Distribution network cluster difference analysis method, device and system based on improved fuzzy evaluation

Abstract

The invention relates to the technical field of power grid management and data analysis, in particular to a power distribution network cluster difference analysis method, device and system based on improved fuzzy evaluation. The method comprises the following steps: constructing a multi-dimensional three-level evaluation system; the third-level indexes are weighted; setting evaluation grades, and calculating the membership degree of each area corresponding to each evaluation grade under different three-level indexes; calculating a weighted value of a third-level index of each region under the same second-level index on the same evaluation level as a membership degree of the second-level index on the evaluation level; and for each third-level index and each second-level index of each region, selecting the evaluation grade corresponding to the maximum membership as the evaluation grade of the corresponding evaluation index, and generating an evaluation report. According to the method, a multi-dimensional three-level evaluation system is constructed, the key dimensions of the power distribution network are comprehensively covered, and refined evaluation is realized. It is ensured that the evaluation system covers core elements such as power distribution network planning, operation and social influence, and the one-sidedness of single index evaluation is avoided. The three-level index structure supports differential analysis of different levels of the power distribution network, and is helpful for accurately positioning weak links or dominant characteristics.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid management and data analysis, and in particular to a method, device and system for analyzing the differences of distribution network clusters based on improved fuzzy evaluation. Background Art

[0002] With the development of distributed power grids, rural distribution networks are no longer solely receiving terminals. Instead, they are evolving from a power network dominated by the main grid and implementing terminal power distribution to a "new rural local power system" dominated by distributed renewable energy and energy storage, interconnected with the main grid. Rural areas vary in their economic development levels, socio-demographic and industrial structures, geographic and natural resource endowments, and the scale of electrical sources, networks, and loads. Therefore, a tailored approach should be adopted to promote the development of new rural local power systems. Furthermore, given the large number of rural areas, typical scenarios for new rural local power systems need to be identified to support research on their development models and planning. Therefore, the classification of typical scenarios for new rural local power systems is of great significance for distribution network planning and operation.

[0003] Existing research primarily focuses on the classification of rural human resources, while less research has focused on the classification of typical RNLPS scenarios. To comprehensively reflect the actual situation of RNLPS (rural new local power system), a comprehensive evaluation from multiple dimensions is necessary. For multidimensional indicator systems, existing literature typically uses fuzzy comprehensive evaluation methods to qualitatively analyze quantitative indicators. However, the indicators used to classify typical RNLPS scenarios are relatively simple and cannot accurately classify rural new local power system scenarios.

[0004] Fuzzy comprehensive evaluation first requires establishing the rating membership of different elements. The standard values ​​for indicator evaluation serve as the specific numerical limits for defining the membership of the indicators. However, there is currently a lack of a unified evaluation system for RNLPS, making it impossible to determine the standard values ​​corresponding to each indicator. Existing systems and evaluation methods are not directly applicable to the accuracy and applicability of rural power grids. Summary of the Invention

[0005] In order to overcome the shortcomings of traditional fuzzy comprehensive evaluation methods in qualitatively analyzing new rural local power systems, which lack accuracy and applicability, this paper proposes a distribution network clustering difference analysis method based on improved fuzzy evaluation, establishes a multidimensional three-level indicator system for measuring new rural local power systems, and evaluates different clustering scenarios, thereby providing a theoretical basis for distribution network planning.

[0006] The present invention proposes a distribution network cluster difference analysis method based on improved fuzzy evaluation. First, a multi-dimensional three-level evaluation system is constructed. Each dimension includes one or more secondary indicators, and each secondary indicator includes one or more tertiary indicators.

[0007] Empowering evaluation indicators at all levels;

[0008] Set the evaluation level and calculate the membership degree of each region corresponding to each evaluation level under different three-level indicators;

[0009] Calculate the weighted value of the third-level indicators of each region under the same second-level indicator at the same evaluation level as the membership degree of the second-level indicator at the evaluation level;

[0010] Calculate the weighted value of the secondary indicators of each region under the same dimension at the same evaluation level as the membership degree of the dimension at the evaluation level;

[0011] Calculate the weighted values ​​of all dimensions of each region on the same evaluation level as the final evaluation membership, and select the evaluation level corresponding to the largest final evaluation membership as the final evaluation level of the region; for each third-level indicator, second-level indicator and dimension of each region, select the evaluation level corresponding to the maximum membership value as the evaluation level of the corresponding evaluation indicator, and generate an evaluation report.

[0012] Preferably, the weighting method for each three-level indicator is as follows: first, different weighting methods are used to obtain the weight vectors of the indicators at all levels in the region, and the weighted value G of the region corresponding to the three-level indicator under each weighting method is calculated;

[0013] For each weighting method, cluster the weighted values ​​G of each region to obtain the performance of the clustering results corresponding to different weighting methods;

[0014] The weight vector obtained by the weighting method corresponding to the optimal clustering result is obtained as the final weight vector.

