A method and system for evaluating the urgency of distribution network project transformation based on cluster analysis

Abstract

This invention relates to the field of distribution network transformation. Specifically, it provides a method and system for evaluating the urgency of distribution network project transformation based on cluster analysis. This method constructs a seven-item failure indicator system, including 10kV feeder trunk length and trunk cross-section. The DEMATEL method and cross-reinforcement matrix method are applied to the analytic hierarchy process to calculate the urgency of project transformation. Secondly, a two-step clustering algorithm is used to identify priority areas for transformation projects in substations.

Description

Technical Field

[0001] The present invention relates to the field of distribution network transformation, and in particular to a method and system for evaluating the urgency of distribution network project transformation based on cluster analysis. Background Art

[0002] With economic development and improved living standards, local electricity load continues to grow rapidly, and the contradiction between the distribution network's power supply capacity and electricity demand is becoming increasingly prominent. At the same time, decision-making for the transformation of medium- and low-voltage distribution network projects faces problems such as the large number of basic data points, the low accuracy and efficiency of manual experience prediction, and limited funds for distribution network transformation. Therefore, it is urgent to conduct urgency analysis of medium- and low-voltage distribution network transformation projects to provide reference for distribution network transformation and grid planning, and continuously improve the distribution network's power supply capacity. Summary of the Invention

[0003] The main purpose of the present invention is to provide a method and system for evaluating the urgency of distribution network project transformation based on cluster analysis, so as to solve the problems in the related art.

[0004] To achieve the above objectives, according to one aspect of the present invention, a method for evaluating the urgency of distribution network project transformation based on cluster analysis is provided, which includes 10kV feeder urgency calculation and 10kV substation urgency calculation. The 10kV feeder urgency calculation includes the following steps:

[0005] Step 1: Data on distribution network feeder trunk length, trunk cross-section, installed capacity, heavy overload, transferable ratio, wiring mode, and line loss rate are packaged and organized to obtain urgency analysis data samples. Five 10kV feeders are selected as test data samples.

[0006] Step 2: Establish a comprehensive evaluation system for distribution network feeder renovation projects through an improved analytic hierarchy process. Under funding constraints, conduct a detailed analysis of the number and severity of existing line problems and propose the concept of line renovation project urgency, which is the conversion value of the severity of the impact of unqualified line indicators on the safe and economic operation of the power grid. An improved analytic hierarchy process based on the decision-making experiment and evaluation laboratory method and the cross-reinforcement matrix method is used to calculate the indicator weights. The line renovation urgency value is calculated by multiplying the unqualified line indicator value with the weight value and accumulating them.

[0007] Step 3: Taking the total project investment as the constraint condition and the maximum improvement of the line urgency as the objective function, an optimization model was established, and a three-layer analytic hierarchy process model was established;

[0008] Step 4: Calculate the comprehensive value of feeder transformation urgency based on the data samples;

[0009] The 10kV area urgency calculation includes the following steps:

[0010] Step 5: Obtain the longitude and latitude coordinates, heavy overload, operating years, low voltage and other information of the substation from the relevant application system, establish a substation problem clustering database, and select the substations with problems as data samples;

[0011] Step 6: Use a two-step clustering algorithm to determine issues such as overload, heavy load, old age, and low voltage in the substation area. This will determine whether low voltage occurs repeatedly, whether low voltage occurs in large areas, whether there are monitoring data quality issues, whether the problem can be solved by management measures, whether it really needs to be included in the distribution network project, and whether the previous investment arrangement strategy is reasonable, in order to determine the urgency of the substation transformation.

[0012] Furthermore, the hierarchical analysis method specifically includes: first splitting the components of a complex problem into an orderly hierarchical structure, then quantifying the subjective judgment of the measurer, establishing the judgment matrix layer by layer, solving the weight of each judgment matrix, and finally obtaining a ranking table of the optimal solution.

[0013] Furthermore, the AHP also includes the decomposition, comparison and synthesis of priorities.

[0014] Furthermore, the decomposition of priorities is applied to construct a hierarchical structure in which elements at the same level are independent of each other; starting from the highest level of goals, to the highest level of criteria, and then from general sub-criteria to the lowest level of alternative measures.

[0015] Furthermore, the decomposition of priorities is applied to construct a hierarchical structure in which elements at the same level are independent of each other; starting from the highest level of goals, to the highest level of criteria, and then from general sub-criteria to the lowest level of alternative measures.

