CN103746366A - Electric power system recovery method based on node importance degree - Google Patents

Abstract

An electric power system recovery method based on node importance degree relates to an electric power system recovery method. At present, the evaluation of the node importance degree does not consider the topology characteristic of a network, so that a decision-making misplay is caused easily. The electric power system recovery method based on the node importance degree of the present invention comprises the following steps of calculating a topology connectivity loss regret value and a node recovery cost increase regret value; calculating the node importance degree; determining the recovery sequence of the generator sets; determining a candidate path; calculating an average importance degree and selecting a recovery path; classifying the recovered node and line into an electrified area; if a segment of line is failed during the recovery process, determining a new recovery path again; when the power supply of the generator sets are recovered, ending work, if not determining the generator sets needing to be recovered at the next step, and determining the candidate paths from the electrified area to the generator sets needing to be recovered. Accordingly to the technical scheme, the node importance degree is judged comprehensively, thereby facilitating improving the rationality of decision, realizing the rapid recovery of the system, and reducing the power-off loss.

Description

A kind of power system recovery method based on node importance degree

Technical field

The present invention relates to power system recovery method.

Background technology

The electric power system accident of having a power failure on a large scale is the serious threat that must face, and the power system recovery after research is had a power failure on a large scale has very important significance.Network reconfiguration is the important stage that after electric power system is had a power failure on a large scale, system is recovered, the main task of this one-phase is as early as possible to dead electricity plant stand power transmission, fast quick-recovery dead electricity unit, and set up a stable grid structure, for next step load restoration is laid a solid foundation.

Wei Zhibo, Liu Yan, Gu Xueping is at < < Automation of Electric Systems > > 2007,31 (1): how the < < delivering on 38-42 has studied optimization aim back bone network in network reconfiguration > > mono-literary composition take load restoration as target based on DPSO algorithm, but how research does not optimize restoration path.Zhou Yunhai, Min Yong is at < < Proceedings of the CSEE > > 2003, 23 (4): the < < delivering on 67-70 recovers to have studied the problem of optimizing restoration path in system reconfiguration optimized algorithm research > > mono-literary composition in control, and provided power transmission optimum path search algorithm, but the method need to judge important node according to expertise, and only considered that line charging electric capacity is on the impact recovering.Lin Zhenzhi, Wen Fushuan is at < < Automation of Electric Systems > > 2009, 33 (6): in restoration path optimization method > > mono-literary composition of the < < delivering on 11-15 based on Weighted Complex Networks model, will determine back bone network and optimize these two links of restoration path to combine research, utilize the betweenness index of weighing network pivotability in Weighted Complex Networks to optimize restoration path, but during evaluation node importance degree, do not consider the topological property of network, easily cause decision-making error, increase system is recovered cost, it is slow that system is recovered, loss of outage is large.

Summary of the invention

The technical assignment of the technical problem to be solved in the present invention and proposition is that prior art scheme is improved and improved, a kind of power system recovery method based on node importance degree is provided, by the importance degree of comprehensive judge node, reach minimizing system recovery time, reduce the object of loss of outage.For this reason, the present invention takes following technical scheme.

A power system recovery method based on node importance degree, is characterized in that it comprises the following steps:

1) calculate the connective loss of the topology regret value of node in electric power system, the recovery cost of node increases regret value, if wherein the connective loss of the topology of node regret value is for do not recover specified node after system blackout, the node logarithm that generation can not power transmission; The recovery cost of node increases regret value need recover the recovery cost that adjacent node increases by other path for do not recover specified node after system blackout;

2) according to topology connective loss regret value and/or recovery cost, increase regret value computing node importance degree;

3) take the available generate output that maximizes recovery system as target, determine the recovery order of each unit;

4) electric power networks is divided into live zone and dead electricity district, determines from live zone to the path candidate of the current generating set of required recovery;

5) each path candidate is calculated to average importance degree, select the path of average importance degree maximum as the restoration path of current generating set;

6) node having recovered and circuit are divided to live zone; If certain section of circuit puts into operation unsuccessfully in recovery process, return to step 4), redefine new restoration path;

7) when generating set all restores electricity, end-of-job, otherwise determine next step generating set that need to recover, and return to step 4), determine from live zone to the path candidate of required recovery generating set.Adopt stepped approach to formulate network reconfiguration strategy: the recovery order of first step optimization generating node, can be by generate output as target to maximize recovery system, and determine serial or parallel reset mode according to recovery; Second step is optimized restoration path, the path candidate of each average importance degree maximum of path selection recovers generating node, has considered the put into operation problem of failure of circuit in recovery process simultaneously.Decision node importance degree, makes node importance degree more objective comprehensively, is conducive to improve the reasonability of decision-making, and the fast quick-recovery of the system that realizes reduces loss of outage.

