Power distribution network carrying capacity evaluation method and system considering multi-scenario charging load

Abstract

The application relates to the technical field of power distribution network carrying capacity evaluation, in particular to a power distribution network carrying capacity evaluation method and system considering multi-scene charging load, which comprises the following steps: collecting power distribution network basic data and charging load characteristic data, generating multiple initial scenes based on a Monte Carlo method, clustering and reducing charging load curves in each scene to obtain a typical charging scene set, performing deterministic power flow calculation, extracting a safety margin index of each scene, determining the comprehensive weight of each index by using an improved sequential relationship weighting method based on the safety margin index of each scene, calculating the carrying capacity comprehensive score of each scene, and taking the minimum value of the carrying capacity comprehensive score as the final carrying capacity evaluation result of the power distribution network. The application realizes accurate quantitative evaluation of the carrying capacity of the power distribution network by combining multi-scene probability modeling and an improved weighting method.

Description

Technical Field

[0001] This application relates to the field of distribution network carrying capacity assessment technology, specifically to a distribution network carrying capacity assessment method and system that takes into account charging loads in multiple scenarios. Background Technology

[0002] With the rapid growth of electric vehicle ownership, the large-scale integration of charging loads poses a severe challenge to the safe operation of the power distribution network. Charging loads exhibit significant spatiotemporal randomness—charging start time, charging duration, and charging power are uncertain due to the influence of user behavior patterns, and load characteristics differ significantly under different scenarios (nighttime charging in residential areas, daytime charging in office areas, and random charging at public fast-charging stations). When charging loads are concentrated, it may lead to problems such as voltage exceeding limits at distribution network nodes, branch power flow overload, and transformer overload. Therefore, accurately assessing the carrying capacity of the power distribution network under the influence of charging loads in multiple scenarios is of great guiding significance for the planning and layout of charging facilities and the upgrading and transformation of the power distribution network.

[0003] Existing methods for assessing the carrying capacity of power distribution networks have the following main shortcomings: First, the charging load modeling is overly simplistic and fails to adequately consider the characteristics of multiple scenarios. Existing methods often use a single probability distribution to describe the spatiotemporal distribution of charging load, mixing the load characteristics of different types of charging scenarios in the model. However, the charging behavior in residential areas, office areas, and public fast charging stations has distinctly different temporal characteristics. Mixed modeling can mask the extreme conditions of each scenario, leading to overly optimistic evaluation results.

[0004] Secondly, the methods for determining the weights of carrying capacity assessment indicators have flaws. Traditional weighting methods often use the analytic hierarchy process (AHP) or entropy weighting to determine weights individually. The AHP is highly subjective, while the entropy weighting method may result in unreasonable weight allocation due to data fluctuations. Existing methods fail to effectively combine the advantages of subjective experience with objective data.

[0005] Third, the comprehensive carrying capacity score does not consider the extreme nature of the scenarios. Existing assessment methods mostly use probability-weighted averaging of the assessment results for each scenario, using expected value to represent carrying capacity. However, the safe operation of the distribution network requires that safety constraints be met even under the most unfavorable scenarios, and expected value assessment may mask safety hazards in extreme scenarios.

[0006] To address the aforementioned issues, this application proposes a distribution network carrying capacity assessment method that considers charging loads in multiple scenarios. Through multi-scenario probabilistic modeling, improved sequential relationship weighting, and comprehensive assessment of extreme scenario values, the method achieves an accurate quantitative assessment of distribution network carrying capacity. Summary of the Invention