[0016] Furthermore, the comparison of priorities includes the application of pairwise comparisons of the relative importance of each element of the upper level to the elements of the lower level to construct an inverse matrix, whose main eigenvectors give the priorities.

[0017] Furthermore, the priority synthesis is to take the priority of the elements of a certain layer relative to the upper layer standards as the weight, and perform weighted summation to obtain the comprehensive index.

[0018] Furthermore, the two-step clustering algorithm is as follows: let cluster i and cluster j, (ij) represents the clusters they merge into, and the number of sample points they contain is n i 、n j 、n (ij) There are K continuous variables in the data set, where the variance estimates of the kth variable, k = 1, 2, ..., K in the entire data set, cluster i, cluster j and (ij) are respectively and Then the log-likelihood distance between cluster i and cluster j is dij The definition is as follows:

[0019]

[0020] Furthermore, if the desired number of clusters is not predetermined for the two-step clustering algorithm, the algorithm can automatically determine the optimal number using Schwarz's Bayesian criterion or Akaike's information criterion. Let the number of clusters be J and the total number of observations in the data set be N, then:

[0021]

[0022]

[0023] On the other hand, a distribution network project transformation urgency assessment system based on cluster analysis is provided, characterized in that it is used to load and execute any of the distribution network project transformation urgency assessment methods based on cluster analysis described above.

[0024] Compared with the existing technology, the present invention has the following beneficial effects: based on the grid parameters and operating data of six major systems, namely the production management information system PMS2.0, the marketing management system SG186, the dispatching automation system iES600, the equipment (asset) operation and maintenance lean management system (GIS system), the electric energy metering system TMR, and the electricity consumption information collection system, the data is pre-processed by using big data collection, cleaning, sorting and classification methods, and then a comprehensive statistical method is applied to establish an optimization model. Among them, the hierarchical analysis method is improved, and the DEMATEL method and the cross-reinforcement matrix method are used to calculate the urgency of the transformation. The main indicators of the hierarchical analysis method model include seven problem unqualified indicators such as the 10kV feeder trunk length and trunk cross-section. At the same time, a two-step clustering algorithm is used to explore the priority areas for substation project transformation. Finally, the project is managed in a granular manner, so that the project reserves correspond to the star-level issues, ensuring the connection between project needs and grid development, and the results are integrated into Superset to realize chart visualization. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] The structures, proportions, sizes, etc. depicted in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not intended to limit the conditions under which the present invention can be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportional relationships, or adjustments in size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and objectives that can be achieved by the present invention.

[0027] Figure 1 This is a hierarchical structure diagram of the AHP sub-model in an embodiment of the present invention;

[0028] Figure 2 This is a diagram showing the urgency analysis results on a satellite map in an embodiment of the present invention;

[0029] Figure 3 This is a sample clustering result diagram in an embodiment of the present invention;

[0030] Figure 4 This is a spatial display diagram of sample clustering results in an embodiment of the present invention;

[0031] Figure 5 This is an analysis chart of the urgency of the second batch of distribution network line projects in Xiaogan in 2017. DETAILED DESCRIPTION

[0032] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0033] The technical solution of the present invention will be further described below with reference to the accompanying drawings and through specific implementation methods.

[0034] The Analytic Hierarchy Process (AHP) is a systematic method for optimizing decision-making on targets and multiple options. It first breaks down the components of a complex problem into an orderly hierarchical structure, then quantifies and measures people's subjective judgments, establishes judgment matrices layer by layer, solves the weights of each judgment matrix, and finally obtains a ranking table of the optimal options.

[0035] In the process of problem solving, AHP uses three principles: priority decomposition, comparison and synthesis.

[0036] The principle of decomposition is applied to construct a hierarchical structure in which elements at the same level are independent of each other. Starting from the highest level, the first level of goals, to the second level of criteria supporting the goals, to the third level of sub-criteria, and so on, and then from general (sometimes indeterminate) sub-criteria to the lowest level of alternative measures.

[0037] The principle of comparative judgment is applied to construct a pairwise comparison reciprocal matrix of the relative importance (priority) of each element on the upper level with respect to each element on the lower level, whose main eigenvector gives the priority.

[0038] The third principle is to take the priority of the elements of a certain layer relative to the upper layer standards as the weight, and perform weighted summation to obtain a comprehensive index.

[0039] As can be seen, the AHP is a nonlinear framework that implements deductive and inductive thinking without using syllogisms. It simultaneously considers multiple factors, allows for dependencies and feedback, and makes numerical tradeoffs to arrive at a synthesis or conclusion. Therefore, its use in decision analysis has reasonable aspects.