[0015] Preferably, the method for clustering the weighted values ​​of each region is: first, normalize the values ​​of the same three-level indicators in different regions, perform weighted operations on the weights of the three-level indicators in each region and the normalized values ​​of the three-level indicators to obtain the weighted values ​​of the regions; then cluster the weighted values ​​of different regions.

[0016] Preferably, the normalization formula for the three-level index value is:

[0017]

[0018] Among them, X ij is the third-level indicator C of region j i The value of T ij For X ij Normalized value of ; min represents the minimum function, and max represents the maximum function.

[0019] Preferably, the weighting methods include: hierarchical analysis method, entropy weight method, multiplication weighting method, additive weighting method and combined weighting method based on game theory; the multiplication weighting method, additive weighting method and combined weighting method based on game theory all use the weight vector obtained by the hierarchical analysis method and the entropy weight method as the benchmark value for calculation.

[0020] Preferably, each three-level indicator is divided into three evaluation levels: "general, good, and excellent". The method for calculating the region's membership to different evaluation levels under any three-level indicator is as follows:

[0021] Set three breakpoints for the three-level indicators i1 <v i2 <v i3 , calculate the region's membership degree corresponding to different evaluation levels under the three-level indicators according to the value x of the region's three-level indicators;

[0022]

[0023] Among them, r i1 Indicates that the area is in the third level indicator C i The membership degree is evaluated as "general", r i2 Indicates that the area is in the third level indicator C i The membership degree is evaluated as “good”, r i3 Indicates that the area is in the third level indicator C i The membership degree of the evaluation is "excellent", x is the third-level indicator C representing the region i The value of .

[0024] Preferably, the breakpoint setting method of each third-level indicator is:

[0025] The third-level indicators C of all regions i Arrange in ascending order and set multiple groups of breakpoints; calculate the membership of each assessment level of each region under different breakpoints, and select the level corresponding to the maximum membership as the final assessment level of the region under the three-level indicator; classify the regions according to the assessment level at the corresponding three-level coordinates; take the minimum internal variance and the maximum inter-class variance as the optimization goals and select the optimal breakpoint.

[0026] Preferably, in the multi-dimensional three-level evaluation system, the dimensions include economic, social, geographical and electrical; the secondary indicators of the economy include the government level, and the tertiary indicators of the government level include the number of new energy vehicles and GDP;

[0027] Secondary indicators of society include population characteristics and industrial proportions; third-level indicators of population characteristics include rural population aging, education level, and population size; third-level indicators of industrial proportions include the proportion of rural primary industry and the proportion of rural secondary industry;

[0028] The second-level indicators of geography include geographical environment and resource endowment; the third-level indicators of geographical environment include altitude and temperature; the third-level indicators of resource endowment include average sunshine hours and average wind intensity;

[0029] The secondary indicators of electricity include source, network and load; the tertiary indicators of source include total photovoltaic installed capacity, total wind power installed capacity and total biomass energy installed capacity; the tertiary indicators of network include power supply voltage qualification rate, N-1 pass rate, rural power grid reliability rate, distribution automation coverage rate, line connection rate, inter-station connection rate, heavy load rate, light load rate, average line length and average power supply radius; the tertiary indicators of load include the number of substations under 10kV voltage level, maximum power load, capacity load ratio, 10kV per household distribution transformer capacity and number of charging piles.

[0030] The present invention proposes a device for implementing the distribution network cluster difference analysis method based on improved fuzzy evaluation, comprising:

[0031] Indicator screening and establishment module, used to establish a four-dimensional three-level evaluation system;

[0032] The indicator weight calculation module is used to calculate the weight vectors of the three-level indicators using different methods and to filter the weight vectors using the weighted value clustering results;

[0033] The fuzzy comprehensive evaluation module is used to establish the breakpoints corresponding to each third-level indicator, and calculate the membership of the region corresponding to different evaluation levels based on the values ​​of the third-level indicators; calculate the weighted values ​​of the membership of different third-level indicators under the second-level indicators as the membership of the second-level indicators under different evaluation levels.

[0034] The present invention proposes a system comprising a memory and a processor, wherein the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement the distribution network cluster difference analysis method based on improved fuzzy evaluation.

[0035] The advantages of the present invention are:

[0036] (1) The present invention constructs a multi-dimensional three-level evaluation system, which is conducive to comprehensively covering the key dimensions of the distribution network and realizing refined evaluation. The present invention ensures that the evaluation system covers core elements such as distribution network planning, operation, and social impact, avoiding the one-sidedness of single indicator evaluation. The three-level indicator structure supports differentiated analysis of different levels of the distribution network, which helps to accurately locate weak links or advantageous features.

