CN105552938A - Voltage sag evaluation method for three-phase asymmetrical power distribution network - Google Patents

Abstract

The invention relates to a voltage sag evaluation method for a three-phase asymmetrical power distribution network. The method comprises the following steps of: establishing a fault random model of the three-phase asymmetrical power distribution network; establishing a DG (Distributed Generation) random model of the three-phase asymmetrical power distribution network; carrying out orthogonal transformation on the fault random model, and acquiring an uncorrelated fault random model corresponding to the fault random model; processing a random variable consisting of the uncorrelated fault random model and the DG random model by adopting a point estimation method, and obtaining random variables of two sets of simulation schemes; carrying out reverse decomposition on the random variables of two sets of simulation schemes, and obtaining random variables of two sets of related simulation schemes, which correspond to the random variables of the two sets of simulation schemes; by a Cornish-Fisher series, respectively determining probability density functions of elements in the random variables of two sets of related simulation schemes; and comprehensively considering correlation of a fault type and a fault line in the power distribution network and various short circuit faults in the power distribution network to simulate characteristics of the integral power distribution network so as to provide reference for adopting measures of inhibiting a voltage sag.

Description

A kind of asymmetrical three-phase distribution network voltage dip appraisal procedure

Technical field

The present invention relates to electric power quality technical field, be specifically related to a kind of asymmetrical three-phase distribution network voltage dip appraisal procedure.

Background technology

Voltage dip, also known as Voltage Drop, refers to that the power-frequency voltage effective value of certain point in system drops to suddenly the 10%-90% of rated value, and recovers normal after the of short duration duration of 10ms-1min subsequently.Along with the technology innovation based on the universal of high-power electric and electronic switchgear and power consumption equipment, especially Numeric Control Technology is widely used in industrial production, and the harm that voltage dip produces is more and more obvious.In power distribution network, the most serious Problem of Voltage Temporary-Drop is caused by short trouble mostly.The impact of voltage dip on sensitive equipment (such as adjustable speed motor, precise hard_drawn tuhes equipment etc.) even can be mentioned in the same breath with power outage.What voltage dip caused industrial cousumer has a strong impact on and endangers the especially outstanding of performance, also may cause casualties and device damage.

The same with other power quality problem, Problem of Voltage Temporary-Drop is not a new problem, and existing a lot of technical data is studied it.By learning the inquiry of prior art data, voltage dip assessment is an importance of power quality analysis.

Along with the development of science and technology, new method is constantly had to be incorporated in quality of voltage analysis.At present, point estimations existing application in quality of power supply field.Technology 1 (Wu Bei, Zhang Yan, Chen Minjiang. the application of point estimations in Voltage stability analysis [J]. Proceedings of the CSEE, 2008,28 (25): 38-43.) point estimations is incorporated in Voltage stability analysis, randomness for branch trouble carries out Voltage Stability Analysis, can unify the uncertain problem of process circuit fault and node injecting power.Technology 2 (Xu Peidong; Xiao Xianyong; Wang Ying. voltage dip frequency two-point estimate stochastic appraisal method [J]. protecting electrical power system and control; 2011; 39 (9): 1-6.) randomness of sensitive equipment withstand voltage and system busbar normal operating voltage is considered, research voltage dip frequency problem.Technology 1 and technology 2 when considering computational accuracy and computing time, can both obtain gratifying result.

The method of present analysis voltage dip mainly contains critical distance method (technology 3:M.N.Moschakis, N.D.Hatziargyriou.Analyticalcalculationandstochasticasse ssmentofvoltagesags [J] .IEEETransactionsonPowerDelivery.2006,21 (3): 1727-1734. technology 4:M.H.J.Bollen.Fastassessmentmethodsforvoltagesagsindist ributionsystems [J] .IEEETransactionsonIndustryApplications, 1996,32 (6): 1414-1423. technology 5:PadmanabhThaku, AsheeshK.Singh, RameshC.Bansal.Ananalyticalapproachforstochasticassessme ntofbalancedandunbalancedvoltagesagsinlargesystems [J] .IEEETransactionsonPowerDelivery, 2006,21 (3): 1493-1500.), fault position method and monte carlo method (MonteCarloSimulation, MCS).Critical distance method is a kind of more classical method being used for assessing the voltage dip that symmetric fault and unbalanced fault cause.Fault position method only utilizes the specific fault of several Chosen Point to emulate the characteristic of whole electric power system.Monte carlo method is a kind of method being widely used in analyzing stochastic problem.The sampling number of monte carlo method and the scale of system have nothing to do, and the complexity of system is little on its impact, but monte carlo method has shortcomings such as nature static, computational efficiency are low, length consuming time.Along with the development of science and technology, new method is constantly had to be incorporated in Voltage Sag Analysis, such as two-point method (technology 6: a kind of active power distribution network Voltage Drop emulation and appraisal procedure, China: CN201410406537.8).Stochastic problem is converted to certain problem by two-point method, achieves the assessment to voltage dip and emulation.But, use above various method carry out voltage dip assessment time be all using power distribution network as three-phase symmetrical network processes, do not consider the asymmetrical three-phase problem of power distribution network, but asymmetrical three-phase phenomenon ubiquity in distribution network.It is less that existing documents and materials fall Study on Problems temporarily to the distribution network voltage of asymmetrical three-phase.

Summary of the invention

The invention provides a kind of asymmetrical three-phase distribution network voltage dip appraisal procedure, its objective is that in the correlation and power distribution network considering fault type and faulty line in power distribution network, various short trouble is to emulate the characteristic of whole power distribution network, for taking to suppress the measure of voltage dip to provide reference, and improve the power supply reliability of power distribution network.

The object of the invention is to adopt following technical proposals to realize:

A kind of asymmetrical three-phase distribution network voltage dip appraisal procedure, its improvements are, comprising:

(1) the fault stochastic model of described asymmetrical three-phase distribution network is set up;

(2) the DG stochastic model of described asymmetrical three-phase distribution network is set up;

(3) orthogonal transform is carried out to described fault stochastic model, obtain the uncorrelated fault stochastic model that described fault stochastic model is corresponding;

(4) adopt the stochastic variable of point estimations to described uncorrelated fault stochastic model and DG stochastic model composition to process, obtain the stochastic variable of two groups of simulating schemes;

(5) adopt that the stochastic variable of Choleskydecomposition mode to described two groups of simulating schemes is counter decomposes, obtain the stochastic variable of two groups of simulating schemes of being correlated with corresponding to the stochastic variable of described two groups of simulating schemes;

(6) probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is determined respectively by Cornish-Fisher progression.

