CN105871594A - Method for calculating important degrees of nodes of power communication network - Google Patents

Abstract

The invention discloses a method for calculating important degrees of nodes of a power communication network, belonging to the technical field of power communication networks. The method comprises the following steps of: constructing a power communication network topological model according to a practical network architecture; calculating the important degrees of operation services of the power communication network; quantifying service types and important degrees thereof born on the link of the power communication network, and introducing quantization values in an adjacent matrix for calculating the natural connectivity in the form of weight values, such that a weighed adjacent matrix is obtained; calculating the weighed natural connectivity of the power communication network; and obtaining the important degrees of various nodes by comparing change of the weighed natural connectivity of the network topological sub-graph obtained after various nodes are sequentially invalid and the weighed natural connectivity of the original graph. According to the method for calculating the important degrees of the nodes provided by the invention, the network topological characteristics and the power service characteristics of the nodes are comprehensively considered; therefore, the method has the practical significance; and furthermore, because the calculation form of the algorithm is concise, the method is applied to the power communication network having a relatively large node scale.

Description

A kind of power telecom network pitch point importance computational methods

Technical field

The present invention relates to field of power communication, be specifically related to a kind of power telecom network pitch point importance Computational methods.

Background technology

Power telecom network is the important component part of power system, and its reliability and safety are to electricity The safe efficient operation of Force system is significant.Communication node is the core of powerline network Heart element, carries upper and lower business and the function of business forwarding.Communication node is in real network In position different, its importance there is also significantly difference, it is generally the case that network core The node importance of heart position is higher.If server node is destroyed, network can Reduce immediately by property, even result in wide area communications and interrupt.Communication net node importance degree is carried out Identifying, the important node in excavation power telecom network is laid equal stress on and is a little protected, and can improve whole The reliability of network and survivability, and then ensure the reliable and stable operation of electrical network.Study joint simultaneously Point is for the influence degree of network virtual condition, it is also possible to carry for reliability of electric force communication network analysis Reference for more practical significance.

Along with the construction of intelligent grid, power communication web frame more sophisticated, power business kind It is on the increase with quantity.Power telecom network interior joint significance level is not only connected with complicated network Topological structure is directly related, and the safe class of the business undertaken, service quality and number of services Closely related.The pitch point importance computational methods of fused business and network topology are generally by electricity at present Power business importance degree is as calculating the important parameter that point is weighed, by network cohesion degree by node in industry The importance of business layer and physical topology layer carries out merging to realize the calculating of pitch point importance.The party Method uses analytic hierarchy process (AHP) in the value of typical services importance degree, cannot to subjective uncertainty Controlling, therefore there is certain deficiency in the method, receives at every minor node in algorithm realizes simultaneously Need after contracting operation again to calculate shortest path matrix with Floyd algorithm, more at number of nodes In the case of, algorithm is difficult to meet practicality demand.

The assessment of pitch point importance comes from the destructiveness that network is caused by node failure, Ke Yitong This destruction result is weighed in the change in excess vol network redundancy path, and this index can be passed through The calculating of Natural communication degree represents.Natural communication degree is estimated as a kind of Survivabilities of Networks, logical Crossing and calculate the weighted sum of different length closed loop number in network, in embodiment network, alternative route is superfluous Yu Xing, by the redundancy reflection network performance of alternate routing in network.Natural communication degree is counting Learn the average characteristics root being formally represented as a kind of specific form, can be special from network adjacent matrix Levy spectrum directly to derive, therefore there is clear and definite physical significance and succinct mathematical form.Due to certainly So the computation complexity of degree of communication is relatively low, can be convenient when interstitial content is more in a network To result of calculation.

Summary of the invention

For the deficiencies in the prior art, it is desirable to provide a kind of power telecom network node is important Degree computational methods, the method use weighting Natural communication degree comprehensively embody node topological property and Power business characteristic, it is possible to objectively calculate pitch point importance and sort, has substantially Low complex degree feature.

