CN117578434A

CN117578434A - Power distribution network flexibility evaluation method and device considering flexible resource adjustability

Info

Publication number: CN117578434A
Application number: CN202311579029.5A
Authority: CN
Inventors: 刘志文; 邵冲; 李岩; 段瑶; 程苒
Original assignee: Guangdong Power Grid Co Ltd; Energy Development Research Institute of China Southern Power Grid Co Ltd
Current assignee: Guangdong Power Grid Co Ltd; Energy Development Research Institute of China Southern Power Grid Co Ltd
Priority date: 2023-11-23
Filing date: 2023-11-23
Publication date: 2024-02-20
Anticipated expiration: 2043-11-23
Also published as: CN117578434B

Abstract

The application relates to a power distribution network flexibility evaluation method and device considering flexible resource adjustability. The method comprises the following steps: acquiring historical operation data of the power distribution network, and determining each typical operation scene of the power distribution network through a clustering algorithm according to the historical operation data; acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index; and quantitatively analyzing the influence of different flexible resource accesses on the flexibility of the power distribution network according to the magnitude of each comprehensive evaluation index value. The method can more comprehensively embody the influence of flexible resource access on the flexibility of the power distribution network.

Description

Power distribution network flexibility evaluation method and device considering flexible resource adjustability

Technical Field

The application relates to the technical field of power distribution network flexibility evaluation, in particular to a power distribution network flexibility evaluation method considering flexible resource adjustability.

Background

With the rapid development of renewable energy sources, flexible power grid technology and scheduling strategies have developed, and flexible power grids convert traditional centralized power systems into distributed and intelligent energy systems by introducing advanced communication, control and information technologies, wherein the distributed energy sources comprise solar energy, wind energy, energy storage systems and the like.

Most of traditional power distribution network performance determining methods consider only a single scene or a relatively conservative power distribution network performance evaluation system when flexible resource access is not considered in various scenes, and consideration of changeable characteristics of different flexible resource cooperative access power distribution networks in different operation scenes is lacking, so that after the power distribution networks are accessed to distributed energy sources, the traditional power distribution network flexibility evaluation method is not flexible and comprehensive enough.

Disclosure of Invention

Based on this, there is a need to provide a power distribution network flexibility assessment method, apparatus, computer device, computer readable storage medium and computer program product, which can more fully embody the influence of flexible resource access on the flexibility of the power distribution network and consider the flexible resource adjustability.

In a first aspect, the present application provides a method for evaluating flexibility of a power distribution network by considering flexible resource adjustability of the power distribution network, including:

Acquiring historical operation data of the power distribution network, and determining each typical operation scene of the power distribution network through a clustering algorithm according to the historical operation data;

acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index;

and quantitatively analyzing the influence of different flexible resource accesses on the flexibility of the power distribution network according to the magnitude of each comprehensive evaluation index value.

In one embodiment, historical operation data of the power distribution network is obtained, and each typical operation scene of the power distribution network is determined through a clustering algorithm according to the historical operation data, including:

determining an initial cluster center of a cluster corresponding to each operation scene in the historical operation data;

respectively calculating the distance between each data point in the historical operation data and each initial clustering center, and distributing each data point to the cluster with the shortest distance between each initial clustering center and each data point;

and updating the clustering center of each cluster according to the data mean value of each cluster so as to form a typical operation scene of the power distribution network by each data point in each cluster corresponding to the clustering center meeting the preset convergence condition.

In one embodiment, acquiring an evaluation index value of each flexible resource access scenario in each typical operation scenario of the power distribution network, and determining a comprehensive evaluation index value of each flexible resource access scenario according to each evaluation index value includes:

respectively constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene, and standardizing the evaluation index matrix to obtain a standardized evaluation index matrix;

creating a sample set according to the standardized index matrix, creating a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance measurement indexes of all indexes through the random forest model, and determining first characteristic weights of all the indexes according to the model performance measurement indexes; randomly resetting the values of the indexes in the standardized index matrix, calculating the difference value of the model performance measurement indexes before and after the resetting arrangement for each index, and determining a second characteristic weight according to the difference value of the model performance measurement indexes;

and calculating the weighted sum of the first characteristic weight and the second characteristic weight to obtain the comprehensive weight of each evaluation index value, and calculating the evaluation index value of each flexible resource access scene under each typical operation scene according to the comprehensive weight.

In one embodiment, an evaluation index matrix is respectively constructed according to evaluation index values of different flexible resource access scenes under each typical operation scene, the evaluation index matrix is standardized, and a standardized evaluation index matrix is obtained, including:

determining the mapping relation between the magnitude of each index value and the performance of the power distribution network, and constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes in each typical operation scene;

under the condition that the larger the index value is, the better the corresponding performance is represented, the index value is standardized according to the preset minimum index value;

and under the condition that the smaller index value represents the better corresponding performance, normalizing the index value according to the preset maximum index value to obtain a normalized index matrix.

In one embodiment, creating a sample set according to the standardized index matrix, creating a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance metrics of each index through the random forest model, and determining a first feature weight of each index according to the model performance metrics, including:

constructing a plurality of decision trees, calculating the coefficient of the foundation of each evaluation factor, sequencing the classified nodes and distributing node weights;

Calculating the reduction amount of the mean square error of each index in the random forest before and after splitting of each node, wherein the mean square error is used for representing the node contribution rate of each index;

and calculating the cumulative sum of the mean square error reduction amounts of all the indexes on all the trees, and carrying out weight normalization on the cumulative sum to obtain the first characteristic weight.

