CN100461578C - Evaluation method for static voltage stability of the power system based on the dynamic continuous tide - Google Patents

Abstract

The invention relates to a evaluation method of electrical power system static state voltage steady which is based on the dynamic continual tidal current, belongs to the electric power system movement and control technogy area. This invention method had considered the net cess not counterbalance on the entire net generator assignment in the existing electrical power system continual tidal current equation, through solve the dynamic continual tidal current equation, reasonably assigns not balanced power, solved the question of the conventional continual tidal current computed result rely on the balance pitch point choice, enhanced the electrical power system quiescent voltage stable appraisal accuracy.

Description

Appraisal procedure based on the power system steady state voltage stability of dynamic continuous tide

Technical field

The present invention relates to a kind of appraisal procedure of the power system steady state voltage stability based on dynamic continuous tide, belong to power system operation and control technology field.

Background technology

Static voltage stability analysis is the critical function in electrical network automatic voltage control system (AVC) and the EMS (EMS), and is significant to the safe operation of electrical network.Continuous tide method (CPF, Continuation Power Flow) is a kind of fundamental analysis method in the power system steady state voltage stability analysis.The continuous tide method makes its Jacobian matrix nonsingular at former power flow equation saddle junction type bifurcation point place, thereby overcome the shortcoming that the ordinary tides flow equation is not restrained near saddle junction type bifurcation point owing to the introducing of the parametric equation of expansion.According to the parametric equation difference of expansion, the continuous tide model can be divided into load type continuous tide, branch road type continuous tide, control type continuous tide, the asymmetric continuous tide of three-phase.Load type continuous tide is the typical model of continuous tide method, the initial proposition of continuous tide problem be exactly for study with analogue system in individual node, a plurality of node, certain zone or the whole network load (and generating) change non-linear effects for system.

A typical load type continuous tide method, its concrete implementation step is:

1, structure spreading parameter power flow equation: introduce load parameter λ, load after the parametrization and generating can be expressed as:

P _d，i(λ)＝P _d，i，0+λΔP _d，i，i∈Ω _d

Q _d，i(λ)＝Q _d，i，0+λΔQ _d，i，i∈Ω _d (1)

P _g，j(λ)＝P _g，j，0+λΔP _g，j，j∈Ω _g

P in the formula _{D, i, 0}, Q _{D, i, 0}, P _{G, j, 0}Be respectively meritorious, idle and generator j meritorious of the load i of initial launch point, Δ P _{D, i}, Δ Q _{D, i}, Δ P _{G, j}Be respectively meritorious, the idle recruitment of predetermined load i and the meritorious recruitment of generator j, Ω _dAnd Ω _gBe respectively and participate in load aggregation and participate in the generator set.

The parametrization power flow equation of conventional power flow equation after introducing parameter lambda can be expressed as,

f(x，λ)＝F(x)+λb＝0 (2)

In the formula: introduce load parameter λ, x is system's static state vector (being node voltage amplitude and phase angle); B is the vector of expression node power injection direction and size; F (x) matching convention power flow equation.

Can construct different spreading parameter power flow equations according to different parametric methods, existing parametric method has local parameterization, arc length parameterized, plan arc length parameterized, quadrature parametrization etc., is example with the local parameter method, introduces the expansion equation

p(x，λ)＝x _k-Δs＝0 (3)

In the formula: Δ s is known quantity in calculating for calculating step-length; x _kBe the component of x, subscript k follow the example of for

$x_k:\left|{\stackrel{.}{x}}_k\right|=\max \{\left|{\stackrel{.}{x}}_1\right|,\left|{\stackrel{.}{x}}_2\right|,\·\·\·,\left|{\stackrel{.}{x}}_n\right|\}---\left(4\right)$

In the formula:

Be variable x ₁, x ₂..., x _nGradient.

The expansion power flow equation that obtains thus is

$\left\{\begin{array}{c}f(x,\mathrm{\λ})=0\\ p(x,\mathrm{\λ})=0\end{array}\right.---\left(5\right)$

2, prediction link: according to current point and what time provided the estimated value of separating on the track next point in the past, thereby help the quick convergence of a bit finding the solution down.Usually the Forecasting Methodology that adopts in the continuous tide has single order differential method (as the tangent predicted method) and multinomial extrapolation method (as dichotomy etc.).With the tangent predicted method is example, is 1 or-1 to calculate tangent vector by specifying in the tangent vector with the corresponding component of Control Parameter at first, is shown below.