To achieve these goals, the present invention adopts the following technical scheme that

A kind of power telecom network pitch point importance computational methods, comprise the steps:

S1 sets up the network topology model of power telecom network according to the actual framework of power telecom network, And build corresponding adjacency matrix；

The business importance degree of the S2 described power telecom network of calculating:

2.1) some characteristic index constitutive characteristic index set is taken, according to all kinds of in power telecom network The business different requirements to selected characteristic index, by the weight of the services sets of all kinds of service constitution The property wanted is mapped to importance value sequence；

2.2) calculating all kinds of business by importance value sequence concentrates each feature to refer in characteristic index Relatively important value matrix under Biao；

2.3) every class business relatively important value matrix under all characteristic indexs is sued for peace, Obtain the comprehensive relatively important value matrix that every class business is corresponding；

2.4) to every class business in step 2.3) row of comprehensive relatively important value matrix that obtains Vector element is sued for peace, and obtains every class business and combines other all kinds business of service set Close relative Link Importance sum；

2.5) to step 2.4) in calculated every class business to service set other own The comprehensive relative Link Importance sum of kind business is normalized, and be mapped to interval [X, 1], on, the importance degree of every class business, wherein 0 ＜ X ＜ 1 are obtained；

The number of all kinds of business run on link is calculated its business in step S2 by S3 The product of importance degree and as link quantized value, the business of link carrying is quantified, will Described link quantized value is incorporated among adjacency matrix with the form of weights, obtains weighted adjacent square Battle array；

S4 calculates the weighting Natural communication degree of power telecom network；

S5 is by the weighting nature of gained network topology subgraph after comparing each node and losing efficacy successively Degree of communication and the weighting Natural communication degree of artwork, it is thus achieved that pitch point importance.

It should be noted that described step S1 is specific as follows:

According to the actual framework of described power telecom network, build corresponding network topology model, Described network topology model includes that node and link, described node number are N, described link Number is M, and the most described network topology model figure G is described as follows:

G=(V, E)；

Wherein V={v₁, v₂..., v_NRepresenting node set, the most each node represents power telecom network In electric power communication device, therefore V is electric power communication device set；E={e₁, e₂..., e_MTable Show that the set on limit, described limit represent the link between electric power communication device；

Corresponding adjacency matrix A (G)=[a_ijElement a in]_ijIt is defined as follows:

It should be noted that described step 2.1) in, note characteristic index collection is Q={q_l(l=1,2 ..., L), L is characterized characteristic index total number in index set；According to business pair The different requirements of selected characteristic index, by services sets B={b_k(k=1,2 ..., K) important Property is mapped to importance value sequence { s_kl∈ 1,2 ..., S_l, wherein K is service set class of business Total number；s_klExpression business b_kIn characteristic index q_lUnder importance value, wherein S_lBy services sets In characteristic index q_lUnder difference degree determine, its numeric representation services sets B is to characteristic index q_l There is S_lPlant different requirements.

It should be noted that described step 2.2) in, by importance value sequence { s_klCalculate respectively Class business b_k(k=1,2 ..., K) in each characteristic index q_l(l=1,2 ..., L, L are that selected feature refers to Mark total number) under relatively important value matrixAs follows:

Business b_kIn characteristic index q_lUnder relatively important value matrixFollowing formula calculates:

$c_{kj}^{\left(q_l\right)}=\left\{\begin{array}{cc}1& s_{kl}/s_{jl}\&GreaterEqual;S_l\\ .& .\\ .& .\\ .& .\\ 2/S_l& 2\&le;s_{kl}/s_{jl}<3\\ 1/S_l& 1\&le;s_{kl}/s_{jl}<2\\ 0& s_{kl}/s_{jl}<1\end{array}\right.;$

Wherein,Represent in characteristic index q_lLower business b_kRelative to business b_j(j=1,2 ..., K And the class of business total number that ≠ k, K are service set) significance level；In characteristic index q_l Lower business b_kRelative to business b_jSignificance level be divided into S_l+ 1 grade, uses numerical value respectively (0,1/S_l, 2/S_l..., 1) represent, numerical value the biggest expression significance level is the highest, and represents important journey The numerical value of degree and S_lRelevant, i.e. services sets is in characteristic index q_lUnder difference degree the biggest, divide The importance degree the most grades gone out, during evaluation assignment importance value more rationally accurately；Interval division Foundation is s_klWith s_jlRatio；s_klExpression business b_kIn characteristic index q_lUnder importance value, s_jl Expression business b_jIn characteristic index q_lUnder importance value, S_lBy services sets in characteristic index q_lUnder Difference degree determine, its numeric representation services sets B is to characteristic index q_lThere is S_lPlant different wanting Ask.

It should be noted that described step 2.3) in, to every class business b_kRefer in all features Relatively important value matrix under BiaoSue for peace, obtain business b_kComprehensive relatively important value Matrix C_k, wherein Matrix C_kElement c_kjCalculated by following formula:

$c_{kj}=\underset{l=1}{\overset{L}{\&Sigma;}}{c}_{kj}^{\left(q_l\right)};$

Wherein, L is characterized index set heat-producing characteristics index total number；J=1,2 ..., K and ≠ k, K is service set class of business total number；Represent in characteristic index q_lLower business b_kRelatively In business b_jSignificance level, i.e. characteristic index q_lCorresponding relatively important value matrixUnit Element, c_kjExpression business b_kRelative to business b_jThe most relatively important under all index feature Value.