Preferably, in described step (1), utilize Monte Carlo simulation mode to set up the fault stochastic model of described asymmetrical three-phase distribution network, comprising:

(1-1) the faulty line model of described asymmetrical three-phase distribution network is set up by following formula:

$F_{Line}=\left\{\begin{array}{c}1,x<P_{Line,1}\\ 2,P_{Line,2}\&le;x<P_{Line,1}+P_{Line,2}\\ .\\ .\\ .\\ M,P_{Line,1}+P_{Line,2}+\mathrm{...}+P_{Line,M-1}\&le;x<1\end{array}\right.---\left(1\right)$

In formula (1), x is for obeying [0,1] equally distributed random number, i.e. x ~ U [0,1], F _linefor fault branch numbering, M is branch road sum in described asymmetrical three-phase distribution network, P _{line, i}for the failure rate of i-th branch road in described asymmetrical three-phase distribution network, { 1, M}, U are uniformly distributed function to i ∈;

(1-2) the abort situation model of described asymmetrical three-phase distribution network is set up by following formula:

F _Loc＝y*100％(2)

In formula (2), y is for obeying [0,1] equally distributed random number, i.e. y ~ U [0,1], F _locfor fault branch F _linefault point before leg length and fault branch F _linethe percentage of the ratio of total length;

(1-3) the fault type model setting up described asymmetrical three-phase distribution network comprises:

If the fault branch F of described asymmetrical three-phase distribution network _linefor three-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$T_{Type}=\left\{\begin{array}{cc}3LG,& z<P_{3LG}\\ 2L,& P_{3LG}\&le;z<P_{3LG}+P_{2L}\\ 2LG,& P_{3LG}+P_{2L}\&le;z<P_{3LG}+P_{2L}+P_{2LG}\\ LG,& P_{3LG}+P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(3\right)$

In formula (3), z is for obeying [0,1] equally distributed random number, i.e. z ~ U [0,1], F _typefor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault type, 3LG is three-phase ground short trouble type, and 2L is two-phase phase fault type, and 2LG is two-phase short circuit and ground fault type, and LG is single-phase grounding fault type, P _3LGfor three-phase ground short trouble probability of happening, P _2Lfor two-phase phase fault probability of happening, P _2LGfor two-phase short circuit and ground fault probability of happening;

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$F_{Type}=\left\{\begin{array}{cc}2L,& z<P_{2L}\\ 2LG,& P_{2L}\&le;z<P_{2L}+P_{2LG}\\ LG,& P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(4\right)$

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

F _Type＝LG(5)。

(1-4) the trouble duration model of described asymmetrical three-phase distribution network is set up by following formula:

F _Dur＝μ(6)

In formula (6), for obeying, μ expects that standard deviation is the random number of the standardized normal distribution of 0.01s, i.e. μ ~ N [0.06,0.01], F for 0.06s _durfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding trouble duration;

(1-5) the fault impedance model of described asymmetrical three-phase distribution network is set up by following formula:

F _Res＝τ(7)

In formula (7), τ is 5 Ω for obeying expectation, and standard deviation is the random number of the standardized normal distribution of 1 Ω, i.e. τ ~ N [5,1], F _resfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault impedance.

Preferably, in described step (2), utilize Monte Carlo simulation mode to set up the DG stochastic model of described asymmetrical three-phase distribution network, comprising:

(2-1) blower fan power generation system adopts a curve model, blower fan power output P _windwith the pass of wind speed v be:

$P_{wind}=\left\{\begin{array}{cc}0,& v<v_{ci},v>v_{co}\\ a+bv,& v_{ci}<v<v_r\\ p_r,& v_r<v<v_{co}\end{array}\right.---\left(8\right)$

In formula (8), be constant, v _rthe rated wind speed of blower fan, P _rthe rated power of blower fan, v _cithe incision wind speed of blower fan, v _coit is the cut-out wind speed of blower fan;

Wind turbines number of units is N _wtgtime, Wind turbines power output P _ωmodel be:

P _ω＝P _windN _wtg(9)

Work as v _ci<v<v _rtime, Wind turbines power output P _ωthe formula of probability density function be:

In formula (10), K is the form parameter of Weibull distribution, and C is the scale parameter of Weibull distribution, f (P _ω) to gain merit the probability density function of exerting oneself for Wind turbines, Q _ωexert oneself for Wind turbines is idle, for power factor;

(2-2) photovoltaic power generation system output power P _solarmodel be:

P _solar＝rAη(11)

In formula (11), r is radiancy, and unit is W/m ², for the gross area of the solar battery of photovoltaic generating system, A _mfor the area of single battery assembly, M is the battery pack number of packages of the solar battery of photovoltaic generating system, for the photoelectric conversion efficiency of the solar battery of photovoltaic generating system, η _mfor the photoelectric conversion efficiency of single battery assembly;

Photovoltaic power generation system output power P _solarprobability density function be:

$f\left(P_{solar}\right)=\frac{\mathrm{\&Gamma;}(\mathrm{\&alpha;}+\mathrm{\&beta;})}{R_{solar}\mathrm{\&Gamma;}\left(\mathrm{\&alpha;}\right)\mathrm{\&Gamma;}\left(\mathrm{\&beta;}\right)}{\left(\frac{P_{solar}}{R_{solar}}\right)}^{\mathrm{\&alpha;}-1}{(1-\frac{P_{solar}}{R_{solar}})}^{\mathrm{\&beta;}-1}---\left(12\right)$

In formula (12), R _solar=r _maxa η is the peak power output of the solar battery of photovoltaic generating system, r _maxfor greatest irradiation degree, α, β are Beta profile shape parameter.