In one embodiment, determining the comprehensive evaluation index value of each flexible resource access scenario according to each evaluation index value includes:

taking evaluation index values of each flexible resource access scene under each typical operation scene as evidence, calculating cosine similarity of each two evidences, calculating mutual support degree between the evidences, and carrying out weighted average treatment on the evidences according to the mutual support degree to obtain a weighted average evidence body;

the collision factor for every two weighted average evidence volumes is calculated,

and for each flexible resource access scene, fusing the evidence bodies in each typical operation scene according to an evidence synthesis formula, and taking the logarithmic amplification value of the fused evidence bodies as a comprehensive evaluation index value.

In a second aspect, the present application further provides a power distribution network flexibility assessment device considering flexible resource adjustability, including:

the data acquisition module is used for acquiring historical operation data of the power distribution network, and determining each typical operation scene of the power distribution network through a clustering algorithm according to the historical operation data;

The index acquisition module is used for acquiring the evaluation index values of the flexible resource access scenes in each typical operation scene of the power distribution network, and determining the comprehensive evaluation index value of each flexible resource access scene according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index;

and the evaluation module is used for quantitatively analyzing the influence of different flexible resource accesses on the flexibility of the power distribution network according to the magnitude of each comprehensive evaluation index value.

In a third aspect, the present application also provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:

In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:

In a fifth aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, performs the steps of:

According to the power distribution network flexibility evaluation method, device, computer equipment, storage medium and computer program product considering the flexible resource adjustability, the historical operation data of the power distribution network are obtained, each typical operation scene of the power distribution network is determined through a clustering algorithm according to the historical operation data, the operation mode of the power distribution network can be better understood through clustering of the historical operation data, and the determination of the typical operation scene enables the system to distinguish and identify different modes of system operation, so that a foundation is provided for subsequent evaluation and optimization; acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index; the performance of the flexible resource access scene under each typical operation scene can be quantified by acquiring each evaluation index value, and each evaluation index is integrated into a comprehensive evaluation index value, so that the contribution of the flexible resource to the overall performance of the system can be more clearly quantified, and the performance influence of each flexible resource access to the power distribution network can be conveniently compared; the influence of different flexible resource accesses on the flexibility of the power distribution network is quantitatively analyzed according to the magnitude of each comprehensive evaluation index value, so that the system can continuously optimize the performance of the system, the configuration of the flexible resources is adjusted according to actual conditions so as to ensure the efficient operation of the system, and the power distribution network can configure the flexible resources according to the magnitude of the comprehensive evaluation index value so as to ensure the efficient operation of the system. By the method, the influence of flexible resource access on the flexibility of the power distribution network is more comprehensively embodied.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the related art, the drawings that are required to be used in the embodiments or the related technical descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to the drawings without inventive effort for a person having ordinary skill in the art.

FIG. 1 is an application environment diagram of a power distribution network flexibility assessment method that considers flexible resource tunability in one embodiment;

FIG. 2 is a flow diagram of a method for evaluating flexibility of a power distribution network in consideration of flexible resource tunability in one embodiment;

FIG. 3 is a flow diagram of a method for evaluating flexibility of a power distribution network in consideration of flexible resource tunability in one embodiment;

FIG. 4 is a schematic diagram of an evaluation index in one embodiment;

FIG. 5 is a block diagram of a power distribution network flexibility assessment device that considers flexible resource tunability in one embodiment;

fig. 6 is an internal structural diagram of a computer device in one embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

The power distribution network flexibility evaluation method considering flexible resource adjustability provided by the embodiment of the application can be applied to an application environment shown in fig. 1. Wherein the terminal 102 communicates with the server 104 via a network. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on a cloud or other network server. The terminal 102 may be a power data acquisition terminal in a power system, such as various sensors, load monitoring devices, smart meters, transformer monitoring devices, line monitoring devices, and SCADA systems (Supervisory Control And Data Acquisition, data acquisition and monitoring systems), etc. The server 104 may be implemented as a stand-alone server or a server cluster including a plurality of servers, and the server 104 may be a control server of the power system.

In an exemplary embodiment, as shown in fig. 2, a method for evaluating flexibility of a power distribution network considering flexible resource adjustability is provided, and an application environment of the method in fig. 1 is taken as an example to describe the method, which includes the following steps 202 to 206. Wherein:

Step 202, acquiring historical operation data of the power distribution network, and determining each typical operation scene of the power distribution network through a clustering algorithm according to the historical operation data.

The distribution network refers to a power network system which is composed of various distribution equipment and settings, converts voltage and distributes electric energy to end users directly. Historical operating data refers to operating data of the power distribution network over a period of time in the past, such as load data, fluctuation data, and the like. Clustering algorithms refer to algorithms for grouping a collection of physical or abstract objects into classes composed of similar objects. Typical operation scenarios are preset operation or response scenarios of the assigned power grid, such as basic power supply, balanced load, fault recovery, and the like.

By way of example, historical operation data of the power distribution network, such as current, voltage, power, equipment state and the like, can be collected through a data collection and monitoring system and the like, the data are cleaned and preprocessed, missing values, abnormal values or noise in the data are processed, the quality of the data is improved, the data can be standardized after preprocessing, different types of data have the same scale, features are extracted from the historical operation data, the features can include statistics of various power parameters, spectrum analysis and the like, key features capable of reflecting the operation state of the system are selected, clustering analysis is conducted on the historical operation data through a selected clustering algorithm according to feature clustering algorithms, such as K-means clustering and hierarchical clustering, the operation data of the system are divided into different clusters according to clustering results, and each cluster represents a typical operation scene, so that typical operation scenes of the power distribution network are obtained.

Step 204, acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; the evaluation indexes comprise a coordination capacity index, a coordination capacity index and a reliability index.