[0037] (2) The present invention calculates the clustering results under different weighting methods to determine which weighting method is more suitable for the clustering effect of the new rural local power system, which is conducive to the reasonable adjustment of the actual system; compares multiple weighting methods and combined weighting methods to avoid the deviation of a single weighting method and improve the scientific nature of the weight. The optimal weight is selected through the clustering performance of the weighted value (G), ensuring that the weight distribution is consistent with the actual data distribution, thereby enhancing the credibility of the evaluation results.

[0038] (3) The present invention establishes the standard values ​​of quantitative indicator evaluation grades, i.e., breakpoints, through breakpoint optimization. This overcomes the problem that traditional fuzzy comprehensive evaluation methods are difficult to deal with when standard values ​​of indicators do not exist or are difficult to obtain, and overcomes the limitation of traditional fuzzy comprehensive evaluation methods that rely on standard values ​​of indicator evaluation grades. This enables the qualitative analysis results of similar typical scenarios in rural local power systems of different dimensions to be optimized.

[0039] (4) Dynamically adjust the membership function of "fair / good / excellent" based on breakpoint optimization to solve the evaluation rigidity problem caused by fixed thresholds in traditional fuzzy evaluation. Automatically optimize the breakpoints with the goal of minimizing the intra-class variance and maximizing the inter-class variance, so that the classification is more consistent with the actual data distribution and reduces the subjectivity caused by human intervention.

[0040] (5) Based on the characteristics of the new rural local power system, the present invention takes into account the scale of electrical sources, grids and loads, and the economic, social and geographical dimensions of different rural areas, and establishes an indicator system for scientifically dividing RNLPS, thereby solving the problem of a single indicator dimension for the division of typical scenarios of the new rural local power system.

[0041] (6) The present invention solves the three major pain points of traditional distribution network evaluation, namely, the strong subjectivity of weights, the rigidity of grade division, and the lack of strength in difference analysis, through the technical chain of "multi-dimensional indicator construction - dynamic weighting optimization - fuzzy clustering verification", and provides a data-driven decision-making tool for differentiated planning, operation optimization and policy formulation of distribution networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is the four-dimensional three-level index system proposed by the present invention;

[0043] Figure 2 This is a flow chart of the distribution network cluster difference analysis method based on improved fuzzy evaluation proposed by the present invention;

[0044] Figure 3 A flow chart of the weighting method for the three-level indicators;

[0045] Figure 4 This is a flow chart of the calculation method for the three-level indicator evaluation grade membership degree;

[0046] Figure 5This is a schematic diagram of the effect of the natural breakpoint method of the present invention;

[0047] Figure 6 This is the result diagram of fuzzy comprehensive evaluation;

[0048] Figure 7 This is a schematic diagram of the structure of a cluster differentiation device for rural distribution network based on an improved fuzzy comprehensive evaluation algorithm;

[0049] Figure 8 A schematic diagram of the structure of a computer device. DETAILED DESCRIPTION

[0050] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0051] Reference Figure 1 、 Figure 2 The distribution network clustering difference analysis method based on improved fuzzy evaluation proposed in the present invention includes the following steps.

[0052] S1. Construct a four-dimensional three-level indicator system.

[0053] The rural electricity dimension reflects how the rural economic dimension affects the investment scale of grid construction and the local electricity consumption level. By understanding factors at the government level, the total rural economy can be considered. This step analyzes the economic level of the RNLPS (rural new local power system) and proposes a second-level indicator at the government level and two third-level indicators: the number of new energy vehicles and GDP. The number of new energy vehicles affects the characteristics and magnitude of rural loads, and GDP affects the investment scale of local grid construction.

[0054] The rural social dimension influences local people's demand for electricity and their willingness to accept new technologies. It also reflects the proportion of different local industries and load characteristics. RNLPS planning needs to be adjusted and revised based on these indicators. Demographic factors represent the demographic structure and dynamics of rural areas, while changes in industrial proportions reflect changes in rural social structure. This paper analyzes the social aspects of RNLPS and proposes two secondary indicators: demographic characteristics and industrial proportions, as well as five tertiary indicators: rural population aging, education level, population size, and the proportion of primary and secondary industries in rural areas. The aging of the rural population reflects the demand for electricity in rural areas. Education level can reflect local people's support for energy-saving policies and products, thereby affecting the level of new energy utilization in rural areas. Population size reflects the level of local electricity consumption. Different rural industrial proportions have different load characteristics, and fluctuations in time and space can affect the source-load balance and even cause large voltage fluctuations.

[0055] Rural geography influences rural energy development and planning through natural resource endowments such as climate, sunlight, and wind power. The geographical environment encompasses factors such as topography, climate conditions, water resources, and soil type, which influence the efficiency of renewable energy utilization. Resource endowment refers to the natural resources available in rural areas, which directly influences the potential for rural renewable energy development and the average annual utilization hours of distributed renewable energy. This step analyzes the geographical aspects of the RNLPS and proposes two secondary indicators: geographical environment and resource endowment, as well as four tertiary indicators: altitude, temperature, average annual sunshine hours, and average daily wind intensity. High altitudes present inconvenient transportation, making equipment selection and grid construction more difficult. Temperature affects the performance of power equipment and electricity demand. Average annual sunshine hours and average daily wind intensity are important criteria for determining whether a region is suitable for renewable energy development.