$\left[\begin{array}{c}\begin{array}{cc}{f}_x(x,\mathrm{\λ})& f_{\mathrm{\λ}}(x,\mathrm{\λ})\end{array}\\ e_k\end{array}\right]\left[\begin{array}{c}\mathrm{dx}\\ \mathrm{d\λ}\end{array}\right]=\left[\begin{array}{c}0\\ \±1\end{array}\right]---\left(6\right)$

E in the formula _kHave only with corresponding k the component of Control Parameter be 1, all the other are 0.The next Control Parameter of calculating the step is got the most violent component dx of variation in the tangent vector _k(absolute value the maximum), corresponding expansion equation right-hand member choosing+1 or-1 is by the corresponding dx of its control component in the following formula _kSymbol decision.The predicted value of solution vector can be calculated by following formula

$\left[\begin{array}{c}{\stackrel{\‾}{x}}^i\\ {\stackrel{\‾}{\mathrm{\λ}}}^i\end{array}\right]=\left[\begin{array}{c}{x}^{i-1}\\ {\mathrm{\λ}}^{i-1}\end{array}\right]+\mathrm{\δ}\left[\begin{array}{c}\mathrm{dx}\\ \mathrm{d\λ}\end{array}\right]---\left(7\right)$

δ is a step-length in the formula.

3, correction link: correction link is exactly to be that initial value calculates the expansion power flow equation with the predicted value, and obviously, predicted value is the closer to solution point, and the convergence of equation is just good more.The calculating that correction is separated can be found the solution by following formula

$\left[\begin{array}{c}f(x,\mathrm{\λ})\\ x_k-\mathrm{\η}\end{array}\right]=\left[0\right]---\left(8\right)$

η is a k component of corresponding Control Parameter during prediction is separated, with

Be initial value, can adopt Newton method or quasi-Newton method to find the solution above-mentioned equation.

4, step-length control: desirable step-length control method should be able to be adjusted with the variation of curve shape, when curve is near flex point, because the convergence domain that may exceed next calculation level is separated in prediction, and make the continuous type computational methods can't be normally enter second of PV curve by flex point, so, must take certain measure to make near program adaptive step-length that reduces flex point, as adopt the method for exploration-rollback, when correction link judges that calculating is dispersed, return a calculation level, step-length is reduced by half, continue to calculate, continue like this along controlling party to souning out and reducing step-length, till the correction calculation convergence.

5, got back to for the 2nd step, the point that has been hopeful to ask for up to having found the solution on the solution curve finishes to calculate.

Because conventional continuous tide when the treatment system load increment, satisfy following formula,

$\underset{j\∈{\mathrm{\Ω}}_G}{\mathrm{\Σ}}\mathrm{\Δ}{P}_{\mathrm{Gj}}=\underset{i\∈{\mathrm{\Ω}}_D}{\mathrm{\Σ}}\mathrm{\Δ}{P}_{\mathrm{Di}}=\mathrm{\λ}---\left(9\right)$

In fact, along with the growth of system loading, the network loss in the system is also increasing, and when system's heavy load, the increment of network loss is especially remarkable, and in conventional continuous tide calculated, this part meritorious amount of unbalance will be born by balance node fully.The processing method that the network loss increment is born by balance node fully, can cause in the computational process of continuous tide, on the one hand, the power delivery direction is along with the increase of network loss constantly changes, under the heavy load situation, because network loss increases significantly, the power delivery direction also will produce bigger deviation with initial power delivery direction; On the other hand, the power delivery direction will change because of the difference of balance node, and when balance node was positioned at load growth center or balance node away from the load growth center, this variation was particularly evident.The variation of power delivery direction on the one hand, makes that the load margin calculate is not the load margin on the initial setting power delivery direction, on the other hand, makes that there is bigger difference in the load margin that calculates when selecting different balance node.Therefore can not carry out accurate assessment to the voltage stability of electric power system.

Summary of the invention

The objective of the invention is to propose a kind of based on dynamic continuous tide (Dynamic Continuation Power Flow, hereinafter to be referred as DCPF) the appraisal procedure of power system steady state voltage stability, the imbalance power in the electrical power distribution system on the whole network generator depends on the problem that balance node is selected to solve conventional continuous tide result of calculation automatically and reasonably.

