criterion

$\left\{\begin{array}{c}\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|<K_1{L}_{n1r}\\ \frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|<K_2{L}_{01r}\end{array}\right.$

Its left part is respectively as two operating values, and the right side part is as two setting values, the nargin COEFFICIENT K ₁And K ₂All get 0.1, calculate inductance sequence length m and get 20.

Table 1, table 2 have provided the simulation analysis result under an end, the two ends shunt reactor circuit different faults situation respectively, wherein abort situation represents that apart from the M end transition resistance is 0 ohm, 300 ohm two kinds of situations with the fault point with respect to the percentage of total track length distance.

Table 1. an end band shunt reactor circuit emulation result of calculation (θ=45 °)

Table 2. two ends band shunt reactor circuit emulation result of calculation (θ=45 °)

By table 1, table 2 as can be seen: no matter be an end band parallel reactance circuit, or two ends band shunt reactor circuit, at the transient fault of circuit diverse location under different transition resistance situations, all less by two operating values that formula (11) is calculated, and all less than the setting value of correspondence, show and calculate inductance parameters, be judged to transient fault very near true inductance parameters; And when permanent fault took place, two operating values of calculating were big and have at least an operating value can't satisfy criterion, show that true fault model and reference model have than big-difference, reliably are judged to permanent fault.And, it can also be seen that transition resistance does not influence the correctness of judged result.

Above-mentioned two example simulation results show, this method has good applicability, be applicable to the permanent failure judgement of an end, two ends band shunt reactor circuit single-phase fault under diverse location, different transition resistance situation, the success rate of transmission line single phase autoreclosing action can be improved, the requirement of band shunt reactor ultra-high/extra-high voltage transmission line system operation high reliability can be satisfied.

Claims (4)

1. permanent failure judgement method for power transmission line with shunt reactor, based on transmission line MN, its a two ends band shunt reactor or an end band shunt reactor; Suppose N side band shunt reactor, this shunt reactor has the little reactance of neutral point, it is characterized in that, may further comprise the steps:

Step 1 is gathered transmission line N side three-phase voltage u _Na, u _Nb, u _Nc, three-phase current i _Al, i _Bl, i _Cl, and shunt reactor three-phase current i _Nax, i _Nbx, i _NcxWith the little reactor current i of neutral point _N0, the single-phase fault of determining transmission line mutually with perfect mutually;

Step 2 will perfect phase voltage, perfect phase current, parallel three phase reactor current and neutral point reactor current and system parameters, substitution parameter recognition equation A ₁L _N1+ A ₂L ₀₁=B, its corresponding discrete parameter recognition equation is A _1kL _N1+ A _2kL ₀₁=B _k, parameter to be asked is L _N1, L ₀₁, wherein: L _N1, L ₀₁Be respectively the little reactor inductance value of N side shunt reactor and neutral point thereof, A _1k, A _2kBe the coefficient value of k sampling instant correspondence, k is the counting in sampling period, k=1, and 2 ..., n, n are the data window length of the required sampled data of least-squares parameter estimation redundant arithmetic, make up matrix equation group AX=B, wherein

$A=\left[\begin{array}{cc}{A}_{11}& A_{21}\\ A_{12}& A_{22}\\ .& .\\ .& .\\ .& .\\ A_{1n}& A_{2n}\end{array}\right],$ $X=\left[\begin{array}{c}{L}_{n1}\\ L_{01}\end{array}\right],$ $B=\left[\begin{array}{c}{B}_1\\ B_2\\ .\\ .\\ .\\ B_n\end{array}\right]$

Can try to achieve X=[A _{2 * n}] ⁺[B], i.e. Dui Ying inductance parameters L _N1And L ₀₁, pass data window in turn, in like manner, can ask for corresponding inductance parameters sequence L _N1(i) and L ₀₁(i), i is the inductance parameters sequence number;

Step 3 is calculated corresponding inductance parameter error estimated value

Wherein m is the inductance parameters sequence length, L _N1(i) and L ₀₁(i) inductance parameters sequence, L _N1rAnd L _01rBe respectively the little reactor inductance value of circuit N side shunt reactor and neutral point thereof;

Step 4 will

Respectively with setting value K ₁L _N1r, K ₂L _01rCompare, wherein K ₁And K ₂Be the nargin coefficient, if $\left\{\begin{array}{c}\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{n1}\left(i\right)-L_{n1r}|<K_1{L}_{n1r}\\ \frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}|L_{01}\left(i\right)-L_{01r}|<K_2{L}_{01r}\end{array}\right.$ When the both set up, line fault was judged to be transient fault; Otherwise line fault is judged to be permanent fault.

2. a kind of permanent failure judgement method for power transmission line according to claim 1 with shunt reactor, it is characterized in that, the described data window of passing in turn, this data window passing time is 20～40ms, the length n of the required sampled data of described least-squares estimation redundant arithmetic is 40～80.

3. a kind of permanent failure judgement method for power transmission line with shunt reactor according to claim 1 is characterized in that, described inductance parameters sequence length m value gets 10～20.

4. a kind of permanent failure judgement method for power transmission line with shunt reactor according to claim 1 is characterized in that, described nargin COEFFICIENT K ₁Get 0.1～0.2, the nargin COEFFICIENT K ₂Get 0.1～0.2.

Images