Preferably, in described step (3), utilize Choleskydecomposition to carry out orthogonal transform to described fault stochastic model and comprise:

5 n-dimensional random variable n p=[p are formed by faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model ₁, p ₂..., p ₅] ^t, the Mean Vector of described stochastic variable p is u _p=[u ₁, u ₂..., u ₅] ^t, the covariance matrix C of described stochastic variable p _pfor:

$C_p=\left[\begin{array}{ccccc}{\mathrm{\&sigma;}}_{p_1}^2& {\mathrm{\&sigma;}}_{p_1{p}_2}& {\mathrm{\&sigma;}}_{p_1{p}_3}& {\mathrm{\&sigma;}}_{p_1{p}_4}& {\mathrm{\&sigma;}}_{p_1{p}_5}\\ {\mathrm{\&sigma;}}_{p_2{p}_1}& {\mathrm{\&sigma;}}_{p_2}^2& {\mathrm{\&sigma;}}_{p_2{p}_3}& {\mathrm{\&sigma;}}_{p_2{p}_4}& {\mathrm{\&sigma;}}_{p_2{p}_5}\\ {\mathrm{\&sigma;}}_{p_3{p}_1}& {\mathrm{\&sigma;}}_{p_3{p}_2}& {\mathrm{\&sigma;}}_{p_3}^2& {\mathrm{\&sigma;}}_{p_3{p}_4}& {\mathrm{\&sigma;}}_{p_3{p}_5}\\ {\mathrm{\&sigma;}}_{p_4{p}_1}& {\mathrm{\&sigma;}}_{p_4{p}_2}& {\mathrm{\&sigma;}}_{p_4{p}_3}& {\mathrm{\&sigma;}}_{p_4}^2& {\mathrm{\&sigma;}}_{p_4{p}_5}\\ {\mathrm{\&sigma;}}_{p_5{p}_1}& {\mathrm{\&sigma;}}_{p_5{p}_2}& {\mathrm{\&sigma;}}_{p_5{p}_3}& {\mathrm{\&sigma;}}_{p_5{p}_4}& {\mathrm{\&sigma;}}_{p_{25}}^2\end{array}\right]---\left(13\right)$

Determine the local derviation coefficient lambda of described stochastic variable p ₃, formula is:

${\mathrm{\&lambda;}}_3=diag({\mathrm{\&lambda;}}_{p_1,3},...,{\mathrm{\&lambda;}}_{p_l,3},...{\mathrm{\&lambda;}}_{p_5,3})---\left(14\right)$

In formula (14), for l stochastic variable p in described stochastic variable p _llocal derviation coefficient, l ∈ [1,5];

Determine that the formula of the uncorrelated matrix q of described stochastic variable p is:

q＝Bp(15)

In formula (15), B is intermediary matrix;

To the covariance matrix C of described stochastic variable p _pcarry out orthogonal transform, formula is:

C _p＝LL ^T(16)

In formula (16), L is described covariance matrix C _pinferior triangular flap;

Then determine the Mean Vector u of described uncorrelated matrix q _qformula be:

u _q＝L ^-1u _p(17)

In formula (17), u _pfor the Mean Vector of described stochastic variable p;

Determine r element q in described uncorrelated matrix q _rlocal derviation coefficient formula be:

${\mathrm{\&lambda;}}_{q_r,3}=\underset{l=1}{\overset{5}{\&Sigma;}}{\left(L_{rl}^{-1}\right)}^3{\mathrm{\&lambda;}}_{p_l,3}{\mathrm{\&sigma;}}_{p_l}^3---\left(18\right)$

In formula (18), for described covariance matrix C _pthe capable l column element of r of inferior triangular flap L, r, l ∈ [1,5].

Preferably, in described step (4), adopt point estimations to carry out process to the fault stochastic model of described asymmetrical three-phase distribution network and DG stochastic model and comprise:

Stochastic variable X=[the X of 7 dimensions is made up of Wind turbines power output and photovoltaic power generation system output power in faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model and described DG stochastic model _1,x ₂..., X ₇] ^t, the nonlinear function that if Y=h (X) is is variable with described stochastic variable X, for the probability density function of i-th stochastic variable in described stochastic variable X, i ∈ [1,7];

I-th stochastic variable X from described stochastic variable X _imiddle extraction two estimation points, wherein, a kth estimation point x _i,kcomputing formula is:

x _i,k＝μ _i+ξ _i,kσ _i(19)

In formula (19), μ _ifor expectation, σ _ifor standard deviation, ξ _i,kfor x _i,kcorresponding position parameter, k=1,2;

Determine described with the estimated value of the 3 rank squares of the described stochastic variable X nonlinear function Y=h (X) that is variable by following formula:

$E\left(Y^j\right)\&cong;\underset{i=1}{\overset{7}{\&Sigma;}}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}\&times;{\&lsqb;h({\mathrm{\&mu;}}_1,{\mathrm{\&mu;}}_2,...{\mathrm{\&mu;}}_{i-1},x_{i,k},{\mathrm{\&mu;}}_{i+1}...,{\mathrm{\&mu;}}_7)\&rsqb;}^j---\left(20\right)$

In formula (20), j=1,2,3, k=1,2, i ∈ [1,7], ω _i,kfor x _i,kcorresponding weight coefficient, μ _i-1for expectation, μ _i+1for expectation;

Wherein, x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{cc}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}{\left({\mathrm{\&xi;}}_{i,k}\right)}^j={\mathrm{\&lambda;}}_{i,j}& j=1,2,3\\ \underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}=\frac{1}{7}& \end{array}\right.---\left(21\right)$

In formula (21), λ _i,jfor i-th stochastic variable X in described stochastic variable X _icenter, j rank square;

Work as j=1, when 2, λ _{i, 1}=0, λ _{i, 2}=1, then x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{c}{\mathrm{\&xi;}}_{i,k}=\frac{{\mathrm{\&lambda;}}_{i,3}+{(-1)}^{3-k}\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}{2}\\ {\mathrm{\&omega;}}_{i,k}=\frac{{(-1)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7({\mathrm{\&xi;}}_{i,2}-{\mathrm{\&xi;}}_{i,1})}=-\frac{{(-1)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}\end{array}\right.---\left(22\right).$

Preferably, in described step (6), if Y _ifor i-th element in the stochastic variable of described two groups of simulating schemes of being correlated with, i ∈ [1,14], then determine by Cornish-Fisher progression that the formula of the probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is respectively:

In formula (23), for ξ _istandardized normal distribution probability density function, χ ₃be three rank cumulant;

Wherein, χ ₃value be Y _ilocal derviation coefficient, ξ _ifor Y _icanonical form, formula is:

ξ _i＝(x-μ _i)/σ _i(24)

In formula (24), μ _ifor Y _iaverage, σ _ifor Y _ivariance.