The flexible resource refers to equipment or a system capable of rapidly adjusting generated or consumed power according to the requirements of a power system, and is beneficial to balancing power supply and power requirements of a power distribution network. The flexible resource access scenario refers to different situations or schemes in power system planning and operation, including how to introduce, manage and coordinate various flexible resources to improve the flexibility and reliability of the system. The evaluation index value refers to a specific value used for measuring the performance of the power system, the effectiveness of flexible resources and other related indexes. The comprehensive evaluation index value refers to a comprehensive index value which considers the influence of the flexible resource on the power distribution network in various aspects, including the aspects of load balancing, power supply reliability, economic benefit and the like, and can be used for evaluating the overall utility of the flexible resource. The co-regulation capability index refers to an index that a metric flexible resource can work cooperatively to meet the requirements of the power system, can reflect interoperability between flexible resources, and how to work cooperatively under different situations to maintain the balance of the power system. The co-operation capability index refers to an index for evaluating the degree of co-operation between a flexible resource and other electric power system participation subjects (such as a power generation subject, a grid operation subject, and a power consumption subject). The reliability index refers to an index for measuring the reliability of the power system, and may include accident frequency, recovery time, equipment reliability, and the like.

For example, evaluation indexes related to the flexible resource access scenario, such as evaluation indexes in aspects of power supply reliability, load balancing, economy, environmental impact and the like, may be first clarified, and each evaluation index of the flexibility of the power system is quantified, for example, interruption time, frequency and the like may be calculated for the power supply reliability; for economy, cost, benefit, etc. may be calculated, and weights may be assigned to each index to reflect importance in the comprehensive evaluation process, the weights may be assigned by an analysis method or experience, and the comprehensive evaluation index value of each flexible resource access scenario may be calculated using the assigned weights, and may be calculated by a weighted summation, or may be calculated by a comprehensive evaluation method, such as an Analytic Hierarchy Process (AHP), etc.

And 206, quantitatively analyzing the influence of different flexible resource accesses on the flexibility of the power distribution network according to the magnitude of each comprehensive evaluation index value.

For example, a threshold value or a standard of the comprehensive evaluation index value may be determined according to a relationship between the magnitude of the comprehensive index value and a change rule of the flexibility of the power distribution network, an actual running condition or experience, so as to measure the influence degree of the flexible resource access scenario on the flexibility of the power distribution network.

According to the power distribution network flexibility evaluation method considering the flexible resource adjustability, the historical operation data of the power distribution network are obtained, each typical operation scene of the power distribution network is determined through a clustering algorithm according to the historical operation data, the operation mode of the power distribution network can be better understood through clustering of the historical operation data, and the determination of the typical operation scene enables the system to distinguish and identify different modes of system operation, so that a foundation is provided for subsequent evaluation and optimization; acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index; the flexibility of the power distribution network of the flexible resource access scene under each typical operation scene can be quantified by acquiring each evaluation index value, and each evaluation index is integrated into a comprehensive evaluation index value, so that the contribution of the flexible resource to the flexibility of the power distribution network can be quantified more clearly, and the influence of each flexible resource access to the flexibility of the power distribution network can be compared conveniently; the influence of different flexible resource accesses on the flexibility of the power distribution network is quantitatively analyzed according to the magnitude of each comprehensive evaluation index value, so that the power distribution network can continuously optimize the flexibility of the power distribution network, the configuration of the flexible resources is adjusted according to actual conditions so as to ensure the efficient operation of the system, and the power distribution network can configure the flexible resources according to the magnitude of the comprehensive evaluation index value so as to ensure the efficient operation of the system, thereby more comprehensively reflecting the influence of the flexible resource accesses on the flexibility of the power distribution network.

In an exemplary embodiment, historical operation data of the power distribution network is obtained, and each typical operation scene of the power distribution network is determined according to the historical operation data through a clustering algorithm, including: determining an initial cluster center of a cluster corresponding to each operation scene in the historical operation data; respectively calculating the distance between each data point in the historical operation data and each initial clustering center, and distributing each data point to the cluster with the shortest distance between each initial clustering center and each data point; and updating the clustering center of each cluster according to the data mean value of each cluster so as to form a typical operation scene of the power distribution network by each data point in each cluster corresponding to the clustering center meeting the preset convergence condition.

Where a cluster refers to a collection of data points in the data that are similar or related to each other, in cluster analysis, a cluster is a population that is partitioned by an algorithm based on the similarity of the data points. The initial cluster center refers to an initial center point of each cluster set at the beginning of the clustering algorithm. Distance refers to a measure of the degree of similarity or difference between two objects, such as Euclidean distance, manhattan distance, minkowski distance, etc., which can be used to measure how far or near between two data points. The data average refers to the average of all values in a set of data, which can be used to describe the center location of the data. The convergence condition refers to a condition that in an iterative algorithm, the algorithm stops iteration.

For example, clustering may be performed using K-means clustering, hierarchical clustering, etc., the number of classes may be determined using heuristic methods, experience, or using some algorithm for determining the number of clusters, the cluster centers of each cluster may be randomly selected from historical operating data or initialized using other methods, the distance between each data point and each cluster center may be calculated, the data point may be assigned to the cluster closest to each cluster, the average of all the data points in each cluster may be calculated, the average may be used as a new cluster center, if the cluster center does not move or the distance moved is small, the algorithm may be considered to have converged, and the final cluster center may be used to represent a typical operating scenario of the power distribution network.

In an exemplary embodiment, obtaining an evaluation index value of each flexible resource access scenario under each typical operation scenario of the power distribution network, and determining, according to each evaluation index value, a comprehensive evaluation index value of each flexible resource access scenario includes: respectively constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene, and standardizing the evaluation index matrix to obtain a standardized evaluation index matrix; creating a sample set according to the standardized index matrix, creating a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance measurement indexes of all indexes through the random forest model, and determining first characteristic weights of all the indexes according to the model performance measurement indexes; randomly resetting the values of the indexes in the standardized index matrix, calculating the difference value of the model performance measurement indexes before and after the resetting arrangement for each index, and determining a second characteristic weight according to the difference value of the model performance measurement indexes; and calculating the weighted sum of the first characteristic weight and the second characteristic weight to obtain the comprehensive weight of each evaluation index value, and calculating the evaluation index value of each flexible resource access scene under each typical operation scene according to the comprehensive weight.