[0040] Two-step cluster analysis is a statistical method that can automatically identify similar people or objects in a data set. Compared with traditional clustering methods (such as mean clustering algorithm), two-step cluster analysis is also considered more reliable and more accurate. It is considered to be the most appropriate clustering technique that is effective for both continuous and categorical data. [8] Two-step clustering allows for the preservation of complete data information, eliminating the need for data conversion prior to analysis, and thus providing rich interpretations. Two-step clustering can handle both numerical and categorical variables, automatically handle outliers, and discard or assign clusters to the nearest cluster. It is more suitable for mining large data sets, so the second part of the optimization model for this project also applied the two-step clustering algorithm and made improvements. Specifically, during the research process, for other non-normal distributions of the data, we used non-parametric tests instead of t-tests to statistically test whether there were significant differences between clusters.

[0041] Two-step cluster analysis involves two steps, pre-clustering and clustering. In the first step, the original cases are clustered by constructing a cluster feature tree. In the second step, a standard hierarchical clustering algorithm is used to form hierarchical clusters, which allows one to explore solutions with different numbers of clusters. For the multiple solutions generated, it can summarize the optimal number of clusters based on Schwarz's Bayesian Information Criterion (BIC). Once the cluster solution is formed, three validation measures are required. First, the silhouette measures of cohesion and separation must be above the required level of 0.05 to indicate that the intra-cluster distance and inter-cluster distance are valid. Second, the χ2 regression model is performed for categorical variables and continuous variables respectively. 2 Test and t-test [9], to identify the importance of each variable in the cluster and point out the significant differences between clusters. If the absolute value of the statistical information of a variable in the cluster is greater than the critical value, the variable is considered important in distinguishing a cluster from other clusters. Third, the final cluster solutions after halving must be similar (for example, the size, number and characteristics of the clusters).

[0042] The statistical indicators of the two-step cluster analysis method are detailed below.

[0043] The log-likelihood criterion is used in the two-step cluster analysis to reveal the natural groupings in the data. Let cluster i and cluster j, (ij) represents the clusters they merge into, and the number of sample points they contain is n. i 、n j 、n (ij) There are K continuous variables in the data set, where the variance estimates of the kth variable, k = 1, 2, ..., K in the entire data set, cluster i, cluster j and (ij) are respectively and Then the log-likelihood distance between cluster i and cluster j is d ij The definition is as follows:

[0044]

[0045] Obviously, the log-likelihood distance is a probability-based distance. When two clusters are merged into one cluster, the distance between them is related to the reduction in log-likelihood. In the pre-clustering step, all cases in the data are scanned and the log-likelihood distance between them is measured to determine whether they will form a pre-cluster based on a certain threshold distance criterion. In the second step, the clusters from the pre-clustering step are clustered into the optimal number of clusters using an agglomerative clustering algorithm. At this time, each sub-cluster in the first step is regarded as a single record, and if they meet the minimum distance threshold in (1), they are merged into a large record until all data records are clustered into one cluster.

[0046] If the desired number of clusters is not predetermined, the algorithm can automatically determine the optimal number using Schwarz's Bayesian Criterion (AIC) or Akaike's Information Criterion (AIC). Let the number of clusters be J and the total number of observations (total records) in the dataset be N, then:

[0047]

[0048]

[0049] BIC is considered one of the most useful and objective selection criteria because it avoids the arbitrariness of traditional clustering techniques. When considering which variables to remove from the analysis, it is best to select those with the lowest BIC. The Student's t-test is used to determine whether the means of two data sets differ from each other. However, the Student's t-test assumes that the data follow a normal distribution. A silhouette measure value less than 0.2 indicates poor clustering; between 0.2 and 0.5, acceptable clustering; and greater than 0.5 indicates excellent clustering.

[0050] The silhouette coefficient is used to measure the cluster density and separation. Suppose a clustering result C = {C1, ..., C k ,…,C m ,…,C C}, i, j are sample points, d(i, j) is a distance between them, |C k | for C k The size of the family. When i∈C k When Then the silhouette coefficient of point i is defined as

[0051] The average silhouette coefficient of all sample points is

[0052] The clustering effect can be seen from the silhouette coefficient. i It measures whether a point i is suitable to be classified into a certain cluster. If s i <0, indicating that point i is in an inaccurate cluster C k s i >0 and close to 1, indicating that point i is accurately classified into cluster C k If most of the sample points have a high silhouette coefficient, that is, If it is relatively large, it means that the clustering is appropriate; otherwise, it means that there are too many or too few categories.