The appraisal procedure based on the power system steady state voltage stability of dynamic continuous tide that the present invention proposes may further comprise the steps:

(1) the dynamically meritorious power flow equation of structure electric power system is P _Gi-β _iΔ P _∑-P _Di-P _i(U, θ)=0,

Wherein, i=1,2 ..., N, P _Gi, P _DiBe respectively current meritorious generating power output and burden with power on the electric power system node i, β _iBe the imbalance power share that generator on the electric power system node i is shared, if node i does not have the sending and receiving motor, or this node output of a generator is non-adjustable, then β _i=0, β _iSatisfy $\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\β}}_i=1,$ U is the Electric Power System Node Voltage amplitude, and θ is the phase angle of node voltage;

(2) according to the dynamically meritorious power flow equation of above-mentioned electric power system, make up an extended dynamic continuous tide equation:

f′(U，θ，λ)＝F′(U，θ)+λb＝0

Wherein, F ' (U, θ) be the above-mentioned dynamically meritorious power flow equation of electric power system and the power flow equation of conventional reactive power flow equation composition, λ is for characterizing the parameter that each node power changes in the electric power system, and b is the vector of each node power injection direction and size in the sign electric power system;

(3) adopt conventional continuous tide predictor corrector method, above-mentioned extended dynamic continuous tide equation is found the solution, the stability margin when obtaining the power system load variation.

In the said method, the process that makes up dynamically meritorious power flow equation may further comprise the steps:

(1) the meritorious power flow equation of establishing electric power system is:

P _Gi-P _Di-P _i(U，θ)＝0，i＝1，2，...，N

(2) share of establishing the imbalance power that generator is born on the node i is β _i, according to the power amount of unbalance of electric power system:

$\mathrm{\Δ}{P}_{\mathrm{\Σ}}=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Gi}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Di}}-P_{\mathrm{Loss}}(U,\mathrm{\θ})$

Obtain dynamic continuous tide equation: P _Gi-β _iΔ P _∑-P _Di-P _i(U, θ)=0.

The appraisal procedure that the present invention proposes based on the power system steady state voltage stability of dynamic continuous tide, it is a kind of improvement on existing continuous tide method basis, in existing electric power system continuous tide equation, considered the distribution of network loss amount of unbalance on the whole network generator, by finding the solution the dynamic continuous tide equation, reasonably distribute imbalance power, solve conventional continuous tide result of calculation and depended on the problem that balance node is selected, improved the accuracy of power system steady state voltage stability assessment.

Description of drawings

Fig. 1 is IEEE 39 node system figure used among the embodiment of the inventive method.

Fig. 2 is the result of calculation comparison diagram of one embodiment of the present of invention and existing continuous tide method.

Embodiment

The appraisal procedure based on the power system steady state voltage stability of dynamic continuous tide that the present invention proposes, the dynamically meritorious power flow equation that at first makes up electric power system is P _Gi-β _iΔ P _∑-P _Di-P _i(U, θ)=0,

Wherein, i=1,2 ..., N, P _Gi, P _DiBe respectively current meritorious generating power output and burden with power on the electric power system node i, β _iBe the imbalance power share that generator on the electric power system node i is shared, if node i does not have the sending and receiving motor, or this node output of a generator is non-adjustable, then β _i=0, β _iSatisfy $\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\β}}_i=1,$ U is the Electric Power System Node Voltage amplitude, and θ is the phase angle of node voltage;

Dynamically meritorious power flow equation according to above-mentioned electric power system makes up an extended dynamic continuous tide equation:

f′(U，θ，λ)＝F′(U，θ)+λb＝0

Wherein, F ' (U, θ) be the above-mentioned dynamically meritorious power flow equation of electric power system and the power flow equation of conventional reactive power flow equation composition, λ is for characterizing the parameter that each node power changes in the electric power system, and b is the vector of each node power injection direction and size in the sign electric power system;

Adopt conventional continuous tide predictor corrector method, above-mentioned extended dynamic continuous tide equation is found the solution, the stability margin when obtaining the power system load variation.

