CN103904643A - Direct current load flow computing method with network loss considered - Google Patents

Abstract

The invention provides a direct current load flow computing method with network loss considered. The method comprises the following steps of collecting data, computing node injecting active power, listing a direct current load flow equation and solving, computing electric grid branch active power loss, judging whether an iterative operation is conducted or not, and computing network loss equivalent load. The method is improved based on a standard direct current load flow computing method, the active power loss generated by branch resistance is considered, the node injecting active power is corrected by the adoption of the network loss equivalent load, and a direct current load flow computing result with the network loss considered is acquired through the iterative operation method. By means of the direct current load flow computing method with the network loss considered, the problems of the unbalanced active power and unreasonable active power distribution due to ignoring the network loss in standard direct current load flow computing are solved, and the direct current load flow computing result more approaches the real load flow result.

Description

A kind of DC power flow computational methods of considering network loss

Technical field

The present invention relates to a kind of electric power system tide computational methods, specifically relate to a kind of DC power flow computational methods of considering network loss.

Background technology

It is the basis of carrying out electric power system computational analysis that trend is calculated, and the Mathematical Modeling that conventional Load Flow calculates is one group of nonlinear equation, and the in the situation that in the quality of data, good and trend being calculated normal convergence, its trend result is general enough accurate.But due to the characteristic of nonlinear equation itself, often there will be trend to calculate situation about not restraining, the generation of this situation generally has two kinds of reasons: power flow equation has solution but do not obtain convergence solution without solution and power flow equation.In addition, the amount of calculation that conventional Load Flow calculates is larger, carries out a large amount of occasions of calculating at need, and the time of expending is also more.

DC power flow calculates and has adopted approximate processing, thinks that the each node voltage of electric power system all gets rated voltage, ignores branch road resistance and parallel branch over the ground, does not consider the relation between idle and voltage, only the problem of the meritorious distribution of research.The Mathematical Modeling that DC power flow calculates is one group of linear equation, and amount of calculation is less, and the convergence problem that does not exist conventional Load Flow to calculate.DC power flow calculates and is often applied to the less demanding occasion of computational accuracy, such as the forecast accident screening in active power adjustment, electrical network Real Time Safety Analysis that in Electric Power Network Planning, design and running, mode arranges, etc.

Due to the active loss of not considering that branch road resistance produces, so the DC power flow computational methods (being standard direct current tidal current computing method) that generally adopt at present exist following shortcoming:

(1) active balance situation and actual conditions do not meet;

No matter analyze the whole network generating and load active balance, or the flow of power between analyzed area electrical network, active loss is all the factor can not be ignored.While analyzing active balance based on standard direct current calculation of tidal current, need computational analysis personnel according to self experience estimation network loss size, and corrected Calculation result, therefore need computational analysis personnel to there is higher specialized capability, more rich practical experience.

(2) meritorious distribution may differ larger with actual trend.

Because generating and load power difference that in esse network loss produces are born by balancing machine completely, in the time ignoring active loss, generating and load are not carried out in the situation of active balance processing, in standard direct current calculation of tidal current balancing machine exert oneself obviously unreasonable, and balancing machine near meritorious distribution also obviously unreasonable., can adopt to reduce generated output or increase the measure of loading and avoid near the phenomenon of the meritorious irrational distribution of balancing machine, but can bring the on the low side or higher phenomenon of wider active power.

Summary of the invention

For the deficiencies in the prior art, the present invention proposes a kind of DC power flow computational methods of considering network loss.In the improved standard direct current of the one tidal current computing method that the method provides, consider the active loss that branch road resistance produces, and utilize the duty values such as network loss to inject active power to node and revise, try to achieve the DC power flow result of calculation of considering network loss by the method for iterative computation.The method can solve standard DC power flow calculate in owing to ignoring the uneven and meritorious more irrational problem that distributes of active power that network loss causes, make DC power flow result of calculation more approach actual trend result.