Beneficial effect of the present invention:

A kind of asymmetrical three-phase distribution network voltage dip appraisal procedure provided by the invention, study for distribution network asymmetry problem, utilize Cholesky factorization and two-point estimate method, random chance problem is converted into multiple certain problem, then existing distribution net analytic al software is adopted to build model, carry out the voltage dip emulation of active power distribution network, the statistic feature of statistics voltage dip, analyze the weak link in electrical network, the voltage dip probability density function of each node is set up based on Cornish-Fisher progression, calculating voltage falls index temporarily, realize assessing the voltage dip of asymmetrical three-phase power distribution network, to consider in the correlation of fault type and faulty line in power distribution network and power distribution network various short trouble to emulate the characteristic of whole power distribution network, asymmetrical three-phase network and three-phase symmetrical network can be applicable to, for taking to suppress the measure of voltage dip to provide reference, improve the power supply reliability of power distribution network.

Accompanying drawing explanation

Fig. 1 is the flow chart of a kind of asymmetrical three-phase distribution network of the present invention voltage dip appraisal procedure.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is elaborated.

For making the object of the embodiment of the present invention, technical scheme and advantage clearly, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making other embodiments all obtained under creative work prerequisite, belong to the scope of protection of the invention.

A kind of asymmetrical three-phase distribution network voltage dip appraisal procedure provided by the invention, study for distribution network asymmetry problem, utilize Cholesky factorization (Choleskydecomposition) and two-point estimate method, random chance problem is converted into multiple certain problem, then existing distribution net analytic al software is adopted to build model, carry out the voltage dip emulation of active power distribution network, the statistic feature of statistics voltage dip, analyze the weak link in electrical network, the voltage dip probability density function of each node is set up based on Cornish-Fisher progression, calculating voltage falls index temporarily, realize assessing the voltage dip of asymmetrical three-phase power distribution network, for taking to suppress the measure of voltage dip to provide reference, improve the power supply reliability of power distribution network, as shown in Figure 1, comprise:

(1) the fault stochastic model of described asymmetrical three-phase distribution network is set up;

(2) the DG stochastic model of described asymmetrical three-phase distribution network is set up;

(3) orthogonal transform is carried out to described fault stochastic model, obtain the uncorrelated fault stochastic model that described fault stochastic model is corresponding;

(4) adopt the stochastic variable of point estimations to described uncorrelated fault stochastic model and DG stochastic model composition to process, obtain the stochastic variable of two groups of simulating schemes;

(5) adopt that the stochastic variable of Choleskydecomposition mode to described two groups of simulating schemes is counter decomposes, obtain the stochastic variable of two groups of simulating schemes of being correlated with corresponding to the stochastic variable of described two groups of simulating schemes;

(6) probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is determined respectively by Cornish-Fisher progression.

Concrete, in described step (1), utilize Monte Carlo simulation mode to set up the fault stochastic model of described asymmetrical three-phase distribution network, comprising:

(1-1) the faulty line model of described asymmetrical three-phase distribution network is set up by following formula:

$F_{Line}=\left\{\begin{array}{c}1,x<P_{Line,1}\\ 2,P_{Line,1}\&le;x<P_{Line,1}+P_{Line,2}\\ .\\ .\\ .\\ M,P_{Line,1}+P_{Line,2}+\mathrm{...}+P_{Line,M-1}\&le;x<1\end{array}\right.---\left(1\right)$

In formula (1), x is for obeying [0,1] equally distributed random number, i.e. x ~ U [0,1], F _linefor fault branch numbering, M is branch road sum in described asymmetrical three-phase distribution network, P _{line, i}for the failure rate of i-th branch road in described asymmetrical three-phase distribution network, { 1, M}, U are uniformly distributed function to i ∈;

(1-2) the abort situation model of described asymmetrical three-phase distribution network is set up by following formula:

F _Loc＝y*100％(2)

In formula (2), y is for obeying [0,1] equally distributed random number, i.e. y ~ U [0,1], F _locfor fault branch F _linefault point before leg length and fault branch F _linethe percentage of the ratio of total length;

(1-3) the fault type model setting up described asymmetrical three-phase distribution network comprises:

If the fault branch F of described asymmetrical three-phase distribution network _linefor three-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$T_{Type}=\left\{\begin{array}{cc}3LG,& z<P_{3LG}\\ 2L,& P_{3LG}\&le;z<P_{3LG}+P_{2L}\\ 2LG,& P_{3LG}+P_{2L}\&le;z<P_{3LG}+P_{2L}+P_{2LG}\\ LG,& P_{3LG}+P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(3\right)$

In formula (3), z is for obeying [0,1] equally distributed random number, i.e. z ~ U [0,1], F _typefor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault type, 3LG is three-phase ground short trouble type, and 2L is two-phase phase fault type, and 2LG is two-phase short circuit and ground fault type, and LG is single-phase grounding fault type, P _3LGfor three-phase ground short trouble probability of happening, P _2Lfor two-phase phase fault probability of happening, P _2LGfor two-phase short circuit and ground fault probability of happening;

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$F_{Type}=\left\{\begin{array}{cc}2L,& z<P_{2L}\\ 2LG,& P_{2L}\&le;z<P_{2L}+P_{2LG}\\ LG,& P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(4\right)$

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

F _Type＝LG(5)。

(1-4) the trouble duration model of described asymmetrical three-phase distribution network is set up by following formula:

F _Dur＝μ(6)

In formula (6), for obeying, μ expects that standard deviation is the random number of the standardized normal distribution of 0.01s, i.e. μ ~ N [0.06,0.01], F for 0.06s _durfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding trouble duration;

(1-5) the fault impedance model of described asymmetrical three-phase distribution network is set up by following formula:

F _Res＝τ(7)

In formula (7), τ is 5 Ω for obeying expectation, and standard deviation is the random number of the standardized normal distribution of 1 Ω, i.e. τ ~ N [5,1], F _resfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault impedance.

