CN106910141A - A kind of complicated active power distribution network decomposing scheme method for optimizing and device - Google Patents

Abstract

The present invention relates to a kind of complicated active power distribution network decomposing scheme method for optimizing and device, methods described includes：Obtain the corresponding evaluation index of each decomposing scheme in the decomposing scheme set and the decomposing scheme set of complicated active power distribution network；According to the corresponding evaluation index of each decomposing scheme in the decomposing scheme set and the decomposing scheme set, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method；Carry out stability analysis to the superiority-inferiority collating sequence, update the superiority-inferiority collating sequence and select to rank in the superiority-inferiority collating sequence decomposing scheme corresponding to most preceding element as optimal Decomposition scheme；The technical scheme that the present invention is provided so that complicated active power distribution network decomposing scheme decision model can simultaneously consider homogeneity, the antagonism of influence factor, for the complicated active power distribution network decomposing scheme Optimal Decision-making problem for the treatment of provides new approaches.

Description

A kind of complicated active power distribution network decomposing scheme method for optimizing and device

Technical field

The present invention relates to field of distribution network, and in particular to a kind of complicated active power distribution network decomposing scheme method for optimizing and dress Put.

Background technology

Set Pair Analysis (Set Pair Analysis, SPA) be it is a kind of indetermination theory, be diligent by Chinese scholar Zhao Ke The one kind proposed in 1989 is on determination, uncertain system-same, different, anti-quantitative analysis systematic analytic method.Its core Thought is that certainty information and uncertain information are included in same system, studied in terms of same, different, anti-three things it Between certainty with it is uncertain, the contact and conversion between things are portrayed comprehensively.

The basic conception of Set Pair Analysis is set pair and its Pair Analysis.So-called set pair, is exactly two collection with certain contact Constituted antithetical phrase is closed, is analyzed according to a certain characteristic spread of set pair, contact of the set pair in the characteristic is carried out to classify quantitative Description, obtaining connection degree representation of the set pair under a certain Question background is：

μ=a+bi+cj

In formula：μ is referred to as Pair Analysis, is normally only a kind of structure function for a Pair Analysis for particular problem, only exists In particular cases it is only a numerical value.A represents two same degree of set, referred to as identical degree；B represents two differences of set Uncertainty degree, referred to as diversity factor；C represents two opposition degree of set, referred to as degree of opposition；I is diversity factor coefficient, [- 1, 1] value.I changes between -1~1, embodies the mutual conversion between certainty and uncertainty, with i → 0, does not know Property substantially increases, and i is when taking -1 and 1, problem all being to determine property；J is opposition label symbol or corresponding coefficient, it is stipulated that value For -1.

Pair Analysis can be uniformly processed the caused uncertainties such as fuzzy, random and INFORMATION OF INCOMPLETE.This delineation is to true It is qualitative with probabilistic quantitative description, wherein a, c is with respect to determining, and b is relatively uncertain, and a, b, c meet as follows Normalizing condition：

A+b+c=1

This relativity is the complexity and changeability due to objective objects, and objective objects are recognized and the master for portraying The uncertainty that the property seen and ambiguity are caused.Thus in formula (1), certainty exists with uncertain, homogeneity and antagonism Relativity, the ambiguity in understanding, the result portrayed is also relative, not exclusive.Set Pair Analysis are effectively featured really The unity of opposites relation of fixed and uncertain system, meets the dialectics of nature and mankind thought mode, with Methodological Significance.

The equipment of power distribution network is numerous, and complex structure is in large scale, and analytical calculation is complex.Especially with intelligent electricity The development of net, substantial amounts of distributed power source accesses power distribution network, and conventional electrical distribution net progressively develops into complicated active power distribution network, further Increased the calculation scale of network analysis calculating.The calculation scale of problem can be substantially reduced by network decomposition, saves a large amount of The calculating time.For complicated active power distribution network, often there is multiple network decomposing scheme, how to be decomposed from numerous feasible networks Preferably go out preferred plan in scheme, in addition to calculating speed to be considered, also need to consider computational accuracy, resource utilization etc. it is many because Element limitation., it is necessary to these factors for considering are both often opposition and unified in actual decision-making, how to make these factors very It is good problem to study to be unified in well in a network decomposition scheme decision model.

The content of the invention

The present invention provides a kind of complicated active power distribution network decomposing scheme method for optimizing and device, the purpose is to cause that complexity has Source power distribution network decomposing scheme decision model can simultaneously consider homogeneity, the antagonism of influence factor, to process complicated active distribution Net decomposing scheme Optimal Decision-making problem provides new approaches.

The purpose of the present invention is realized using following technical proposals：

A kind of complicated active power distribution network decomposing scheme method for optimizing, it is theed improvement is that, including：

Obtain each decomposing scheme correspondence in the decomposing scheme set and the decomposing scheme set of complicated active power distribution network Evaluation index；

According to the corresponding evaluation index of each decomposing scheme in the decomposing scheme set and the decomposing scheme set, adopt The superiority-inferiority collating sequence of each decomposing scheme is determined with Set Pair Analysis Method；

Stability analysis is carried out to the superiority-inferiority collating sequence, the superiority-inferiority collating sequence is updated and is selected described excellent The decomposing scheme corresponding to most preceding element is ranked in pessimum collating sequence as optimal Decomposition scheme.