In order to realize foregoing invention object, the present invention takes following technical scheme:

DC power flow computational methods of considering network loss, the method comprises:

(1) collect data;

(2) computing node injects active power;

(3) list DC power flow equation and solve;

It is characterized in that, described injection active power comprises the duty values such as generator output, load and network loss; Said method comprising the steps of:

(4) calculate grid branch active loss;

(5) judge whether interative computation;

(6) calculate the duty values such as network loss.

Preferably, described step (1) comprising:

Obtain the parameter of electrical network to be calculated, ask the imaginary part B of node admittance matrix by DC power flow computational methods, i.e. susceptance matrix, and this susceptance matrix is carried out to LU decomposition; Make current iteration number of times g=1, and initialization branch road active loss vector P _loss.0=0, and the duty value such as network loss vector P _equal.0=0.

Preferably, the meritorious vector power of node injection is expressed as P _g=P _gen-P _load-P _equal.g-1, wherein, P _genfor generator output vector, P _loadfor the vector of loading, P _equal.g-1the duty value vectors such as the network loss obtaining for a front iterative computation.

Preferably, described step (3) comprising:

Utilize the LU decomposition result of the susceptance matrix B that step (1) obtains and node that described step (2) obtains to inject meritorious vector power P _g, list described DC power flow equation, solve and obtain node voltage phase angle vector θ _g; This equation is expressed as: P _g=B θ _g.

Preferably, described step (4) comprising: utilize the reactance of the each branch road of electrical network and the node voltage phase angle that described step (3) obtains, calculate grid branch active power vector P _branch.g, and calculate grid branch active loss vector P _loss.g.

Preferably, described step (5) comprising: the P that this iterative step (4) is obtained _loss.gthe P obtaining with last iteration step (4) _loss.g-1difference as branch road active loss incremental vector, this incremental vector is expressed as: Δ P _loss.g=P _loss.g-P _loss.g-1; Judge inequality | Δ P _loss.g|≤ε P _loss.gwhether set up, wherein, ε is not more than 0.1 less coefficient; If this inequality is set up, or iterations reaches the upper limit, finishing iteration process, the θ now obtaining _gand P _branch.gbe final DC power flow result; If this inequality is false and iterations does not reach the upper limit, continue step (6).

Preferably, described step (6) comprising:

The grid branch active loss vector P obtaining according to described step (4) _loss.gcalculate the duty value vector P such as network loss _equal.g; P _equal.gin the equivalent load of the network loss computing formula of arbitrary node i be:

wherein, P _loss.ij.gfor the active power loss of branch road i-j, i.e. grid branch active loss vector P _loss.gin element; Make current iteration number of times g add 1, and proceed to step (2), enter new round iterative computation.

Compared with the prior art, beneficial effect of the present invention is:

Method of the present invention is carried out the research of active loss problem in DC power flow calculating under actual electric network data, standard direct current tidal current computing method is improved, for electric power system computational analysis provides a kind of DC power flow computational methods of considering network loss, arrange with aspects such as trend adjustment all significant to the mode in power system planning, design and running.

The present invention has overcome standard DC power flow and has calculated the defect of bringing owing to ignoring network loss, can solve in the calculating of standard DC power flow owing to ignoring network loss the active power imbalance causing and the more irrational problem that distributes of gaining merit, make DC power flow result of calculation more approach actual trend result, contribute to improve the accuracy of every electric power system computational analysis of carrying out based on DC power flow result.

The present invention, in improving DC power flow result of calculation accuracy, has retained the fast advantage of standard DC power flow computational speed.The inventive method only need be carried out one time matrix decomposition operation, then utilizes the method for revising node injection active power to carry out iteration, and amount of calculation there is no obvious increase compared with standard direct current tidal current computing method, thereby is still applicable to the occasion that needs calculate fast.

Brief description of the drawings

Fig. 1 is the DC power flow calculation flow chart according to method of the present invention.

Embodiment

DC power flow computational methods of the present invention, consider the active loss that branch road resistance produces, and utilize the duty values such as network loss to inject active power to node and revise, and try to achieve the DC power flow result of calculation of considering network loss by the method for iterative computation.