The evaluation index matrix refers to a matrix taking values of a plurality of evaluation indexes under different conditions or models as elements, and can be used for comparing and evaluating performances of different models, wherein each row can represent one model, and each column represents one evaluation index. Normalization refers to the process of converting data into 0 as the mean and 1 as the standard deviation, eliminating the influence of different scales, so that comparison and analysis between different features or variables can be more easily performed. A sample set refers to a data sample used to train or test a model. The random forest model is a learning model constructed based on a plurality of decision trees, each decision tree is trained on a sample subset and a feature subset selected randomly, and finally, a comprehensive result is obtained through voting or averaging. Decision trees refer to tree structures used to make decisions about a problem, each node of the decision tree representing an attribute or feature, each branch representing a possible decision result, and each leaf node representing a category or result. Model performance metrics refer to the degree of influence of each input feature on the model output in a model, and can be measured by observing the contribution of the feature when the node splits. The weighted sum is a result of adding a set of values according to a certain weight. The integrated weight is an integrated weight calculated based on a plurality of weights or evaluation indexes.

For example, an evaluation index matrix may be respectively constructed for each flexible resource access scenario under each typical operation scenario, each evaluation index matrix is standardized by using a minimum-maximum standardization or other suitable standardization method to obtain a standardized evaluation index matrix, each row in the standardized evaluation index matrix is used as a sample to create a sample set, a random forest model is trained by using the sample set, a model performance measurement index of each evaluation index may be obtained in the trained random forest model by using a machine learning library in Python, this may be achieved by checking an attribute of the model, determining a first feature weight of each index according to the model performance measurement index, randomly resetting each index in the standardized evaluation index matrix, calculating a difference value of the model performance measurement index before and after resetting the arrangement for each index, determining a second feature weight of each index according to the difference value of the model performance measurement index, calculating a weighted sum of the first feature weight and the second feature weight, obtaining a comprehensive weight of each evaluation, and calculating a typical evaluation index under each flexible resource access scenario according to the comprehensive weight.

In an exemplary embodiment, an evaluation index matrix is respectively constructed according to evaluation index values of different flexible resource access scenarios in each typical operation scenario, the evaluation index matrix is standardized, and a standardized evaluation index matrix is obtained, including: determining the mapping relation between the magnitude of each index value and the flexibility of the power distribution network, and constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes in each typical operation scene; under the condition that the larger the index value is, the better the corresponding flexibility is, the index value is standardized according to the preset minimum index value; under the condition that the smaller index value represents the better corresponding flexibility, the index value is standardized according to the preset maximum index value, and a standardized index matrix is obtained.

The mapping relation refers to the association between the sizes of different index values and the flexibility of the power distribution network.

The mapping relation between each evaluation index and the flexibility of the power distribution network can be determined based on experimental data or experience, an evaluation index matrix is constructed, rows of the matrix represent each flexible resource access scene, columns represent each evaluation index, and according to the mapping relation, the larger each index value represents the better flexibility or the smaller each index value represents the better flexibility, the maximum-minimum standardization can be used, and a standardized formula is applied to each evaluation index value, so that a standardized evaluation index value matrix is obtained.

In an exemplary embodiment, creating a sample set according to a standardized index matrix, creating a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance metrics of each index through the random forest model, and determining a first feature weight of each index according to the model performance metrics, including: constructing a plurality of decision trees, calculating the coefficient of the foundation of each evaluation factor, sequencing the classified nodes and distributing node weights; calculating the reduction amount of the mean square error of each index in the random forest before and after splitting of each node, wherein the mean square error is used for representing the node contribution rate of each index; and calculating the cumulative sum of the mean square error reduction amounts of all the indexes on all the trees, and carrying out weight normalization on the cumulative sum to obtain the first characteristic weight.

Wherein, the evaluation factor refers to various factors for weighing, comparing or judging when evaluating or deciding, and can comprise various indexes or standards affecting the performance of the model or the decision quality. The coefficient of kunity refers to an index for measuring unequal distribution or uncertainty, and is used in decision trees for measuring the degree of confounding of different categories in a node. Classification nodes refer to a node in a tree structure that is responsible for partitioning a data set into two or more subsets. Node weights refer to the weights assigned to each node in a decision tree. The mean square error is an indicator for measuring the difference between the predicted value and the actual value, and is typically the average value of the square error. The node contribution rate refers to the contribution degree of each node to final prediction or classification, and is related to factors such as purity of the node, sample number and the like.

For each decision tree, the base coefficient of each evaluation factor is calculated on each classification node, the base coefficients of each evaluation factor in all decision trees are ordered, corresponding node weights are allocated, after the ordering, the first several evaluation factors can be selected as characteristics, for each node, the mean square error reduction amount of each evaluation factor is calculated on each tree in the random forest to measure the contribution of each index to the model before and after node splitting, for each evaluation factor, the cumulative sum of the mean square error reduction amounts of each index on all nodes is calculated to be used as the node contribution rate, so that the contribution of each index to the whole random forest can be measured, the node contribution rate is normalized in weight, and the first characteristic weight is obtained to ensure that the weights among different indexes have comparability.

