CN107292534A - The yardstick competition evaluation method and device of urban power distribution network long term dynamics investment - Google Patents

Abstract

The invention relates to a scale competition evaluation method and device for medium and long-term dynamic investment in urban distribution networks. According to the weight distribution of each type of evaluation criteria; through the data envelopment analysis method to analyze the long-term statistical data of each evaluation index, each evaluation index in each type of evaluation criteria is divided into input indicators and output indicators; according to the input indicators and output indicators, use The CCR model of the data envelopment analysis method, and then obtain the efficiency evaluation value of each planning cycle of various criteria; use the dynamic weighting vector to carry out linear dynamic weighted aggregation on the efficiency evaluation value of each planning cycle to obtain the dynamic evaluation result; according to the dynamic evaluation result , establish a dynamic assessment scale, and conduct scale assessment on construction cost and construction efficiency respectively. The invention is suitable for indirect competition among similar enterprises in different regions, and realizes the maximization of resource allocation efficiency.

Description

Scale competition evaluation method and device for medium and long-term dynamic investment in urban distribution network

technical field

The invention relates to technical issues such as power distribution system planning and operations research, and specifically relates to a scale competition evaluation method and device for medium and long-term dynamic investment in urban power distribution networks.

Background technique

With the access of large-scale clean energy, the deepening of supply and demand interaction, the liberalization of incremental power distribution business, and the implementation of power substitution strategy, the distribution network, as an important link to ensure high-quality and high-reliability power consumption of users, has been difficult to adapt to the new situation Under the development needs of power supply reliability and distribution intelligence, my country's annual investment in distribution networks is huge and is increasing year by year. Upgrading and transforming weak links in power distribution systems and making precise investments are important problems in power grid construction. Therefore, it is of great significance to study a set of reasonable, scientific and comprehensive medium and long-term dynamic investment strategy evaluation models for distribution network planning, construction and transformation.

The medium and long-term dynamic investment strategy evaluation of distribution network is a typical dynamic comprehensive evaluation problem involving multiple objects, multiple indicators and multiple time periods. There are large differences due to different operational objectives. It is difficult to objectively understand the spatio-temporal characteristics and regional differences of distribution network construction and development only by observing and analyzing a large amount of data. In the actual application of existing projects, most of the expert scoring methods are used. The evaluation results are greatly affected by the subjectivity of the experts, and the amount of information reflected in the evaluation results is relatively small. It is difficult to objectively, comprehensively, and dynamically reflect the actual development of the distribution network. Condition. At the same time, some scholars have proposed the post-evaluation model of distribution network investment effect, which mainly evaluates the operation effect and investment efficiency of distribution network investment and construction from the perspective of operation, and lacks the optimization of several construction schemes from the perspective of planning. The evaluation is poor to ensure the accuracy of distribution network investment planning, and there is also a lack of dynamic rolling evaluation of distribution network planning strategies for several planning cycles. The evaluation model only reflects the construction benefits of distribution network, and lacks incentives for evaluation results. In order to promote the benefits of distribution network planning and construction; and for distribution network performance indicators, some scholars have constructed evaluation index systems from distribution network operation capacity, structure, reliability, etc., but ignored the impact of investment construction on distribution network. The comprehensiveness of the evaluation needs to be further strengthened; some scholars have proposed a three-level comprehensive evaluation index system for smart distribution networks that considers distribution network characteristics, economics, and social benefits, but ignores the calculation and evaluation of indicators. The study of the method reduces the actual operability.

It is concluded that the existing evaluation models for the efficiency of distribution network investment and construction mainly have the following three deficiencies: the evaluation method is highly subjective, the index system is relatively low in comprehensiveness, and the evaluation time scale is single. Distribution network investment planning is a medium- and long-term dynamic rolling planning problem. The construction scale is huge, and the construction goals vary greatly with regions. The accuracy of investment strategies will directly affect the efficiency of distribution network upgrading and transformation.

Contents of the invention

The purpose of the present invention is to address the above deficiencies and provide a scale competition evaluation method and device for medium and long-term dynamic investment in urban distribution networks, so as to realize optimal allocation of resources between regions and improve the accuracy of investment in distribution networks.

The solution adopted by the present invention to solve the technical problem is: a scale competition evaluation method for medium and long-term dynamic investment in urban power distribution network, comprising the following steps:

Step S1: select multiple evaluation indicators, and classify the evaluation indicators into six types of evaluation criteria. The six types of evaluation criteria are: power supply quality, grid structure, equipment level, power supply capacity, informatization level, and investment capacity;

Step S2: assign weights to the six types of evaluation criteria through fuzzy analytic hierarchy process, and obtain the weights of various evaluation criteria;

Step S3: Analyze the long-term statistical data of each evaluation index by the data envelopment analysis method, and divide each evaluation index in each type of evaluation criteria into input index and output index, where the input index refers to the decision-maker engaged in distribution network investment The input amount of the construction, the output index refers to the effective output obtained by the decision-maker through the investment and construction of the distribution network;

Step S4: according to the input index and the output index, utilize the CCR model of the data envelopment analysis method, and then obtain the efficiency evaluation value of each planning cycle of various criteria;

Step S5: Using the dynamic weighting vector, the efficiency evaluation values of each planning cycle are linearly and dynamically weighted to obtain the dynamic evaluation results.

Further, the decision-making method of the weight of the six types of evaluation criteria is evaluated by several experts, and the weight vector of each expert is obtained through the fuzzy analytic hierarchy process, as follows:

Let the fuzzy complementary matrix F=(f _ij ) _n×n (f _ij ∈[0,1]), if f _ij =F(a _i ,a _j ), then f _ij means a _i and a _j "... ratio... much more important” fuzzy membership relationship, f _ij uses 0.1-0.9 to give a quantitative scale, and F has the following properties:

(1) f _ii =0.5, i=1,2,...,n;

(2)f _ij +f _ji =1,i,j=1,2,...,n;

(3) There is a normalized vector U _z =(u _z1 ,u _z2 ,…,u _zn ) and α(α>1), for any i, j, f _ij =log _α u _zi -log _α u _zj +0.5; Among them, α represents the resolution ability of the decision maker, U _z represents the weight decision vector of the zth expert on the evaluation criteria, u _zi and u _zj are two elements of U _z , which represent the expert z’s evaluation criteria i and Evaluate the weight decision result of j, n is the number of evaluation criteria, n=6;

The weight vector U _z =(u _z1 ,u _z2 ,...,u _zn ) is determined by the solution of the following constraint programming problem:

Using the Lagrange multiplier method to transform it into an unconstrained programming problem, we can obtain:

The weight vector of each expert is obtained through the above formula (2).

Further, the decision-making results of each expert are clustered and analyzed by using the cluster analysis method, and the decision weight vector is uniformly obtained, which specifically includes the following steps:

Step S21: Assuming that the weight distribution of various criteria in the distribution network is completed by N experts, input _N 6-dimensional expert decision vector sets D=(U ₁ , U ₂ ,...,U _z ,...,UN ), Among them, U _z =(u _z1 ,u _z2 ,…,u _z6 ) represents the weight distribution result of expert z to the 6 types of evaluation criteria, and the number of clusters to be classified is k;

Step S22: Randomly select the decision vectors of k experts to assign weights to the 6 types of distribution network evaluation criteria as the initial clustering centers {p ₁ ,p ₂ ,…,p _k }, where p _i ={p _i1 ,p _i2 , ..., p _i6 } represents the weight decision vector of the i-th clustering center expert; select the maximum number of clustering iterations P; determine the maximum convergence coefficient M at the end of the iteration;

Step S23: Calculate the Euclidean distances from each of the k decision vectors to each cluster, and divide each decision vector into the cluster with the smallest distance. The formula for calculating the Euclidean distance is:

In formula (3), dist(U _z , p _j ) represents the distance from the zth expert decision vector to the jth cluster;

Step S24: recalculate the center values {p ₁ , p ₂ ,...,p _k } of the k clusters, where p _j ={p _j1 ,p _j2 ,...,p _j6 }, and L represents the center of the cluster The number of decision vectors of , the calculation formula is:

Step S25: Check whether the clustering operation is over: if the number of iterations is equal to P, then end the clustering; otherwise, calculate the convergence distance of each cluster in this iteration, if the convergence distance is less than the given parameter M, then end, otherwise continue to iterate , the formula for calculating the convergence distance of the mth iteration is:

Step S26: Assuming that category p _l includes individual sorting vectors n _p , the weight η _l of this type of expert decision vector is calculated using the ratio of the expert decision vector to the total number of decision vectors:

From this, the final weight vector is obtained as:

In formula (7), p _l represents the cluster center value of the lth class.