It should be noted that described step 2.4) method particularly includes: to business b_kComprehensive Relatively important value matrix C_kRow vector element sue for peace, obtain business b_kTo service set The comprehensive relative Link Importance sum of other all business:

$c_k=\underset{j=1}{\overset{K}{\&Sigma;}}{c}_{kj};$

K is service set class of business total number, c_kjFor business b_kComprehensive relatively important value square Battle array C_kElement.

It should be noted that described step 2.5) method particularly includes:

Note business b_kComprehensive relative Link Importance sum to other all classs of business of service set For c_k, then business b_kImportance degree α_k:

${\mathrm{\&alpha;}}_k=(1-X)\frac{c_k-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}{\underset{k=1}{\overset{K}{\max }}\left\{c_k\right\}-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}+X;$

Between the excessive then business of X value, importance degree difference is too small, and the differentiation of business can be made to spend Low, therefore take X=0.1, the distance of importance degree maxima and minima is set as 10 times；K For service set class of business total number.

It should be noted that in described step S3, note weighted adjacent matrix is H (G), H (G) Element h_ijRepresent node v_iWith node v_jThe quantized value of business is carried in institute's connected link, I, j ∈ 1,2 ..., N}, N are network topology model interior joint sum；h_ijCalculated by following formula:

$h_{ij}=\underset{k=1}{\overset{M_{ij}}{\&Sigma;}}{\mathrm{\&alpha;}}_k{\mathrm{\&beta;}}_{ijk};$

M_ijRepresent node v_iWith node v_jClass of business sum, α is run in institute's connected link_kRepresent industry Business b_kImportance value, β_ijkRepresent node v_iWith node v_jBusiness b in institute's connected link_kQuantity.

It should be noted that in described step S4, the weighting Natural communication of powerline network Degree is calculated by following formula:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}\left(G\right)=ln\left(\frac{1}{N}\underset{j=1}{\overset{N}{\&Sigma;}}{e}^{{\mathrm{\&lambda;}}_j\left(G\right)}\right);$

WhereinFor the weighting Natural communication degree of described power telecom network, λ_j(G) it is jth joint The eigenvalue of the weighted adjacent matrix H (G) that point is corresponding, N is network topology model interior joint Sum.

It should be noted that the concrete grammar of described step S5 is as follows:

Make G-v_iRepresent node v_iGained powerline network topology subgraph after inefficacy, due to nature Degree of communication is strictly monotone about adding limit and removing limit, it is therefore assumed that node failure is and this All links that node connects all lost efficacy, and calculated powerline network topology subgraph by following formula G-v_iWeighting Natural communication degree:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}(G-v_i)=ln\left(\frac{1}{N}\underset{j=1}{\overset{N}{\&Sigma;}}{e}^{{\mathrm{\&lambda;}}_j(G-v_i)}\right);$

WhereinFor node v_iThe weighting of the powerline network topology subgraph after inefficacy is certainly So degree of communication, λ_j(G-v_i) it is the weighted adjacent matrix that this topology subgraph jth node is corresponding Eigenvalue, N is network topology model interior joint sum；

When network node lost efficacy, powerline network survivability declines, and network operation business is subject to To impact, weighting Natural communication angle value reduces；By comparing each node v_iInstitute after losing efficacy successively Obtain the weighting Natural communication degree of network topology subgraph and changing of artwork weighting Natural communication degree To pitch point importance, it is shown below:

$r_i=1-\frac{\stackrel{\&OverBar;}{\mathrm{\&lambda;}}(G-v_i)}{\stackrel{\&OverBar;}{\mathrm{\&lambda;}}\left(G\right)};$

Wherein r_iIt is node v_iNormalization importance,Represent original network topology illustraton of model The weighting Natural communication degree of G,It is figure G-v_iWeighting Natural communication degree.

The beneficial effects of the present invention is: the pitch point importance computational methods that the present invention provides are comprehensive Consider network topology characteristic and the power business characteristic of node, have more practical significance.Business The computational methods of importance degree require to substitute expert's subjective scoring with objective indicator, eliminate evaluation result In subjective uncertainty.And the present invention provides algorithm form of calculation succinct, it is adaptable to node is advised The powerline network that mould is bigger.

Accompanying drawing explanation

Fig. 1 is the flow chart of the present invention；

Fig. 2 is the network topology structure schematic diagram of the example of the present invention；

Fig. 3 is the importance degree calculation flow chart of power telecom network.