As the further of technique scheme improved and supplemented, the present invention also comprises following additional technical feature.

By passing judgment on node, be adjacent the importance degree of internodal topological relationship calculation node.

By passing judgment on node, be adjacent the importance degree of internodal electrical link computing node.

Importance degree using the charging capacitor of circuit as the weight calculation node of circuit.

In to generating set recovery process, first recover black startup unit, think that system provides black startup power supply; When arranging unit starting, secondly the preferential unit with warm start condition that starts is the unit that creep speed is large, if current recovery power is enough to start multiple units, adopt the mode of parallel recovery unit, otherwise adopt build down mode to recover successively each generating set.

During optimal route selection, average regret value and recovery cost to the connective loss of each path candidate calculating path topology increase average regret value, according to descending sequence of average regret value of the connective loss of topology, select the path of the connective average regret value maximum of losing of topology as restoration path, if this value is identical, by recovering cost, increase the descending sequence of average regret value again, the path that selection recovery cost increases average regret value maximum is as restoration path.

The computing formula of average importance degree is: λ _Γ=P _Γ+ μ C _Γ; Wherein λ _Γfor average importance degree, P _Γfor the average regret value of the connective loss of topology, C _Γfor recovering cost, increase average regret value, μ is less than 1 coefficient.

Beneficial effect: the comprehensive decision node importance degree of the technical program, make node importance degree more objective, be conducive to improve the reasonability of decision-making, realize the fast quick-recovery of system, reduce loss of outage.

Accompanying drawing explanation

Fig. 1 is flow chart of the present invention.

Fig. 2 is four node weighted network schematic diagrames.

Fig. 3 is the time response that fired power generating unit starts.

Tu4Shi IEEE New England 10 machine 39 node system figure.

Fig. 5 is final recovery rack.

Embodiment

Below in conjunction with Figure of description, technical scheme of the present invention is described in further detail.

As shown in Figure 1, a kind of power system recovery method based on node importance degree, comprises the following steps:

1) calculate the connective loss of the topology regret value of node in electric power system, the recovery cost of node increases regret value, if wherein the connective loss of the topology of node regret value is for do not recover specified node after system blackout, the node logarithm that generation can not power transmission; The recovery cost of node increases regret value need recover the recovery cost that adjacent node increases by other path for do not recover specified node after system blackout;

2) according to topology connective loss regret value and/or recovery cost, increase regret value computing node importance degree;

3) take the available generate output that maximizes recovery system as target, determine the recovery order of each unit;

4) electric power networks is divided into live zone and dead electricity district, determines from live zone to the path candidate of the current generating set of required recovery;

5) each path candidate is calculated to average importance degree, select the path of average importance degree maximum as the restoration path of current generating set;

6) node having recovered and circuit are divided to live zone; If certain section of circuit puts into operation unsuccessfully in recovery process, return to step 4), redefine new restoration path;

7) when generating set all restores electricity, power cut-off, otherwise determine next step generating set that need to recover, and return to step 4), determine from live zone to the path candidate of required recovery generating set.Adopt stepped approach to formulate network reconfiguration strategy: the recovery order of first step optimization generating node, can be by generate output as target to maximize recovery system, and determine serial or parallel reset mode according to recovery; Second step is optimized restoration path, the path candidate of each average importance degree maximum of path selection recovers generating node, has considered the put into operation problem of failure of circuit in recovery process simultaneously.

Wherein: node importance degree can by passing judgment on, node be adjacent internodal topological relation, node is adjacent internodal electrical link and determines; Weight using the charging capacitor of circuit as circuit.”

Node importance degree evaluation method describes in detail:

If there is interconnection l between the node i in network and node j _ij, this circuit weight is w _ij, M in connection matrix M _ij=M _ji=w _ij; Otherwise, if node i is not directly connected with node j, M _ij=M _ji=0.