[0007] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for assessing the carrying capacity of distribution networks that considers charging loads in multiple scenarios. The specific technical solution adopted is as follows: This application provides a method for assessing the carrying capacity of a distribution network that takes into account charging loads in multiple scenarios, including the following steps: Step 1: Collect basic data of the power distribution network and characteristic data of the charging load. The basic data of the power distribution network includes topology parameters, line impedance parameters, transformer rated capacity and load data. The characteristic data of the charging load includes charging start time distribution parameters, charging duration distribution parameters and charging power distribution parameters under various scenarios. Step 2: Based on the charging load characteristic data, generate multiple initial scenarios using the Monte Carlo method, cluster the generated initial scenarios, and obtain each typical charging scenario and its corresponding weight coefficient. Step 3: For each typical charging scenario, the charging load of each scenario is superimposed on the basic load of the distribution network, and deterministic power flow calculation is performed to extract the safety margin index under each scenario. The safety margin index includes node voltage margin, branch power flow margin and transformer load margin. Step 4: Based on the safety margin indicators in each scenario, the comprehensive weight of each indicator is determined by an improved sequential relationship weighting method. In the improved sequential relationship weighting method, the subjective weights are corrected by the dispersion of each indicator, and then the comprehensive weight of each indicator is obtained. Step 5: Calculate the comprehensive carrying capacity score for each scenario based on the safety margin index and the comprehensive weight, and take the minimum value of the comprehensive carrying capacity score as the final carrying capacity assessment result of the distribution network.

[0008] Preferably, the charging load characteristic data in step 1 is collected by obtaining the charging start time distribution, charging duration distribution and charging power distribution under each charging scenario through historical operation data of the charging operation platform; wherein, the charging start time of each scenario is fitted with a normal distribution or a uniform distribution, and the charging duration is fitted with a log-normal distribution.

[0009] Preferably, the specific method for generating multiple initial scenarios using the Monte Carlo method in step 2 is as follows: based on the charging start time distribution, charging duration distribution, and charging power distribution of each scenario, each scenario is independently sampled to generate the start charging time, charging duration, and charging power value of each charging device in each scenario; the sampling results are superimposed in time sequence to obtain the charging load time sequence curve of each scenario over a preset duration; and the sampling is repeated to generate a preset number of initial scenarios.

[0010] Preferably, the specific method for clustering and reducing the generated initial scenes in step 2 is as follows: a clustering algorithm is used to cluster a preset number of initial scenes into multiple clusters, wherein each cluster corresponds to a typical charging scene, and the proportion of the number of scenes in each cluster to the total number of initial scenes is used as the weight coefficient of each typical charging scene.

[0011] Preferably, the node voltage margin in step 3 is calculated as follows: the absolute value of the deviation between the voltage amplitude of each node and the rated voltage under a typical charging scenario is calculated, and the difference between 1 and the maximum absolute value of the voltage deviation among all nodes in the entire power grid is used as the node voltage margin index.

[0012] Preferably, the branch power flow margin in step 3 is calculated as follows: the ratio of the power flow value of each branch under a typical charging scenario to the rated capacity of the branch is calculated, and the difference between 1 and the maximum value of the ratio among all branches in the entire power grid is used as the branch power flow margin index.

[0013] Preferably, the transformer load margin in step 3 is calculated as follows: the ratio of the transformer's load power to its rated capacity under a typical charging scenario is calculated, and 1 minus this ratio is used as the transformer load margin index.

[0014] Preferably, the improved sequential relation weighting method in step 4 is implemented as follows: First, the subjective weights of each safety margin index are determined using sequential relation analysis. The indexes are ranked and assigned values ​​based on their relative importance to the safe operation of the distribution network, and the subjective weight vectors of each index are calculated. Second, the dispersion of each safety margin index under each typical charging scenario is calculated to determine the objective correction coefficients of each index. The objective correction coefficients are negatively correlated with the dispersion. Third, the subjective weight vectors are multiplied by the objective correction coefficients and then normalized to obtain the comprehensive weight vector of each index.

[0015] Preferably, the calculation method for the comprehensive carrying capacity score in step 5 is as follows: for each typical charging scenario, the safety margin index values ​​under that scenario are multiplied by their corresponding comprehensive weights and then summed to obtain the comprehensive carrying capacity score for that scenario; the minimum value of the comprehensive carrying capacity score among all typical charging scenarios is taken as the final carrying capacity assessment result of the distribution network; when the final carrying capacity assessment result is lower than a preset threshold, it is determined that the distribution network faces the risk of insufficient carrying capacity under the current scenario, and an alarm signal is output.

[0016] This application also provides a distribution network carrying capacity assessment system that takes into account charging loads in multiple scenarios, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described distribution network carrying capacity assessment methods that take into account charging loads in multiple scenarios.