The flow chart that the DC power flow of the present embodiment calculates as shown in Figure 1, specifically comprises the steps:

(1) collect data;

(2) computing node injects active power;

(3) list DC power flow equation and solve;

(4) calculate grid branch active loss;

(5) judge whether interative computation;

(6) calculate the duty values such as network loss.

In standard DC flow model, approximate think that the each node voltage of electric power system all gets rated voltage, i.e. V=1(perunit value, lower with); Consider branch road i-j both end voltage phase angle difference θ _ijless, approximate think sin θ _ij=θ _ij, cos θ _ij=1; Because the branch road resistance ratio reactance of high-voltage fence is little a lot, so ignore branch road resistance, i.e. r _ij=0, only consider branch road reactance x _ij; Ignore all nodes parallel branch over the ground.Standard direct current power flow equation is

P＝Bθ

In formula: P=P _gen-P _loadfor node injects meritorious vector power, P _genfor generator output vector, P _loadfor the vector of loading, the injecting power P of node i _i=P _gen.i-P _load.i, P _gen.iand P _load.ibe respectively the generator output and the load value that are positioned at node i.θ is node voltage phase angle vector, and the phase angle difference of node i and reference point (generally select balance node as a reference point) is θ _i0=θ _i-θ ₀=θ _i.B is the susceptance matrix of network,

represent the arbitrary node j being connected with node i by branch road.

The active power of branch road i-j is:

$P_{\mathrm{ij}}=\frac{{\mathrm{\θ}}_i-{\mathrm{\θ}}_j}{x_{\mathrm{ij}}}$

If at this moment consider the resistance r of this branch road _ij, in branch road, can produce active power loss:

$P_{\mathrm{loss}.\mathrm{ij}}=P_{\mathrm{ij}}^2{r}_{\mathrm{ij}}$

One of assumed condition of considering DC power flow equation P=B θ is to ignore branch road resistance, and therefore, branch road active loss can be by representing at duty values such as branch road two ends increases.The equivalent load of network loss of node i is calculated as follows:

$P_{\mathrm{equal}.i}=\underset{j\∈i,j\≠i}{\mathrm{\Σ}}\frac{P_{\mathrm{loss}.\mathrm{ij}}}{2}$

For all branches, active power vector is P _branch, active loss vector is P _loss.For whole nodes, the duty value vectors such as network loss are P _equal.

Recalculate the meritorious vector power of node injection as follows:

P＝P _gen-P _load-P _equal

Can calculate active loss vector P by again solving DC power flow equation P=B θ _losswith the duty value such as network loss vector P _equal, and computing node injects meritorious vector power P again.Therefore, can by as shown in Figure 1 repeatedly loop iterative computation.

Current iteration number of times is designated as g, P _equal.g-1the duty value vectors such as the network loss calculating for last iteration, to inject meritorious vector power be P to this iteration node _g=P _gen-P _load-P _equal.g-1, solving equation P _g=B θ _gobtain the node voltage phase angle vector θ of this iteration _g, and then calculate branch road active power vector P _branch.gand branch road active loss vector P _loss.g.P _loss.gthe branch road active loss vector P calculating with last iteration _loss.g-1difference be branch road active loss incremental vector Δ P _loss.g=P _loss.g-P _loss.g-1.If | Δ P _loss.g|≤ε P _loss.g, wherein ε is a less coefficient (between desirable 0.001 to 0.1), finishing iteration process, the θ now obtaining _gand P _branch.gbe final DC power flow result; Otherwise calculate the duty value vector P such as network loss _equal.g, and make current iteration number of times g add 1, then carry out new round iteration.If iterations reaches the upper limit, i.e. g=g _max, and this inequality is still false, and also needs finishing iteration process.3 to 5 left and right of general iteration can meet | Δ P _loss.g|≤ε P _loss.g, therefore g _maxdesirable 10 left and right.

Repeatedly solve P for improving _g=B θ _gcomputational efficiency, can before iteration, first calculate susceptance matrix B and carry out LU decomposition, in each iteration, solve θ like this _grequired amount of calculation will be very little, can obviously improve the speed of solving.In addition, before iteration for the first time, need initialization branch road active loss vector P _loss.0=0, and the duty value such as network loss vector P _equal.0=0.