Preferably, each point in the decomposing scheme set and the decomposing scheme set for obtaining complicated active power distribution network The corresponding evaluation index of solution scheme, including：

Determine that the corresponding resource utilization index of each decomposing scheme, parallel computation are complicated in the decomposing scheme set respectively Degree index, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and parallel computation cost refer to Mark.

Further, the corresponding resource utilization index of each decomposing scheme in the decomposing scheme set is determined as the following formula y₁：

In above formula, n is the number of partitions of network, P_i ²It is i-th calculation scale of subregion in network, i ∈ [1, n]；

The corresponding parallel computation complexity index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₂：

In above formula, O_minIt is theoretical minimum parallel computation complexity,It is i-th computation complexity of subregion, l in network To coordinate level number of servers, M is calculation cost coefficient, and i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

The corresponding parallel computation precision index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₃：

In above formula, U_i' it is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m It is nodes sum；

The corresponding parallel computation speed-up ratio index S of each decomposing scheme in the decomposing scheme set is determined as the following formula_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

Preferably, it is described corresponding according to each decomposing scheme in the decomposing scheme set and the decomposing scheme set Evaluation index, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method, including：

Non- income type index in the corresponding evaluation index of each decomposing scheme is converted into income type index；

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme, the corresponding rule of each decomposing scheme are obtained Generalized desired value；

Select the optimum specification desired value collection U=(u in the corresponding standardization desired value of each decomposing scheme₁, u₂,...u_n) and most bad standardization desired value collection V=(v₁,v₂,...v_n), build each decomposing scheme comparing space [V, U]；

In the comparing space [V, U] of each decomposing scheme, each decomposing scheme and the optimum specification are determined The relative similarity degree of desired value collection U；

By the descending order pair of each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U Each decomposing scheme is ranked up, and obtains the superiority-inferiority collating sequence of each decomposing scheme.

Further, the non-income type index in the corresponding evaluation index by each decomposing scheme is converted to income type and refers to Mark, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula will Be converted to income type index：

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th maximum of index Desired value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

Further, it is described that nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme, obtain each point The corresponding standardization desired value of solution scheme, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,It is k-th decomposition side On r-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum to case, and n is evaluation index sum.

Further, it is described in the comparing space [V, U] of each decomposing scheme, determine each decomposing scheme and institute The relative similarity degree of optimum specification desired value collection U is stated, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th point Solution scheme is d on r-th standardization desired value of index_kr, each decomposing scheme is in r-th standardization desired value of index Optimum specification desired value be u_r, each decomposing scheme on r-th standardization desired value of index in it is most bad standardization refer to Scale value is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimal rule as the following formula The Pair Analysis of generalized desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kFor k-th decomposing scheme and optimum specification refer to The identical degree of scale value collection U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kIt is k-th decomposition side The opposition degree of case and optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

Preferably, it is described that stability analysis is carried out to the superiority-inferiority collating sequence, update the superiority-inferiority collating sequence And select to rank in the superiority-inferiority collating sequence decomposing scheme corresponding to most preceding element as optimal Decomposition scheme, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, and selection is described γ in superiority-inferiority collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme, when 0≤i≤1 and c_kb_p-c_pb_k≤0 When, i need to meet i ∈ [0,1], if meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if discontented Foot, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

When 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meetIf meeting, institute State γ in superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kWith γ_pSorting position exchange；

As -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meeting, superiority-inferiority sequence γ in sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position Exchange；

As -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_k And γ_pSorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th point The diversity factor of solution scheme and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and optimum specification desired value collection U Opposition degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pFor p-th decomposing scheme with it is optimal The diversity factor of standardization desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.

A kind of complicated active power distribution network decomposing scheme preferred embodiment, it is theed improvement is that, described device includes：

Acquisition module, for each in the decomposing scheme set and the decomposing scheme set that obtain complicated active power distribution network The corresponding evaluation index of decomposing scheme；

Determining module, for according to each decomposing scheme correspondence in the decomposing scheme set and the decomposing scheme set Evaluation index, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method；

Analysis module, for carrying out stability analysis to the superiority-inferiority collating sequence, updates the superiority-inferiority sequence sequence Arrange and select to rank in the superiority-inferiority collating sequence decomposing scheme corresponding to most preceding element as optimal Decomposition scheme.

Preferably, the acquisition module, including：

First determining unit, determines that the corresponding resource utilization of each decomposing scheme refers in the decomposing scheme set respectively Mark, parallel computation complexity index, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and The parallel computation indicator of costs.