In the said method, the process that makes up dynamically meritorious power flow equation is that the meritorious power flow equation of establishing electric power system is: P _Gi-P _Di-P _i(U, θ)=0, i=1,2 ..., N

If the share of the imbalance power that generator is born on the node i is β _i, according to the power amount of unbalance of electric power system:

$\mathrm{\Δ}{P}_{\mathrm{\Σ}}=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Gi}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Di}}-P_{\mathrm{Loss}}(U,\mathrm{\θ})$

Obtain the dynamic continuous tide equation; P _Gi-β _iΔ P _∑-P _Di-P _i(U, θ)=0.

Below introduce the detailed process of the inventive method in detail.

Form the dynamic power flow equation: the meritorious power flow equation of the system of setting up departments,

P _Gi-P _Di-P _i(U，θ)＝0，i＝1，2，...，N (10)

The equation that comprises balance node in the formula (10).

If the power amount of unbalance that occurs in the system is,

$\mathrm{\Δ}{P}_{\mathrm{\Σ}}=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Gi}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Di}}-P_{\mathrm{Loss}}(U,\mathrm{\θ})---\left(11\right)$

P _Gi, P _DiBe current meritorious generating power output and burden with power of node i, P _LossIt is the total network loss of system.This difference should be shared jointly by all generators, and formula (10) becomes

P _Gi-β _iΔP _∑-P _Di-P _i(U，θ)＝0，i＝1，2，...，N (12)

Wherein, β _iBe the imbalance power share that generator on the node i is shared, if node i does not have the sending and receiving motor, or this node output of a generator is non-adjustable, then β _i=0.β _iSatisfy $\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\β}}_i=1.$

By (12) formula and original reactive power flow equation together, just obtained the dynamic power flow equation:

F′(U，θ)＝0 (13)

As seen, the generator active power of the V θ node in the conventional trend also needs given in advance.The power difference of whole system comprises that trend can't know that trend calculates the network loss that could determine when finishing and will come balance by all generators in calculating in advance.

The extended dynamic power flow equation of constructing variableization: based on formula (13), introduce parameter lambda, can obtain parameterized dynamic power flow equation:

f′(U，θ，λ)＝F′(U，θ)+λb＝0 (14)

According to the definition of transmit power calculation, the increment of network loss is born by the generator of power output area, therefore, and β _iCan describe fully by direction vector b.Like this, the power amount of unbalance,

ΔP _∑＝-λ-ΔP _Loss(U，θ) (15)

In the formula, Δ P _Loss(U θ) is the variable quantity of current network loss and ground state network loss, promptly

ΔP _Loss(U，θ)＝P _Loss(U，θ)-P _Loss(U ₀，θ ₀) (16)

Therefore, the generator node equation of gaining merit is rewritten as

P _Gi-b Δ P _∑-P _Di-P _i(U, θ)=0, i ∈ generator node (17)

With the local parameter method is example, introduces the expansion equation

p(U，θ，λ)＝u _k-Δs＝0 (18)

In the formula, Δ s is known quantity in calculating for calculating step-length.u _kBe the component of U, subscript k follow the example of for:

$u_k:\left|{\stackrel{.}{u}}_k\right|=\max \{\left|{\stackrel{.}{u}}_1\right|,\left|{\stackrel{.}{u}}_2\right|,...,\left|{\stackrel{.}{u}}_n\right|\}---\left(19\right)$

In the formula,

Be variable u ₁..., u _nGradient.

The extended dynamic power flow equation that obtains,

$\left\{\begin{array}{c}f\′(U,\mathrm{\θ},\mathrm{\λ})=0\\ p(U,\mathrm{\θ},\mathrm{\λ})=0\end{array}\right.---\left(20\right)$

As seen, compare with the expansion power flow equation based on conventional trend, the unknown number number is all consistent with the equation number, and difference only is the meritorious equation of the generator node of formula (17), needs to add network loss variation delta P _Loss(U, sendout θ).

Network loss P _Loss(U, expression formula θ) can be written as,

$\begin{array}{cc}{P}_{\mathrm{Loss}}(U,\mathrm{\θ})=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_i=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{U}_i\underset{j\∈i}{\mathrm{\Σ}}{U}_j{G}_{\mathrm{ij}}\cos {\mathrm{\θ}}_{\mathrm{ij}}& i=\mathrm{1,2},...,N\end{array}---\left(21\right)$

In the formula, j ∈ i represents the node j that all and i link, and comprises j=i.