In described step (2), utilize Monte Carlo simulation mode to set up the DG stochastic model of described asymmetrical three-phase distribution network, comprising:

(2-1) blower fan power generation system adopts a curve model, blower fan power output P _windwith the pass of wind speed v be:

$P_{wind}=\left\{\begin{array}{cc}0,& v<v_{ci},v>v_{co}\\ a+bv,& v_{ci}<v<v_r\\ p_r,& v_r<v<v_{co}\end{array}\right.---\left(8\right)$

In formula (8), be constant, v _rthe rated wind speed of blower fan, P _rthe rated power of blower fan, v _cithe incision wind speed of blower fan, v _coit is the cut-out wind speed of blower fan;

Wind turbines number of units is N _wtgtime, Wind turbines power output P _ωmodel be:

P _ω＝P _windN _wtg(9)

Work as v _ci<v<v _rtime, Wind turbines power output P _ωthe formula of probability density function be:

In formula (10), K is the form parameter of Weibull distribution, and C is the scale parameter of Weibull distribution, f (P _ω) to gain merit the probability density function of exerting oneself for Wind turbines, Q _ωexert oneself for Wind turbines is idle, for power factor;

(2-2) photovoltaic power generation system output power P _solarmodel be:

P _solar＝rAη(11)

In formula (11), r is radiancy, and unit is W/m ², for the gross area of the solar battery of photovoltaic generating system, A _mfor the area of single battery assembly, M is the battery pack number of packages of the solar battery of photovoltaic generating system, for the photoelectric conversion efficiency of the solar battery of photovoltaic generating system, η _mfor the photoelectric conversion efficiency of single battery assembly;

Photovoltaic power generation system output power P _solarprobability density function be:

$f\left(P_{solar}\right)=\frac{\mathrm{\&Gamma;}(\mathrm{\&alpha;}+\mathrm{\&beta;})}{R_{solar}\mathrm{\&Gamma;}\left(\mathrm{\&alpha;}\right)\mathrm{\&Gamma;}\left(\mathrm{\&beta;}\right)}{\left(\frac{P_{solar}}{R_{solar}}\right)}^{\mathrm{\&alpha;}-1}{(1-\frac{P_{solar}}{R_{solar}})}^{\mathrm{\&beta;}-1}---\left(12\right)$

In formula (12), R _solar=r _maxa η is the peak power output of the solar battery of photovoltaic generating system, r _maxfor greatest irradiation degree, α, β are Beta profile shape parameter.

In described step (3), utilize Choleskydecomposition to carry out orthogonal transform to described fault stochastic model and comprise:

5 n-dimensional random variable n p=[p are formed by faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model ₁, p ₂..., p ₅] ^t, the Mean Vector of described stochastic variable p is u _p=[u ₁, u ₂..., u ₅] ^t, the covariance matrix C of described stochastic variable p _pfor:

$C_p=\left[\begin{array}{ccccc}{\mathrm{\&sigma;}}_{p_1}^2& {\mathrm{\&sigma;}}_{p_1{p}_2}& {\mathrm{\&sigma;}}_{p_1{p}_3}& {\mathrm{\&sigma;}}_{p_1{p}_4}& {\mathrm{\&sigma;}}_{p_1{p}_5}\\ {\mathrm{\&sigma;}}_{p_2{p}_1}& {\mathrm{\&sigma;}}_{p_2}^2& {\mathrm{\&sigma;}}_{p_2{p}_3}& {\mathrm{\&sigma;}}_{p_2{p}_4}& {\mathrm{\&sigma;}}_{p_2{p}_5}\\ {\mathrm{\&sigma;}}_{p_3{p}_1}& {\mathrm{\&sigma;}}_{p_3{p}_2}& {\mathrm{\&sigma;}}_{p_3}^2& {\mathrm{\&sigma;}}_{p_3{p}_4}& {\mathrm{\&sigma;}}_{p_3{p}_5}\\ {\mathrm{\&sigma;}}_{p_4{p}_1}& {\mathrm{\&sigma;}}_{p_4{p}_2}& {\mathrm{\&sigma;}}_{p_4{p}_3}& {\mathrm{\&sigma;}}_{p_4}^2& {\mathrm{\&sigma;}}_{p_4{p}_5}\\ {\mathrm{\&sigma;}}_{p_5{p}_1}& {\mathrm{\&sigma;}}_{p_5{p}_2}& {\mathrm{\&sigma;}}_{p_5{p}_3}& {\mathrm{\&sigma;}}_{p_5{p}_4}& {\mathrm{\&sigma;}}_{p_{25}}^2\end{array}\right]---\left(13\right)$

Determine the local derviation coefficient lambda of described stochastic variable p ₃, formula is:

${\mathrm{\&lambda;}}_3=diag({\mathrm{\&lambda;}}_{p_1,3},...,{\mathrm{\&lambda;}}_{p_l,3},...{\mathrm{\&lambda;}}_{p_5,3})---\left(14\right)$

In formula (14), for l stochastic variable p in described stochastic variable p _llocal derviation coefficient, l ∈ [1,5];

Determine that the formula of the uncorrelated matrix q of described stochastic variable p is:

q＝Bp(15)

In formula (15), B is intermediary matrix;

To the covariance matrix C of described stochastic variable p _pcarry out orthogonal transform, formula is:

C _p＝LL ^T(16)

In formula (16), L is described covariance matrix C _pinferior triangular flap;

Then determine the Mean Vector u of described uncorrelated matrix q _qformula be:

u _q＝L ^-1u _p(17)

In formula (17), u _pfor the Mean Vector of described stochastic variable p;

Determine r element q in described uncorrelated matrix q _rlocal derviation coefficient formula be:

${\mathrm{\&lambda;}}_{q_r,3}=\underset{l=1}{\overset{5}{\&Sigma;}}{\left(L_{rl}^{-1}\right)}^3{\mathrm{\&lambda;}}_{p_l,3}{\mathrm{\&sigma;}}_{p_l}^3---\left(18\right)$

In formula (18), for described covariance matrix C _pthe capable l column element of r of inferior triangular flap L, r, l ∈ [1,5].