Further, the corresponding resource utilization index of each decomposing scheme in the decomposing scheme set is determined as the following formula y₁：

In above formula, n is the number of partitions of network, P_i ²It is i-th calculation scale of subregion in network, i ∈ [1, n]；

The corresponding parallel computation complexity index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₂：

In above formula, O_minIt is theoretical minimum parallel computation complexity,It is i-th computation complexity of subregion, l in network To coordinate level number of servers, M is calculation cost coefficient, and i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

The corresponding parallel computation precision index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₃：

In above formula, U_i' it is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m It is nodes sum；

The corresponding parallel computation speed-up ratio index S of each decomposing scheme in the decomposing scheme set is determined as the following formula_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

Preferably, the determining module, including：

Converting unit, refers to for the non-income type index in the corresponding evaluation index of each decomposing scheme to be converted into income type Mark；

Specification unit, for carrying out nondimensionalization treatment to the corresponding evaluation index of each decomposing scheme, obtains each point The corresponding standardization desired value of solution scheme；

Select unit, for selecting the optimum specification desired value in the corresponding standardization desired value of each decomposing scheme Collection U=(u₁,u₂,...u_n) and most bad standardization desired value collection V=(v₁,v₂,...v_n), build the comparing of each decomposing scheme Space [V, U]；

Second determining unit, in the comparing space [V, U] of each decomposing scheme, determining each decomposing scheme With the relative similarity degree of the optimum specification desired value collection U；

Sequencing unit, for by each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U Descending order is ranked up to each decomposing scheme, obtains the superiority-inferiority collating sequence of each decomposing scheme.

Further, the converting unit, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula will Be converted to income type index：

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th maximum of index Desired value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

Further, the specification unit, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,It is k-th decomposition side On r-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum to case, and n is evaluation index sum.

Further, second determining unit, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th point Solution scheme is d on r-th standardization desired value of index_kr, each decomposing scheme is in r-th standardization desired value of index Optimum specification desired value be u_r, each decomposing scheme on r-th standardization desired value of index in it is most bad standardization refer to Scale value is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimal rule as the following formula The Pair Analysis of generalized desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kFor k-th decomposing scheme and optimum specification refer to The identical degree of scale value collection U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kIt is k-th decomposition side The opposition degree of case and optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

Preferably, the analysis module, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, and selection is described γ in superiority-inferiority collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme；

First judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kWhen≤0, i need to meet i ∈ [0,1], if meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_k And γ_pSorting position exchange；

Second judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if It is unsatisfactory for, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

3rd judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meet, Then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then in the superiority-inferiority collating sequence γ_kAnd γ_pSorting position exchange；

4th judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, If it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th point The diversity factor of solution scheme and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and optimum specification desired value collection U Opposition degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pFor p-th decomposing scheme with it is optimal The diversity factor of standardization desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.

Beneficial effects of the present invention：

The technical scheme that the present invention is provided, complicated active power distribution network decomposing scheme Optimal Decision-making is carried out using Set Pair Analysis, Using the method clear concept of relative similarity degree evaluation of programme quality degree, calculate simple, be easy to programming realization；By set pair point Analysis has obtained confidence level preferably complexity active power distribution network decomposed decision scheme, meets and reaches resource utilization, parallel computation The composite factors such as complexity, parallel computation precision are optimal to be actually needed；The good and bad judge of scheme is carried out under the conditions of relative determination While, recycle relative unascertained information that ranking results are carried out with the analysis of stability, i stability regions are given, find it Its ranking results, never can determine metastable sequence in stable sort.

Brief description of the drawings

Fig. 1 is a kind of flow chart of complicated active power distribution network decomposing scheme method for optimizing of the present invention；

Fig. 2 is a kind of structural representation of complicated active power distribution network decomposing scheme preferred embodiment of the present invention.

Specific embodiment

Specific embodiment of the invention is elaborated below in conjunction with the accompanying drawings.

To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is A part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art The all other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.

A kind of complicated active power distribution network decomposing scheme method for optimizing that the present invention is provided, as shown in figure 1, including：

Each decomposing scheme in the 101. decomposing scheme set and the decomposing scheme set for obtaining complicated active power distribution network Corresponding evaluation index；

102. refer to according to the corresponding evaluation of each decomposing scheme in the decomposing scheme set and the decomposing scheme set Mark, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method；

103. pairs of superiority-inferiority collating sequences carry out stability analysis, update the superiority-inferiority collating sequence and select institute State and decomposing scheme corresponding to most preceding element is ranked in superiority-inferiority collating sequence as optimal Decomposition scheme.

Specifically, the step 101, including：

Determine that the corresponding resource utilization index of each decomposing scheme, parallel computation are complicated in the decomposing scheme set respectively Degree index, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and parallel computation cost refer to Mark.

The corresponding resource utilization index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₁：

In above formula, n is the number of partitions of network,It is i-th calculation scale of subregion in network, i ∈ [1, n]；

What resource utilization was characterized is server whole utilization efficiency, and each subregion is needed by coordinating after the completion of each calculating The calculating of each child partition Server Calculates Data amount of coordinating of server by utilizing.If each child partition calculation scale was differed Greatly, each subregion can be caused to wait mutually and causes the waste of resource.y₁Represented closer to 1 each subregion calculation scale closer to, The level of resources utilization is also higher.The index is income type index.