In an exemplary embodiment, determining a comprehensive evaluation index value of each flexible resource access scenario according to each evaluation index value includes: taking evaluation index values of each flexible resource access scene under each typical operation scene as evidence, calculating cosine similarity of each two evidences, calculating mutual support degree between the evidences, and carrying out weighted average treatment on the evidences according to the mutual support degree to obtain a weighted average evidence body; and calculating conflict factors of every two weighted average evidence bodies, fusing the evidence bodies under each typical operation scene according to an evidence synthesis formula for each flexible resource access scene, and taking the logarithmic amplification value of the fused evidence bodies as a comprehensive evaluation index value.

Where evidence refers to data that supports or opposes a hypothesis, proposition, or decision in evidence theory. Cosine similarity refers to the cosine value of the included angle between two vectors, and ranges from-1 to 1, and the closer the value is to 1, the higher the similarity is. The degree of mutual support refers to the degree of mutual support between two or more pieces of evidence in evidence theory. The weighted average evidence body refers to a more representative evidence body generated by taking the weight of each evidence into consideration and performing weighted average on the evidence according to the credibility of the evidence. The conflict factor refers to a value used to measure the degree of inconsistency or conflict between evidence. The logarithmic amplification value is a value obtained by taking the logarithm of the evidence credibility and amplifying the high-credibility evidence, and can enhance the contribution of the high-credibility evidence to the whole evidence body.

For example, since the D-S evidence theory may cause a problem of an anti-intuitive fusion result, on the basis of the D-S evidence theory, an index value under each typical scene output by a random forest may be used as an evidence, the cosine similarity of each evidence is calculated first, the cosine similarity is converted into the mutual support degree between the evidences, the cosine similarity is accumulated and normalized to obtain the credibility of the evidence, the evidence is weighted and averaged by using the credibility as a weight coefficient, and the weighted average evidence replaces the original evidence, so as to improve the D-S evidence theory. The steps can be realized by using a preset correlation or similarity measurement method, for each two weighted average evidence bodies, conflict factors such as a dissimilarity measurement or a difference measurement can be calculated, for example, euclidean distance or cosine similarity can be calculated, for each flexible resource access scene, evidence bodies under each typical operation scene can be fused by using an evidence synthesis formula, and finally, a logarithmic amplification value of the fused evidence bodies is calculated by using a preset logarithmic function and is used as a comprehensive evaluation index value, and the fused evidence bodies can infer higher-level conclusions to obtain flexible resource access scheme flexibility index values under all typical scenes.

In an exemplary embodiment, a method for evaluating flexibility of a power distribution network by considering flexible resource adjustability is provided, wherein the method considers related flexibility indexes of the flexible resource after the flexible resource is accessed into the power distribution network, and after different indexes are weighted by a weighting method, a power distribution network flexibility evaluation model is built, and influences of different flexible resource accesses on the flexibility of the power distribution network are quantitatively analyzed. As shown in fig. 3, the method comprises the steps of:

step 302: and determining flexibility evaluation parameters, and respectively calculating the flexibility evaluation parameters. In order to comprehensively evaluate the influence of the flexible resources on the flexibility of the power system in the aspects of 3 aspects of coordination capacity, coordination benefit quality and power grid reliability through a demand response model, the flexibility evaluation is divided into 2 layers to establish a flexible resource flexibility evaluation index system, and the coordination regulation capacity and interaction capacity of the power grid are respectively evaluated, wherein the regulation characteristic refers to indexes capable of reflecting the power grid, and the interaction effect refers to indexes capable of reflecting the coordination benefit quality and the power grid reliability of the power grid.

The index for measuring the coordination capacity comprises the following components:

(1) Index of renewable energy consumption rate

The renewable energy consumption index reflects the ability of the distribution network to absorb the output power of the distributed new energy, and can be expressed as follows:

Wherein: p (P) _RENi Representing the theoretical active output of renewable energy accessed by the ith node; p (P) _RECi The active power consumption of renewable energy sources accessed by the ith node in the power distribution network is shown; n is n _DG Representing the total number of nodes accessing the renewable energy source.

(2) New energy fluctuation support level index

In order to promote the large-scale grid connection of renewable energy sources, the utilization of flexible resources to stabilize the output fluctuation of the renewable energy sources is one of effective solutions. The new energy fluctuation support level is an index for measuring the interaction effect of the flexible resource and the new energy, and is defined as the ratio of the interaction amount of the flexible resource for eliminating the fluctuation of the new energy to the fluctuation amount of the new energy, as shown in the following:

wherein:

wherein:the actual interaction amount of the flexible resource at the moment t;And->The actual power and the initial natural power of the flexible resource at the moment t are respectively;The power fluctuation amount of the new energy at the moment t; n is n _FL Representing the total number of nodes accessing the flexible resource.

(3) Peak-to-valley ratio change rate

The peak-to-valley ratio fluctuation rate refers to the ratio of the variation of the peak-to-valley difference of the system to the peak-to-valley difference of the system without interaction under the two conditions of interaction and no interaction, and reflects the effect of flexible resource interaction on the peak-to-valley adjustment of the system:

wherein: p (P) _p-v Andand respectively representing load peak-valley differences of the non-flexible resource cooperative access and the flexible resource cooperative access.

Quality index of cooperative benefit:

(1) Network loss

In order to improve the use efficiency of energy sources, it is very important to reduce the line loss of the power distribution network. Thus, the following network loss indicators of the distribution network are defined:

wherein: lambda (lambda) _loss Representing network loss, lambda _loss The smaller the value of the flexible resource interaction is, the better the energy utilization rate improving effect of the flexible resource interaction on the power distribution network is. P (P) _loss Active loss of the power distribution network; p (P) _loss,k Active loss of the kth branch; n is n _l Representing the total number of branches.