[0056] The rural electrical dimension reflects the planning, construction, and operation of rural power grids and is a key factor influencing the absorption of new energy and power grid planning. The power supply side reflects the composition of rural renewable energy, the power grid side reflects the degree of rural electrical automation and network reliability, and the load side reflects the power demand pattern in rural areas. This step analyzes the electrical level of the RNLPS and evaluates the electrical dimension using three secondary indicators: source, network, and load. Three tertiary indicators, 10 tertiary indicators, and five tertiary indicators are set, respectively. The installed capacity of new energy reflects the current scale of new energy development and affects the power supply side planning; the power supply voltage qualification rate, N-1 pass rate, and rural power grid reliability rate reflect the power supply level of the local power grid; the distribution automation coverage rate, line connection rate, and inter-station connection rate represent the flexible adjustment level of the local power grid; the heavy load rate and light load rate reflect the economy and efficiency of operation; the average line length and average power supply radius reflect the power supply range of the local substation; the above indicators affect the power grid side planning from different angles; the number of substations under the 10kV voltage level, the maximum power load, the capacity load ratio, and the 10kV per household distribution and transformation capacity reflect the local power supply demand, and the number of charging piles reflects the magnitude of new load factors in rural areas.

[0057] Therefore, this implementation method proposes a four-dimensional three-level indicator system for measuring the new rural local power system, which proposes four dimensions: economic, social, geographical, and electrical. Under each dimension, multiple second-level indicators are set, and under each second-level indicator, multiple third-level indicators are set, as shown in Table 1:

[0058] Table 1: Four-dimensional three-level indicator system

[0059]

[0060]

[0061] S2. Empower the three-level indicators, such as Figure 3 As shown, the specific steps include the following steps.

[0062] S21. First, a variety of different methods are used to assign weights to the three-level indicators. The specific weighting methods can be based on the Analytic Hierarchy Process (AHP), entropy weight method, multiplication weighting method, additive weighting method, combined weighting method based on game theory, and expert weighting method.

[0063] Each weighting method can obtain the weights of all dimensions, secondary indicators and tertiary indicators, and the sum of the weights of the dimensions is 1, the sum of the weights of the secondary indicators is 1, and the sum of the weights of the tertiary indicators is 1.

[0064] Assume that the set of all three-level indicators is recorded as C = {C1, C2, C3, ..., C i ,…,C 29};Ci represents the i-th third-level indicator, 1≤i≤29;

[0065] Assume that the three-level index weight vector established by AHP method is recorded as w1 = [α1, α2, α3, ..., α i ,…,α 29 ]; α i represents C determined by AHP method i The weight of .

[0066] The three-level weight vector established by the entropy weight method is recorded as w2 = [β1, β2, β3, ..., β i ,…,β 29 ]; β i represents C determined by entropy weight method i The weight of

[0067]

[0068] g i =1-E i

[0069]

[0070] Where m represents the number of regions; g i Indicates index C i The coefficient of variation, E i Indicates index C i Information entropy of X ij Represents the third-level index C of region j i The value of n represents the number of third-level indicators. In this embodiment, n=29.

[0071] The combined weighting rule can comprehensively consider the role of AHP method and entropy weight method in the weighting process, while reflecting human subjective will, it also ensures the objectivity of indicator weights and avoids the influence of a single weighting method; the commonly used combined weighting methods include additive weighting method, multiplicative weighting method and combined weighting method based on game theory. Among them, multiplicative weighting has a multiplier effect that the larger the better and the smaller the smaller, and is more suitable for data sets with uniform weights; additive weighting overcomes the multiplier effect and has better actual effects, but it is more dependent on the judgment level of decision makers and has strong subjective will; the combined weighting rule of game theory seeks to coordinate the conflict between the two with the goal of minimizing deviation.

[0072] The weight vector of the three-level indicators established by multiplication weighting is recorded as w3=[w i_mult |1≤i≤n];w i_mult C is determined by multiplication weights i The weight of P ij Indicates index C i characteristic proportion.