[0017] As can be seen from the above, the distribution network carrying capacity assessment method and system considering charging loads in multiple scenarios provided in this application have at least the following beneficial effects: (i) A scenario generation method for probabilistic modeling and cluster reduction of charging load in multiple scenarios was established. This application addresses three typical charging scenarios: residential nighttime charging, office daytime charging, and public fast-charging stations. Based on charging load characteristic data, a large number of initial scenarios are generated using the Monte Carlo method, followed by cluster reduction. This method preserves the randomness of each scenario while controlling the computational scale. Unlike existing methods that use a simplified single probability distribution, this scenario generation method more accurately reflects the multimodal distribution characteristics of actual charging load.

[0018] (II) An improved sequential relation weighting method is proposed, achieving an effective combination of subjective experience and objective data. Based on the traditional sequential relation analysis method for determining subjective weights, this application introduces information entropy to quantify the objective dispersion of each indicator. The entropy value reflects the distinguishability of the indicator across different scenarios, and the subjective weights are adjusted accordingly. Compared to the purely subjective analytic hierarchy process or the purely objective entropy weighting method, this method takes into account both expert experience and data characteristics, resulting in a more reasonable weight allocation.

[0019] (iii) Using the minimum value of extreme scenarios as the final evaluation result ensures the safety margin of the distribution network under the most unfavorable operating conditions. This application uses the minimum value of the comprehensive carrying capacity score in each typical charging scenario as the final evaluation result, which effectively identifies the operational risks under extreme charging scenarios and avoids the defect of the probability weighted average method in traditional methods, which causes extreme operating conditions to be smoothed out.

[0020] (iv) A complete evaluation chain has been formed, from multi-scenario probabilistic modeling to improved weighted evaluation and extreme scenario value selection. This application organically links multi-scenario charging load generation, safety margin index extraction, improved sequential relationship weighting, and extreme scenario evaluation, forming a complete distribution network carrying capacity evaluation technology system, providing a scientific decision-making basis for the planning and access of charging facilities. Attached Figure Description

[0021] Figure 1 A flowchart illustrating the steps of the distribution network carrying capacity assessment method for charging loads in multiple scenarios provided in this application. Detailed Implementation

[0022] The following, in conjunction with the accompanying drawings, details the specific scheme of the distribution network carrying capacity assessment method and system for charging loads in multiple scenarios provided in this application.

[0023] Please see Figure 1 The document illustrates a flowchart of a distribution network carrying capacity assessment method considering charging loads in multiple scenarios, provided in one embodiment of this application. The method includes the following steps: Specifically, a detailed explanation will be provided using a 10kV distribution network with a large-scale electric vehicle charging load as an example. This distribution network has a typical radial structure, comprising 33 nodes and 32 branches, with a total load of approximately 8MW, powered by a 10MVA transformer. The area where the distribution network is located includes three typical charging scenarios: residential nighttime charging (approximately 300 private charging piles), office daytime charging (approximately 200 charging piles), and public fast charging station scenarios (20 DC fast charging piles). The basic data of the distribution network was obtained from the GIS system and equipment ledger system, while the charging load characteristic data was extracted from the historical operating data of the charging operation platform, spanning the past 12 months.

[0024] Step 1: Collect basic data of the power distribution network and characteristic data of the charging load. The basic data of the power distribution network includes topology parameters, line impedance parameters, transformer rated capacity and load data. The characteristic data of the charging load includes charging start time distribution parameters, charging duration distribution parameters and charging power distribution parameters under various scenarios.

[0025] The purpose of this step is to obtain basic data on the distribution network and characteristic data on charging loads, so as to provide a data foundation for subsequent scenario construction and carrying capacity assessment.

[0026] Distribution network basic data collection: Obtain the topology parameters of the distribution network, the impedance parameters of each line (resistance and reactance per unit value), the rated capacity and short-circuit impedance of transformers, and the basic load curve data of each node (typical daily load curve, sampling interval is 15 minutes) through the distribution GIS platform and equipment ledger system.

[0027] Charging load characteristic data collection: Extracting charging behavior characteristic parameters under various charging scenarios from the historical operation database of the charging operation platform: • Nighttime charging scenario in residential areas: The charging start time follows a normal distribution with a mean of 19:30 and a standard deviation of 60 minutes; the charging duration follows a log-normal distribution (mean of 180 minutes and log-standard deviation of 0.3); the charging power is 7kW (single-phase AC slow charging).