Take four meshed networks shown in Fig. 2 as example, wherein w _ijfor the weight of line related, connection matrix is:

$M=\left[\begin{array}{cccc}0& w_{12}& w_{13}& 0\\ w_{21}& 0& w_{23}& 0\\ w_{31}& w_{32}& 0& w_{34}\\ 0& 0& w_{43}& 0\end{array}\right]---\left(1\right)$

In formula: w ₁₂=w ₂₁, w ₁₃=w ₃₁, w ₂₃=w ₃₂, w ₃₄=w ₄₃.

Definition logic connection matrix:

$J=\left[\begin{array}{cccc}0& T\left(w_{12}\right)& T\left(w_{13}\right)& 0\\ T\left(w_{21}\right)& 0& T\left(w_{23}\right)& 0\\ T\left(w_{31}\right)& T\left(w_{32}\right)& 0& T\left(w_{34}\right)\\ 0& 0& T\left(w_{43}\right)& 0\end{array}\right]---\left(2\right)$

In formula: T () is the true and false computing of logic, if the several non-zeros in bracket, its value is true (value is 1); If the number in bracket is zero, its value is false (value is 0).

Definition path discrimination matrix:

$J^{*}={\mathrm{<figure-callout\; id="1"\; label="J"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">J</figure-callout>}}^1\&CirclePlus;{\mathrm{<figure-callout\; id="2"\; label="J"\; filenames="121129100841.png,121129100943.png"\; state="\{\{state\}\}">J</figure-callout>}}^2\&CirclePlus;{\mathrm{<figure-callout\; id="3"\; label="J"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">J</figure-callout>}}^3\&CirclePlus;\&CenterDot;\&CenterDot;\&CenterDot;\&CirclePlus;J^L---\left(3\right)$

In formula: J ^dfor the path matrix that between any two nodes, distance is d; D is the euclidean distance between node pair using the number of lines of path process as distance; L is any two internodal ultimate ranges;

for logic OR computing.Can adopt path discrimination matrix to judge between any two nodes whether have access path, if

show to have access path between node i and node j, otherwise do not exist.As an example of Fig. 2 example, describe:

${\mathrm{<figure-callout\; id="12"\; label="J"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">J</figure-callout>}}_{12}^{*}=T\left(w_{12}\right)\&CirclePlus;T\left(w_{13}\right)\&CircleTimes;T\left(w_{23}\right)---\left(4\right)$

${\mathrm{<figure-callout\; id="14"\; label="J"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">J</figure-callout>}}_{14}^{*}=T\left(w_{13}\right)\&CircleTimes;T\left(w_{34}\right)\&CirclePlus;T\left(w_{12}\right)\&CircleTimes;T\left(w_{23}\right)\&CircleTimes;\left(w_{34}\right)---\left(5\right)$

In formula: symbol

for logic and operation, its priority of operations is higher than logic OR computing.For node to (i, j),

formula (4) represents that node is to existing two access path: l between (1,2) ₁₂and l ₁₃l ₃₂; Formula (5) represents that node is to existing two access path: l between (Isosorbide-5-Nitrae) ₁₃l ₃₄and l ₁₂l ₂₃l ₃₄.For network shown in Fig. 2, the result of calculation of formula (4) and formula (5) is all 1, and this shows to exist between corresponding node access path.

In Fig. 2, if reject node 3, node 3 and coupled circuit are rejected, corresponding to formula (1), be that the element of the row and column at node 3 places is all set to zero.Suppose that amended network path discrimination matrix is J ^*', matrix element this shows to reject after node 3, and node is to still existing access path between (1,2); after this shows to reject node 3, to there is not access path in node between (Isosorbide-5-Nitrae).For node to (i, j), ${J_{(i,j)}^{*}}^{\&prime;}={J_{\mathrm{ij}}^{*}}^{\&prime;}={J_{\mathrm{ji}}^{*}}^{\&prime;}.$

For the network topology characteristic in Power System Restoration Process, for node m to be evaluated, define two " regretting " evaluation indexes here, i.e. the regret value P of the connective loss of topology _mincrease regret value C with recovery cost _m.The regret value of the connective loss of topology is defined as:

$P_m=\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}(J_{(i,j)}^{*}-{J_{(i,j)}^{*}}^{\&prime;})---\left(6\right)$

In formula:

for the node pair set being directly connected with node m;

for the path discriminant value of former network node to (i, j),

for removing after node m, the path discriminant value of node to (i, j).In power system recovery, if there is access path between two nodes, power transmission mutually; Otherwise, just cannot power transmission between two nodes, can not recover another node by any node wherein.In the network reconfiguration stage, sometimes need first to recover non-generating node (also representing with m without loss of generality); The main purpose of doing is like this to recover other coupled node by node m.If not recovery nodes m, likely affects whole system recovery process, because some node may can only pass through node m power transmission.The node logarithm of the power transmission mutually producing is in this case the regret value of the connective loss of topology.