(2) Voltage deviation index

The access of the flexible resource can effectively relieve voltage deviation caused by new energy output, meets the power quality requirement of users, and is an index for reflecting the running condition of the power distribution network:

(3) Maximum allowable fluctuation rate of payload

The maximum allowable fluctuation rate of the net load reflects the adjustment capacity, namely the climbing capacity, of the flexible resources after being cooperatively connected into the power distribution network:

wherein: lambda (lambda) _FR,max Representing a maximum allowable volatility of the payload t period;the allowable climbing rate of the distributed power supply in the t period is represented;The climbing rate allowed by the distributed energy storage in the t period is represented;The allowable climbing rate of the power distribution network in the t period is represented; / >A payload representing a period t; n is n _DG And n _ESS The distributed power supply quantity and the distributed energy storage quantity are respectively represented.

Grid reliability index:

(1) Line overload rate index

The line overload rate index shows the residual degree of the transmission capacity of the line, the higher the line overload rate is, the lower the residual degree is, the new energy consumption is limited, and the regulation function after the flexible resource is accessed is beneficial to reducing the line overload rate, so that the new energy consumption capacity of the power distribution network is enhanced:

wherein: lambda (lambda) _LR,i Representing the load factor of line i; f (F) _i,max Representing the maximum active transmission capacity of line i; f (F) _i,0 Representing the power flow of line i when the system is running; n is n _l Indicating the number of lines.

λ _LR >Line out-of-limit is indicated at 1, assuming that lambda is _LR When a certain threshold value is exceeded, the line is considered to be overloaded, the overload rate level of the power grid refers to the proportion of the number of lines exceeding rated capacity in the power grid to the number of bus lines, and the overload rate level is used for measuring the running safety of the power grid and can be expressed by the following formula:

(2) Power grid tide entropy index

Entropy is a function describing the state of a system and can be used to measure the uncertainty, degree of confusion, disorder or information content of the system, the more chaotic the system, the greater the entropy; the more ordered the system, the less entropy. Defining a constant sequence u= { U ₁ ,U ₂ ,…,U _k ,…,U _n }. By l _k Representing liability rate lambda _LF ∈(U _k ,U _k+1 ]The number of the branches of the circuit in different load intervals is subjected to probability processing, and the obtained number can be obtained:

the power grid tide entropy is as follows:

if the load rates of all the lines are in the same interval, the uncertainty of the line power flow distribution is minimum, and the ordering degree of the power flow distribution is highest. If no liability rates of any two lines are in the same interval, the power flow entropy of the power grid reaches the maximum value.

Step 304:

the typical operation scene of the power distribution network is constructed by using a k-means clustering algorithm, and the clustering steps are as follows:

(1) The initial center of the classes is randomly selected.

(2) In the ith iteration, the distances from the respective cluster centers are calculated for all samples, and are divided into categories in which the centers with the shortest distances are located.

(3) And updating the central value of each class by using the average value of each class.

(4) Judging whether the convergence condition is satisfied, and if not, returning to the step 2 iteration.

Step 306:

the index weighting is carried out by using a random forest:

(1) Constructing an index matrix according to a flexible resource flexibility evaluation index system, and calculating a matrix R 'formed by m typical scenes and n indexes for a kth flexible resource access scene' _k 。

(2) Normalization processing is carried out on each index element to obtain a standardized index matrix R _k 。

1) For the index that the larger the attribute value is, the better:

2) For the index that the smaller the attribute value is, the better:

wherein: mak (g) _(:,j) ) And min (g) _(:,j) ) The maximum value and the minimum value of the j-th index in the physical sense and within the national standard specified range are respectively represented.

3) Obtaining an evaluation matrix R after the k flexible resource access scene is standardized _k 。

(3) In a standardized evaluation matrix R _k Constructing a data set M, randomly and uniformly extracting N samples from the data set M by using a Bagging sampling method, and forming a training sample set T by using the extracted sample data set.

(4) A plurality of decision trees are constructed from the training sample set T. And calculating the coefficient of the kunity of each evaluation factor, sequencing the classified nodes and distributing node weights. Randomly selecting S characteristic variables on each node of the decision tree, and performing node splitting

And selecting the characteristic variable with the minimum base index to split, and constructing a plurality of classification decision trees.

The base index is an index for measuring the performance of a classifier in machine learning, and the smaller the value is, the better the performance is, and the importance of a certain index is measured here:

wherein: k represents the number of categories in the dataset; p (P) _i The proportion of the class to which the i-th class evaluation factor belongs in the data set T.

If the sample set T is divided into subsets T by a feature ₁ And T is ₂ The coefficient of the kunity is:

(5) Calculating a mean square error (Mean Squared Error, MSE) for each node on each tree in the random forest, wherein the mean square error for the node s is:

wherein: n (N) _s The number of the node samples; y is _i Is the true value of the ith sample;is the average of the true values of all samples on that node.

(6) The node contribution rate of each index is calculated, the contribution rate of the index j at the node s is represented by the reduction amount of the mean square error caused by feature splitting, and the larger the reduction amount of the mean square error is before and after feature splitting, the better the effect of feature splitting is represented, and the better the effect is represented as the node of the decision tree. The calculation formula of the reduction amount of the mean square error is:

ΔMSE _s ＝MSE _s -p _l MSE _l -ρ _r MSE _r

wherein: MSE (mean square error) _s Mean square error for the node before splitting; p is p _l And p _r Sample ratios respectively representing the left child node and the post-splitting node; MSE (mean square error) _l And MSE _r Respectively mean square error of both.

(7) The overall importance of each index is calculated and the importance of index j is expressed as the cumulative sum of the mean square error reduction of the feature over all trees.

Wherein: s is a node set of the index j in the t th regression tree.

(8) The importance value of the index j is converted into weight, and the weight is normalized:

in order to observe the change of the model performance, if the model effect is reduced greatly, the effect of the index on the target value is obvious, and a larger weight is given, wherein the weight is beta. The calculation method comprises the following steps:

(1) Selecting R for trained random forest model ² As a measure of model performance, the calculation formula is:

wherein: f (f) _i Is a model regression value.