Detailed description of the invention

Below with reference to accompanying drawing, the invention will be further described, it should be noted that this reality Execute example premised on the technical program, give detailed embodiment and concrete operation Journey, but protection scope of the present invention is not limited to the present embodiment.

As it is shown in figure 1, a kind of power telecom network pitch point importance computational methods include walking as follows Rapid:

Step 1: set up the network topology model of power telecom network:

Described network topology model includes that node and link, described node number are N, described chain Road number is M, and the most described network topology model figure G is described as follows:

G=(V, E)；

Wherein V={v₁, v₂..., v_NRepresenting node set, the most each node represents power telecom network In electric power communication device, therefore V is electric power communication device set；E={e₁, e₂..., e_MTable Show that the set on limit, described limit represent the link between electric power communication device；

Corresponding adjacency matrix A (G)=[a_ijElement a in]_ijIt is defined as follows:

As in figure 2 it is shown, as a example by certain power telecom network localized network, carry out simulating, verifying.Should Topological structure comprises 14 nodes and 16 links, and wherein No. 1 node is provincial control centre (middle tune), No. 13 nodes are regional dispatching center (adjust), and No. 14 nodes are 220kV power transformation Standing, remaining node is 500kV transformer station, and No. 2, No. 5 and No. 7 nodes are aggregation node.

The adjacency matrix of figure G is expressed as follows:

$A\left(G\right)=\left[\begin{array}{cccccccccccccc}0& 1& 0& 0& 0& 0& 0& 0& 0& 1& 1& 0& 0& 0\\ 1& 0& 1& 0& 0& 0& 0& 0& 0& 0& 0& 1& 1& 0\\ 0& 1& 0& 1& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1& 0& 1& 0& 0& 0& 0& 0& 1& 0& 0& 0\\ 0& 0& 0& 1& 0& 1& 0& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 1& 0& 1& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 1& 0& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 1& 0& 1& 0& 1& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 1& 0& 1& 0& 0& 0& 0\\ 1& 0& 0& 0& 0& 0& 0& 0& 1& 0& 0& 0& 0& 0\\ 1& 0& 0& 1& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 1& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 1& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 1\\ 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 1& 0\end{array}\right]$

Step 2: calculating power communication network service importance degree, flow process is as shown in Figure 3.

In power communication network operation, the different business importance degree of carrying is the most different, and business importance degree is Business occurs to interrupt or the extent of injury to electrical network in the case of defect, and the extent of injury is the highest, industry Business importance degree is the highest.Power business importance degree is to evaluate as the description method of power business feature The basic foundation of powerline network overall performance under multiple power business.The present invention uses feature Metrics evaluation method substitutes the marking of expert's subjectivity and is evaluated power business importance degree, selected characteristic Index, calculates business importance degree.

Selected characteristic index set Q={q_l(l=1,2 ..., L), according to business, selected feature is referred to Target difference requirement, by services sets B={b_k(k=1,2 ..., K) importance be mapped to importance value sequence Row { s_kl, s_klExpression business b_kIn characteristic index q_lUnder importance value.{s_kl∈ 1,2 ..., S_l, Wherein S_lBy services sets in characteristic index q_lUnder difference degree determine, its numeric representation services sets B is to index q_lThere is S_lPlant different requirements.By importance value sequence { s_klEvery class industry can be calculated Business b_kIn characteristic index q_lUnder relatively important value matrixElement

$c_{kj}^{\left(q_l\right)}=\left\{\begin{array}{cc}1& s_{kl}/s_{jl}\&GreaterEqual;S_l\\ .& .\\ .& .\\ .& .\\ 2/S_l& 2\&le;s_{kl}/s_{jl}<3\\ 1/S_l& 1\&le;s_{kl}/s_{jl}<2\\ 0& s_{kl}/s_{jl}<1\end{array}\right.;$

Wherein,Represent in characteristic index q_lLower business b_kRelative to business b_jSignificance level. In characteristic index q_lLower business b_kRelative to business b_jSignificance level be divided into S_l+ 1 grade, point Yong numerical value (0,1/S_l, 2/S_l..., 1) represent, numerical value the biggest expression significance level is the highest.And table Show numerical value and the S of significance level_lRelevant, i.e. services sets is in characteristic index q_lUnder difference degree more Greatly, the importance degree the most grades marked off, during evaluation assignment importance value more rationally accurately. Interval division foundation is s_klWith s_jlRatio.