If not restoration evaluation node m, just likely cannot restore electricity to adjacent node; These adjacent nodes also likely restore electricity by other path, but the cost needing than restoring electricity by node m is high and have a big risk, because need to select circuit that weight is larger or longer path to obtain electric energy.Using total weight of restoration path as recovering cost, in recovery process, select the path of recovering cost minimization to carry out recovery nodes.Here adopting the recovery cost added value that mode of recovery nodes m does not cause to call recovery cost increase regret value, be expressed as:

$C_m=\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}({H_{(i,j)}^{\min }}^{\&prime;}-H_{(i,j)}^{\min })---\left(7\right)$

In formula:

be illustrated in evaluation node m around in adjacent node, while taking the mode of recovery nodes m not, can realize the node pair set of mutual power transmission.

when taking the mode of recovery nodes m not, node is to the recovery cost minimum value between (i, j), i.e. the recovery cost of shortest path. for selecting under the mode of recovery nodes m, node is to the recovery cost minimum value between (i, j).If take the not mode of recovery nodes m, reduce finding available circuit during shortest path, the shortest path therefore obtaining is in this case not less than the shortest path under the mode of recovery nodes m,

for determine in network the node of power transmission mutually between minimum restoring cost, definition recovery cost minimization matrix:

In formula:

for the recovery Cost matrix that is d of the distance between any two nodes, L' by using path the number of lines of process as ultimate range between the node of path.The total weight that it may be noted that beeline path is not necessarily minimum.The loss that circuit weight is brought more greatly and risk generally also can be larger, so can select the path of total weight minimum as the right recovery cost minimal path of node.Symbol ⊙, for getting minimum operation symbol, gets minimum value to the element of same position in all matrixes.

Take Fig. 2 as example, node is respectively the minimum restoring cost of (1,2) and (Isosorbide-5-Nitrae): $H_{\left(\mathrm{1,2}\right)}^{\min }=\min ({\mathrm{<figure-callout\; id="12"\; label="w"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">w</figure-callout>}}_{12},{\mathrm{<figure-callout\; id="13"\; label="w"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">w</figure-callout>}}_{13}+w_{23})=\min \left(\mathrm{0.4,0.3}\right)=0.3$ $H_{\left(\mathrm{1,4}\right)}^{\min }=\min ({\mathrm{<figure-callout\; id="13"\; label="w"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">w</figure-callout>}}_{13}+w_{34},{\mathrm{<figure-callout\; id="12"\; label="w"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">w</figure-callout>}}_{12}+{\mathrm{<figure-callout\; id="23"\; label="w"\; filenames="121129100943.png,121129101023.png"\; state="\{\{state\}\}">w</figure-callout>}}_{23}+w_{34})=\min \left(\mathrm{0.4,0.9}\right)=0.4.$ When the importance degree of evaluation node 3, because reject the node that can realize mutual power transmission after node 3 to only having (1,2), therefore

the recovery cost that can be tried to achieve node 3 by formula (7) increases regret value C ₃=0.4-0.3=0.1.This value is greater than zero explanation and takes the not strategy of recovery nodes 3 will cause recovery cost to increase, and the cost increasing is a standard of evaluation node 3 importance degrees.Increase cost more, illustrate that this node more should be restored, the importance degree of node is also just larger.

The regret value of the connective loss of the topology of evaluation node importance degree and recovery cost increase this two aspect of regret value are integrated to defined node importance degree comprehensive evaluation index.The importance degree comprehensive evaluation index of node m is:

${\mathrm{\&beta;}}_m=\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}(J_{(i,j)}^{*}-{J_{(i,j)}^{*}}^{\&prime;})+u\&CenterDot;\underset{(i,j)\&Element;{\tilde{V}}_m}{\mathrm{\&Sigma;}}({H_{(i,j)}^{\min }}^{\&prime;}-H_{(i,j)}^{\min })---\left(9\right)$

In formula: u is proportionality coefficient, for adjusting P _mand C _mrelative Link Importance.