It is three steps that the step of parallel computation can sketch：Child servers each first carry out the calculating in respective region, and The relevant data result of Coordination is sent to coordination service device, the number that then coordination service device is sent by each child servers According to the correction value for calculating each Coordination, will finally coordinate the data is activation of value to each child servers, each child servers are to each Child partition carries out parallel computation.Calculating in general power system needs to carry out successive ignition, it is only necessary to repeat above-mentioned steps Until the precision required for meeting, determine that the corresponding parallel computation of each decomposing scheme is complicated in the decomposing scheme set as the following formula Degree index y₂：

In above formula, O_minIt is theoretical minimum parallel computation complexity,It is i-th computation complexity of subregion, l in network To coordinate level number of servers, M is calculation cost coefficient, and i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

Coordinating level server Main Resources is used to be responsible for the communication with each child partition, and calculation scale is compared each child partition and calculated Server should be small, and the Section 2 to denominator in formula (4) is multiplied by a coefficient M (2~5) to characterize cooperation layer clothes herein Business device is higher than the calculation cost of each child servers, the smaller parallel computation rule for representing used partition method of the value of formula (4) Mould is bigger.Work as y₂When=1, parallel computation complexity reaches theoretical minimum parallel computation complexity, and now parallel computation can be obtained Maximum speed-up ratio.The index is income type index.

By to known network be calculated the magnitude of voltage of each node in network, so as to obtain the voltage of Study system Level.The voltage change situation of each node before and after being decomposed by voltage level comparing cell, parallel computation is to whole after determining subregion The influence of body computational accuracy.The corresponding parallel computation precision index of each decomposing scheme in the decomposing scheme set is determined as the following formula y₃：

In above formula, U_i' it is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m It is nodes sum；

Speed-up ratio is the degree that system concurrency is played when reflecting parallel system operation concurrent program, it with hardware, it is soft Part and the characteristic of application have relation.Determine that the corresponding parallel computation of each decomposing scheme adds in the decomposing scheme set as the following formula Speed compares index S_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

After obtaining the corresponding evaluation index of each decomposing scheme in decomposing scheme set and the decomposing scheme set, institute Step 102 is stated, including：

Non- income type index in the corresponding evaluation index of each decomposing scheme is converted into income type index；

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme, the corresponding rule of each decomposing scheme are obtained Generalized desired value；

Select the optimum specification desired value collection U=(u in the corresponding standardization desired value of each decomposing scheme₁, u₂,...u_n) and most bad standardization desired value collection V=(v₁,v₂,...v_n), build each decomposing scheme comparing space [V, U]；

Wherein, it is to be analyzed in same scope, optimum scheme comparison and Worst scheme, both can result from scheme collection Inside, may also come from the outside of scheme collection, can be determined according to aims of systems and objective condition.

In the comparing space [V, U] of each decomposing scheme, each decomposing scheme and the optimum specification are determined The relative similarity degree of desired value collection U；

By the descending order pair of each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U Each decomposing scheme is ranked up, and obtains the superiority-inferiority collating sequence of each decomposing scheme.

Specifically, the non-income type index in the corresponding evaluation index by each decomposing scheme is converted to income type and refers to Mark, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula will Be converted to income type index：

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th maximum of index Desired value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

It is described that nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme, obtain each decomposing scheme correspondence Standardization desired value, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,It is k-th decomposition side On r-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum to case, and n is evaluation index sum.

It is described in the comparing space [V, U] of each decomposing scheme, determine each decomposing scheme with the optimal rule The relative similarity degree of generalized desired value collection U, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th point Solution scheme is d on r-th standardization desired value of index_kr, each decomposing scheme is in r-th standardization desired value of index Optimum specification desired value be u_r, each decomposing scheme on r-th standardization desired value of index in it is most bad standardization refer to Scale value is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimal rule as the following formula The Pair Analysis of generalized desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kFor k-th decomposing scheme and optimum specification refer to The identical degree of scale value collection U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kIt is k-th decomposition side The opposition degree of case and optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

Wherein, the present invention is it is determined that the relative similarity degree of each decomposing scheme and the optimum specification desired value collection U Before, need to first determine that each decomposing scheme concentrates the relative exchange premium degree of single optimal index, bag with the optimum specification desired value Include：

Note estimated plan is s_k=(d_k1,d_k2,...,d_kn) (k=1,2 ..., m), in e_rInterval [the v of comparing_r,u_r] Middle determination set pair { d_kr,u_rPair Analysis.

D can be represented_krAnd u_rDegree of closeness；D can be represented_krAnd v_rDegree of closeness.

In d_kr∈[v_r,u_r] when discussNumerical value：WhenWhen take minimum valueWork as d_kr=v_r Or u_rWhen take maximum

To makeIt is normalized, that is, usesRemoveWithRespectively obtain The two can be considered to d_krWith u_rThey can be respectively defined as set pair { d by the affirmation and negation of degree of closeness_kr, u_rIdentical degree and opposition degree.