Further, the clustering quality is judged by the silhouette coefficient method, and the weight vector with the best clustering quality is selected, which specifically includes the following steps:

Step S27: Assuming that the expert decision vector set D is divided into k clusters p ₁ , p ₂ ,...,p _k , for each decision vector u∈D, calculate the average distance b(u ), similarly, c(u) represents the minimum average distance from u to all clusters that do not belong to u.

The calculation formula of the silhouette coefficient of vector u is:

Step S28: Judging the clustering quality by the silhouette coefficient, and obtaining the weight vector with the best clustering quality.

Further, in step S4, the construction of the CCR model includes the following steps:

Step S41: Assuming that distribution networks in h regions are planned and constructed at the same time, the original input indicators and output indicators of the planned input construction amount and expected effective output of the j-th regional distribution network are respectively with Each regional distribution network is expected to have s types of input and r types of output;

Step S42: Carry out normalization processing on the input index and output index, wherein, the positive input index is processed by formula (11) and formula (12), and the negative input index is processed by formula (13) and formula (14):

x _ij represents the input amount of the j-th regional distribution network to the i-th type of input index; y _rj represents the effective output of the r-th type of output index obtained by the j-th regional distribution network; with are the original input and output data respectively;

Step S43: Let v _i be the weight coefficient variable of the i-th type of input, f _r be the weight coefficient variable of the r-th type of output, and the ratio of the output volume to the input volume of various criteria is defined as the distribution network investment construction period t Efficiency index:

Step S44: Taking the benefit index of the distribution network in the j _0th region as the target and the benefit indices of all distribution networks as constraints, construct an optimal CCR model:

θ _{t, j0} = 1 indicates that the benefits of distribution network construction in this area are relatively high, and θ _{t, j0} < 1 indicates that the benefits of distribution network construction in this area are relatively low.

Further, in step S5, introduce information entropy and time degree to calculate dynamic weight vector; the definition formula of dynamic weight vector τ _t and time degree β is as follows:

In the formula, τ=[τ ₁ , τ ₂ ,...,τ _t ] represents the time series weighted vector, which reflects the difference in the contribution of investment strategies in different planning cycles to the dynamic evaluation, τ _t ∈ [0, 1], the information of the dynamic weighted vector Entropy reflects the extent to which the weight contains information in the dynamic weighting process of the investment efficiency evaluation value of each period. The smaller the entropy value, the greater the amount of information it acquires; The degree of emphasis on each cycle in , the smaller the value, the more attention to recent data; the time is predetermined by the decision maker, and the mathematical model of dynamic weighted vector calculation can be expressed as:

Further, according to the evaluation result of step S5, a dynamic assessment scale is established, and the scale assessment is carried out on the construction cost and construction efficiency respectively. The scale assessment is as follows: the regulatory agency gives corresponding assessment rewards to regions with high distribution network investment efficiency , on the contrary, for areas with low distribution network investment efficiency, assessment penalties are given, and the mathematical model is as follows:

r _j,t = ρ _j,t + g _j,t (21)

In formula (21), r _j , t is the return obtained by region j for distribution network construction within the construction period t; ρ _j , t is the construction income obtained by region j within the construction period t, positive numbers indicate rewards, and negative numbers Indicates the penalty; g _j , t is the cost subsidy obtained by region j in the construction period t.

Further, the calculation of the construction income includes the following steps:

Step S61: Calculate the efficiency evaluation value θ _j ' and dynamic weighting vector τ _t of distribution network construction in each region;

Step S62: Carry out linear weighting on the distribution network investment efficiency evaluation value of each period, and obtain the distribution network investment efficiency dynamic evaluation value θ _t,j of region j in the construction period t, namely:

θ _t,j ^dynamic ＝τ ₁ θ _1,j +τ ₂ θ _2,j +…+τ _t θ _t,j (22)

Step S63: Define the reward and punishment coefficient ξ _t,j of region j in the construction period t as:

Calculate the construction income of distribution network in region j in construction period t:

In the formula, P _t , j represents the added value of distribution network investment in region j during construction period t compared with period t-1.

Furthermore, the construction cost subsidy of the distribution network in each region is jointly determined by its own construction cost and the construction cost of other regions, and its calculation method is shown in the following formula:

In formula (25), L _{j, t} represents the investment in the distribution network construction of region j in the construction period t; c _{j, t} is the unit construction cost of the distribution network in region j in the construction period t, and ε _j is the area The proportion of j's own cost, l _i , _t is the weight of the distribution network construction cost of company i in the observation period t.

The present invention also provides a device based on the above-mentioned scale competition evaluation method for medium and long-term dynamic investment in urban distribution networks, including a central processing unit, a storage module, and a display module;

The central processing unit is used to perform the following steps:

Step S1: select multiple evaluation indicators, and classify the evaluation indicators into six types of evaluation criteria. The six types of evaluation criteria are: power supply quality, grid structure, equipment level, power supply capacity, informatization level, and investment capacity;

Step S2: assign weights to the six types of evaluation criteria through fuzzy analytic hierarchy process, and obtain the weights of various evaluation criteria;

Step S3: Analyze the long-term statistical data of each evaluation index by the data envelopment analysis method, and divide each evaluation index in each type of evaluation criteria into input index and output index, where the input index refers to the decision-maker engaged in distribution network investment The input amount of the construction, the output index refers to the effective output obtained by the decision-maker through the investment and construction of the distribution network;

Step S4: according to the input index and the output index, utilize the CCR model of the data envelopment analysis method, and then obtain the efficiency evaluation value of each planning cycle of various criteria;

Step S5: Using the dynamic weighting vector, the efficiency evaluation values of each planning cycle are linearly and dynamically weighted to obtain the dynamic evaluation results.

Compared with the prior art, the present invention has the following beneficial effects: the present invention proposes an evaluation model, selects similar enterprises as research objects, and utilizes the idea of inter-regional comparison and evaluation to accurately evaluate distribution network investment strategies from space and time dimensions. Judgment, select areas with better investment efficiency as benchmark areas, and evaluate the evaluation values of other areas with reference to benchmark areas, which overcomes the shortcomings of difficult to unify evaluation standards and relatively subjective evaluation methods, and the evaluation results can objectively reflect the various aspects of the distribution network. Investment planning efficiency, so as to provide guidance for precise investment in distribution network. This model is suitable for indirect competition between similar enterprises in different regions, and has obvious advantages in minimizing construction costs and maximizing resource allocation efficiency.

Description of drawings

Below in conjunction with accompanying drawing, the patent of the present invention is further described.

Figure 1 is a flow chart of the evaluation model of the medium and long-term dynamic investment strategy of the distribution network.

Figure 2 is a diagram of the distribution network investment strategy scale evaluation index system.

Figure 3 is a bar chart of the province's 2016-2022 distribution network investment planning efficiency evaluation value.

Figure 4 is a radar map of the investment efficiency evaluation values of each criterion in typical regions of the province from 2016 to 2022.