[0028] • Daytime charging scenario in office area: The charging start time follows a normal distribution with a mean of 8:30 and a standard deviation of 45 minutes; the charging duration follows a log-normal distribution (mean of 240 minutes and log-standard deviation of 0.25); the charging power is 7kW.

[0029] • Public fast charging station scenario: The charging start time approximately follows a uniform distribution (covering 7:00 to 22:00); the charging duration follows a log-normal distribution (mean 40 minutes, logarithmic standard deviation 0.35); the charging power is 60kW (DC fast charging).

[0030] Step 2: Based on the charging load characteristic data, generate multiple initial scenarios using the Monte Carlo method, cluster the generated initial scenarios, and obtain each typical charging scenario and its corresponding weight coefficient.

[0031] The purpose of this step is to generate typical charging scenarios based on the collected charging load characteristic data through probabilistic modeling and Monte Carlo methods, providing input for subsequent power flow calculations and carrying capacity assessments.

[0032] The spatiotemporal distribution of charging load exhibits significant randomness. Different users have different charging start times, charging durations, and charging powers. These random factors collectively determine the charging load level experienced by the distribution network at any given time. By generating a large number of random scenarios using the Monte Carlo method, various possible combinations of charging loads can be covered, avoiding the evaluation bias caused by relying solely on a single deterministic scenario.

[0033] The specific generation method is as follows: The first step involves sampling the nighttime charging scenario in residential areas. Based on the distribution of charging start time, charging duration, and charging power parameters of this scenario, charging behavior parameters for 300 charging devices are generated. Specifically, the starting charging time tstart for each device is sampled from a normal distribution N(19:30, 60min²); the charging duration Tcharge for each device is sampled from a log-normal distribution; and the charging power is set to 7kW. Based on the starting time and charging duration, the charging interval for each device on the 24-hour time axis is determined, and the power consumption of 7kW is included within this interval.

[0034] The second step is to perform the same sampling operation on daytime charging scenarios in office areas and public fast charging station scenarios.

[0035] The third step is to overlay the charging loads of the three scenarios on the time axis to obtain the 24-hour charging load time-series curve corresponding to this sampling, with a sampling interval of 15 minutes (96 sampling points).

[0036] The fourth step is to repeat the above sampling process 1000 times to obtain 1000 initial scenes, which constitute the initial scene set.

[0037] Scene Clustering Reduction: A clustering algorithm is used to cluster a preset number of initial scenes into multiple clusters, where each cluster corresponds to a typical charging scene. The proportion of scenes within each cluster to the total number of initial scenes is used as the weight coefficient for each typical charging scene. Specifically, in this embodiment, the initial scene set contains 1000 scenes; too many scenes would significantly increase the burden of subsequent power flow calculations. The K-means clustering algorithm is used to reduce the initial scene set. The 96-dimensional time-series data of the 1000 scenes are used as input, and the number of clusters K=6 is set. After the algorithm iterates and converges, 6 clusters are obtained, each cluster corresponding to a typical charging scene. The proportion of scenes within a cluster to the total number of initial scenes is used as the weight coefficient ωk (k=1,...,6) for that typical charging scene.

[0038] It is understandable that many of the 1000 initial scenes have similar temporal patterns (such as peaks and troughs being close in position). Clustering can merge redundant scenes and significantly reduce the computational scale while preserving the main distribution characteristics.

[0039] Step 3: For each typical charging scenario, the charging load of each scenario is superimposed onto the basic load of the distribution network, and deterministic power flow calculation is performed to extract the safety margin index under each scenario. The safety margin index includes node voltage margin, branch power flow margin and transformer load margin.

[0040] This step performs deterministic power flow calculations for six typical charging scenarios and extracts safety margin indicators for each scenario.

[0041] Power flow calculation: The charging load curves of various typical charging scenarios are superimposed with the basic load curve of the distribution network to obtain the load value of each node at each sampling time. The forward-backward substitution method is used for three-phase unbalanced power flow calculation, with a convergence accuracy set to... pu.

[0042] The theoretical significance of extracting safety margin indicators: In this embodiment, each indicator is constructed in the form of "1 minus the maximum degree of exceeding the limit", so that the indicator values ​​all fall in the range of [0,1]. The larger the value, the larger the safety margin and the healthier the operating status, which is convenient for subsequent weighted comprehensive scoring.