[0073] The weight vector of the three-level indicators established by the addition weighting is recorded as w4=[w i_add |1≤i≤n];w i_add C is determined by additive weighting i The weight of

[0074] The three-level index weight vector established by the combined weighting method based on game theory is denoted as w5=[w i_game |1≤i≤n];w i_game C represents the combination weighted i The weight of

[0075]

[0076] w i_add =λ 1_add α i +λ 2_add β i

[0077] w i_game =λ 1_game α i +λ 2_game β i

[0078] Wherein, n represents the number of indicators, and in this embodiment, n=29; λ 1_add and λ 2_add Represents the weight coefficient of additive weighting, the sum of the two is 1, which is determined manually; 1_game and λ 2_game represents the weight coefficient of the combined weighting method based on game theory, and the sum of the two is 1. Specifically, it can be obtained from formula (11): and Established, and and It is established by the subjective weight vector w1 and the objective weight vector w2 in formula (12);

[0079]

[0080] S22. Calculate the weighted values ​​of each level 3 indicator in each region obtained by each weighting method;

[0081] For example, the weight vector w1=[α1, α2, α3, ..., α i ,…,α 29 ] The weighted value G of a certain area calculated;

[0082]

[0083] Among them, T i C is the third-level indicatori The normalized value of can be calculated by Min-Max normalization, Z-core normalization, maximum value normalization and other methods.

[0084] In this embodiment, when normalizing the three-level indicators, the positive and negative directions of the three-level indicator values ​​are considered, and then the data normalization processing is performed on the three-level indicator values; let the three-level indicator C of region j be i The value of X ij , and its normalized value is denoted as T ij ;

[0085]

[0086] min represents the minimum function, and max represents the maximum function.

[0087] S23. For the indicator weight vector obtained by each weighting method, calculate the weight value G of each region and perform clustering, and evaluate the performance of the clustering results corresponding to different weighting methods.

[0088] The weighted value clustering in this embodiment is essentially to divide the RNLPS typical scene; first, the number K of typical scene division clusters of RNLPS is determined by the elbow rule; then the K-means clustering algorithm is used to cluster the weighted values ​​of each region calculated under the same weight vector.

[0089] In this embodiment, the elbow method is used to calculate the error sum of squares (SSE) within different clusters to evaluate the compactness of the clustering. Specifically, the clustering result corresponding to the maximum SSE value is taken as the final clustering result.

[0090]

[0091] Q q represents the qth cluster, K is the total number of clusters, Represents cluster Q q The cluster center, G j represents the weight value G of region j; Indicates G j and The distance between them can be specifically the Euclidean distance.

[0092] The advantage of the K-means clustering algorithm lies in its rapid processing speed for large sample data. It also performs well for high-dimensional datasets. Therefore, the K-means clustering algorithm is suitable for multidimensional, large-scale data types within a four-dimensional, three-level index system. The K-means clustering algorithm divides the study area into K clusters, ensuring that regions within the same cluster are consistent while regions within different clusters exhibit significant differences. The distance between different objects is used as a similarity metric: closer distances between objects indicate higher similarity, and closer distances indicate lower similarity.

[0093] Assuming that all regions are finally clustered into their own clusters, the average silhouette coefficient based on the median and the variance of the silhouette coefficient s 2 Expressed as:

[0094]

[0095] Where s j represents the silhouette coefficient of region j, a j represents the median distance between region j and other regions in the same cluster, b j represents the median of the distances between region j and all regions in the nearest cluster;

[0096] That is: a j =median j' (D(G j ,G j' ))=median j' |G j -G j' |

[0097] b j =min z [median k (D(G j ,G z,k ))]=min z [median k |G j -G z,k |]

[0098] Among them, median means taking the median; min means taking the minimum value; G j represents the weight value G of region j, G j’ represents the weight value G of region j', and j and j' belong to the same cluster; G z,k The weight G represents the weight value of interval k in cluster z, and region j does not belong to cluster z.

[0099] S24, obtain the weight vector obtained by the weighting method corresponding to the optimal clustering result as the final weight vector. The evaluation index of the clustering result is the average silhouette coefficient The larger the better, the silhouette coefficient variance s 2 The smaller the better. Therefore, you can take The largest clustering result is taken as the optimal clustering result.

[0100] Thus, in this embodiment, the weight vectors corresponding to different weighting methods are first obtained, and then the weight value G of each region is calculated based on the three-level indicator weight vector; then the elbow method is used to determine the clustering result of the weight value G of the region under the same weighting method; and then the clustering result is used to determine the weight value G of the region under the same weighting method. As the weight vector screening condition, select The largest weight vector is used as the final weight vector.

[0101] It is worth noting that in this embodiment, only the weights of the third-level indicators are considered during clustering, but the weighting method corresponding to the weight vector of the screened third-level indicators also gives the weights of the dimensions and second-level indicators.

[0102] That is, the final weight vector corresponds one-to-one to the third-level indicator, the second-level indicator and the weight, and each final weight vector contains the weights of different regions obtained using the same weighting method under the third-level indicator.

[0103] In this way, the weight vector of the final three-level index is determined as w0 = [h1, h2, h3, ..., h i ,…,h 29 ]; w0∈{w1,w2,w3,w4,w5,w6}, w6 is the weight vector of the third-level indicators marked by experts; let the weight vector of the second-level indicators obtained by the screening weighting method be recorded as {h 30 , h 31 , h 32 , h 33 , h 34 , h 35 , h 36 , h 37}, the dimension weight vector is recorded as {h 38 , h 39 , h 40 , h 41}.