Formula (17) is a key content of the present invention, from the angle that calculating is found the solution, can carry out to a certain degree abbreviation or expansion on this basis, for example, ignores Δ P when forming Jacobian matrix _Loss(U, introducing θ) directly uses former Jacobian matrix to carry out iterative to the correction of Jacobian matrix; Introduce the network loss state variable, expand a network loss equation again and find the solution.

Employing is found the solution as the prediction-bearing calibration of conventional continuous tide method.

Below be an embodiment of the inventive method: carry out emulation experiment with the IEEE39 node system and make embodiment, further specify as follows:

IEEE39 node system structure as shown in Figure 1, heavy line is represented bus (being the node in the nodal analysis method) among the figure, the other numeral node serial number of bus, downward arrow is represented load, band ' G ' character representation generator in the symbol circle.Its DCPF method specifically may further comprise the steps:

1) forms the dynamic power flow equation;

For present embodiment, according to formula row formula.

2) the extended dynamic power flow equation of constructing variableization:

In the present embodiment, adopt the method for expansion network loss equation, list parameterized extended dynamic power flow equation according to formula.

3) employing is found the solution as the prediction-bearing calibration of conventional continuous tide method:

In the present embodiment, employing is found the solution as the prediction-bearing calibration of conventional continuous tide method.The comparison of solving result and prior art, as shown in Figure 2.Among the figure, abscissa is different balance node, and ordinate is the whole network stability margin that calculates, and the column with slant lines bar is the result of calculation that adopts conventional continuous tide method, and some post bar is the result of calculation that adopts the dynamic continuous tide method.As seen, when adopting conventional continuous tide, select different generators as balance node, the stability margin difference that calculates is very big; And adopt improved continuous tide method based on the dynamic power flow equation, and select different generator nodes as balance node, the stability margin that is calculated is consistent.Therefore, adopt the dynamic continuous tide method to calculate, the result of stability margin does not rely on the selection of balance node, has improved the accuracy of power system steady state voltage stability assessment.

Claims (2)

1. appraisal procedure based on the power system steady state voltage stability of dynamic continuous tide is characterized in that this method may further comprise the steps:

(1) the dynamically meritorious power flow equation of structure electric power system is P _Gi-β _iΔ P _Σ-P _Di-P _i(U, θ)=0,

Wherein, i=1,2 ..., N, P _Gi, P _DiBe respectively current meritorious generating power output and burden with power on the electric power system node i, Δ P _ΣBe the power amount of unbalance of electric power system, β _iBe the imbalance power share that generator on the electric power system node i is shared, β _iSatisfy $\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\β}}_i=1,$ If node i does not have the sending and receiving motor, or this node output of a generator is non-adjustable, then β _i=0, U is the Electric Power System Node Voltage amplitude, and θ is the phase angle of node voltage;

(2) according to the dynamically meritorious power flow equation of above-mentioned electric power system, make up an extended dynamic continuous tide equation:

f′(U，θ，λ)＝F′(U，θ)+λb＝0

Wherein, F ' (U, θ) be the above-mentioned dynamically meritorious power flow equation of electric power system and the power flow equation of conventional reactive power flow equation composition, λ is for characterizing the parameter that each node power changes in the electric power system, and b is the vector of each node power injection direction and size in the sign electric power system;

(3) adopt conventional continuous tide predictor corrector method, above-mentioned extended dynamic continuous tide equation is found the solution, the stability margin when obtaining the power system load variation.

2. the method for claim 1, it is characterized in that wherein said structure dynamically the process of meritorious power flow equation may further comprise the steps:

(1) the meritorious power flow equation of establishing electric power system is:

P _Gi-P _Di-P _i(U，θ)＝0，i＝1，2，...，N

(2) share of establishing the imbalance power that generator is born on the node i is β _i, according to the power amount of unbalance of electric power system:

${\mathrm{\ΔP}}_{\mathrm{\Σ}}=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Gi}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_{\mathrm{Di}}-P_{\mathrm{Loss}}(U,\mathrm{\θ})$

Obtain the dynamic continuous tide equation:

P _Gi-β _iΔP _Σ-P _Di-P _i(U，θ)＝0

P wherein _Loss(U θ) is the total network loss of electric power system.

Images