According to a+b+c=1, set pair { d is calculated_kr,u_rDiversity factor be：

Thus { d_kr,u_rPair Analysis be：

From above formula, work as d_kr=u_rOr v_rWhen, diversity factor minimum zero；WhenWhen, diversity factor is to the maximum

In the complicated active power distribution network decomposing scheme decision-making technique of this paper, using μ_k=a_k+b_ki+c_kStablize relatively in j A_kAnd c_kConstitute and press close to degree γ relatively_kCarry out the good and bad situation of evaluation of programme.But b_kIt is relatively uncertain, its value size Probabilistic size is indicated, and the symbol and value of i can be considered b_kTo a_kOr c_kAmendment direction and amendment degree, will be right The evaluation result of scheme produces influence.It is ranked up surely it is therefore necessary to the evaluation result to complicated active power distribution network decomposing scheme Qualitative analysis, and find other ranking results in addition to basic label as much as possible, that is, be expanded sequence.

Analysis mode μ_k=a_k+b_ki+c_kJ can be obtained, and work as i>When 0, as to a_kPositive amendment, indicate to S_kClose to ideal The affirmative attitude of optimal case U, and i is closer to 1, this correcting action is stronger.Conversely, as i ＜ 0, as to c_kForward direction Amendment, indicates to S_kClose to the negative attitude of U, and i is closer to -1, and this correcting action is stronger.Therefore, on i [- 1, 1] change in, can carry out the stability analysis of schemes ranking, therefore, the step 103, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, and selection is described γ in superiority-inferiority collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme, when 0≤i≤1 and c_kb_p-c_pb_k≤0 When, i need to meet i ∈ [0,1], if meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if discontented Foot, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

When 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meetIf meeting, institute State γ in superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kWith γ_pSorting position exchange；

As -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meeting, superiority-inferiority sequence γ in sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position Exchange；

As -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_k And γ_pSorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th point The diversity factor of solution scheme and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and optimum specification desired value collection U Opposition degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pFor p-th decomposing scheme with it is optimal The diversity factor of standardization desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.

The present invention also provides a kind of complicated active power distribution network decomposing scheme preferred embodiment, as shown in Fig. 2 described device bag Include：

Acquisition module, for each in the decomposing scheme set and the decomposing scheme set that obtain complicated active power distribution network The corresponding evaluation index of decomposing scheme；

Determining module, for according to each decomposing scheme correspondence in the decomposing scheme set and the decomposing scheme set Evaluation index, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method；

Analysis module, for carrying out stability analysis to the superiority-inferiority collating sequence, updates the superiority-inferiority sequence sequence Arrange and select to rank in the superiority-inferiority collating sequence decomposing scheme corresponding to most preceding element as optimal Decomposition scheme.

The acquisition module, including：

First determining unit, determines that the corresponding resource utilization of each decomposing scheme refers in the decomposing scheme set respectively Mark, parallel computation complexity index, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and The parallel computation indicator of costs.

Wherein, the corresponding resource utilization index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₁：

In above formula, n is the number of partitions of network,It is i-th calculation scale of subregion in network, i ∈ [1, n]；

The corresponding parallel computation complexity index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₂：

In above formula, O_minIt is theoretical minimum parallel computation complexity,It is i-th computation complexity of subregion, l in network To coordinate level number of servers, M is calculation cost coefficient, and i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

The corresponding parallel computation precision index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₃：

In above formula, U_i' it is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m It is nodes sum；

The corresponding parallel computation speed-up ratio index S of each decomposing scheme in the decomposing scheme set is determined as the following formula_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

The determining module, including：

Converting unit, refers to for the non-income type index in the corresponding evaluation index of each decomposing scheme to be converted into income type Mark；

Specification unit, for carrying out nondimensionalization treatment to the corresponding evaluation index of each decomposing scheme, obtains each point The corresponding standardization desired value of solution scheme；

Select unit, for selecting the optimum specification desired value in the corresponding standardization desired value of each decomposing scheme Collection U=(u₁,u₂,...u_n) and most bad standardization desired value collection V=(v₁,v₂,...v_n), build the comparing of each decomposing scheme Space [V, U]；

Second determining unit, in the comparing space [V, U] of each decomposing scheme, determining each decomposing scheme With the relative similarity degree of the optimum specification desired value collection U；

Sequencing unit, for by each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U Descending order is ranked up to each decomposing scheme, obtains the superiority-inferiority collating sequence of each decomposing scheme.

The converting unit, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula will Be converted to income type index：

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th maximum of index Desired value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

The specification unit, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,It is k-th decomposition side On r-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum to case, and n is evaluation index sum.