Characteristic index is concentrated the relatively important value matrix of business under all indexsAsk With, obtain comprehensive relatively important value matrix C_k, wherein Matrix C_kElement c_kjBy following formula meter Calculate:

$c_{kj}=\underset{l=1}{\overset{L}{\&Sigma;}}{c}_{kj}^{\left(q_l\right)};$

To comprehensive relatively important value matrix C_kRow vector element sue for peace, obtain business b_kRight The comprehensive relative Link Importance sum of other all business of service set:

$c_k=\underset{j=1}{\overset{K}{\&Sigma;}}{c}_{kj};$

To c_kIt is normalized, and is mapped on interval [X, 1], obtain business b_k's Importance degree α_k:

${\mathrm{\&alpha;}}_k=(1-X)\frac{c_k-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}{\underset{k=1}{\overset{K}{\max }}\left\{c_k\right\}-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}+X;$

Wherein, take X=0.1, the distance of importance degree maxima and minima is set as 10 times.

Choose representative power business according to technical field of electric power production run relevant regulations, feature refers to Mark and requirement are as shown in table 1:

Table 1

Gained business importance degree evaluation result such as table 2 institute is calculated by above-mentioned computational methods and upper table Show:

Table 2

Step 3: the business of link carrying is quantified, quantized value is drawn with the form of weights Enter among adjacency matrix, calculating weighted adjacent matrix H (G):

The element h of weighted adjacent matrix H (G)_ijRepresent node v_iWith node v_jHold in institute's connected link The quantized value of load business.h_ijCalculated by following formula:

$h_{ij}=\underset{k=1}{\overset{M_{ij}}{\&Sigma;}}{\mathrm{\&alpha;}}_k{\mathrm{\&beta;}}_{ijk};$

M_ijRepresent node v_iWith node v_jClass of service sum, α is run in institute's connected link_kRepresent industry Business b_kImportance value, β_ijkRepresent business b run on link_kNumber.Electric power shown in Fig. 2 In communication network topology fabric link, the distribution situation of all kinds of business is as shown in table 3:

Table 3

Calculate the weighted adjacent matrix H (G) of powerline network topological structure as shown in Figure 2:

$H\left(G\right)=\left[\begin{array}{cccccccccccccc}0& 55.428& 0& 0& 0& 0& 0& 0& 0& 60.762& 50.406& 0& 0& 0\\ 55.428& 0& 14.024& 0& 0& 0& 0& 0& 0& 0& 0& 7.513& 13.822& 0\\ 0& 14.024& 0& 1.238& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 1.238& 0& 32.805& 0& 0& 0& 0& 0& 41.652& 0& 0& 0\\ 0& 0& 0& 32.805& 0& 9.488& 0& 1.238& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 9.488& 0& 1.238& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 1.238& 0& 28.572& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 1.238& 0& 28.572& 0& 42.161& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 42.161& 0& 51.253& 0& 0& 0& 0\\ 60.762& 0& 0& 0& 0& 0& 0& 0& 51.253& 0& 0& 0& 0& 0\\ 50.406& 0& 0& 41.652& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 7.513& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 13.822& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 22.501\\ 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 0& 22.501& 0\end{array}\right];$

Step 4: the weighting Natural communication degree of calculating powerline network:

The importance degree calculating static topological connection interior joint is typically to consider net before and after knot removal The situation of change of network connection situation, is equivalent to the importance of node after this node is deleted net The destructiveness of network.The destructiveness that network topology is caused by node failure can pass through many aspects Weighing, the present invention uses and the impact of alternative route redundancy in network is weighed node failure to net The destruction result of network.

The present invention use Natural communication degree measure the redundancy of alternative route in network.Natural Degree of communication analyzes the node disturbance degree for network communications capability from topological layer, it is possible to the most anti- Reflect the node significance level in topological layer.The present invention utilizes the weighting obtained based on operation layer analysis Adjacency matrix, provides the definition of power telecom network weighting Natural communication degree.Described power telecom network The weighting Natural communication degree computing formula of network topological diagram G is as follows:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}\left(G\right)=ln\left(\frac{1}{N}\underset{j=1}{\overset{N}{\&Sigma;}}{e}^{{\mathrm{\&lambda;}}_j\left(G\right)}\right);$

WhereinFor the weighting Natural communication degree of network G, λ_j(G) it is that jth node is corresponding The eigenvalue of weighted adjacent matrix H (G).

The Natural communication degree computing formula of described powerline network topological diagram G is as follows:

$\stackrel{\&OverBar;}{{\mathrm{\&lambda;}}^{\&prime;}}\left(G\right)=ln\left(\frac{1}{N}\underset{j=1}{\overset{N}{\&Sigma;}}{e}^{{\mathrm{\&lambda;}}_j^{\&prime;}\left(G\right)}\right);$

WhereinFor scheming the Natural communication degree of G, λ '_j(G) it is corresponding the adjoining of jth node The eigenvalue of matrix A (G).