Figure 5 is a schematic diagram of the reward and punishment coefficient of the province from 2016 to 2022.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figures 1-2, a scale competition evaluation method for mid- and long-term dynamic investment in urban distribution networks in this embodiment includes the following steps:

Step S1: select multiple evaluation indicators, and classify the evaluation indicators into six types of evaluation criteria. The six types of evaluation criteria are: power supply quality, grid structure, equipment level, power supply capacity, informatization level, and investment capacity;

Step S2: assign weights to the six types of evaluation criteria through fuzzy analytic hierarchy process, and obtain the weights of various evaluation criteria;

Step S3: Analyze the long-term statistical data of each evaluation index by the data envelopment analysis method, and divide each evaluation index in each type of evaluation criteria into input index and output index, where the input index refers to the decision-maker engaged in distribution network investment The input amount of the construction, the output index refers to the effective output obtained by the decision-maker through the investment and construction of the distribution network;

Step S4: according to the input index and the output index, utilize the CCR model of the data envelopment analysis method, and then obtain the efficiency evaluation value of each planning cycle of various criteria;

Step S5: Using the dynamic weighting vector, the efficiency evaluation values of each planning cycle are linearly and dynamically weighted to obtain the dynamic evaluation results.

In this embodiment, in step S1, the evaluation index is selected based on the reality of distribution network planning and construction, combined with industry standards implemented within the electric power enterprise.

Figure 1 shows the scale evaluation model of medium and long-term dynamic investment strategy of distribution network. The evaluation of medium and long-term dynamic investment strategy of distribution network is mainly divided into two stages: index system construction and evaluation model construction. The construction of the index system needs to determine the evaluation index, evaluation criterion and evaluation standard according to the overall goal of the regional distribution network construction in the planning cycle, and assign the weight of the index according to the key points of the planning and construction; the evaluation stage mainly completes the quantitative evaluation of each index and the planning cycle. Weighting of evaluation values.

Proceeding from the reality of distribution network planning and construction, combined with the industry standards implemented by power companies, the following index system is selected: power outage time, number of main transformers passing through N-1 in the power grid, number of main transformers, number of lines passing through N-1 lines in the power grid, The number of cable lines, the number of overhead lines, the number of heavy-duty lines, the number of heavy-duty main transformers, the sum of power supply radius, the number of low-voltage power supply users, the capacity of public distribution transformers, the number of distribution transformers for information collection, the number of smart meters, net assets etc. The above index system is divided into six categories of criteria for evaluating the accuracy of investment strategies, namely power supply quality, grid structure, equipment level, power supply capacity, informatization level, and investment capacity. Taking the relative construction efficiency of the distribution network as the overall goal, the six types of evaluation criteria as the middle layer, and each evaluation index into the corresponding criterion layer, a hierarchical structure model for evaluating the accuracy of distribution network investment strategies can be established, as shown in Figure 2 .

The present invention adopts the data envelopment analysis method for the evaluation of the index layer, and its basic idea is to divide the evaluation index into "input index" and "output index", and determine the DEA evaluation value by analyzing the ratio relationship between the two, wherein, " "Input index" refers to the amount of input that decision makers engage in investment and construction of distribution network, and "output index" refers to the effective output obtained by decision makers through investment and construction of distribution network. The input and output indicators can be analyzed and determined from the perspective of data correlation analysis on long-term statistical data, and there is a certain qualitative positive correlation between the two. The DEA input-output relationship of each criterion is shown in Table 1:

Table 1 DEA input and output indicators of each criterion

Distribution network investment planning aims to improve the quality of distribution network power supply, optimize the structure of distribution network lines, improve the level of distribution network equipment, enhance the power supply capacity of distribution network, and improve the information level of distribution network. The evaluation of long-term dynamic investment strategies is a typical multi-indicator, multi-disciplinary and multi-dimensional unstructured index system evaluation problem. The distribution of index weights is one of the key issues in the evaluation, which needs to be assisted by mathematical analysis methods. Fuzzy analytic hierarchy process is a multi-index weight distribution method combining fuzzy mathematics and analytic hierarchy process. It improves the problems that the analytic hierarchy process judgment matrix consistency index is difficult to achieve and the consistency of the judgment matrix is different from the consistency of people's decision-making thinking. The situation that "A is more important than B, B is more important than C, and C is more important than A" violates common sense. The application of fuzzy analytic hierarchy process can eliminate the uncertainty in the distribution of index weights, and decision makers can distribute power according to different planning cycles Reasonably change the fuzzy matrix according to the goal of network construction, so that the evaluation results can better reflect the actual situation of the current distribution network development.

In this embodiment, the decision-making method of the weight of the 6 types of evaluation criteria is evaluated by several experts, and the weight vector of each expert is obtained through the fuzzy analytic hierarchy process, as follows:

Let the fuzzy complementary matrix F=(f _ij ) _n×n (f _ij ∈[0,1]), if f _ij =F(a _i ,a _j ), then f _ij means a _i and a _j "... ratio... much more important” fuzzy membership relationship, f _ij uses 0.1-0.9 to give a quantitative scale, and the data scale is shown in Table 2.

Table 2 0.1-0.9 Quantity Scale

F has the following properties:

(1) f _ii =0.5, i=1,2,...,n;

(2)f _ij +f _ji =1,i,j=1,2,...,n;

(3) There is a normalized vector U _z =(u _z1 ,u _z2 ,…,u _zn ) and α(α>1), for any i, j, f _ij =log _α u _zi -log _α u _zj +0.5; Among them, α represents the resolution ability of the decision maker, U _z represents the weight decision vector of the zth expert on the evaluation criteria, u _zi and u _zj are two elements of U _z , which represent the expert z’s evaluation criteria i and Evaluate the weight decision result of j, n is the number of evaluation criteria, n=6;

The weight vector U _z =(u _z1 ,u _z2 ,...,u _zn ) is determined by the solution of the following constraint programming problem:

Using the Lagrange multiplier method to transform it into an unconstrained programming problem, we can obtain:

The weight vector of each expert is obtained through the above formula (2).

In order to obtain a more scientific decision-making, the fuzzy matrix is jointly constructed by multiple experts. Different knowledge backgrounds and experiences of experts may lead to different decisions for the same object. In order to extract a unified decision-making result from the decision-making results of multiple experts Weight vector, while reducing the subjectivity of decision-making, so this article introduces the cluster analysis method. The existing relatively mature clustering methods include partition method, hierarchical method, density-based method and grid-based method, and have been widely used in business intelligence, image pattern recognition, biology and security and other fields. In this paper, k-means algorithm is introduced to cluster analysis of expert decision vectors.

Assume that the decision-making results of experts on the weight distribution of each criterion index of the distribution network constitute a vector set D, and each decision vector includes 6 objects in the Euclidean space, namely, power supply quality, power grid structure, equipment level, power supply capacity, information level and Investment capacity, given the number k of clusters of each criterion weight vector set, randomly create an initial partition, and use an iterative method to try to improve by continuously moving the cluster center and aiming at high intra-cluster similarity and low inter-cluster similarity divided.

In this embodiment, the decision-making results of each expert are clustered and analyzed by using the cluster analysis method, and the decision-making weight vector is uniformly obtained, which specifically includes the following steps:

Step S21: Assuming that the weight distribution of various criteria in the distribution network is completed by N experts, input _N 6-dimensional expert decision vector sets D=(U ₁ , U ₂ ,...,U _z ,...,UN ), Among them, U _z =(u _z1 ,u _z2 ,…,u _z6 ) represents the weight distribution result of expert z to the 6 types of evaluation criteria, and the number of clusters to be classified is k;

Step S22: Randomly select the decision vectors of k experts to assign weights to the 6 types of distribution network evaluation criteria as the initial clustering centers {p ₁ ,p ₂ ,…,p _k }, where p _i ={p _i1 ,p _i2 , ..., p _i6 } represents the weight decision vector of the i-th clustering center expert; select the maximum number of clustering iterations P; determine the maximum convergence coefficient M at the end of the iteration;

Step S23: Calculate the Euclidean distances from each of the k decision vectors to each cluster, and divide each decision vector into the cluster with the smallest distance. The formula for calculating the Euclidean distance is:

In formula (3), dist(U _z , p _j ) represents the distance from the zth expert decision vector to the jth cluster;

Step S24: recalculate the center values {p ₁ , p ₂ ,...,p _k } of the k clusters, where p _j ={p _j1 ,p _j2 ,...,p _j6 }, and L represents the center of the cluster The number of decision vectors of , the calculation formula is:

Step S25: Check whether the clustering operation is over: if the number of iterations is equal to P, then end the clustering; otherwise, calculate the convergence distance of each cluster in this iteration, if the convergence distance is less than the given parameter M, then end, otherwise continue to iterate , the formula for calculating the convergence distance of the mth iteration is:

Step S26: Assuming that category p _l includes individual sorting vectors n _p , the weight η _l of this type of expert decision vector is calculated using the ratio of the expert decision vector to the total number of decision vectors:

From this, the final weight vector is obtained as:

In formula (7), p _l represents the cluster center value of the lth class.