Second determining unit, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th point Solution scheme is d on r-th standardization desired value of index_kr, each decomposing scheme is in r-th standardization desired value of index Optimum specification desired value be u_r, each decomposing scheme on r-th standardization desired value of index in it is most bad standardization refer to Scale value is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimal rule as the following formula The Pair Analysis of generalized desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kFor k-th decomposing scheme and optimum specification refer to The identical degree of scale value collection U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kIt is k-th decomposition side The opposition degree of case and optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

The analysis module, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, and selection is described γ in superiority-inferiority collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme；

First judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kWhen≤0, i need to meet i ∈ [0,1], if meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_k And γ_pSorting position exchange；

Second judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if It is unsatisfactory for, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

3rd judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meet, Then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then in the superiority-inferiority collating sequence γ_kAnd γ_pSorting position exchange；

4th judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, If it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th point The diversity factor of solution scheme and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and optimum specification desired value collection U Opposition degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pFor p-th decomposing scheme with it is optimal The diversity factor of standardization desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.

Finally it should be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention rather than its limitations, to the greatest extent Pipe has been described in detail with reference to above-described embodiment to the present invention, and those of ordinary skill in the art should be understood：Still Specific embodiment of the invention can be modified or equivalent, and without departing from any of spirit and scope of the invention Modification or equivalent, it all should cover within claims of the invention.

Claims (16)

1. a kind of complicated active power distribution network decomposing scheme method for optimizing, it is characterised in that methods described includes：

Each decomposing scheme is corresponding in the decomposing scheme set and the decomposing scheme set of the complicated active power distribution network of acquisition comments Valency index；

According to the corresponding evaluation index of each decomposing scheme in the decomposing scheme set and the decomposing scheme set, using collection The superiority-inferiority collating sequence of each decomposing scheme is determined to analytic approach；

Stability analysis is carried out to the superiority-inferiority collating sequence, the superiority-inferiority collating sequence is updated and is selected the superiority-inferiority The decomposing scheme corresponding to most preceding element is ranked in collating sequence as optimal Decomposition scheme.

2. the method for claim 1, it is characterised in that the decomposing scheme set of the acquisition complexity active power distribution network with And the corresponding evaluation index of each decomposing scheme in the decomposing scheme set, including：

Determine that the corresponding resource utilization index of each decomposing scheme, parallel computation complexity refer in the decomposing scheme set respectively Mark, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and the parallel computation indicator of costs.

3. method as claimed in claim 2, it is characterised in that determine each decomposing scheme in the decomposing scheme set as the following formula Corresponding resource utilization index y₁：

$y_1=\left\{\frac{min(P_1^2,P_2^2...P_{n-1}^2,P_n^2)}{max(P_1^2,P_2^2...P_{n-1}^2,P_n^2)}\right\}\×100\%$

In above formula, n is the number of partitions of network, P_i ²It is i-th calculation scale of subregion in network, i ∈ [1, n]；

The corresponding parallel computation complexity index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₂：

$y_2=\left(\frac{O_{\min }}{max\left(P_i^2\right)+M\×l^2}\right)\×100\%$

In above formula, O_minIt is theoretical minimum parallel computation complexity, P_i ²It is i-th computation complexity of subregion in network, l is association Adjust a wage scale number of servers, M is calculation cost coefficient, i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

The corresponding parallel computation precision index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₃：

$y_3=max(1-|\frac{U_i^{\′}-U_i}{U_i}\left|\right)\×100\%$

In above formula, U '_iIt is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m is net Network interior joint sum；

The corresponding parallel computation speed-up ratio index S of each decomposing scheme in the decomposing scheme set is determined as the following formula_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

4. the method for claim 1, it is characterised in that described according to the decomposing scheme set and the decomposition side The corresponding evaluation index of each decomposing scheme in case set, determines that the superiority-inferiority of each decomposing scheme sorts sequence using Set Pair Analysis Method Row, including：

Non- income type index in the corresponding evaluation index of each decomposing scheme is converted into income type index；

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme, the corresponding standardization of each decomposing scheme is obtained Desired value；

Select the optimum specification desired value collection U=(u in the corresponding standardization desired value of each decomposing scheme₁,u₂,...u_n) Most bad standardization desired value collection V=(v₁,v₂,...v_n), build the comparing space [V, U] of each decomposing scheme；

In the comparing space [V, U] of each decomposing scheme, determine each decomposing scheme with the optimum specification index The relative similarity degree of value collection U；

By the descending order of each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U to described Each decomposing scheme is ranked up, and obtains the superiority-inferiority collating sequence of each decomposing scheme.

5. method as claimed in claim 4, it is characterised in that the non-receipts in the corresponding evaluation index by each decomposing scheme Beneficial type index is converted to income type index, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula willConversion It is income type index：

${{\stackrel{\‾}{d}}_{kr}}^{\′}={\stackrel{\‾}{d}}_{krmax}-{\stackrel{\‾}{d}}_{kr}$

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th Maximum Index of index Value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

6. method as claimed in claim 4, it is characterised in that described to be carried out to the corresponding evaluation index of each decomposing scheme Nondimensionalization treatment, obtains the corresponding standardization desired value of each decomposing scheme, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

$d_{kr}=\frac{{\stackrel{\‾}{d}}_{kr}}{\sqrt{\underset{k=1}{\overset{m}{\Σ}}{\stackrel{\‾}{d}}_{kr}^2}}$

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,For k-th decomposing scheme is closed In r-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