[0104] S3. Calculate the degree of membership of each region corresponding to each evaluation level under different three-level indicators.

[0105] In this implementation, each three-level indicator is divided into three evaluation levels: "general, good, and excellent".

[0106] Reference Figure 4 , Level 3 indicator C iThe membership calculation of specifically includes the following steps.

[0107] S31. All the third-level indicators C of the region i Arrange in ascending order, then set three breakpoints to set the third-level indicator C i Divide into four intervals, take the minimum internal variance and the maximum inter-class variance as the optimization goal, optimize the breakpoint position, and get C i The corresponding breakpoint (v i1 , v i2 , v i3 ), v i1 <v i2 <v i3 .

[0108] Data breakpoints appropriately group similar values ​​in the data and determine the optimal data segmentation point; the natural breakpoint method better maintains the statistical characteristics of the data when the number of levels is determined. In a specific embodiment, taking the new energy vehicle ownership index as an example, the number of new energy vehicles in 59 county-level regions is first sorted in ascending order. Then, by setting different breakpoints in sequence, the internal variance and inter-class variance at different breakpoints are found. The breakpoint that minimizes the internal variance and maximizes the inter-class variance is selected as the final standard value for the evaluation level of the new energy vehicle ownership index. Figure 5 In the figure, the horizontal axis represents the 59 county-level regions arranged in ascending order of the number of new energy vehicles, and the vertical axis represents the number of new energy vehicles in each county-level region. The natural breakpoint method is used to divide them, forming three breakpoints: (33, 3355), (50, 7536), and (56, 13107). The three breakpoints correspond to the standard values ​​of the "general, good, and excellent" evaluation levels of the new energy vehicle ownership indicators in the 59 county-level regions of Anhui Province.

[0109] S32. Calculate the third-level index C of each region i The membership degree of each evaluation level below;

[0110]

[0111] Among them, r i1 Indicates that the area is in the third level indicator C i The membership degree is evaluated as "general", r i2 Indicates that the area is in the third level indicator C i The membership degree is evaluated as “good”, r i3 Indicates that the area is in the third level indicator C i The membership degree of the evaluation is "excellent", x is the third-level indicator C representing the region i The value of .

[0112] S4. Calculate the secondary index B of each region u and dimension A yThe membership degree of each evaluation level below;

[0113] Let w0 = [h1, h2, h3, ..., h i ,…,h 29 ] is the weight vector corresponding to the three-level indicators, h i Indicates the third-level indicator C i The weight of

[0114] Let d u1 Indicates that the area is in the secondary indicator B u The membership degree of the next evaluation is "average", d u2 Indicates that the area is in the secondary indicator B u The membership degree is evaluated as “good”, d u3 Indicates that the area is in the secondary indicator B u The next assessment is “excellent” in terms of membership;

[0115]

[0116] Among them, CB u Indicates secondary indicator B u The third level indicator C i A collection of .

[0117] Let {h 30 , h 31 , h 32 , h 33 , h 34 , h 35 , h 36 , h 37} is the weight vector corresponding to the secondary index, F y1 Indicates that the area is in dimension A y The membership degree F is evaluated as "general". y2 Indicates that the area is in dimension A y The membership degree F is evaluated as "good". y3 Indicates that the area is in dimension A y The next assessment is “excellent” in terms of membership;

[0118]

[0119] Among them, BA y Represents dimension A y Secondary indicator B u The collection of h u Indicates secondary indicator B u The weight, h u ∈{h 30 , h 31 , h 32 , h 33 , h 34 , h35 , h 36 , h 37}.

[0120] S5. Calculate the final evaluation level of the region and obtain the region's third-level indicator membership matrix C, second-level indicator membership matrix B, and dimension membership matrix A, as shown in Table 2, Table 3, and Table 4, respectively.

[0121] The final evaluation grade of a region is calculated by first calculating the dimension weighted values ​​of each evaluation grade, and then selecting the evaluation grade corresponding to the largest dimension weighted value as the final evaluation grade.

[0122] Let {h 38 , h 39 , h 40 , h 41} is the weight vector corresponding to the dimension; R1 is the membership degree of the regional evaluation grade of “general”, R2 is the membership degree of the regional evaluation grade of “good”, and R3 is the membership degree of the regional evaluation grade of “excellent”;

[0123]

[0124] h y Represents dimension A y The weight, h y ∈{h 38 , h 39 , h 40 , h 41};A y ∈{economic, social, geographical, electrical};

[0125] At this time, if MAX{R1, R2, R3} = R1, it means that the final evaluation grade of the area is "general"; if MAX{R1, R2, R3} = R2, it means that the final evaluation grade of the area is "good"; if MAX{R1, R2, R3} = R3, it means that the final evaluation grade of the area is "excellent".