[0043] The node voltage margin is calculated as follows: the absolute value of the deviation between the voltage amplitude of each node and the rated voltage under a typical charging scenario is taken, and the difference between 1 and the maximum absolute value of the voltage deviation among all nodes in the entire power grid is used as the node voltage margin index.

[0044] The branch power flow margin is calculated as follows: the ratio of the power flow value of each branch under a typical charging scenario to the rated capacity of the branch is calculated, and the difference between 1 and the maximum value of this ratio among all branches in the entire power grid is used as the branch power flow margin index.

[0045] The transformer load margin is calculated as follows: the ratio of the transformer's load power to its rated capacity under a typical charging scenario is calculated, and 1 minus this ratio is used as the transformer load margin index.

[0046] In this embodiment, taking typical charging scenario 1 as an example, after the power flow calculation is completed, the specific calculations of each safety margin index are as follows: Node voltage margin: Extract the voltage amplitude of each node, calculate the deviation from the rated voltage of 1.0 pu, and take the maximum absolute value. In this embodiment, the voltage of node 18 is 0.932 pu (low), the absolute value of the deviation is 0.068, and the node voltage margin = 1 - 0.068 = 0.932.

[0047] Branch power flow margin: Extract the power flow value of each branch, calculate the ratio of it to its rated capacity, and take the maximum value. In this embodiment, the power flow value of branch 2-3 is 87% of the rated capacity, and the branch power flow margin = 1 - 0.87 = 0.13.

[0048] Transformer load margin: Extract the ratio of transformer load power to rated capacity. In this example, it is 78%, and the transformer load margin = 1 - 0.78 = 0.22.

[0049] The column vector of the three indicators is [0.932, 0.13, 0.22].

[0050] Step 4: Based on the safety margin indicators in each scenario, an improved sequential relation weighting method is used to determine the comprehensive weight of each indicator. In the improved sequential relation weighting method, the subjective weights are corrected by the dispersion of each indicator, thereby obtaining the comprehensive weight of each indicator.

[0051] This step is one of the core elements demonstrating the inventiveness of this application. The theoretical basis for introducing the improved sequential relationship weighting method lies in the fact that carrying capacity assessment involves multiple safety margin indicators, each with different importance to the safe operation of the distribution network, requiring reasonable weighting. Traditional sequential relationship analysis methods rely solely on expert experience to determine subjective weights, which may ignore the actual dispersion of each indicator across scenarios; while this application introduces information entropy for objective correction based on subjective weights, achieving an effective combination of subjective experience and objective data.

[0052] Step 1: Determine subjective weights (sequential relationship analysis method).

[0053] The importance of each safety margin indicator for the safe operation of the distribution network is ranked. In this embodiment, experts determine that node voltage margin is the most important for capacity assessment (voltage exceeding limits is the most direct hazard), followed by branch power flow margin, and then transformer load margin. A sequential relationship analysis method is used to determine the subjective weight of each safety margin indicator. Based on the relative importance of each indicator to the safe operation of the distribution network, values ​​are assigned, and the subjective weight of each indicator is calculated to form a subjective weight vector.

[0054] Specifically, the order of importance is: node voltage margin > branch power flow margin > transformer load margin.

[0055] Based on the importance ratio of adjacent indicators (1.0 indicates equal importance, 1.2 indicates slightly important, and 1.5 indicates significantly important), the values ​​in this embodiment are as follows: the importance ratio of the first adjacent indicator = importance of node voltage margin / importance of branch power flow margin = 1.3; the importance ratio of the second adjacent indicator = importance of branch power flow margin / importance of transformer load margin = 1.2. Further, the subjective weight is: Wsub = [0.45, 0.35, 0.20].

[0056] Step 2: Determine the objective correction coefficient (information entropy method).