Since the input parameter k will have a certain impact on the clustering results, in the actual clustering, several k values will be selected for multiple clustering and the clustering results will be analyzed, and the optimal result will be selected as the final weight vector. This paper introduces The silhouette coefficient method evaluates the quality of clustering.

In this embodiment, the clustering quality is judged by the silhouette coefficient method, and the weight vector with the best clustering quality is selected, which specifically includes the following steps:

Step S27: Assuming that the expert decision vector set D is divided into k clusters p ₁ , p ₂ ,...,p _k , for each decision vector u∈D, calculate the average distance b(u ), similarly, c(u) represents the minimum average distance from u to all clusters that do not belong to u.

The calculation formula of the silhouette coefficient of vector u is:

Step S28: Judging the clustering quality by the silhouette coefficient, and obtaining the weight vector with the best clustering quality.

The silhouette coefficient method combines the degree of cohesion and separation, which can be used to judge the quality of the clustering. It takes a value between -1 and +1, and the larger the value, the better the clustering effect.

Distribution network planning and construction objectives are quite different from the structural characteristics, weak links, and phased operation objectives of regional distribution networks. It is only difficult to understand the distribution networks in various regions by simply analyzing a large number of statistical data. actual conditions of construction. Therefore, the idea of "relatively effective" data envelopment method is introduced into the distribution network investment planning efficiency evaluation model, and the scale effectiveness and technical Effectiveness, the model has advantages that cannot be underestimated in avoiding subjective factors, simplifying algorithms and reducing errors, etc., and has been maturely applied to many fields such as resource allocation and productivity improvement.

The basic idea of the data envelopment method (DEA) is to regard a regional distribution network as a decision-making unit (Decision Making Unit, DMU), and then form an overall evaluation by many DMUs, and use the "input indicators" and " The weight of "output index" is a variable, and the ratio of "output index" and "input index" is maximized as the objective function to build a DEA evaluation model, and calculate the effective production edge of all distribution network investment construction, so as to determine whether the distribution network investment construction is DEA efficient.

The CCR model is a general classic model of DEA.

In this embodiment, in step S4, the construction of the CCR model includes the following steps:

Step S41: Assuming that distribution networks in h regions are planned and constructed at the same time, the original input indicators and output indicators of the planned input construction amount and expected effective output of the j-th regional distribution network are respectively with Each regional distribution network is expected to have s types of input and r types of output;

Step S42: Carry out normalization processing on the input index and output index, wherein, the positive input index is processed by formula (11) and formula (12), and the negative input index is processed by formula (13) and formula (14):

x _ij represents the input amount of the j-th regional distribution network to the i-th type of input index; y _rj represents the effective output of the r-th type of output index obtained by the j-th regional distribution network; with are the original input and output data respectively;

The x _ij in formulas (11) and (13) are the original input data after standardization processing of each index, which can be obtained through historical data using statistical analysis and data mining techniques, and the positive input index (that is, the larger the index value reflects the evaluation index). The better the value) is normalized by formula (11), and the negative input index (that is, the larger the index value is, the worse the evaluation value is) is normalized by formula (13); similarly, in formulas (12) and (14) y and _ij are the original output data after standardization processing of each index, formulas (12) and (14) are positive output index (that is, the larger the index value, the better the evaluation value) and negative output index (that is, the index value The larger the value, the worse the evaluation value) the normalized processing formula; with They are the original input and output data, that is, the input index data and expected effective output data of distribution network investment planning.

Step S43: Let v _i be the weight coefficient variable of the i-th type of input, f _r be the weight coefficient variable of the r-th type of output, and the ratio of the output volume to the input volume of various criteria is defined as the distribution network investment construction period t Efficiency index:

Step S44: Taking the benefit index of the distribution network in the j _0th region as the target and the benefit indices of all distribution networks as constraints, construct an optimal CCR model:

θ _{t, j0} = 1 indicates that the benefits of distribution network construction in this area are relatively high, and θ _{t, j0} < 1 indicates that the benefits of distribution network construction in this area are relatively low.

In this embodiment, in step S5, the dynamic weight vector is calculated by introducing information entropy and time degree; the definition formula of dynamic weight vector τ _t and time degree β is as follows:

In the formula, τ=[τ ₁ , τ ₂ ,...,τ _t ] represents the time series weighted vector, which reflects the difference in the contribution of investment strategies in different planning cycles to the dynamic evaluation, τ _t ∈ [0, 1], the information of the dynamic weighted vector Entropy reflects the extent to which the weight contains information in the dynamic weighting process of the investment efficiency evaluation value of each period. The smaller the entropy value, the greater the amount of information it acquires; The degree of emphasis on each cycle in , the smaller the value, the more attention to recent data, see Table 3 for specific meanings.

Table 3 Scale reference table of time scale

The time is given in advance by the decision maker, and the mathematical model of dynamic weight vector calculation can be expressed as:

In this embodiment, according to the evaluation result of step S5, a dynamic assessment scale is established, and the scale assessment is performed on the construction cost and construction efficiency respectively, and the scale assessment is specifically as follows:

Regulatory agencies give corresponding assessment rewards to areas with high distribution network investment efficiency. On the contrary, they give assessment penalties to areas with low distribution network investment efficiency. The mathematical model is as follows:

r _j,t = ρ _j,t +g _j,t (21)

In formula (21), r _j , t is the reward obtained by region j for distribution network construction within the construction period t; ρ _j , t is the construction income obtained by region j within the construction period t, positive numbers indicate rewards, and negative numbers Indicates the penalty; g _j , t is the cost subsidy obtained by region j in the construction period t.

In this embodiment, the calculation of the construction income includes the following steps:

Step S61: Calculate the efficiency evaluation value θ _j ' and dynamic weighting vector τ _t of distribution network construction in each region;

Step S62: Carry out linear weighting on the distribution network investment efficiency evaluation value of each period, and obtain the distribution network investment efficiency dynamic evaluation value θ _t,j of region j in the construction period t, namely:

θ _t,j ^dynamic ＝τ ₁ θ _1,j +τ ₂ θ _2,j +…+τ _t θ _t,j (22)

Step S63: Define the reward and punishment coefficient ξ _t,j of region j in the construction period t as:

Calculate the construction income of distribution network in region j in construction period t:

In the formula, P _t , j represents the added value of distribution network investment in region j during construction period t compared with period t-1.

In this embodiment, the construction cost subsidy of the distribution network in each region is jointly determined by its own construction cost and the construction cost of other regions, and its calculation method is shown in the following formula:

In formula (25), L _{j, t} represents the investment in the distribution network construction of region j within the construction period t; c _{j, t} is the unit construction cost of the distribution network in region j within the construction period t, and ε _j is the area The proportion of j's own cost, l _i , _t is the weight of the distribution network construction cost of company i in the observation period t.

Concrete implementation process of the present invention:

The data of 4 planning cycles in 9 municipal areas of a certain provincial network are selected as the research object to verify the rationality of the model embodied in the present invention.