Step 5: calculate node v_iThe weighting Natural communication degree of network after inefficacy:

Make G-v_iRepresent node v_iGained powerline network topology subgraph after inefficacy.Due to nature Degree of communication is strictly monotone about adding limit and removing limit, therefore present invention assumes that node failure Be that all links being connected with this node all lost efficacy, i.e. node failure refers to weighted adjacent matrix In this node place row element and column element zero setting.Powerline network topology is calculated by following formula Subgraph G-v_iWeighting Natural communication degree:

$\stackrel{\&OverBar;}{\mathrm{\&lambda;}}(G-v_i)=ln\left(\frac{1}{N}\underset{j=1}{\overset{N}{\&Sigma;}}{e}^{{\mathrm{\&lambda;}}_j(G-v_i)}\right);$

WhereinFor the weighting Natural communication degree of network, λ_j(G-v_i) it is node v_iAfter inefficacy The eigenvalue of powerline network topology subgraph weighted adjacent matrix.

Step 5: calculate node v_iImportance degree:

As network node v_iBreak down, when all links that is i.e. connected with this node all lost efficacy, net Network remains node communication path necessarily affected, under the redundancy of whole network alternate routing Fall, owing to remaining origin node, therefore under conditions of node total number is constant, weighting nature is even Logical degree declines therewith, and the decline of weighting Natural communication degree simultaneously also is able to embody failure node to net Network runs the impact of business.It is the most that the value of weighting Natural communication degree declines, and shows this node pair The impact of whole network is the biggest.By the value of comparing cell weighting Natural communication degree, can obtain Node cross-layer merges importance degree, thus judges the relative of any two node in powerline network Importance.

Described powerline network node v_iIt is as follows that cross-layer merges importance degree computing formula:

$r_i=1-\frac{\stackrel{\&OverBar;}{\mathrm{\&lambda;}}(G-v_i)}{\stackrel{\&OverBar;}{\mathrm{\&lambda;}}\left(G\right)};$

Wherein r_iIt is node v_iNormalization importance degree,Represent original network topology figure G's Weighting Natural communication degree,It is figure G-v_iWeighting Natural communication degree.r_iValue the biggest, Node significance level is the highest.

By the Natural communication degree of gained network topology subgraph after comparing each node and losing efficacy successively With artwork Natural communication degree, topological layer pitch point importance can be obtained.Described powerline network Topological layer pitch point importance computing formula is as follows:

$r_i^{\&prime;}=1-\frac{{\stackrel{\&OverBar;}{\mathrm{\&lambda;}}}^{\&prime;}(G-v_i)}{\stackrel{\&OverBar;}{{\mathrm{\&lambda;}}^{\&prime;}}\left(G\right)};$

Wherein r '_iIt is node v_iNormalization topology importance degree,Represent original network topology figure The Natural communication degree of G,It is figure G-v_iNatural communication degree.Pitch point importance calculates Comparative result is as shown in table 4 below:

Table 4

By simulation result, fused business of the present invention and the computational methods of topology and topological layer Computational methods evaluation result differs greatly, and improves node v₁, v₁₀, v₁₁Node importance, Owing in actual power communication backbone network, core province point of adjustment has higher importance, therefore Blending algorithm of the present invention more conforms to practical situation, and topological layer algorithm lays particular emphasis on network topology and belongs to Property；Blending algorithm method of the present invention is basically identical with the assessment result of weighting node contraction algorithm, Illustrate the effectiveness of blending algorithm of the present invention, but from algorithm complex, weighting joint Point contraction algorithm needs again to calculate the shortest with Floyd algorithm after each node contraction operates Path matrix, blending algorithm utilizes the change of Natural communication degree to weigh pitch point importance, calculates shape Formula is succinct and complexity is relatively low.And weighting node contraction algorithm determines node weight by expert analysis mode Value, has certain subjectivity.Therefore by cross-layer aggregators importance degree algorithm of the present invention, can Effectively to determine network core node, it is adaptable to the powerline network that node is larger.

For a person skilled in the art, can according to above technical scheme and design, Make various corresponding change and deformation, and all these changes and deformation should be construed as being included in Within the protection domain of the claims in the present invention.