In described step (4), adopt point estimations to carry out process to the fault stochastic model of described asymmetrical three-phase distribution network and DG stochastic model and comprise:

Stochastic variable X=[the X of 7 dimensions is made up of Wind turbines power output and photovoltaic power generation system output power in faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model and described DG stochastic model ₁, X ₂..., X ₇] ^t, the nonlinear function that if Y=h (X) is is variable with described stochastic variable X, for the probability density function of i-th stochastic variable in described stochastic variable X, i ∈ [1,7];

I-th stochastic variable X from described stochastic variable X _imiddle extraction two estimation points, wherein, a kth estimation point x _i,kcomputing formula is:

x _i,k＝μ _i+ξ _i,kσ _i(19)

In formula (19), μ _ifor expectation, σ _ifor standard deviation, ξ _i,kfor x _i,kcorresponding position parameter, k=1,2;

Determine described with the estimated value of the 3 rank squares of the described stochastic variable X nonlinear function Y=h (X) that is variable by following formula:

$E\left(Y^j\right)\&cong;\underset{i=1}{\overset{7}{\&Sigma;}}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}\&times;{\&lsqb;h({\mathrm{\&mu;}}_1,{\mathrm{\&mu;}}_2,...{\mathrm{\&mu;}}_{i-1},x_{i,k},{\mathrm{\&mu;}}_{i+1}...,{\mathrm{\&mu;}}_7)\&rsqb;}^j---\left(20\right)$

In formula (20), j=1,2,3, k=1,2, i ∈ [1,7], ω _i,kfor x _i,kcorresponding weight coefficient, μ _i-1for expectation, μ _i+1for expectation;

Wherein, x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{cc}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}{\left({\mathrm{\&xi;}}_{i,k}\right)}^j={\mathrm{\&lambda;}}_{i,j}& j=1,2,3\\ \underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}=\frac{1}{7}& \end{array}\right.---\left(21\right)$

In formula (21), λ _i,jfor i-th stochastic variable X in described stochastic variable X _icenter, j rank square;

Work as j=1, when 2, λ _{i, 1}=0, λ _{i, 2}=1, then x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{c}{\mathrm{\&xi;}}_{i,k}=\frac{{\mathrm{\&lambda;}}_{i,3}+{(-1)}^{3-k}\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}{2}\\ {\mathrm{\&omega;}}_{i,k}=\frac{{(-1)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7({\mathrm{\&xi;}}_{i,2}-{\mathrm{\&xi;}}_{i,1})}=-\frac{{(-1)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}\end{array}\right.---\left(22\right).$

In described step (5), adopt that the stochastic variable of Choleskydecomposition mode to described two groups of simulating schemes is counter decomposes, obtain the stochastic variable of two groups of simulating schemes of being correlated with corresponding to the stochastic variable of described two groups of simulating schemes;

Wherein, the stochastic variable of described two groups of simulating schemes is uncorrelated random variables, uncorrelated random variables cannot carry out simulation calculation, therefore need to carry out inverse transformation, by the stochastic variable of two groups of corresponding for the stochastic variable of described two groups of simulating schemes simulating schemes of being correlated with, such as: only have two-phase, after Cholesky factorization, fault type may be three-phase ground, but this is impossible, inverse transformation is therefore wanted to go back.

In described step (6), if Y _ifor i-th element in the stochastic variable of described two groups of simulating schemes of being correlated with, i ∈ [1,14], then determine by Cornish-Fisher progression that the formula of the probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is respectively:

In formula (23), for ξ _istandardized normal distribution probability density function, χ ₃be three rank cumulant;

Wherein, χ ₃value be Y _ilocal derviation coefficient, ξ _ifor Y _icanonical form, formula is:

ξ _i＝(x-μ _i)/σ _i(24)

In formula (24), μ _ifor Y _iaverage, σ _ifor Y _ivariance.

Finally should be noted that: above embodiment is only in order to illustrate that technical scheme of the present invention is not intended to limit; although with reference to above-described embodiment to invention has been detailed description; those of ordinary skill in the field are to be understood that: still can modify to the specific embodiment of the present invention or equivalent replacement; and not departing from any amendment of spirit and scope of the invention or equivalent replacement, it all should be encompassed within claims of the present invention.

Claims (6)

1. an asymmetrical three-phase distribution network voltage dip appraisal procedure, is characterized in that, described method comprises:

(1) the fault stochastic model of described asymmetrical three-phase distribution network is set up;

(2) the DG stochastic model of described asymmetrical three-phase distribution network is set up;

(3) orthogonal transform is carried out to described fault stochastic model, obtain the uncorrelated fault stochastic model that described fault stochastic model is corresponding;

(4) adopt the stochastic variable of point estimations to described uncorrelated fault stochastic model and DG stochastic model composition to process, obtain the stochastic variable of two groups of simulating schemes;

(5) adopt that the stochastic variable of Choleskydecomposition mode to described two groups of simulating schemes is counter decomposes, obtain the stochastic variable of two groups of simulating schemes of being correlated with corresponding to the stochastic variable of described two groups of simulating schemes;

(6) probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is determined respectively by Cornish-Fisher progression.

2. the method for claim 1, is characterized in that, in described step (1), utilizes Monte Carlo simulation mode to set up the fault stochastic model of described asymmetrical three-phase distribution network, comprising:

(1-1) the faulty line model of described asymmetrical three-phase distribution network is set up by following formula:

$F_{Line}=\left\{\begin{array}{c}1,x<P_{Line,1}\\ 2,P_{Line,1}\&le;x<P_{Line,1}+P_{Line,2}\\ .\\ .\\ .\\ M,P_{Line,1}+P_{Line,2}+...+P_{Line,M-1}\&le;x<1\end{array}\right.---\left(1\right)$

In formula (1), x is for obeying [0,1] equally distributed random number, i.e. x ~ U [0,1], F _linefor fault branch numbering, M is branch road sum in described asymmetrical three-phase distribution network, P _{line, i}for the failure rate of i-th branch road in described asymmetrical three-phase distribution network, { 1, M}, U are uniformly distributed function to i ∈;