[0057] The dispersion of each safety margin index under six typical charging scenarios is calculated. In this embodiment, information entropy is used as the measure of dispersion. In actual application scenarios, implementers can choose their own method, and this embodiment does not impose any special restrictions. Information entropy reflects the dispersion of index values ​​across different scenarios. If an index has similar values ​​across scenarios (low dispersion), it indicates that the index is not sensitive to changes in scenarios and has low distinguishability in carrying capacity assessment. If the value fluctuates greatly (high dispersion), it indicates that the index is sensitive to changes in scenarios, has high distinguishability, and should be given higher weight. According to the above analysis, the objective correction coefficient is negatively correlated with this dispersion. Preferably, in this embodiment, the objective correction coefficient is calculated as the difference between the numerical value 1 and the information entropy value.

[0058] Calculations show that the entropy value of node voltage margin is 0.85 (low dispersion) across various scenarios, the entropy value of branch power flow margin is 0.65 (medium dispersion), and the entropy value of transformer load margin is 0.50 (highest dispersion). A smaller information entropy indicates a greater degree of dispersion. The correction coefficient vector is: Emod=[0.15,0.35,0.50].

[0059] Step 3: Calculate the overall weight.

[0060] By multiplying the subjective weight vector and the objective correction coefficient vector term by term and then normalizing, the comprehensive weight of each indicator can be obtained, thus forming the comprehensive weight vector.

[0061] In this embodiment, the comprehensive weight vector is: [node voltage margin 0.233, branch power flow margin 0.422, transformer load margin 0.345]. Compared with the purely subjective weight, the weights of branch power flow margin and transformer load margin are increased because they have a higher degree of dispersion in different scenarios and are more valuable for distinguishing the carrying capacity of different scenarios.

[0062] Step 5: Calculate the comprehensive carrying capacity score for each scenario based on the safety margin index and the comprehensive weight, and take the minimum value of the comprehensive carrying capacity score as the final carrying capacity assessment result of the distribution network.

[0063] This step calculates the comprehensive carrying capacity score for each typical charging scenario, and uses the minimum value as the final evaluation result. For each typical charging scenario, the safety margin index values ​​under that scenario are multiplied by their corresponding comprehensive weights and then summed to obtain the comprehensive carrying capacity score for that scenario; the minimum comprehensive carrying capacity score among all typical charging scenarios is used as the final carrying capacity evaluation result of the distribution network.

[0064] Taking typical charging scenario 1 as an example, its safety margin index vector is [0.932, 0.13, 0.22], the comprehensive weight vector is [0.233, 0.422, 0.345], and the comprehensive load-bearing capacity score is 0.932×0.233+0.13×0.422+0.22×0.345=0.348.

[0065] The comprehensive scores were calculated for all six scenarios: Scenario 1 = 0.348, Scenario 2 = 0.412, Scenario 3 = 0.295 (worst-case scenario), Scenario 4 = 0.381, Scenario 5 = 0.436, and Scenario 6 = 0.403. The minimum value of 0.295 was taken as the final carrying capacity assessment result for the distribution network.

[0066] The safe operation of the distribution network requires meeting safety constraints even under the most unfavorable scenario. Therefore, the minimum score (rather than the average) of each scenario is taken as the final evaluation result to reflect the capacity bottleneck under extreme conditions. When the final capacity evaluation result is lower than a preset threshold, it is determined that the distribution network faces a risk of insufficient capacity in the current scenario, and an alarm signal is output. In this embodiment, a score of 0.295 corresponds to a percentage of approximately 29.5 points, which is lower than the preset threshold of 0.4, thus determining that the distribution network faces a risk of insufficient capacity in the current scenario, and an alarm signal is output.

[0067] Based on the same inventive concept as the above method, this application embodiment also provides a distribution network carrying capacity assessment system that takes into account charging loads in multiple scenarios, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described distribution network carrying capacity assessment methods that take into account charging loads in multiple scenarios.

[0068] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.