The following describes substep network reconfiguration strategy:

The first step: in the network reconfiguration stage, determine generating set recovery order take the available generate output that maximizes recovery system as target.At the available generate output P of network reconfiguration stage whole system _totalfor all generating sets of the system sum of exerting oneself deducts unit starting power demand:

$P_{\mathrm{total}}=\underset{n=1}{\overset{N_{\mathrm{sb}}}{\mathrm{\&Sigma;}}}{P}_n^{\mathrm{sb}}+\underset{k=1}{\overset{N_{\mathrm{smb}}}{\mathrm{\&Sigma;}}}{P}_k^{\mathrm{snb}}-\underset{k=1}{\overset{N_{\mathrm{snb}}}{\mathrm{\&Sigma;}}}{P}_k^{\mathrm{snbj}}---\left(10\right)$

In formula: N _sband N _snbbe respectively black startup unit number and non-black startup unit number that the network reconfiguration stage starts;

with

be respectively the black startup unit of n platform and k Tai Feihei and start the generated output that unit recovered in this stage;

be that k Tai Feihei starts the required power of unit starting.The start-up course of fired power generating unit is comparatively complicated, adopts simplified model as shown in Figure 3 here.

At T ₁moment generating set obtains startup power supply to start to start; Elapsed boot time T _gpafter, generating set starts power output, progressively increases and exerts oneself until T subsequently by certain creep speed ₂moment reaches unit maximum output P _max.

When determining the recovery order of generating set, should be according to the Startup time of the non-black startup unit of each machine unit characteristic reasonable arrangement, to reach the target of maximization network available generate output that reconstruction stage is recovered.Determine that generating set optimized database restore order can be described as on mathematics:

$\max P_{\mathrm{total}}=\underset{n=1}{\overset{N_{\mathrm{sb}}}{\mathrm{\&Sigma;}}}{P}_n^{\mathrm{sb}}+\underset{k=1}{\overset{N_{\mathrm{snb}}}{\mathrm{\&Sigma;}}}{P}_k^{\mathrm{snb}}-\underset{k=1}{\overset{N_{\mathrm{snb}}}{\mathrm{\&Sigma;}}}{P}_k^{\mathrm{snbj}}---\left(11\right)$

The fast quick-recovery of system is the main target in network reconfiguration stage.Due to generating set, there is the restrictions such as different starting powers, start-up time, creep speed, maximum heat startup time limit, therefore study rational generating set recovery order significant to maximize the available generate output of recovering.Therefore, for each generating set, can adopt following recovery policy:

(1) first recover black startup unit, to provide black startup power supply for system;

(2) arranging during unit starting, the unit that overall arrangement possesses warm start condition preferentially starts, and strives for allowing unit warm start as much as possible (unit warm start as less in the warm start time limit of giving priority in arranging for) as far as possible;

(3), for realizing the fast quick-recovery of system, considering under the prerequisite of (2) individual strategy the unit that priority restores creep speed is larger;

(4) if the load power recovered of certain is not enough to recover to have maximum heat starts the unit in time limit, can consider first to recover other type unit in moment, but will make as far as possible this unit subsequently can warm start;

(5) if current recovery power is enough to start multiple units, adopt parallel recovery plan, otherwise adopt build down strategy to recover successively each generating set.

Second step: how determining after the recovery order of generating set, be that the suitable restoration path of these machine group selections is next step problem that need to solve.Here with the average importance degree in path, be target selection restoration path to the maximum.The average importance degree of definition restoration path is:

${\mathrm{\&lambda;}}_{\mathrm{\&Gamma;}}=\underset{m={\mathrm{\&Psi;}}_{\mathrm{\&Gamma;}}}{\mathrm{\&Sigma;}}[\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}(J_{(i,j)}^{*}-{J_{(i,j)}^{*}}^{\&prime;})+u\&CenterDot;\underset{(i,j)\&Element;{\tilde{V}}_m}{\mathrm{\&Sigma;}}({H_{(i,j)}^{\min }}^{\&prime;}-H_{(i,j)}^{\min })]/N_{\mathrm{\&Gamma;}}---\left(12\right)$