(2) Randomly resetting and arranging the data values of the index j, and calculating R after resetting ² The importance score for j is obtained by taking the difference:

wherein:performance scores before reset;Is the performance score after reset.

(3) Traversing all indexes to obtain the importance score of each index. The weight of index j is:

finally, the alpha and the beta are weighted and synthesized to obtain the index weight as follows:

ε _j ＝ωα _j +(1-ω)β _j

wherein: omega is an adjustment coefficient and takes a value of 0 to 1.

(4) Comprehensive evaluation index value Z obtained in ith typical scene by kth flexible resource access scene _ki The method comprises the following steps:

step 308:

cosine similarity is a measure of similarity between two non-zero vectors, and is used herein to calculate the degree of mutual support between evidence bodies. Accumulating and normalizing the similarity of the cosines to obtain the credibility of the evidence; taking the credibility as a weight coefficient to carry out weighted average on the evidence; replacing the original evidence body with the weighted average evidence body; finally, the evidence synthesis formula is adopted to fuse the evidence bodies, so that a flexible resource access scenario flexibility index value in all typical scenes is obtained. The flexible resource access scenario flexibility index value flow under all typical scenes is obtained by using the improved D-S evidence theory as follows:

(1) Obtaining index value Z of kth flexible resource access scene of power distribution network in ith typical scene by adopting random forest algorithm _ki The basic probability distribution conditions (basic probability assignment, BPA) of the evidence synthesis formula are considered, and the similarity of Cosine between the evidences is calculated.

(2) The mutual support degree between the evidences is obtained, and the weight of the evidence is normalized.

(3) And calculating a conflict factor K of the evidence synthesis formula between the two evidence bodies.

(4) According to the evidence synthesis formula, merging evidence bodies in the two different typical scenes to infer indexes of the power distribution network in the k flexible resource access scene.

And (3) repeating the steps (3) and (4) for each flexible resource access scene until obtaining a corresponding unique index value.

The obtained P' _k Usually less than 1, P 'is used for the convenience of artificial habit' _k And properly amplified.

In the above, P ₀ The method is used for calculating the flexibility index value of the power distribution network according to the previous steps without considering any flexible resource access, and the flexibility index value is used as a comparison.

The flexibility evaluation method for accessing the flexible resources into the power distribution network based on the random forest and the improved D-S evidence theory. Firstly, a flexibility evaluation index system is constructed from three aspects of coordination capacity, coordination benefit quality and power grid reliability. Then, based on a large number of historical operating scenes, a typical operating scene set of renewable energy sources and loads is extracted through a k-means clustering algorithm. And calculating flexibility evaluation index values of the flexible resource access scene under each typical scene, and performing dimension reduction processing on the index matrix of each scene by using a random forest. And finally, utilizing an improved D-S evidence theory, regarding the comprehensive flexibility index value under each typical scene as an evidence body, and obtaining a unified flexibility index by fusion reasoning, thereby quantitatively analyzing the influence of different flexible resource accesses on the flexibility of the power distribution network.

It should be understood that, although the steps in the flowcharts related to the embodiments described above are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a power distribution network flexibility evaluation device considering the flexible resource adjustability, which is used for realizing the power distribution network flexibility evaluation method considering the flexible resource adjustability. The implementation scheme of the solution provided by the device is similar to the implementation scheme described in the above method, so the specific limitation in the embodiment of the power distribution network flexibility assessment device considering the flexible resource adjustability provided below can be referred to the limitation of the power distribution network flexibility assessment method considering the flexible resource adjustability, which is not described herein.

In an exemplary embodiment, as shown in fig. 5, there is provided a power distribution network flexibility assessment apparatus 500 considering flexible resource adjustability, including: a data acquisition module 502, an index acquisition module 504, and an evaluation module 506, wherein:

the data acquisition module 502 is configured to acquire historical operation data of the power distribution network, and determine each typical operation scenario of the power distribution network according to the historical operation data through a clustering algorithm;

the index obtaining module 504 is configured to obtain an evaluation index value of each flexible resource access scenario in each typical operation scenario of the power distribution network, and determine a comprehensive evaluation index value of each flexible resource access scenario according to each evaluation index value; wherein, the evaluation index comprises a coordination ability index, a coordination ability index and a reliability index;

and the evaluation module 506 is configured to quantitatively analyze the influence of different flexible resource accesses on the flexibility of the power distribution network according to the magnitude of each comprehensive evaluation index value.

In one embodiment, the data acquisition module 502 is configured to: determining an initial cluster center of a cluster corresponding to each operation scene in the historical operation data; respectively calculating the distance between each data point in the historical operation data and each initial clustering center, and distributing each data point to the cluster with the shortest distance between each initial clustering center and each data point; and updating the clustering center of each cluster according to the data mean value of each cluster so as to form a typical operation scene of the power distribution network by each data point in each cluster corresponding to the clustering center meeting the preset convergence condition.

In one embodiment, the index acquisition module 504 is configured to: respectively constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene, and standardizing the evaluation index matrix to obtain a standardized evaluation index matrix; creating a sample set according to the standardized index matrix, creating a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance measurement indexes of all indexes through the random forest model, and determining first characteristic weights of all the indexes according to the model performance measurement indexes; randomly resetting the values of the indexes in the standardized index matrix, calculating the difference value of the model performance measurement indexes before and after the resetting arrangement for each index, and determining a second characteristic weight according to the difference value of the model performance measurement indexes; and calculating the weighted sum of the first characteristic weight and the second characteristic weight to obtain the comprehensive weight of each evaluation index value, and calculating the evaluation index value of each flexible resource access scene under each typical operation scene according to the comprehensive weight.