(1-2) the abort situation model of described asymmetrical three-phase distribution network is set up by following formula:

F _Loc＝y*100％(2)

In formula (2), y is for obeying [0,1] equally distributed random number, i.e. y ~ U [0,1], F _locfor fault branch F _linefault point before leg length and fault branch F _linethe percentage of the ratio of total length;

(1-3) the fault type model setting up described asymmetrical three-phase distribution network comprises:

If the fault branch F of described asymmetrical three-phase distribution network _linefor three-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$F_{Type}=\left\{\begin{array}{cc}3LG,& z<P_{3LG}\\ 2L,& P_{3LG}\&le;z<P_{3LG}+P_{2L}\\ 2LG,& P_{3LG}+P_{2L}\&le;z<P_{3LG}+P_{2L}+P_{2LG}\\ LG,& P_{3LG}+P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(3\right)$

In formula (3), z is for obeying [0,1] equally distributed random number, i.e. z ~ U [0,1], F _typefor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault type, 3LG is three-phase ground short trouble type, and 2L is two-phase phase fault type, and 2LG is two-phase short circuit and ground fault type, and LG is single-phase grounding fault type, P _3LGfor three-phase ground short trouble probability of happening, P _2Lfor two-phase phase fault probability of happening, P _2LGfor two-phase short circuit and ground fault probability of happening;

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

$F_{Type}=\left\{\begin{array}{cc}2L,& z<P_{2L}\\ 2LG,& P_{2L}\&le;z<P_{2L}+P_{2LG}\\ LG,& P_{2L}+P_{2LG}\&le;z<1\end{array}\right.---\left(4\right)$

If the fault branch F of described asymmetrical three-phase distribution network _linefor two-phase, then set up the fault branch F of described asymmetrical three-phase distribution network by following formula _linecorresponding fault type model:

F _Type＝LG(5)。

(1-4) the trouble duration model of described asymmetrical three-phase distribution network is set up by following formula:

F _Dur＝μ(6)

In formula (6), for obeying, μ expects that standard deviation is the random number of the standardized normal distribution of 0.01s, i.e. μ ~ N [0.06,0.01], F for 0.06s _durfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding trouble duration;

(1-5) the fault impedance model of described asymmetrical three-phase distribution network is set up by following formula:

F _Res＝τ(7)

In formula (7), τ is 5 Ω for obeying expectation, and standard deviation is the random number of the standardized normal distribution of 1 Ω, i.e. τ ~ N [5,1], F _resfor the fault branch F of described asymmetrical three-phase distribution network _linecorresponding fault impedance.

3. the method for claim 1, is characterized in that, in described step (2), utilizes Monte Carlo simulation mode to set up the DG stochastic model of described asymmetrical three-phase distribution network, comprising:

(2-1) blower fan power generation system adopts a curve model, blower fan power output P _windwith the pass of wind speed v be:

$P_{wind}=\left\{\begin{array}{cc}0,& v<v_{ci},v>v_{co}\\ a+bv,& v_{ci}<v<v_r\\ p_r,& v_r<v<v_{co}\end{array}\right.---\left(8\right)$

In formula (8), be constant, v _rthe rated wind speed of blower fan, P _rthe rated power of blower fan, v _cithe incision wind speed of blower fan, v _coit is the cut-out wind speed of blower fan;

Wind turbines number of units is N _wtgtime, Wind turbines power output P _ωmodel be:

P _ω＝P _windN _wtg(9)

Work as v _ci<v<v _rtime, Wind turbines power output P _ωthe formula of probability density function be:

In formula (10), K is the form parameter of Weibull distribution, and C is the scale parameter of Weibull distribution, f (P _ω) to gain merit the probability density function of exerting oneself for Wind turbines, Q _ωexert oneself for Wind turbines is idle, for power factor;

(2-2) photovoltaic power generation system output power P _solarmodel be:

P _solar＝rAη(11)

In formula (11), r is radiancy, and unit is W/m ², for the gross area of the solar battery of photovoltaic generating system, A _mfor the area of single battery assembly, M is the battery pack number of packages of the solar battery of photovoltaic generating system, for the photoelectric conversion efficiency of the solar battery of photovoltaic generating system, η _mfor the photoelectric conversion efficiency of single battery assembly;

Photovoltaic power generation system output power P _solarprobability density function be:

$f\left(P_{solar}\right)=\frac{\mathrm{\&Gamma;}\left(\mathrm{\&alpha;}+\mathrm{\&beta;}\right)}{R_{solar}\mathrm{\&Gamma;}\left(\mathrm{\&alpha;}\right)\mathrm{\&Gamma;}\left(\mathrm{\&beta;}\right)}{\left(\frac{P_{solar}}{R_{solar}}\right)}^{\mathrm{\&alpha;}-1}{\left(1-\frac{P_{solar}}{R_{solar}}\right)}^{\mathrm{\&beta;}-1}---\left(12\right)$

In formula (12), R _solar=r _maxa η is the peak power output of the solar battery of photovoltaic generating system, r _maxfor greatest irradiation degree, α, β are Beta profile shape parameter.

4. the method for claim 1, is characterized in that, in described step (3), utilizes Choleskydecomposition to carry out orthogonal transform to described fault stochastic model and comprises:

5 n-dimensional random variable n p=[p are formed by faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model ₁, p ₂..., p ₅] ^t, the Mean Vector of described stochastic variable p is u _p=[u ₁, u ₂..., u ₅] ^t, the covariance matrix C of described stochastic variable p _pfor:

$C_p=\left[\begin{array}{ccccc}{\mathrm{\&sigma;}}_{p_1}^2& {\mathrm{\&sigma;}}_{p_1{p}_2}& {\mathrm{\&sigma;}}_{p_1{p}_3}& {\mathrm{\&sigma;}}_{p_1{p}_4}& {\mathrm{\&sigma;}}_{p_1{p}_5}\\ {\mathrm{\&sigma;}}_{p_2{p}_1}& {\mathrm{\&sigma;}}_{p_2}^2& {\mathrm{\&sigma;}}_{p_2{p}_3}& {\mathrm{\&sigma;}}_{p_2{p}_4}& {\mathrm{\&sigma;}}_{p_2{p}_5}\\ {\mathrm{\&sigma;}}_{p_3{p}_1}& {\mathrm{\&sigma;}}_{p_3{p}_2}& {\mathrm{\&sigma;}}_{p_3}^2& {\mathrm{\&sigma;}}_{p_3{p}_4}& {\mathrm{\&sigma;}}_{p_3{p}_5}\\ {\mathrm{\&sigma;}}_{p_4{p}_1}& {\mathrm{\&sigma;}}_{p_4{p}_2}& {\mathrm{\&sigma;}}_{p_4{p}_3}& {\mathrm{\&sigma;}}_{p_4}^2& {\mathrm{\&sigma;}}_{p_4{p}_5}\\ {\mathrm{\&sigma;}}_{p_5{p}_1}& {\mathrm{\&sigma;}}_{p_5{p}_2}& {\mathrm{\&sigma;}}_{p_5{p}_3}& {\mathrm{\&sigma;}}_{p_5{p}_4}& {\mathrm{\&sigma;}}_{p_{25}}^2\end{array}\right]---\left(13\right)$

Determine the local derviation coefficient lambda of described stochastic variable p ₃, formula is:

λ ₃＝diag(λ _p1,3,…,λ _pl,3,…λ _p5,3)(14)

In formula (14), for l stochastic variable p in described stochastic variable p _llocal derviation coefficient, l ∈ [1,5];

Determine that the formula of the uncorrelated matrix q of described stochastic variable p is:

q＝Bp(15)

In formula (15), B is intermediary matrix;

To the covariance matrix C of described stochastic variable p _pcarry out orthogonal transform, formula is:

C _p＝LL ^T(16)

In formula (16), L is described covariance matrix C _pinferior triangular flap;

Then determine the Mean Vector u of described uncorrelated matrix q _qformula be:

u _q＝L ^-1u _p(17)

In formula (17), u _pfor the Mean Vector of described stochastic variable p;

Determine r element q in described uncorrelated matrix q _rlocal derviation coefficient formula be:

${\mathrm{\&lambda;}}_{q_r,3}=\underset{l=1}{\overset{5}{\&Sigma;}}{\left(L_{rl}^{-1}\right)}^3{\mathrm{\&lambda;}}_{p_l,3}{\mathrm{\&sigma;}}_{p_l}^3---\left(18\right)$

In formula (18), for described covariance matrix C _pthe capable l column element of r of inferior triangular flap L, r, l ∈ [1,5].

5. the method for claim 1, is characterized in that, in described step (4), adopts point estimations to carry out process to the fault stochastic model of described asymmetrical three-phase distribution network and DG stochastic model and comprises:

Stochastic variable X=[the X of 7 dimensions is made up of Wind turbines power output and photovoltaic power generation system output power in faulty line model, abort situation model, fault type model, trouble duration model and fault impedance model in described fault stochastic model and described DG stochastic model ₁, X ₂..., X ₇] ^t, the nonlinear function that if Y=h (X) is is variable with described stochastic variable X, for the probability density function of i-th stochastic variable in described stochastic variable X, i ∈ [1,7];

I-th stochastic variable X from described stochastic variable X _imiddle extraction two estimation points, wherein, a kth estimation point x _i,kcomputing formula is:

x _i,k＝μ _i+ξ _i,kσ _i(19)

In formula (19), μ _ifor expectation, σ _ifor standard deviation, ξ _i,kfor x _i,kcorresponding position parameter, k=1,2;

Determine described with the estimated value of the 3 rank squares of the described stochastic variable X nonlinear function Y=h (X) that is variable by following formula:

$E\left(Y^j\right)\&cong;\underset{i=1}{\overset{7}{\&Sigma;}}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}\&times;{\&lsqb;h\left({\mathrm{\&mu;}}_1,{\mathrm{\&mu;}}_2,...{\mathrm{\&mu;}}_{i-1},x_{i,k},{\mathrm{\&mu;}}_{i+1}...,{\mathrm{\&mu;}}_7\right)\&rsqb;}^j---\left(20\right)$

In formula (20), j=1,2,3, k=1,2, i ∈ [1,7], ω _i,kfor x _i,kcorresponding weight coefficient, μ _i-1for expectation, μ _i+1for expectation;

Wherein, x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{c}\underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}{\left({\mathrm{\&xi;}}_{i,k}\right)}^j={\mathrm{\&lambda;}}_{i,j}\\ \underset{k=1}{\overset{2}{\&Sigma;}}{\mathrm{\&omega;}}_{i,k}=\frac{1}{7}\end{array}\right.,j=1,2,3---\left(21\right)$

In formula (21), λ _i,jfor i-th stochastic variable X in described stochastic variable X _icenter, j rank square;

Work as j=1, when 2, λ _{i, 1}=0, λ _{i, 2}=1, then x _i,kcorresponding weight coefficient ω _i,kand position parameter ξ _i,kmeet equation:

$\left\{\begin{array}{c}{\mathrm{\&xi;}}_{i,k}=\frac{{\mathrm{\&lambda;}}_{i,3}+{\left(-1\right)}^{3-k}\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}{2}\\ {\mathrm{\&omega;}}_{i,k}=\frac{{\left(-1\right)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7\left({\mathrm{\&xi;}}_{i,2}-{\mathrm{\&xi;}}_{i,1}\right)}=-\frac{{\left(-1\right)}^{3-k}{\mathrm{\&xi;}}_{i,3-k}^j}{7\sqrt{28+{\mathrm{\&lambda;}}_{i,3}^2}}\end{array}\right.---\left(22\right).$

6. the method for claim 1, is characterized in that, in described step (6), if Y _ifor i-th element in the stochastic variable of described two groups of simulating schemes of being correlated with, i ∈ [1,14], then determine by Cornish-Fisher progression that the formula of the probability density function of element in the stochastic variable of described two groups of simulating schemes of being correlated with is respectively:

In formula (23), for ξ _istandardized normal distribution probability density function, χ ₃be three rank cumulant;

Wherein, χ ₃value be Y _ilocal derviation coefficient, ξ _ifor Y _icanonical form, formula is:

ξ _i＝(x-μ _i)/σ _i(24)

In formula (24), μ _ifor Y _iaverage, σ _ifor Y _ivariance.