Claims (10)

1. power telecom network pitch point importance computational methods, it is characterised in that include as Lower step:

S1 sets up the network topology model of power telecom network according to the actual framework of power telecom network, And build corresponding adjacency matrix；

The business importance degree of the S2 described power telecom network of calculating:

2.1) some characteristic index constitutive characteristic index set is taken, according to all kinds of in power telecom network The business different requirements to selected characteristic index, by the weight of the services sets of all kinds of service constitution The property wanted is mapped to importance value sequence；

2.2) calculating all kinds of business by importance value sequence concentrates each feature to refer in characteristic index Relatively important value matrix under Biao；

2.3) every class business relatively important value matrix under all characteristic indexs is sued for peace, Obtain the comprehensive relatively important value matrix that every class business is corresponding；

2.4) to every class business in step 2.3) row of comprehensive relatively important value matrix that obtains Vector element is sued for peace, and obtains every class business and combines other all kinds business of service set Close relative Link Importance sum；

2.5) to step 2.4) in calculated every class business to service set other own The comprehensive relative Link Importance sum of kind business is normalized, and be mapped to interval [X, 1], on, the importance degree of every class business, wherein 0 ＜ X ＜ 1 are obtained；

The number of all kinds of business run on link is calculated its business in step S2 by S3 The product of importance degree and as link quantized value, the business of link carrying is quantified, will Described link quantized value is incorporated among adjacency matrix with the form of weights, obtains weighted adjacent square Battle array；

S4 calculates the weighting Natural communication degree of power telecom network；

S5 is by the weighting nature of gained network topology subgraph after comparing each node and losing efficacy successively Degree of communication and the weighting Natural communication degree of artwork, it is thus achieved that pitch point importance.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step S1 specific as follows:

According to the actual framework of described power telecom network, build corresponding network topology model, Described network topology model includes that node and link, described node number are N, described link Number is M, and the most described network topology model figure G is described as follows:

G=(V, E)；

Wherein V={v₁, v₂..., v_NRepresenting node set, the most each node represents power telecom network In electric power communication device, therefore V is electric power communication device set；E={e₁, e₂..., e_MTable Show that the set on limit, described limit represent the link between electric power communication device；

Corresponding adjacency matrix A (G)=[a_ijElement a in]_ijIt is defined as follows:

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step 2.1) in, note characteristic index integrates as Q={q_l(l=1,2 ..., L), L It is characterized characteristic index total number in index set；According to business to selected characteristic index not Same requirement, by services sets B={b_k(k=1,2 ..., K) importance be mapped to importance value sequence {s_kl∈ 1,2 ..., S_l, wherein K is the total number of service set class of business；s_klExpression business b_k In characteristic index q_lUnder importance value, wherein S_lBy services sets in characteristic index q_lUnder difference Degree determines, its numeric representation services sets B is to characteristic index q_lThere is S_lPlant different requirements.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step 2.2) in, by importance value sequence { S_klCalculate all kinds of business b_k(k=1,2 ..., K) in each characteristic index q_l(l=1,2 ..., L, L are selected characteristic index sum Mesh) under relatively important value matrixAs follows:

Business b_kIn characteristic index q_lUnder relatively important value matrixElementUnder employing Formula calculates:

$c_{kj}^{\left(q_l\right)}=\left\{\begin{array}{cc}1& s_{kl}/s_{jl}\&GreaterEqual;S_l\\ \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}\\ 2/S_l& 2\&le;s_{kl}/s_{jl}<3\\ 1/S_l& 1\&le;s_{kl}/s_{jl}<2\\ 0& s_{kl}/s_{jl}<1\end{array}\right.;$

Wherein,Represent in characteristic index q_lLower business b_kRelative to business b_j(j=1,2 ..., K And the class of business total number that ≠ k, K are service set) significance level；In characteristic index q_l Lower business b_kRelative to business b_jSignificance level be divided into S_l+ 1 grade, uses numerical value respectively (0,1/S_l, 2/S_l..., 1) represent, numerical value the biggest expression significance level is the highest, and represents important journey The numerical value of degree and S_lRelevant, i.e. services sets is in characteristic index q_lUnder difference degree the biggest, divide The importance degree the most grades gone out, during evaluation assignment importance value more rationally accurately；Interval division Foundation is s_klWith s_jlRatio；s_klExpression business b_kIn characteristic index q_lUnder importance value, s_jl Expression business b_jIn characteristic index q_lUnder importance value, S_lBy services sets in characteristic index q_lUnder Difference degree determine, its numeric representation services sets B is to characteristic index q_lThere is S_lPlant different wanting Ask.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step 2.3) in, to every class business b_kPhase under all characteristic indexs To important value matrixSue for peace, obtain business b_kComprehensive relatively important value matrix C_k, Wherein Matrix C_kElement c_kjCalculated by following formula:

$c_{\mathrm{kj}}=\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{c}_{\mathrm{kj}}^{\left(q_l\right)};$

Wherein, L is characterized index set heat-producing characteristics index total number；J=1,2 ..., K and ≠ k, K is service set class of business total number；Represent in characteristic index q_lLower business b_kRelatively In business b_jSignificance level, i.e. characteristic index q_lCorresponding relatively important value matrixUnit Element, c_kjExpression business b_kRelative to business b_jThe most relatively important under all index feature Value.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step 2.4) method particularly includes: to business b_kThe most relatively important Value matrix C_kRow vector element sue for peace, obtain business b_kTo service set, other own The comprehensive relative Link Importance sum of business:

$c_k=\underset{l=1}{\overset{K}{\mathrm{\&Sigma;}}}{c}_{\mathrm{kj}};$

K is service set class of business total number, c_kjFor business b_kComprehensive relatively important value square Battle array C_kElement.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, described step 2.5) method particularly includes:

Note business b_kComprehensive relative Link Importance sum to other all classs of business of service set For c_k, then business b_kImportance degree α_k:

${\mathrm{\&alpha;}}_k=(1-X)\frac{c_k-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}{\underset{k=1}{\overset{K}{\max }}\left\{c_k\right\}-\underset{k=1}{\overset{K}{\min }}\left\{c_k\right\}}+X;$

Between the excessive then business of X value, importance degree difference is too small, and the differentiation of business can be made to spend Low, therefore take X=0.1, the distance of importance degree maxima and minima is set as 10 times；K For service set class of business total number.

Power telecom network pitch point importance computational methods the most according to claim 1, its Being characterised by, in described step S3, note weighted adjacent matrix is the element h of H (G), H (G)_ij Represent node v_iWith node v_jThe quantized value of business is carried in institute's connected link, i, j ∈ 1,2 ..., N}, N For network topology model interior joint sum；h_ijCalculated by following formula:

$h_{\mathrm{ij}}=\underset{k=1}{\overset{M_{\mathrm{ij}}}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{\mathrm{\&beta;}}_{\mathrm{ijk}};$

M_ijRepresent node v_iWith node v_jClass of business sum, α is run in institute's connected link_kRepresent industry Business b_kImportance value, β_ijkRepresent node v_iWith node v_jBusiness b in institute's connected link_kQuantity.

Power telecom network pitch point importance computational methods the most according to claim 1, its It is characterised by, in described step S4, under the weighting Natural communication degree of powerline network passes through Formula calculates:

$\stackrel{-}{\mathrm{\&lambda;}}\left(G\right)=\ln \left(\frac{1}{N}\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}^{{\mathrm{\&lambda;}}_j\left(G\right)}\right);$

WhereinFor the weighting Natural communication degree of described power telecom network, λ_j(G) it is jth joint The eigenvalue of the weighted adjacent matrix H (G) that point is corresponding, N is network topology model interior joint Sum.

Power telecom network pitch point importance computational methods the most according to claim 1, its Being characterised by, the concrete grammar of described step S5 is as follows:

Make G-v_iRepresent node v_iGained powerline network topology subgraph after inefficacy, due to nature Degree of communication is strictly monotone about adding limit and removing limit, it is therefore assumed that node failure is and this All links that node connects all lost efficacy, and calculated powerline network topology subgraph by following formula G-v_iWeighting Natural communication degree:

$\stackrel{-}{\mathrm{\&lambda;}}(G-v_i)=\ln \left(\frac{1}{N}\underset{j=1}{\overset{N}{\mathrm{\&Sigma;}}}{e}^{{\mathrm{\&lambda;}}_j(G-v_i)}\right);$

WhereinFor node v_iThe weighting of the powerline network topology subgraph after inefficacy is certainly So degree of communication, λ_j(G-v_i) it is the weighted adjacent matrix that this topology subgraph jth node is corresponding Eigenvalue, N is network topology model interior joint sum；

When network node lost efficacy, powerline network survivability declines, and network operation business is subject to To impact, weighting Natural communication angle value reduces；By comparing each node v_iInstitute after losing efficacy successively Obtain the weighting Natural communication degree of network topology subgraph and changing of artwork weighting Natural communication degree To pitch point importance, it is shown below:

$r_i=1-\frac{\stackrel{-}{\mathrm{\&lambda;}}(G-v_i)}{\stackrel{-}{\mathrm{\&lambda;}}\left(G\right)};$

Wherein r_iIt is node v_iNormalization importance,Represent original network topology illustraton of model The weighting Natural communication degree of G,It is figure G-v_iWeighting Natural communication degree.