In formula: Ψ _Γfor the set of the not recovery nodes of restoration path Γ process; N _Γfor the not recovery nodes number of path Γ process.The implication of formula (12) is to ask for the mean value of restoration path through the importance degree of node.In recovery process, should recover important node as far as possible, also need to consider operating time that circuit recovers to bring, the factors such as risk put into operation simultaneously.Optimal mode is the important node more than the restoration path of selection can be tried one's best with few operation recovery of trying one's best.Select the longer path may be through more more important node, but action required number of times also can increase simultaneously, thereby strengthen corresponding risk.Therefore, here using path average importance degree as the standard of choosing restoration path.

The optimization aim of determining restoration path is:

$\max {\mathrm{\&lambda;}}_{\mathrm{\&Gamma;}}=\underset{m\&Element;{\mathrm{\&Psi;}}_{\mathrm{\&Gamma;}}}{\mathrm{\&Sigma;}}[\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}(J_{(i,j)}^{*}-{J_{(i,j)}^{*}}^{\&prime;})+u\&CenterDot;\underset{(i,j)\&Element;{\tilde{V}}_m}{\mathrm{\&Sigma;}}({H_{(i,j)}^{\min }}^{\&prime;}-H_{(i,j)}^{\min })]/N_{\mathrm{\&Gamma;}}---\left(13\right)$

Alternative restoration path is the restoration path of the generating node that recovers to needs of charging zone from having recovered.Calculate the average importance degree λ of alternative each path candidate _Γ, select afterwards the restoration path of average importance degree maximum to next one generating node power transmission.

The average regret value of the connective loss of topology of definition restoration path is:

$P_{\mathrm{\&Gamma;}}=\underset{m\&Element;{\mathrm{\&Psi;}}_{\mathrm{\&Gamma;}}}{\mathrm{\&Sigma;}}\left[\underset{(i,j)\&Element;{\hat{V}}_m}{\mathrm{\&Sigma;}}(J_{(i,j)}^{*}-{J_{(i,j)}^{*}}^{\&prime;})\right]/N_{\mathrm{\&Gamma;}}---\left(14\right)$

Definition recovery cost increases average regret value and is:

$C_{\mathrm{\&Gamma;}}=\underset{m\&Element;{\mathrm{\&Psi;}}_{\mathrm{\&Gamma;}}}{\mathrm{\&Sigma;}}\left[\underset{(i,j)\&Element;{\tilde{V}}_m}{\mathrm{\&Sigma;}}({H_{(i,j)}^{\min }}^{\&prime;}-H_{(i,j)}^{\min })\right]/N_{\mathrm{\&Gamma;}}---\left(15\right)$

Select the average importance degree evaluation method in path with priority: for each path candidate, first calculate the average regret value P of the connective loss of this path topology _Γand recovery cost increases average regret value C _Γ, then according to descending sequence of average regret value of the connective loss of topology, if this value is identical, by recovering cost, increase the descending sequence of average regret value again, finally selecting optimal path is restoration path.If put into operation unsuccessfully because a variety of causes causes certain section of circuit on restoration path in recovery process, can again by above-mentioned steps, find path candidate and calculate corresponding path regret value having restored electricity at present on node and line tower foundation, and then selecting the path of average importance degree maximum as new restoration path.

Illustrate:

As shown in Figure 4, New England's 10 machine 39 node systems.Node 33 is black startup unit place node.Suppose: after having a power failure on a large scale, black startup unit starts immediately, for system recovers to provide initial power; Current recovery unit starting is complete, starts to restart next unit after power output.Owing to being mainly in order to guarantee system stability in the input of this stage load, when formulating generating set recovery order, put aside load Input.Recovery order and the recovery moment of each generating node of formulating according to the each generating set recovery policy described in summary of the invention are as shown in table 1.