In one embodiment, the index acquisition module 504 is configured to: determining the mapping relation between the magnitude of each index value and the flexibility of the power distribution network, and constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene; under the condition that the larger the index value is, the better the corresponding flexibility is, the index value is standardized according to the preset minimum index value; under the condition that the smaller index value represents the better corresponding flexibility, the index value is standardized according to the preset maximum index value, and a standardized index matrix is obtained.

In one embodiment, the index acquisition module 504 is configured to: constructing a plurality of decision trees, calculating the coefficient of the foundation of each evaluation factor, sequencing the classified nodes and distributing node weights; calculating the reduction amount of the mean square error of each index in the random forest before and after splitting of each node, wherein the mean square error is used for representing the node contribution rate of each index; and calculating the cumulative sum of the mean square error reduction amounts of all the indexes on all the trees, and carrying out weight normalization on the cumulative sum to obtain the first characteristic weight.

In one embodiment, the index acquisition module 504 is configured to: taking evaluation index values of each flexible resource access scene under each typical operation scene as evidence, calculating cosine similarity of each two evidences, calculating mutual support degree between the evidences, and carrying out weighted average treatment on the evidences according to the mutual support degree to obtain a weighted average evidence body; and calculating conflict factors of every two weighted average evidence bodies, fusing the evidence bodies under each typical operation scene according to an evidence synthesis formula for each flexible resource access scene, and taking the logarithmic amplification value of the fused evidence bodies as a comprehensive evaluation index value.

The modules in the power distribution network flexibility evaluation device considering the flexible resource adjustability can be all or partially implemented by software, hardware and combinations thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In one exemplary embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 6. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing historical operation data of the power distribution network, typical operation scene data of the power distribution network, flexible resource access scene data, calculation model or function data, calculated index values and other data possibly related to the index values. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by a processor, implements a method for evaluating flexibility of a power distribution network taking flexible resource adjustability into account.

It will be appreciated by those skilled in the art that the structure shown in fig. 6 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements the steps of the method embodiments described above.

In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method embodiments described above.

It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, etc.) and the data (including, but not limited to, data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use, and processing of the related data are required to meet the related regulations.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the various embodiments provided herein may include at least one of relational databases and non-relational databases. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic units, quantum computing-based data processing logic units, etc., without being limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples only represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the present application. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application shall be subject to the appended claims.

Claims

1. A power distribution network flexibility assessment method considering flexible resource adjustability, the method comprising:

acquiring historical operation data of a power distribution network, and determining each typical operation scene of the power distribution network through a clustering algorithm according to the historical operation data;

acquiring evaluation index values of each flexible resource access scene in each typical operation scene of the power distribution network, and determining comprehensive evaluation index values of each flexible resource access scene according to each evaluation index value; wherein the evaluation index comprises a coordination capacity index, a coordination capacity index and a reliability index;

2. The method according to claim 1, wherein the obtaining historical operation data of the power distribution network, and determining each typical operation scenario of the power distribution network according to the historical operation data through a clustering algorithm, comprises:

updating the clustering centers of the clusters according to the data average value of the clusters so as to form a typical operation scene of the power distribution network by each data point in each cluster corresponding to the clustering center meeting the preset convergence condition.

3. The method according to claim 2, wherein the obtaining the evaluation index value of each flexible resource access scenario in each typical operation scenario of the power distribution network, and determining the comprehensive evaluation index value of each flexible resource access scenario according to each evaluation index value, includes:

Respectively constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene, and normalizing the evaluation index matrix to obtain a normalized evaluation index matrix;

creating a sample set according to the standardized index matrix, building a random forest model according to the sample set, training a decision tree in the random forest model, obtaining model performance measurement indexes of all indexes through the random forest model, and determining first characteristic weights of all the indexes according to the model performance measurement indexes; randomly resetting the values of the indexes in the standardized index matrix, calculating the difference value of the model performance measurement indexes before and after resetting the arrangement for each index, and determining a second characteristic weight according to the difference value of the model performance measurement indexes;

4. The method of claim 3, wherein the constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenarios in each typical operation scenario, normalizing the evaluation index matrix, and obtaining a normalized evaluation index matrix includes:

Determining the mapping relation between the magnitude of each index value and the flexibility of the power distribution network, and constructing an evaluation index matrix according to the evaluation index values of different flexible resource access scenes under each typical operation scene;

under the condition that the larger the index value is, the better the corresponding flexibility is, the index value is standardized according to a preset minimum index value;

and under the condition that the smaller index value represents the better corresponding flexibility, normalizing the index value according to the preset maximum index value to obtain a normalized index matrix.

5. A method according to claim 3, wherein creating a sample set from the standardized index matrix, creating a random forest model from the sample set, training a decision tree in the random forest model, obtaining a model performance metric for each index through the random forest model, and determining a first feature weight for each index according to the model performance metrics comprises:

And calculating the cumulative sum of the mean square error reduction amounts of all the indexes on all the trees, and carrying out weight normalization on the cumulative sum to obtain a first characteristic weight.

6. The method according to claim 1, wherein the determining the integrated evaluation index value for each of the flexible resource access scenarios from each of the evaluation index values comprises:

taking evaluation index values of the flexible resource access scenes under the typical operation scenes as evidence, calculating cosine similarity of every two evidences, calculating mutual support degree between the evidences, and carrying out weighted average processing on the evidences according to the mutual support degree to obtain a weighted average evidence body;

7. A power distribution network flexibility assessment device considering flexible resource adjustability, the device comprising:

The index acquisition module is used for acquiring the evaluation index values of the flexible resource access scenes in the typical operation scenes of the power distribution network, and determining the comprehensive evaluation index value of each flexible resource access scene according to each evaluation index value; wherein the evaluation index comprises a coordination capacity index, a coordination capacity index and a reliability index;

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.

9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.