7. method as claimed in claim 4, it is characterised in that described in the comparing space [V, U] of each decomposing scheme, Determine the relative similarity degree of each decomposing scheme and the optimum specification desired value collection U, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th decomposition side Case is d on r-th standardization desired value of index_kr, each decomposing scheme in r-th standardization desired value of index most Excellent standardization desired value is u_r, each decomposing scheme is on the most bad standardization desired value in r-th standardization desired value of index It is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimum specification as the following formula The Pair Analysis of desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kIt is k-th decomposing scheme and optimum specification desired value Collect the identical degree of U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kFor k-th decomposing scheme with The opposition degree of optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

${\mathrm{\γ}}_k=\frac{a_k}{a_k+c_k}.$

8. the method for claim 1, it is characterised in that described that analysis of stability is carried out to the superiority-inferiority collating sequence Analysis, updates the superiority-inferiority collating sequence and selects to rank the decomposition side corresponding to most preceding element in the superiority-inferiority collating sequence Case as optimal Decomposition scheme, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, selects the quality γ in property collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme, when 0≤i≤1 and c_kb_p-c_pb_kWhen≤0, i I ∈ [0,1] need to be met, if meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

When 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meetIf meeting, the quality γ in property collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pRow Sequence location swap；

As -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meet, in the superiority-inferiority collating sequence γ_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

As -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIt is described if meeting γ in superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_p Sorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th decomposition side The diversity factor of case and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and the opposition of optimum specification desired value collection U Degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pIt is p-th decomposing scheme and optimum specification Change the diversity factor of desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.

9. a kind of complicated active power distribution network decomposing scheme preferred embodiment, it is characterised in that described device includes：

Acquisition module, for respectively being decomposed in the decomposing scheme set and the decomposing scheme set that obtain complicated active power distribution network The corresponding evaluation index of scheme；

Determining module, for commenting according to each decomposing scheme is corresponding in the decomposing scheme set and the decomposing scheme set Valency index, the superiority-inferiority collating sequence of each decomposing scheme is determined using Set Pair Analysis Method；

Analysis module, for carrying out stability analysis to the superiority-inferiority collating sequence, updates the superiority-inferiority collating sequence simultaneously Select to rank in the superiority-inferiority collating sequence decomposing scheme corresponding to most preceding element as optimal Decomposition scheme.

10. device as claimed in claim 9, it is characterised in that the acquisition module, including：

First determining unit, determine respectively in the decomposing scheme set the corresponding resource utilization index of each decomposing scheme and Row computation complexity index, parallel computation precision index, parallel computation speed-up ratio index, parallel efficiency calculation index and parallel meter Calculate the indicator of costs.

11. devices as claimed in claim 10, it is characterised in that determine each decomposition side in the decomposing scheme set as the following formula The corresponding resource utilization index y of case₁：

$y_1=\left\{\frac{min(P_1^2,P_2^2...P_{n-1}^2,P_n^2)}{max(P_1^2,P_2^2...P_{n-1}^2,P_n^2)}\right\}\×100\%$

In above formula, n is the number of partitions of network, P_i ²It is i-th calculation scale of subregion in network, i ∈ [1, n]；

The corresponding parallel computation complexity index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₂：

$y_2=\left(\frac{O_{\min }}{max\left(P_i^2\right)+M\×l^2}\right)\×100\%$

In above formula, O_minIt is theoretical minimum parallel computation complexity, P_i ²It is i-th computation complexity of subregion in network, l is association Adjust a wage scale number of servers, M is calculation cost coefficient, i ∈ [1, n], M ∈ [2,5], n are the number of partitions of network；

The corresponding parallel computation precision index y of each decomposing scheme in the decomposing scheme set is determined as the following formula₃：

$y_3=max(1-|\frac{U_i^{\′}-U_i}{U_i}\left|\right)\×100\%$

In above formula, U_i' it is network decomposition posterior nodal point i magnitudes of voltage, U_iIt is network decomposition front nodal point i magnitudes of voltage, i ∈ [1, m], m is net Network interior joint sum；

The corresponding parallel computation speed-up ratio index S of each decomposing scheme in the decomposing scheme set is determined as the following formula_P：

S_P=T_S/T_P

In above formula, T_STime required for serial Solve problems, T_PTime required for Parallel implementation problem；

The corresponding parallel efficiency calculation index E of each decomposing scheme in the decomposing scheme set is determined as the following formula：

E=S_P/P

In above formula, S_PIt is parallel computation speed-up ratio, P is parallel computation required service device quantity；

The corresponding parallel computation indicator of costs C of each decomposing scheme in the decomposing scheme set is determined as the following formula：

C=T_P*P

In above formula, T_PTime required for Parallel implementation problem, P is parallel computation required service device quantity.