[0053] The application results of the above statistical methods are introduced below at three levels for the urgency of 10kV feeders and 10kV substations.

[0054] 10kV feeder urgency calculation:

[0055] 1. Data layer

[0056] The data flow of the distribution network mainly consists of real-time data flow, offline data flow and online data query. The data such as the main length, main cross-section, installation capacity, heavy overload, transferable ratio, wiring mode and line loss rate of the distribution network feeder are packaged and organized to obtain the urgency analysis data sample, and 5 10kV feeders are selected as test data samples.

[0057] 2. Method layer

[0058] Medium-voltage distribution network projects involve numerous lines and equipment, and require comprehensive consideration of factors such as transformation effects, funding plans, power supply reliability, power quality, and economic benefits. Traditional mathematical optimization methods are difficult to model them in a unified and holistic manner. The analytic hierarchy process and its improved methods are the most widely used optimization methods.

[0059] By improving the analytic hierarchy process (AHP), a comprehensive evaluation system for distribution network feeder transformation projects was established. Under funding constraints, the number and severity of existing line problems were analyzed in detail, and the concept of urgency of line transformation projects was proposed, that is, the severity conversion value of the impact of unqualified line indicators on the safe and economic operation of the power grid. The improved AHP based on the Decision Experiment and Evaluation Laboratory (DEMATEL) method and the cross-reinforcement matrix method was used to calculate the indicator weights. The line indicator unqualified value was multiplied by the weight value and the sum was accumulated to calculate the line transformation urgency value.

[0060] An optimization model was established with the total project investment as a constraint and the maximum improvement of line urgency as the objective function. This can arrange the priorities of construction and renovation projects more scientifically and effectively under the constraints of limited funds. A three-layer hierarchical analysis method model was established, and its structure is as follows: Figure 1 shown.

[0061] The comprehensive value of feeder transformation urgency calculated for the data samples is shown in Table 1.

[0062] Table 1 Feeder index unqualified value conversion table

[0063]

[0064]

[0065] 3. Application layer

[0066] The urgency calculation results are displayed on the satellite map in a data visualization way. The order of urgency of feeder line reconstruction from high to low is patrol 58 Yangpo line, patrol 55 Sangshu line, patrol 53 Bali line, patrol 56 Xundian line, patrol 59 Gaodian line, and the display effect is as follows: Figure 2 shown.

[0067] Calculation of urgency for 10kV substations:

[0068] 1. Data layer

[0069] The latitude and longitude coordinates, heavy overload, operating years, low voltage and other information of the substation are obtained from the relevant application systems, a substation problem clustering database is established, and the substations with problems in Yunmeng Company are selected as data samples.

[0070] 2. Method layer

[0071] A two-step clustering algorithm is used to determine whether low voltage occurs repeatedly, whether low voltage occurs in clusters, whether there are monitoring data quality issues, whether they are problems that can be solved by management measures, whether they need to be included in the distribution network project, and whether the previous investment arrangement strategy is reasonable. The urgency of the transformation of the substation is determined. The results are shown in the figure. Figure 3 .

[0072] 3. Application layer

[0073] The 375 substations of Yunmeng Company are grouped into three categories: Figure 4 .

[0074] from Figure 4 The middle white line is the administrative boundary of the district and county. It can be seen that the problem areas are mainly distributed in the west and south of the county company.

[0075] From this we can draw the following conclusions:

[0076] The substations with both old and overload problems are cluster-1, accounting for 18.4%. This type of problem is the most urgent, with a total of 69 substations. Priority should be given to resolving this type of substation, requiring approximately 17 million yuan in renovation funds.

[0077] The aging problem is the most prominent among the sample substations. Cluster-3 is the substation with only aging problems, accounting for as high as 67.7%. For substations with such problems, it is advisable to gradually solve them through an overall coordination plan in combination with load development.

[0078] Application results:

[0079] 1. This project is applied to the second batch of distribution network project evaluation in Xiaogan in 2017. The urgency of 101 line projects is calculated by analytic hierarchy process, such as Figure 5 .

[0080] The implementation of this project can provide a more objective and investment-oriented urgency evaluation for distribution network line project review based on expert experience.

[0081] 2. Distribution network planning: Grid planning is problem- and demand-oriented. Through project urgency data analysis, the areas where problems and demands are located are identified, and the key areas for distribution network planning are determined.