(1) Calculate the weight coefficient of each criterion

The fuzzy complementary matrix is obtained by expert evaluation, and the weight vector of each expert can be calculated by referring to the formula (2) of the fuzzy analytic hierarchy process, and the cluster analysis of the expert ranking vector is carried out by the formula (3)-(7) of the cluster analysis method And calculate the unified decision vector, and finally test the quality of clustering by the formula (8)-(10). The value comparison of the resolution and the results of the clustering quality inspection are shown in Tables 4 and 5, respectively, thus verifying the rationality of the value of each parameter in this paper. In this paper, α = e, k = 3, and the evaluation model of the criterion layer is obtained as follows:

θ＝0.1969θ ₁ +0.1722θ ₂ +0.1835θ ₃ +0.2031θ ₄ +0.1335θ ₅ +0.1108θ ₆

In the formula, θ ₁ , θ ₂ ,..., θ ₆ represent the efficiency evaluation values of the input and output quantities of various criteria (namely power supply quality, power grid structure, equipment level, power supply capacity, informatization level and investment capacity) respectively, and are determined by the model ( 16) Calculated; θ represents the evaluation value of grid investment efficiency.

Table 4 Weight distribution under different α values

Table 5 Rules for selecting the number k of cluster centers

(2) Scale evaluation model

The input and output data of each index are the construction planning data of a power distribution system in the next four cycles. The data of each planning cycle can be obtained by using statistics or data mining methods for historical data. This paper assumes that all planning decision makers are rational decisions. The planning data is calculated according to the CCR model of the data envelopment method to obtain the relative efficiency evaluation value of each criterion, and the comprehensive evaluation value of the investment efficiency of the distribution network is obtained through weight linear weighting.

(3) Build a dynamic scale

The dynamic weighting vector τ=[τ ₂₀₁₆ , τ ₂₀₁₈ , τ ₂₀₂₀ , τ ₂₀₂₂ ] is calculated by the model (20), where the time degree β=0.40; the calculation results are shown in Table 6.

Table 6 Timing weighting vector

The dynamic scale evaluation model can be obtained by weighting the static model of each planning cycle according to the dynamic weight vector:

θ _2022,j =0.141θ' _2016,j +0.203θ' _2018,j +0.306θ' _2020,j +0.350θ' _2022,j

Based on the data of 6 criteria and 4 planning cycles in 9 regions of the province, the dynamic scale of distribution network investment and construction efficiency can be established. The regions and criteria above the boundary have relatively high investment and construction benefits. On the contrary, the area The investment and construction benefits of the guidelines are relatively low, and assessment incentives are introduced for the benefit evaluation value. Regions with higher efficiency can obtain assessment income, and on the contrary, they will receive assessment penalties, which can stimulate various regions to improve the efficiency of investment and construction of distribution networks .

The calculation results are shown in Table 7:

Table 7 Efficiency Evaluation Values of Each Standard in 2016

It can be seen from Table 7 that the overall efficiency of distribution network construction investment in 9 cities in the province in 2016 was at the upper-middle level, and the power supply quality, grid structure and power supply capacity were relatively high, and most areas exceeded However, the level of equipment and informatization is still relatively low, with an average of 0.547 and 0.764 respectively. The investment capacity of each region is relatively high, with an average of 0.871.

Then the investment efficiency evaluation value of the province's distribution network is compared horizontally, and the evaluation value of the investment efficiency of the distribution network in each region of the province is shown in Figure 3. It can be seen from the figure that the development level of the distribution network in the province varies greatly depending on the region. The investment efficiency of the distribution network in the eastern region (regions 2, 3, and 8) is relatively high, and the evaluation value in 2016 is mostly at 0.85 And above the level, the efficiency of the western region (regions 4, 5, 9) is relatively low, and most of the evaluation values in 2016 are below 0.8. There are similar laws between the development level of the province's distribution network and the regional economic development level. The region is relatively advanced, and the western region is relatively lagging behind. Decision makers can focus on areas with relatively low construction efficiency or guidelines to adjust the construction goals for the next cycle, and take assessment and punishment for areas with relatively low efficiency to stimulate them to improve construction efficiency, thereby improving the overall level of the province's distribution network .

A longitudinal comparison (time dimension) analysis is carried out on the investment efficiency evaluation value of the distribution network in the province. Figure 4 and Figure 5 respectively show the distribution network investment planning efficiency evaluation values of 9 city-level regions in the province during the four rolling planning cycles from 2016 to 2022, and A, B, C, D, E, and F represent power supply respectively Quality, power grid structure, equipment level, power supply capacity, informatization level, and investment level. Different identifiers in the figure represent the evaluation value of distribution network investment efficiency in different construction periods, which can describe the distribution network investment in various regions of the province from a municipal perspective. Building progress. From Figure 4, it can be clearly seen that the distribution network investment efficiency in each typical region gradually rises from a low level to a high level from 2016 to 2022, reflecting the investment in the distribution network during the four planning cycles of the province. Efficiency is gradually improving, and the efficiency increase in some areas has even reached 30%. The development of short boards in different areas also presents different distributions. Area 1 and area 5 show a similar pattern, and the efficiency of equipment level construction is relatively poor, at 0.6 and The quality of power supply in region 3 is at a relatively low level compared to other regions, at 0.9 or below. The development of region 7 is relatively balanced, and all indicators are at relatively high levels.

In actual projects, regulators can use this model to evaluate the construction of distribution networks and introduce an assessment and incentive system to minimize construction costs while maximizing construction efficiency and precise investment and construction, which is conducive to improving the overall level of distribution networks. In order to adapt to more power distribution needs under the new situation.

This embodiment also provides a device based on the scale competition evaluation method for medium and long-term dynamic investment in the urban distribution network described above, including a central processing unit, a storage module, and a display module;

The central processing unit is used to perform the following steps:

Step S1: select multiple evaluation indicators, and classify the evaluation indicators into six types of evaluation criteria. The six types of evaluation criteria are: power supply quality, grid structure, equipment level, power supply capacity, informatization level, and investment capacity;

Step S2: assign weights to the six types of evaluation criteria through fuzzy analytic hierarchy process, and obtain the weights of various evaluation criteria;

Step S3: Analyze the long-term statistical data of each evaluation index by the data envelopment analysis method, and divide each evaluation index in each type of evaluation criteria into input index and output index, where the input index refers to the decision-maker engaged in distribution network investment The input amount of the construction, the output index refers to the effective output obtained by the decision-maker through the investment and construction of the distribution network;

Step S4: according to the input index and the output index, utilize the CCR model of the data envelopment analysis method, and then obtain the efficiency evaluation value of each planning cycle of various criteria;

Step S5: Using the dynamic weighting vector, the efficiency evaluation values of each planning cycle are linearly and dynamically weighted to obtain the dynamic evaluation results.

In summary, the present invention provides a method and device for evaluating the scale competition of medium and long-term dynamic investment in urban distribution networks, which is suitable for indirect competition between similar enterprises in different regions, and minimizes construction costs and maximizes resource allocation efficiency. has obvious advantages.

The above-listed preferred embodiments have further described the purpose, technical solutions and advantages of the present invention in detail. It should be understood that the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Within the spirit and principles of the present invention, any modifications, equivalent replacements, improvements, etc., shall be included within the protection scope of the present invention.

Claims (10)

1. a kind of yardstick competition evaluation method of urban power distribution network long term dynamics investment, it is characterised in that comprise the following steps：

Step S1：Multiple evaluation indexes are chosen, by evaluation index to belonging in 6 class interpretational criterias, 6 class interpretational criterias are：For Electricity quality, electric network composition, equipment, power supply capacity, the level of IT application and the investment capacity；

Step S2：Weight distribution is carried out to 6 class interpretational criterias by Fuzzy AHP, the power of all kinds of interpretational criterias is obtained Weight；

Step S3：Analyzed by the DEA Method statistics long-term to each evaluation index, will be per class evaluation Each evaluation index in criterion is divided into input pointer and output-index, and wherein input pointer refers to that policymaker is engaged in power distribution network throwing The input amount built is provided, output-index refers to that policymaker passes through the throughput that is obtained to power distribution network investment construction；

Step S4:According to input pointer and output-index, using the CCR models of DEA Method, and then all kinds of standards are obtained The then efficiency rating value of each planning horizon；

Step S5：Using dynamic weight vectors, the efficiency rating value to each planning horizon carries out linear dynamic weighting assembly, obtained Dynamic evaluation result.

2. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 1, its feature exists In：The decision mode of 6 class interpretational criteria weights is that some experts assess, and obtains every expert's by Fuzzy AHP Weight vectors, it is specific as follows：

If fuzzy complementary matrix F=(f_ij)_n×n(f_ij∈ [0,1]), if f_ij=F (a_i,a_j), then f_ijRepresent a_iWith a_j" ... ratio ... weight Will be much " fuzzy membership relation, f_ijQuantity scale is given using 0.1-0.9, F has the following properties that：

(1)f_ii=0.5, i=1,2 ..., n；

(2)f_ij+f_ji=1, i, j=1,2 ..., n；

(3) there is normalized vector U_z=(u_z1,u_z2,…,u_zn) and α (α>1), to arbitrary i, j, f is met_ij=log_αu_zi- log_αu_zj+0.5；Wherein, α represents the resolution capability of policymaker, U_zRepresent z-th of expert to the Weight Decision-making of interpretational criteria to Amount, u_ziAnd u_zjFor U_zTwo elements, represent expert z to interpretational criteria i and evaluate j Weight Decision-making result, n is interpretational criteria Number, n=6；

Weight vectors U_z=(u_z1,u_z2,…,u_zn) determined by the solution of following constraint planning problem：

<mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mi>min</mi> <mi> </mi> <mi>Z</mi> <mo>=</mo> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>log</mi> <mi>&amp;alpha;</mi> </msub> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>i</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>log</mi> <mi>&amp;alpha;</mi> </msub> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>j</mi> </mrow> </msub> <mo>+</mo> <mn>0.5</mn> <mo>-</mo> <msub> <mi>f</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <munderover> <mi>&amp;Sigma;</mi> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mn>1</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>j</mi> </mrow> </msub> <mo>&gt;</mo> <mn>0</mn> <mo>,</mo> <mi>j</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mn>...</mn> <mo>,</mo> <mi>n</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>

Unconstrained optimization problem is turned to using method of Lagrange multipliers, can be tried to achieve：

<mrow> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>i</mi> </mrow> </msub> <mo>=</mo> <mfrac> <msup> <mi>&amp;alpha;</mi> <mrow> <mfrac> <mn>1</mn> <mi>n</mi> </mfrac> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msub> <mi>f</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> </mrow> </msup> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>k</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msup> <mi>&amp;alpha;</mi> <mrow> <mfrac> <mn>1</mn> <mi>n</mi> </mfrac> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msub> <mi>f</mi> <mrow> <mi>k</mi> <mi>j</mi> </mrow> </msub> </mrow> </msup> </mrow> </mfrac> <mo>,</mo> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>n</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow>

The weight vectors of every expert are obtained by above formula (2).

3. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 2, its feature exists In：Clustering is carried out to the result of decision of every expert using clustering methodology, and uniformly obtains decision weights vector, specifically Comprise the following steps：

Step S21：It is assumed that all kinds of criterion weight distribution of power distribution network are completed by N experts, input N number of 6 dimension expert to be sorted and determine Plan vector set D=(U₁,U₂,…,U_z,…,U_N), wherein U_z=(u_z1,u_z2,…,u_z6) represent power of the expert z to 6 class interpretational criterias Result is reassigned, number of clusters to be sorted is k；

Step S22：K expert of random selection is used as initial clustering to the decision vector of the class interpretational criteria weight distribution of power distribution network 6 Center { p₁,p₂,…,p_k, wherein p_i={ p_i1,p_i2,…,p_i6Represent that the Weight Decision-making of ith cluster center expert is vectorial；Choosing Select cluster maximum iteration P；Determine the maximum convergence coefficient M that iteration terminates；

Step S23：The Euclidean distance that k decision vector each arrives each cluster is calculated, each decision vector is assigned to most narrow spacing From cluster in, the calculation formula of Euclidean distance is：

<mrow> <mi>d</mi> <mi>i</mi> <mi>s</mi> <mi>t</mi> <mrow> <mo>(</mo> <msub> <mi>U</mi> <mi>z</mi> </msub> <mo>,</mo> <msub> <mi>p</mi> <mi>j</mi> </msub> <mo>)</mo> </mrow> <mo>=</mo> <msqrt> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>l</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>6</mn> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>l</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>p</mi> <mrow> <mi>j</mi> <mi>l</mi> </mrow> </msub> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow>

In formula (3), dist (U_z,p_j) represent z-th of expert decision-making vector to the distance of j-th of cluster；

Step S24:Recalculate the central value { p of k cluster₁,p₂,…,p_k, wherein, p_j={ p_j1,p_j2,…,p_j6, L is represented The decision vector number at such center is included into, calculation formula is：

<mrow> <msub> <mi>p</mi> <mrow> <mi>j</mi> <mi>l</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mn>1</mn> <mi>L</mi> </mfrac> <munder> <mo>&amp;Sigma;</mo> <mrow> <msub> <mi>u</mi> <mi>z</mi> </msub> <mo>&amp;Element;</mo> <msub> <mi>p</mi> <mi>j</mi> </msub> </mrow> </munder> <msub> <mi>u</mi> <mrow> <mi>z</mi> <mi>l</mi> </mrow> </msub> <mo>,</mo> <mi>l</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mn>6</mn> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow>

Step S25:Examine whether cluster operation terminates：If iterations is equal to P, terminate cluster；Otherwise this iteration is calculated The convergence distance each clustered, terminates if given parameter M is both less than if convergence distance, otherwise continues iteration, the m times iteration is received Hold back and be apart from calculation formula：

<mrow> <msub> <mi>d</mi> <mrow> <mi>j</mi> <mi>l</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>m</mi> <mo>)</mo> </mrow> <mo>=</mo> <msqrt> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>l</mi> <mo>=</mo> <mn>1</mn> </mrow> <mn>6</mn> </munderover> <msup> <mrow> <mo>(</mo> <msub> <mi>p</mi> <mrow> <mi>j</mi> <mi>l</mi> </mrow> </msub> <mo>(</mo> <mi>m</mi> <mo>)</mo> <mo>-</mo> <msub> <mi>p</mi> <mrow> <mi>j</mi> <mi>l</mi> </mrow> </msub> <mo>(</mo> <mi>m</mi> <mo>-</mo> <mn>1</mn> <mo>)</mo> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow>

Step S26:Assuming that classification p_lIncluding individual ordering vector n_pIt is individual, utilize such expert decision-making vector and decision vector sum Ratio calculation such expert decision-making vector weight η_l：

<mrow> <msub> <mi>&amp;eta;</mi> <mi>l</mi> </msub> <mo>=</mo> <mfrac> <msub> <mi>n</mi> <mi>p</mi> </msub> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>q</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msub> <mi>n</mi> <mi>q</mi> </msub> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>6</mn> <mo>)</mo> </mrow> </mrow>

Thus obtaining final weight vector is：

<mrow> <mi>&amp;omega;</mi> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>l</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>k</mi> </munderover> <msub> <mi>&amp;eta;</mi> <mi>l</mi> </msub> <msub> <mi>p</mi> <mi>l</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow>

In formula (7), p_lRepresent the cluster centre value of l classes.

4. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 3, its feature exists In：Clustering result quality is judged using silhouette coefficient method, the optimal weight vectors of clustering result quality is chosen, specifically includes following step Suddenly：

Step S27：Assuming that expert decision-making vector set D is divided into k cluster p₁,p₂,…,p_k, for each decision vector u ∈ D, U and the affiliated clusters of u other vectorial average distance b (u) are calculated, similar, c (u) represents u to the minimum for all clusters for being not belonging to u Average distance；

<mrow> <mi>b</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <munder> <mo>&amp;Sigma;</mo> <mrow> <mi>u</mi> <mo>,</mo> <msup> <mi>u</mi> <mo>&amp;prime;</mo> </msup> <mo>&amp;Element;</mo> <msub> <mi>p</mi> <mi>i</mi> </msub> <mo>,</mo> <mi>u</mi> <mo>&amp;NotEqual;</mo> <msup> <mi>u</mi> <mo>&amp;prime;</mo> </msup> </mrow> </munder> <mi>d</mi> <mi>i</mi> <mi>s</mi> <mi>t</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>,</mo> <msup> <mi>u</mi> <mo>&amp;prime;</mo> </msup> <mo>)</mo> </mrow> </mrow> <mrow> <mo>|</mo> <msub> <mi>p</mi> <mi>i</mi> </msub> <mo>|</mo> <mo>-</mo> <mn>1</mn> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>8</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <mi>c</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>=</mo> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mrow> <msub> <mi>p</mi> <mi>j</mi> </msub> <mo>:</mo> <mn>1</mn> <mo>&amp;le;</mo> <mi>j</mi> <mo>&amp;le;</mo> <mi>k</mi> <mo>,</mo> <mi>j</mi> <mo>&amp;NotEqual;</mo> <mi>k</mi> </mrow> </munder> <mo>{</mo> <mfrac> <mrow> <munder> <mo>&amp;Sigma;</mo> <mrow> <msup> <mi>u</mi> <mo>&amp;prime;</mo> </msup> <mo>&amp;Element;</mo> <msub> <mi>p</mi> <mi>j</mi> </msub> </mrow> </munder> <mi>d</mi> <mi>i</mi> <mi>s</mi> <mi>t</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>,</mo> <msup> <mi>u</mi> <mo>&amp;prime;</mo> </msup> <mo>)</mo> </mrow> </mrow> <mrow> <mo>|</mo> <msub> <mi>p</mi> <mi>j</mi> </msub> <mo>|</mo> </mrow> </mfrac> <mo>}</mo> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>9</mn> <mo>)</mo> </mrow> </mrow>

Vectorial u silhouette coefficient calculation formula is：

<mrow> <mi>s</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfrac> <mrow> <mi>c</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>-</mo> <mi>b</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> </mrow> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mo>{</mo> <mi>b</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>,</mo> <mi>c</mi> <mrow> <mo>(</mo> <mi>u</mi> <mo>)</mo> </mrow> <mo>}</mo> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>10</mn> <mo>)</mo> </mrow> </mrow>

Step S28：Clustering result quality is judged by silhouette coefficient, the optimal weight vectors of clustering result quality are obtained.

5. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 1, its feature exists In：In step s 4, the structure of CCR models comprises the following steps：

Step S41：Assuming that have h local distribution network while carrying out planning construction, j-th of local distribution network intend input construction amount and It is expected that throughput amount is originally inputted index and output-index respectively x_j ⁰=(x_1j ⁰,x_1j ⁰,…,x_sj ⁰)^T>0 and y_j ⁰= (y_1j ⁰,y_1j ⁰,…,y_rj ⁰)^T>0, j=1,2 ..., h；Each local distribution network is pre- in respect of s kinds input quantity and r kind output quantities；

Step S42：Standardization processing is carried out to input pointer and output-index, wherein, positive input pointer using formula (11) and Formula (12) is handled, and negative sense input pointer is handled using formula (13) and formula (14)：

<mrow> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> <mrow> <munder> <mi>max</mi> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>11</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <msub> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> <mrow> <munder> <mi>max</mi> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>12</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <munder> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> <mrow> <munder> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mi>min</mi> <mi>i</mi> </munder> <msubsup> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>13</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <msub> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <munder> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> <mrow> <munder> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> <mo>-</mo> <munder> <mi>min</mi> <mi>r</mi> </munder> <msubsup> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> <mn>0</mn> </msubsup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>14</mn> <mo>)</mo> </mrow> </mrow>

x_ijRepresent input amount of j-th of local distribution network to the i-th type input pointer；y_rjRepresent that j-th of local distribution network is obtained The r type output-index throughput amounts obtained；

Step S43：If v_iFor the weight coefficient variable of i-th kind of input, f_rFor r kinds export weight coefficient variable, all kinds of criterions it is defeated The ratio between output and input quantity are defined as the efficiency index of cycle t power distribution network investment construction：

<mrow> <msub> <mi>&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>r</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>a</mi> </munderover> <msub> <mi>f</mi> <mi>r</mi> </msub> <msub> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> </msub> </mrow> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>s</mi> </munderover> <msub> <mi>v</mi> <mi>i</mi> </msub> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>15</mn> <mo>)</mo> </mrow> </mrow>

Step S44：With jth₀The performance index of individual local distribution network is target, and the performance index of all power distribution networks is constraint, is built Optimize CCR models：

<mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msub> <mi>max&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> <mn>0</mn> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>r</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>a</mi> </munderover> <msub> <mi>f</mi> <mi>r</mi> </msub> <msub> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> <mn>0</mn> </mrow> </msub> </mrow> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>s</mi> </munderover> <msub> <mi>v</mi> <mi>i</mi> </msub> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> <mn>0</mn> </mrow> </msub> </mrow> </mfrac> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mfrac> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>r</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>z</mi> </munderover> <msub> <mi>f</mi> <mi>r</mi> </msub> <msub> <mi>y</mi> <mrow> <mi>r</mi> <mi>j</mi> </mrow> </msub> </mrow> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>s</mi> </munderover> <msub> <mi>v</mi> <mi>i</mi> </msub> <msub> <mi>x</mi> <mrow> <mi>i</mi> <mi>j</mi> </mrow> </msub> </mrow> </mfrac> <mo>&amp;le;</mo> <mn>1</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>V</mi> <mo>=</mo> <mo>&amp;lsqb;</mo> <msub> <mi>v</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>v</mi> <mn>2</mn> </msub> <mo>,</mo> <mo>...</mo> <mo>,</mo> <msub> <mi>v</mi> <mi>s</mi> </msub> <mo>&amp;rsqb;</mo> <mo>&amp;GreaterEqual;</mo> <mn>0</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>F</mi> <mo>=</mo> <mo>&amp;lsqb;</mo> <msub> <mi>f</mi> <mn>1</mn> </msub> <mo>,</mo> <msub> <mi>f</mi> <mn>2</mn> </msub> <mo>,</mo> <mo>...</mo> <mo>,</mo> <msub> <mi>f</mi> <mi>a</mi> </msub> <mo>&amp;rsqb;</mo> <mo>&amp;GreaterEqual;</mo> <mn>0</mn> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>16</mn> <mo>)</mo> </mrow> </mrow>

θ_{T, j0}=1 shows that this area's distribution network construction benefit is of a relatively high, θ_{T, j0}<1 shows that this area's distribution network construction benefit is relative It is relatively low.

6. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 5, its feature exists In：In step s 5, introduce comentropy and time degree calculates dynamic weight vectors；Dynamic weight vectors τ_tWith determining for time degree β Adopted formula is as follows：

<mrow> <mi>I</mi> <mo>=</mo> <mo>-</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>T</mi> </munderover> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <msub> <mi>ln&amp;tau;</mi> <mi>t</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>17</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>T</mi> </munderover> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <mo>=</mo> <mn>1</mn> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>18</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <mi>&amp;beta;</mi> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>T</mi> </munderover> <mfrac> <mrow> <mi>T</mi> <mo>-</mo> <mi>t</mi> </mrow> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> </mfrac> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>19</mn> <mo>)</mo> </mrow> </mrow>

In formula, τ=[τ₁, τ₂..., τ_t] sequential weighing vector is represented, reflect different investment tacticses planning horizon to dynamic evaluation Contribute otherness, τ_t∈ [0,1], the comentropy of dynamic weight vectors is reflected dynamically to be added to each cycle investments efficiency rating value Weight includes the degree of information during power, and entropy is smaller, represents that the information content that it is obtained is bigger；β represents time degree, its size The attention degree in each periodic samples assembling process to each cycle is represented, its value is smaller, represented to recent data more Pay attention to；Time degree is previously given by policymaker, and the mathematical modeling that dynamic weight vectors are calculated is represented by：

<mrow> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> <mrow> <mo>(</mo> <mo>-</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>T</mi> </munderover> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <msub> <mi>ln&amp;tau;</mi> <mi>t</mi> </msub> <mo>)</mo> </mrow> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>&amp;beta;</mi> <mo>=</mo> <mi>&amp;Sigma;</mi> <mfrac> <mrow> <mi>T</mi> <mo>-</mo> <mi>t</mi> </mrow> <mrow> <mi>t</mi> <mo>-</mo> <mn>1</mn> </mrow> </mfrac> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>T</mi> </munderover> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <mo>=</mo> <mn>1</mn> <mo>,</mo> <msub> <mi>&amp;tau;</mi> <mi>t</mi> </msub> <mo>&amp;Element;</mo> <mo>&amp;lsqb;</mo> <mn>0</mn> <mo>,</mo> <mn>1</mn> <mo>&amp;rsqb;</mo> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mi>t</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>T</mi> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>20</mn> <mo>)</mo> </mrow> <mo>.</mo> </mrow>

7. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 1, its feature exists In：According to step S5 evaluation result, dynamic examination scale is set up, carrying out scale to construction cost and construction efficiency respectively examines Core, the scale examination is specific as follows：

Corresponding examination reward is given in the area higher to power distribution network efficiency of investment by regulator, on the contrary, being invested for power distribution network Examination punishment is given in less efficient area, and mathematical modeling is as follows：

r_j,t=ρ_j,t+g_j,t (21)

In formula (21), r_j, t is the return that area j is obtained in construction period t to distribution network construction；ρ_j, t is that area j is being built The construction income obtained in cycle t, positive number represents reward, negative number representation punishment；g_j, t is that area j is obtained in construction period t Effort Cost Subsidy.

8. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 7, its feature exists In：The calculating for building income comprises the following steps：

Step S61:Calculate the efficiency rating value θ of each department distribution network construction_j' and dynamic weight vectors τ_t；

Step S62：Power distribution network efficiency of investment evaluation of estimate to each cycle carries out linear weighted function, obtains regional j in construction period t Power distribution network efficiency of investment dynamic evaluation value θ_{T, j}, i.e.,：

θ_t,j ^dynamic=τ₁θ_1,j+τ₂θ_2,j+…+τ_tθ_t,j (22)

Step S63：Define coefficient of rewards and punishment ξs of the area j in construction period t_{T, j}For：

<mrow> <msub> <mi>&amp;xi;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <msup> <msub> <mi>&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mrow> <mi>d</mi> <mi>y</mi> <mi>n</mi> <mi>a</mi> <mi>m</mi> <mi>i</mi> <mi>c</mi> </mrow> </msup> <mo>-</mo> <mfrac> <mn>1</mn> <mi>n</mi> </mfrac> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>n</mi> </munderover> <msup> <msub> <mi>&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>i</mi> </mrow> </msub> <mrow> <mi>d</mi> <mi>y</mi> <mi>n</mi> <mi>a</mi> <mi>m</mi> <mi>i</mi> <mi>c</mi> </mrow> </msup> </mrow> <mrow> <msup> <msub> <mi>&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> <mi>max</mi> </mrow> </msub> <mrow> <mi>d</mi> <mi>y</mi> <mi>n</mi> <mi>a</mi> <mi>m</mi> <mi>i</mi> <mi>c</mi> </mrow> </msup> <mo>-</mo> <msup> <msub> <mi>&amp;theta;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> <mi>min</mi> </mrow> </msub> <mrow> <mi>d</mi> <mi>y</mi> <mi>n</mi> <mi>a</mi> <mi>m</mi> <mi>i</mi> <mi>c</mi> </mrow> </msup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>23</mn> <mo>)</mo> </mrow> </mrow>

Calculate the construction income for obtaining regional j power distribution networks in construction period t：

<mrow> <msub> <mi>&amp;rho;</mi> <mrow> <mi>t</mi> <mo>,</mo> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfenced open = "{" close = ""> <mtable> <mtr> <mtd> <mrow> <msub> <mi>P</mi> <mi>t</mi> </msub> <msub> <mi>&amp;xi;</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>,</mo> <msub> <mi>P</mi> <mi>t</mi> </msub> <mo>&gt;</mo> <mn>0</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mn>0</mn> <mo>,</mo> <msub> <mi>P</mi> <mi>t</mi> </msub> <mo>&lt;</mo> <mn>0</mn> <mo>,</mo> <msub> <mi>&amp;xi;</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>&gt;</mo> <mn>0</mn> </mrow> </mtd> </mtr> <mtr> <mtd> <mrow> <mo>-</mo> <mo>|</mo> <msub> <mi>P</mi> <mi>t</mi> </msub> <msub> <mi>&amp;xi;</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>|</mo> <mo>,</mo> <msub> <mi>P</mi> <mi>t</mi> </msub> <mo>&lt;</mo> <mn>0</mn> <mo>,</mo> <msub> <mi>&amp;xi;</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>&lt;</mo> <mn>0</mn> </mrow> </mtd> </mtr> </mtable> </mfenced> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>24</mn> <mo>)</mo> </mrow> </mrow>

In formula, P_t, j represents regional j in construction period t compared to the value added invested to power distribution network in cycle t-1.

9. the yardstick competition evaluation method of urban power distribution network long term dynamics investment according to claim 8, its feature exists In：The construction cost subsidy of each local distribution network is together decided on by the construction cost of itself and other regional construction costs, Its computational methods is shown below：

<mrow> <msub> <mi>g</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>=</mo> <msub> <mi>L</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>{</mo> <msub> <mi>&amp;epsiv;</mi> <mi>j</mi> </msub> <msub> <mi>c</mi> <mrow> <mi>j</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>+</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>-</mo> <msub> <mi>&amp;epsiv;</mi> <mi>j</mi> </msub> <mo>)</mo> </mrow> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mi>i</mi> <mo>&amp;NotEqual;</mo> <mi>j</mi> </mrow> <mi>n</mi> </munderover> <mrow> <mo>(</mo> <msub> <mi>l</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <msub> <mi>c</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>t</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>}</mo> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>25</mn> <mo>)</mo> </mrow> </mrow>

In formula (25), L_{J, t}Represent distribution network construction input amounts of the area j in construction period t；c_j,tIt is regional j in the construction period The unit construction cost of power distribution network, ε in t_jFor regional j cost taken by themselves proportion, l_{I, t}It is company i in observation in construction period t Weight shared by interior distribution network construction cost.

10. a kind of dress of the yardstick competition evaluation method of the urban power distribution network long term dynamics investment based on described in claim 1 Put, it is characterised in that：Including central processing unit, memory module, display module；

The central processing unit is to carry out following step：

Step S1：Multiple evaluation indexes are chosen, by evaluation index to belonging in 6 class interpretational criterias, 6 class interpretational criterias are：For Electricity quality, electric network composition, equipment, power supply capacity, the level of IT application and the investment capacity；

Step S2：Weight distribution is carried out to 6 class interpretational criterias by Fuzzy AHP, the power of all kinds of interpretational criterias is obtained Weight；

Step S3：Analyzed by the DEA Method statistics long-term to each evaluation index, will be per class evaluation Each evaluation index in criterion is divided into input pointer and output-index, and wherein input pointer refers to that policymaker is engaged in power distribution network throwing The input amount built is provided, output-index refers to that policymaker passes through the throughput that is obtained to power distribution network investment construction；

Step S4:According to input pointer and output-index, using the CCR models of DEA Method, and then all kinds of standards are obtained The then efficiency rating value of each planning horizon；

Step S5：Using dynamic weight vectors, the efficiency rating value to each planning horizon carries out linear dynamic weighting assembly, obtained Dynamic evaluation result.