Claims

1. A method for assessing the carrying capacity of a distribution network considering charging loads in multiple scenarios, characterized in that, Includes the following steps: Step 1: Collect basic data of the power distribution network and characteristic data of the charging load. The basic data of the power distribution network includes topology parameters, line impedance parameters, transformer rated capacity and load data. The characteristic data of the charging load includes charging start time distribution parameters, charging duration distribution parameters and charging power distribution parameters under various scenarios. Step 2: Based on the charging load characteristic data, generate multiple initial scenarios using the Monte Carlo method, cluster the generated initial scenarios, and obtain each typical charging scenario and its corresponding weight coefficient. Step 3: For each typical charging scenario, the charging load of each scenario is superimposed on the basic load of the distribution network, and deterministic power flow calculation is performed to extract the safety margin index under each scenario. The safety margin index includes node voltage margin, branch power flow margin and transformer load margin. Step 4: Based on the safety margin indicators in each scenario, the comprehensive weight of each indicator is determined by an improved sequential relationship weighting method. In the improved sequential relationship weighting method, the subjective weights are corrected by the dispersion of each indicator, and then the comprehensive weight of each indicator is obtained. Step 5: Calculate the comprehensive carrying capacity score for each scenario based on the safety margin index and the comprehensive weight, and take the minimum value of the comprehensive carrying capacity score as the final carrying capacity assessment result of the distribution network.

2. The method according to claim 1, characterized in that, The method for collecting charging load characteristic data in step 1 is as follows: the charging start time distribution, charging duration distribution and charging power distribution under each charging scenario are obtained through the historical operation data of the charging operation platform; wherein, the charging start time of each scenario is fitted with a normal distribution or a uniform distribution, and the charging duration is fitted with a log-normal distribution.

3. The method according to claim 2, characterized in that, The specific method for generating multiple initial scenarios using the Monte Carlo method in step 2 is as follows: based on the charging start time distribution, charging duration distribution, and charging power distribution of each scenario, each scenario is sampled independently to generate the start charging time, charging duration, and charging power value of each charging device in each scenario; the sampling results are superimposed in time sequence to obtain the charging load time sequence curve of each scenario over a preset duration; and the preset number of initial scenarios are generated by repeated sampling.

4. The method according to claim 3, characterized in that, The specific method for clustering and reducing the generated initial scenes in step 2 is as follows: a clustering algorithm is used to cluster a preset number of initial scenes into multiple clusters, where each cluster corresponds to a typical charging scene, and the proportion of the number of scenes in each cluster to the total number of initial scenes is used as the weight coefficient of each typical charging scene.

5. The method according to claim 1, characterized in that, The node voltage margin in step 3 is calculated as follows: the absolute value of the deviation between the voltage amplitude of each node and the rated voltage under a typical charging scenario is taken, and the difference between 1 and the maximum absolute value of the voltage deviation among all nodes in the entire power grid is used as the node voltage margin index.

6. The method according to claim 1, characterized in that, The branch power flow margin in step 3 is calculated as follows: the ratio of the power flow value of each branch under a typical charging scenario to the rated capacity of the branch is calculated, and the difference between 1 and the maximum value of this ratio among all branches in the entire power grid is used as the branch power flow margin index.

7. The method according to claim 1, characterized in that, The method for calculating the transformer load margin in step 3 is as follows: calculate the ratio of the transformer's load power to its rated capacity under a typical charging scenario, and use 1 minus this ratio as the transformer load margin index.

8. The method according to claim 1, characterized in that, The specific implementation of the improved sequential relation weighting method in step 4 is as follows: First, the subjective weights of each safety margin index are determined by sequential relation analysis. The indexes are ranked and assigned values ​​according to their relative importance to the safe operation of the distribution network, and the subjective weight vectors of each index are calculated. The second step is to calculate the degree of dispersion of each safety margin index under each typical charging scenario, in order to determine the objective correction coefficient of each index. The objective correction coefficient is negatively correlated with the degree of dispersion. The third step is to normalize the subjective weight vector after multiplying it with the objective correction coefficient to obtain the comprehensive weight vector of each index.

9. The method according to claim 1, characterized in that, The calculation method for the comprehensive load-bearing capacity score in step 5 is as follows: for each typical charging scenario, the safety margin index values ​​under that scenario are multiplied by their corresponding comprehensive weights and then summed to obtain the comprehensive load-bearing capacity score for that scenario. The minimum comprehensive carrying capacity score in all typical charging scenarios is taken as the final carrying capacity assessment result of the distribution network; when the final carrying capacity assessment result is lower than the preset threshold, it is determined that the distribution network faces the risk of insufficient carrying capacity in the current scenario, and an alarm signal is output.

10. A distribution network carrying capacity assessment system considering charging loads in multiple scenarios, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the distribution network carrying capacity assessment method that takes into account charging loads in multiple scenarios as described in any one of claims 1-9.

Images