[0082] 3. The evaluation of medium and low voltage projects takes the maximum improvement of feeder urgency as the objective function and the total project investment as the constraint conditions to establish a distribution network optimization and transformation model to maximize the use of funds and effectively arrange the project transformation sequence.

[0083] 4. Post-evaluation of medium and low voltage projects Post-evaluation of medium and low voltage projects that have been put into production can improve investment decision-making management, improve relevant policies and measures, enhance scientific management level, and provide experience for future medium and low voltage project construction.

[0084] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0085] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can, for example, be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0086] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for evaluating the urgency of distribution network project transformation based on cluster analysis, characterized in that: The 10kV feeder urgency calculation and the 10kV substation urgency calculation are included. The 10kV feeder urgency calculation includes the following steps: Step 1: Data on distribution network feeder trunk length, trunk cross-section, installed capacity, heavy overload, transferable ratio, wiring mode, and line loss rate were packaged and organized to obtain urgency analysis data samples. Five 10kV feeders were selected as test data samples. Step 2: Establish a comprehensive evaluation system for distribution network feeder renovation projects through an improved analytic hierarchy process. Under funding constraints, conduct a detailed analysis of the number and severity of existing line problems and propose the concept of line renovation project urgency, which is the conversion value of the severity of the impact of unqualified line indicators on the safe and economic operation of the power grid. An improved analytic hierarchy process based on the decision-making experiment and evaluation laboratory method and the cross-reinforcement matrix method is used to calculate the indicator weights. The line renovation urgency value is calculated by multiplying the unqualified line indicator value with the weight value and accumulating them. The AHP method specifically includes: first, breaking down the components of a complex problem into an orderly, hierarchical structure; then, quantifying and measuring people's subjective judgments; establishing judgment matrices layer by layer; solving for the weights of each judgment matrix; and finally, obtaining a ranking table of optimal solutions. The AHP method also includes priority decomposition, comparison, and synthesis. Step 3: Taking the total project investment as the constraint condition and the maximum improvement of the line urgency as the objective function, an optimization model was established, and a three-layer analytic hierarchy process model was established; Step 4: Calculate the comprehensive value of feeder transformation urgency based on the data samples; The 10kV area urgency calculation includes the following steps: Step 5: Obtain the longitude and latitude coordinates, heavy overload, operating years, and low voltage information of the substation from the relevant application system, establish a substation problem clustering database, and select substations with problems as data samples; Step 6: A two-step clustering algorithm is used to determine whether low voltage occurs repeatedly, whether low voltage occurs in patches, whether there are monitoring data quality issues, whether they are problems that can be solved by management measures, whether they need to be included in the distribution network project, and whether the previous investment arrangement strategy is reasonable, so as to determine the urgency of the transformation of the substation; the two-step clustering algorithm is specifically as follows: set cluster and clustering , Indicates the clusters they are merged into, and the number of sample points they contain is 、 、 , there are A continuous variable, where variables, In the entire data set, clustering , clustering and The variance estimates of 、 、 and , then clustering and clustering The log-likelihood distance between The definition is as follows: ； If the desired number of clusters is not predetermined in the two-step clustering algorithm, the algorithm can automatically determine the optimal number using Schwarz's Bayesian criterion or Akaike's information criterion; let the number of clusters be , the total number of observations in the dataset , then: ； 。 2. The method for evaluating the urgency of distribution network project transformation based on cluster analysis according to claim 1, characterized in that: The decomposition of priorities is applied to construct a hierarchical structure in which elements of the same level are independent of each other; starting from the highest level of goals, to the highest level of criteria, and then from general sub-criteria to the lowest level of alternative measures.

3. The method for evaluating the urgency of distribution network project transformation based on cluster analysis according to claim 1, characterized in that: The comparison of the priorities includes the application of pairwise comparisons of the relative importance of each element of the upper level to the elements of the lower level to construct an inverse matrix, whose main eigenvectors give the priorities.

4. The method for evaluating the urgency of distribution network project transformation based on cluster analysis according to claim 1, characterized in that: The priority synthesis is to take the priority of the elements of a certain layer relative to the upper layer standard as the weight, and perform weighted summation to obtain a comprehensive index.

5. A distribution network project transformation urgency assessment system based on cluster analysis, characterized in that: Used to load and execute the distribution network project transformation urgency assessment method based on cluster analysis as described in any one of claims 1-4.

Images