The present embodiment is applied in certain actual electric network, and has obtained Expected Results.

Method of the present invention is carried out the research of active loss problem in DC power flow calculating under actual electric network data, has proposed a kind of DC power flow computational methods of considering network loss.The method is improved standard direct current tidal current computing method, considers the active loss that branch road resistance produces, and utilizes the duty values such as network loss to inject active power to node and revise, and tries to achieve the DC power flow result of calculation of considering network loss by the method for iterative computation.The method can solve standard DC power flow calculate in owing to ignoring the uneven and meritorious more irrational problem that distributes of active power that network loss causes, make DC power flow result of calculation more approach actual trend result.The method arranges with aspects such as trend adjustment all significant to the mode in power system planning, design and running, have popularizing application prospect widely.

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 the present invention is had been described in detail with reference to above-described embodiment, those of ordinary skill in the field are to be understood that: still can modify or be equal to replacement the specific embodiment of the present invention, and do not depart from any amendment of spirit and scope of the invention or be equal to replacement, it all should be encompassed in the middle of claim scope of the present invention.

Claims (7)

1. DC power flow computational methods of considering network loss, the method comprises:

(1) collect data;

(2) computing node injects active power;

(3) list DC power flow equation and solve;

It is characterized in that, described injection active power comprises the duty values such as generator output, load and network loss; Said method comprising the steps of:

(4) calculate grid branch active loss;

(5) judge whether interative computation;

(6) calculate the duty values such as network loss.

2. the method for claim 1, is characterized in that, described step (1) comprising:

Obtain the parameter of electrical network to be calculated, ask the imaginary part B of node admittance matrix by DC power flow computational methods, i.e. susceptance matrix, and this susceptance matrix is carried out to LU decomposition; Make current iteration number of times g=1, and initialization branch road active loss vector P _loss.0=0, and the duty value such as network loss vector P _equal.0=0.

3. the method for claim 1, is characterized in that, node injects meritorious vector power and is expressed as P _g=P _gen-P _load-P _equal.g-1, wherein, P _genfor generator output vector, P _loadfor the vector of loading, P _equal.g-1the duty value vectors such as the network loss obtaining for a front iterative computation.

4. the method for claim 1, is characterized in that, described step (3) comprising:

Utilize the LU decomposition result of the susceptance matrix B that step (1) obtains and node that described step (2) obtains to inject meritorious vector power P _g, list described DC power flow equation, solve and obtain node voltage phase angle vector θ _g; This equation is expressed as: P _g=B θ _g.

5. the method for claim 1, is characterized in that, described step (4) comprising: utilize the reactance of the each branch road of electrical network and the node voltage phase angle that described step (3) obtains, calculate grid branch active power vector P _branch.g, and calculate grid branch active loss vector P _loss.g.

6. the method for claim 1, is characterized in that, described step (5) comprising: the P that this iterative step (4) is obtained _loss.gthe P obtaining with last iteration step (4) _loss.g-1difference as branch road active loss incremental vector, this incremental vector is expressed as: Δ P _loss.g=P _loss.g-P _loss.g-1; Judge inequality | Δ P _loss.g|≤ε P _loss.gwhether set up, wherein, ε is not more than 0.1 less coefficient; If this inequality is set up, or iterations reaches the upper limit, finishing iteration process, the θ now obtaining _gand P _branch.gbe final DC power flow result; If this inequality is false and iterations does not reach the upper limit, continue step (6).

7. the method for claim 1, is characterized in that, described step (6) comprising:

The grid branch active loss vector P obtaining according to described step (4) _loss.gcalculate the duty value vector P such as network loss _equal.g; P _equal.gin the equivalent load of the network loss computing formula of arbitrary node i be:

wherein, P _loss.ij.gfor the active power loss of branch road i-j, i.e. grid branch active loss vector P _loss.gin element; Make current iteration number of times g add 1, and proceed to step (2), enter new round iterative computation.

Images