The each generating of table 1 node recovery order and moment

The each generating of table 2 node restoration path

Having determined after each generating set recovery order, choose the path of average importance degree maximum as restoration path.According to selecting the restoration path that obtains each generating set with the average importance degree evaluation method in path of priority as shown in table 2 described in summary of the invention.The choosing method of explanation restoration path as an example of recovery nodes 39 example, the path candidate of recovery nodes 39 has 3, and the average importance degree of these 3 path candidates is more as shown in table 3.Article 3, the average regret value P of the connective loss of the topology of path candidate _Γbe 0, now cannot distinguish quality, therefore comparative costs increases average regret value C again _Γ, finally choose C _Γ maximum Article 3 path candidate (5 → 8 → 9 → 39) recovery nodes 39.As can be seen from Table 2: when recovery nodes 37, node 35 and node 31, the P in selected path _Γand C _Γbe 0, this is because be all restored with the node of these node direct neighbors before this.(in figure, solid line represents to recover circuit, and the node that its two ends connect is recovery nodes finally to have obtained recovery rack as shown in Figure 5; Dotted line represents not recover circuit).

The average importance degree of path candidate of table 3 recovery nodes 39

In recovery process, may run into the put into operation situation of failure of circuit.Here take recovery nodes 32 as example, restoration path is 2 → 3 → 4 → 5 → 6 → 11 → 10 → 32, suppose in recovery process, the circuit that connects these two nodes during from node 6 recovery nodes 11 puts into operation unsuccessfully, now, using the node that recovered and circuit as charging zone, reselect the path of recovery nodes 32.Now path candidate has 4, and the average importance degree of path candidate is more as shown in table 4, through alternative the 4th article of path candidate (4 → 14 → 13 → 10 → 32) recovery nodes 32.

The average importance degree of path candidate (circuit 6-11 puts into operation after failure) of table 4 recovery nodes 32

Claims (7)

1. the power system recovery method based on node importance degree, is characterized in that it comprises the following steps:

1) calculate the connective loss of the topology regret value of node in electric power system, the recovery cost of node increases regret value, if wherein the connective loss of the topology of node regret value is for do not recover specified node after system blackout, the node logarithm that generation can not power transmission; The recovery cost of node increases regret value need recover the recovery cost that adjacent node increases by other path for do not recover specified node after system blackout;

2) according to topology connective loss regret value and/or recovery cost, increase regret value computing node importance degree;

3) take the available generate output that maximizes recovery system as target, determine the recovery order of each unit;

4) electric power networks is divided into live zone and dead electricity district, determines from live zone to the path candidate of the current generating set of required recovery;

5) each path candidate is calculated to average importance degree, select the path of average importance degree maximum as the restoration path of current generating set;

6) node having recovered and circuit are divided to live zone; If certain section of circuit puts into operation unsuccessfully in recovery process, return to step 4), redefine new restoration path;

7) when generating set all restores electricity, end-of-job, otherwise determine next step generating set that need to recover, and return to step 4), determine from live zone to the path candidate of required recovery generating set.

2. a kind of power system recovery method based on node importance degree according to claim 1, is characterized in that: the importance degree that is adjacent internodal topological relationship calculation node by passing judgment on node.

3. a kind of power system recovery method based on node importance degree according to claim 1, is characterized in that: the importance degree that is adjacent internodal electrical link computing node by passing judgment on node.

4. a kind of power system recovery method based on node importance degree according to claim 1, is characterized in that: the importance degree using the charging capacitor of circuit as the weight calculation node of circuit.

5. according to a kind of power system recovery method based on node importance degree described in the arbitrary claim of claim 1-4, it is characterized in that: in to generating set recovery process, first recover black startup unit, think that system provides black startup power supply; When arranging unit starting, secondly the preferential unit with warm start condition that starts is the unit that creep speed is large, if current recovery power is enough to start multiple units, adopt the mode of parallel recovery unit, otherwise adopt build down mode to recover successively each generating set.

6. a kind of power system recovery method based on node importance degree according to claim 1, it is characterized in that: during optimal route selection, average regret value and recovery cost to the connective loss of each path candidate calculating path topology increase average regret value, according to descending sequence of average regret value of the connective loss of topology, select the path of the connective average regret value maximum of losing of topology as restoration path, if this value is identical, by recovering cost, increase the descending sequence of average regret value again, the path that selection recovery cost increases average regret value maximum is as restoration path.

7. a kind of power system recovery method based on node importance degree according to claim 1, is characterized in that: the computing formula of average importance degree is: λ _Γ=P _Γ+ μ C _Γ; Wherein λ _Γfor average importance degree, P _Γfor the average regret value of the connective loss of topology, C _Γfor recovering cost, increase average regret value, μ is less than 1 coefficient.

Images