[0126] Table 2: Regional three-level indicator membership matrix C

[0127] generally good excellent <![CDATA[Third-level indicator C1]]> <![CDATA[r 11 ]]> <![CDATA[r 12 ]]> <![CDATA[r 13 ]]> <![CDATA[Third-level indicator C2]]> <![CDATA[r 21 ]]> <![CDATA[r 22 ]]> <![CDATA[r 23 ]]> …… <![CDATA[Third-level indicator C i > <![CDATA[r i1 ]]> <![CDATA[r i2 ]]> <![CDATA[r i3 ]]>

[0128] Table 3: Regional secondary indicator membership matrix B

[0129] generally good excellent <![CDATA[Secondary indicator B1]]> <![CDATA[d 11 ]]> <![CDATA[d 12 ]]> <![CDATA[d 13 ]]> <![CDATA[Secondary indicator B2]]> <![CDATA[d 21 ]]> <![CDATA[d 22 ]]> <![CDATA[d 23 ]]> …… <![CDATA[Secondary indicator B u > <![CDATA[d u1 ]]> <![CDATA[d u2 ]]> <![CDATA[d u3 ]]>

[0130] Table 4: Dimensional membership matrix A of regions

[0131] generally good excellent <![CDATA[Dimension A1]]> <![CDATA[F 11 ]]> <![CDATA[F 12 ]]> <![CDATA[F 13 ]]> <![CDATA[Dimension A2]]> <![CDATA[F 21 ]]> <![CDATA[F 22 ]]> <![CDATA[F 23 ]]> <![CDATA[Dimension A3]]> <![CDATA[F 31 ]]> <![CDATA[F 32 ]]> <![CDATA[F 33 ]]> <![CDATA[Dimension A4]]> <![CDATA[F 41 ]]> <![CDATA[F 42 ]]> <![CDATA[F 43 ]]>

[0132] In specific implementation, you can use Figure 6 The bar chart shown shows the evaluation levels of different secondary and tertiary indicators.

[0133] Reference Figure 7 The present invention also includes a distribution network cluster difference analysis device based on improved fuzzy evaluation, which includes:

[0134] An indicator screening and establishment module is used to establish a four-dimensional, three-level indicator system for evaluating new rural local power systems. This module analyzes the specific interpretations of indicators in each dimension in practice to understand the factors that influence planning and development when dividing different scenarios in new rural local power systems.

[0135] The indicator weighting calculation module is used to analyze the specific impact of different weighting methods on the clustering results based on the silhouette coefficient of the clustering results, and determine the best weighting method for the rural new local power system scenario division to ensure that the clustering results obtained by this weighting method are optimal;

[0136] The fuzzy comprehensive evaluation module analyzes the corresponding evaluation level standard values, i.e., breakpoints, for different indicators, ensuring that the membership degree of each indicator can be established when conducting fuzzy comprehensive evaluation, thereby enhancing the applicability of fuzzy comprehensive evaluation to the qualitative analysis of typical scenarios of new rural local power systems.

[0137] Reference Figure 8 A computer device provided in an embodiment of the present application includes: a processor and a memory, the memory storing a computer program executable by the processor, and the computer program performs the above method when executed by the processor.

[0138] An embodiment of the present application further provides a storage medium, on which a computer program is stored. When the computer program is run by a processor, the above method is executed.

[0139] Among them, the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0140] In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "connect," "fixed," etc. should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integration; mechanical connection or electrical connection; direct connection or indirect connection through an intermediate medium; internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0141] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0142] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0143] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection having one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0144] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0145] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0146] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and are not to be construed as limiting the present invention. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

[0147] Of course, it will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, but also encompasses the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and it is intended that all variations that fall within the meaning and range of equivalents of the claims be encompassed within the present invention. Any reference signs in the claims should not be construed as limiting the claim to which they relate.

[0148] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0149] The technology, shape, and structure not described in detail in the present invention are all well-known technologies.

Claims

1. A distribution network clustering difference analysis method based on improved fuzzy evaluation, characterized by: First, a multi-dimensional three-level evaluation system is constructed, where each dimension includes one or more secondary indicators, and each secondary indicator includes one or more tertiary indicators; Empowering evaluation indicators at all levels; Set the evaluation level and calculate the membership degree of each region corresponding to each evaluation level under different three-level indicators; Calculate the weighted value of the third-level indicators of each region under the same second-level indicator at the same evaluation level as the membership degree of the second-level indicator at the evaluation level; Calculate the weighted value of the secondary indicators of each region under the same dimension at the same evaluation level as the membership degree of the dimension at the evaluation level; Calculate the weighted values ​​of all dimensions of each region on the same evaluation level as the final evaluation membership, and select the evaluation level corresponding to the largest final evaluation membership as the final evaluation level of the region; for each third-level indicator, second-level indicator and dimension of each region, select the evaluation level corresponding to the maximum membership value as the evaluation level of the corresponding evaluation indicator, and generate an evaluation report.

2. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 1 is characterized in that: The way to weight each three-level indicator is as follows: first, different weighting methods are used to obtain the weight vectors of the indicators at all levels in the region, and then the weighted value G of the region corresponding to the three-level indicator under each weighting method is calculated; For each weighting method, cluster the weighted values ​​G of each region to obtain the performance of the clustering results corresponding to different weighting methods; The weight vector obtained by the weighting method corresponding to the optimal clustering result is obtained as the final weight vector.

3. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 2 is characterized in that: The method for clustering the weighted values ​​of each region is as follows: first, normalize the values ​​of the same three-level indicators in different regions, perform weighted operation on the weights of the three-level indicators in each region and the normalized values ​​of the three-level indicators to obtain the weighted values ​​of the regions; then cluster the weighted values ​​of different regions.

4. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 3 is characterized in that: The normalization formula of the three-level indicator value is: Among them, X ij is the third-level indicator C of region j i The value of T ij For X ij Normalized value of ; min represents the minimum function, and max represents the maximum function.

5. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 2, characterized in that: The weighting methods include: hierarchical analysis method, entropy weight method, multiplication weighting method, additive weighting method and combination weighting method based on game theory; the multiplication weighting method, additive weighting method and combination weighting method based on game theory all use the weight vectors obtained by the hierarchical analysis method and entropy weight method as the benchmark value for calculation.

6. The distribution network cluster difference analysis method based on improved fuzzy evaluation according to claim 1, characterized in that: Each three-level indicator is divided into three evaluation levels: "general, good, and excellent". The method for calculating the region's membership in different evaluation levels under any three-level indicator is as follows: Set three breakpoints for the three-level indicators i1 <v i2 <v i3 , calculate the region's membership degree corresponding to different evaluation levels under the three-level indicators according to the value x of the region's three-level indicators; Among them, r i1 Indicates that the area is in the third level indicator C i The membership degree is evaluated as "general", r i2 Indicates that the area is in the third level indicator C i The membership degree is evaluated as "good", r i3 Indicates that the area is in the third level indicator C i The membership degree of the evaluation is "excellent", x is the third-level indicator C representing the region i value.

7. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 6, characterized in that: The breakpoint setting method for each third-level indicator is: The third-level indicators C i Arrange in ascending order and set multiple groups of breakpoints; calculate the membership of each assessment level of each region under different breakpoints, and select the level corresponding to the maximum membership as the final assessment level of the region under the three-level indicator; classify the regions according to the assessment level at the corresponding three-level coordinates; take the minimum internal variance and the maximum inter-class variance as the optimization goals and select the optimal breakpoint.

8. The distribution network clustering difference analysis method based on improved fuzzy evaluation according to claim 1 is characterized in that: The multi-dimensional three-level evaluation system includes economic, social, geographical and electrical dimensions; the second-level indicators of the economy include the government level, and the third-level indicators of the government level include the number of new energy vehicles and GDP; Secondary indicators of society include population characteristics and industrial proportions; third-level indicators of population characteristics include rural population aging, education level, and population size; third-level indicators of industrial proportions include the proportion of rural primary industry and the proportion of rural secondary industry; The second-level indicators of geography include geographical environment and resource endowment; the third-level indicators of geographical environment include altitude and temperature; the third-level indicators of resource endowment include average sunshine hours and average wind intensity; The secondary indicators of electricity include source, network and load; the tertiary indicators of source include total photovoltaic installed capacity, total wind power installed capacity and total biomass energy installed capacity; the tertiary indicators of network include power supply voltage qualification rate, N-1 pass rate, rural power grid reliability rate, distribution automation coverage rate, line connection rate, inter-station connection rate, heavy load rate, light load rate, average line length and average power supply radius; the tertiary indicators of load include the number of substations under 10kV voltage level, maximum power load, capacity load ratio, 10kV per household distribution transformer capacity and number of charging piles.

9. A device for implementing the distribution network cluster difference analysis method based on improved fuzzy evaluation according to any one of claims 1 to 8, characterized in that: include: Indicator screening and establishment module, used to establish a four-dimensional three-level evaluation system; The indicator weight calculation module is used to calculate the weight vectors of the three-level indicators using different methods and to filter the weight vectors using the weighted value clustering results; The fuzzy comprehensive evaluation module is used to establish the breakpoints corresponding to each three-level indicator and calculate the degree of membership of the region corresponding to different evaluation levels based on the values ​​of the three-level indicators; The weighted values ​​of the membership degrees of different third-level indicators under the second-level indicators are calculated as the membership degrees under different evaluation levels of the second-level indicators.

10. A system, characterized in that: The invention comprises a memory and a processor, wherein a computer program is stored in the memory, the processor is connected to the memory, and the processor is used to execute the computer program to implement the distribution network cluster difference analysis method based on improved fuzzy evaluation as described in any one of claims 1 to 8.

Images