12. devices as claimed in claim 9, it is characterised in that the determining module, including：

Converting unit, for the non-income type index in the corresponding evaluation index of each decomposing scheme to be converted into income type index；

Specification unit, for carrying out nondimensionalization treatment to the corresponding evaluation index of each decomposing scheme, obtains each decomposition side The corresponding standardization desired value of case；

Select unit, for selecting the optimum specification desired value collection U in the corresponding standardization desired value of each decomposing scheme =(u₁,u₂,...u_n) and most bad standardization desired value collection V=(v₁,v₂,...v_n), the comparing for building each decomposing scheme is empty Between [V, U]；

Second determining unit, in the comparing space [V, U] of each decomposing scheme, determining each decomposing scheme and institute State the relative similarity degree of optimum specification desired value collection U；

Sequencing unit, for by each decomposing scheme and the relative similarity degree of the optimum specification desired value collection U from big Each decomposing scheme is ranked up to small order, obtains the superiority-inferiority collating sequence of each decomposing scheme.

13. devices as claimed in claim 12, it is characterised in that the converting unit, including：

Assuming that k-th decomposing scheme is on r-th desired value of indexIt is non-income type index, then as the following formula willConversion It is income type index：

${{\stackrel{\‾}{d}}_{kr}}^{\′}={\stackrel{\‾}{d}}_{krmax}-{\stackrel{\‾}{d}}_{kr}$

In above formula,ForIncome type desired value,It is k-th decomposing scheme on r-th Maximum Index of index Value, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

14. devices as claimed in claim 12, it is characterised in that the specification unit, including：

Nondimensionalization treatment is carried out to the corresponding evaluation index of each decomposing scheme as the following formula：

$d_{kr}=\frac{{\stackrel{\‾}{d}}_{kr}}{\sqrt{\underset{k=1}{\overset{m}{\Σ}}{\stackrel{\‾}{d}}_{kr}^2}}$

In above formula, d_krIt is k-th decomposing scheme on r-th standardization desired value of index,For k-th decomposing scheme on R-th desired value of index, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum.

15. devices as claimed in claim 12, it is characterised in that second determining unit, including：

If decomposing scheme set S={ s₁,s₂,...,s_m, evaluation index set E={ e₁,e₂,...,e_n, remember k-th decomposition side Case is d on r-th standardization desired value of index_kr, each decomposing scheme in r-th standardization desired value of index most Excellent standardization desired value is u_r, each decomposing scheme is on the most bad standardization desired value in r-th standardization desired value of index It is v_rWherein, k ∈ [1, m], r ∈ [1, n], m are decomposing scheme sum, and n is evaluation index sum；

In the comparing space [V, U] of each decomposing scheme, determine k-th decomposing scheme with the optimum specification as the following formula The Pair Analysis of desired value collection U, i.e. set pair { s_k, U } Pair Analysis u { s_k,U}：

u{s_k, U } and=a_k+b_ki+c_kj

In above formula, s_kIt is k-th decomposing scheme, a in decomposing scheme set_kIt is k-th decomposing scheme and optimum specification desired value Collect the identical degree of U, b_kIt is k-th decomposing scheme and the diversity factor of optimum specification desired value collection U, c_kFor k-th decomposing scheme with The opposition degree of optimum specification desired value collection U, i is diversity factor coefficient, and j is opposition label symbol；

Wherein, i ∈ [- 1,1], j=1,

The relative similarity degree γ of k-th decomposing scheme and the optimum specification desired value collection U is determined as the following formula_k：

${\mathrm{\γ}}_k=\frac{a_k}{a_k+c_k}.$

16. devices as claimed in claim 9, it is characterised in that the analysis module, including：

Make γ_k＞ γ_p, then with γ_pCompare, γ_kIt is the forward element that sorted in the superiority-inferiority collating sequence；

The diversity factor coefficient i of the corresponding decomposing scheme of each element is equal in the superiority-inferiority collating sequence, selects the quality γ in property collating sequence_kCorresponding decomposing scheme and γ_pCorresponding decomposing scheme；

First judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kWhen≤0, i need to meet i ∈ [0,1], described if meeting γ in superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_p Sorting position exchange；

Second judging unit, for when 0≤i≤1 and c_kb_p-c_pb_kDuring ＞ 0, i needs to meet If meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then superiority-inferiority sequence γ in sequence_kAnd γ_pSorting position exchange；

3rd judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kWhen >=0, i need to meet i ∈ [- 1,0), if meet, institute State γ in superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, if it is not satisfied, then γ in the superiority-inferiority collating sequence_kWith γ_pSorting position exchange；

4th judging unit, for as -1≤i ＜ 0 and a_kb_p-a_pb_kDuring ＜ 0, i needs to meetIf meeting, γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position it is constant, If it is not satisfied, then γ in the superiority-inferiority collating sequence_kAnd γ_pSorting position exchange；

In said process, a_kIt is k-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_kIt is k-th decomposition side The diversity factor of case and optimum specification desired value collection U, c_kIt is k-th decomposing scheme and the opposition of optimum specification desired value collection U Degree, a_pIt is p-th decomposing scheme and the identical degree of optimum specification desired value collection U, b_pIt is p-th decomposing scheme and optimum specification Change the diversity factor of desired value collection U, c_pIt is p-th decomposing scheme and the opposition degree of optimum specification desired value collection U.