Method and system for evaluating openable capacity of distribution network considering access of large-scale charging facilities

Abstract

The present application relates to a kind of distribution network openable capacity evaluation method and system considering the access of large-scale charging facilities, the method obtains the historical charging load and basic load data of distribution transformer under multiple typical days, analyzes the correlation between charging simultaneous rate and charging setting number, power, distribution transformer location attribute and commissioning year and other influence factors, and with historical simultaneous rate data and distribution transformer characteristic information as input, generates dynamic simultaneous rate prediction result using conditional variational autoencoder model, and then combines dynamic simultaneous coefficient to calculate the openable capacity of charging facility of distribution transformer in each time period.The scheme can reflect the actual bearing bottleneck of distribution transformer under different time periods and regional characteristics, make the evaluation result more accurate, and provide decision basis for operation scheduling and equipment installation.

Description

Technical Field

[0001] This invention relates to distribution network capacity assessment, and in particular to a method and system for assessing the open capacity of a distribution network that takes into account the access of large-scale charging facilities. Background Technology

[0002] With the rapid growth in the number of electric vehicles, the large-scale integration of charging facilities has placed significant pressure on the safe operation of the power distribution network. During peak periods and times, such as holidays at highway service areas and summer in residential areas, charging loads surge, coinciding with seasonal peak electricity demand, which can easily lead to overload or heavy load issues on some distribution transformers. Against this backdrop, accurately assessing the available capacity of the power distribution network for charging facilities has become a key issue in ensuring the safe and reliable operation of the power distribution network.

[0003] Currently, the assessment of available capacity is typically based on the static simultaneity factor, combined with the remaining capacity of distribution transformers. However, this assessment method has the following limitations:

[0004] First, it fails to adequately distinguish the dynamic characteristics of base load and charging load at different times, making it difficult to adapt to load surges in specific time and space scenarios, which can easily lead to a time and space mismatch between capacity assessment results and actual operational needs.

[0005] Second, determining the number of charging facilities that can be connected by statically comparing the rated capacity of the charging facilities with the remaining capacity of the transformer does not take into account the dynamic changes in the charging load simultaneity rate, and therefore cannot accurately reflect the actual load-bearing pressure of the distribution transformer.

[0006] Third, although some methods take into account the differences in load types in different scenarios, a differentiated analysis system for diverse scenario characteristics has not yet been formed, making it difficult to identify and locate the shortcomings in power distribution capacity under specific scenarios.

[0007] In summary, existing methods for assessing the available capacity based on the static simultaneity factor are insufficient to accurately reflect the actual load-bearing capacity bottleneck of distribution transformers under different time periods and regional characteristics. Summary of the Invention

[0008] Purpose of the invention: The purpose of this invention is to provide a method and system for assessing the open capacity of distribution networks that can more accurately reflect the characteristics of different time periods and regions, taking into account the access of large-scale charging facilities.

[0009] Technical solution: The method for assessing the open capacity of distribution networks considering the access of large-scale charging facilities, as described in this invention, includes:

[0010] Acquire historical charging load data and base load data of the distribution transformer on multiple typical days;

[0011] Based on the acquired data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor was analyzed. The influencing factors include the number of charging facilities, the power of the charging facilities, the location attributes of the distribution transformers, and the year the distribution transformers were put into operation.

[0012] Using historical charging simultaneity rate data and distribution transformer characteristic information as input, a conditional variational autoencoder model is used to generate dynamic simultaneity rate prediction results for distribution transformer charging under specific scenarios.

[0013] Based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and the available capacity of the charging facilities of the distribution transformer is determined by combining the dynamic simultaneity coefficient.

[0014] By acquiring historical charging load and base load data under multiple typical days, this study analyzes the correlation between charging simultaneity rate and influencing factors such as the number and power of charging facilities, transformer location attributes, and commissioning year. A conditional variational autoencoder model is then used to generate dynamic simultaneity rate prediction results, ultimately calculating the available capacity. This approach overcomes the limitations of existing static simultaneity rate coefficient assessment methods, reflecting the actual carrying capacity of transformers under different time periods and regional characteristics, making the available capacity assessment results closer to actual operating conditions.

[0015] Preferably, the acquisition of historical charging load data and base load data of the distribution transformer on multiple typical days includes:

[0016] The operation data of the distribution transformer in the analyzed area within a preset historical period are obtained. According to the season, the data of three typical days in each season, namely the weekdays, weekends and statutory holidays with the largest electricity consumption, the largest electricity load, the largest charging volume and the largest charging load, are selected. The average value of the data of the three typical days is calculated for the corresponding multiple sampling periods to obtain the four load types of the three typical days in each season.

[0017] Select typical power supply areas corresponding to different area types, and obtain the electricity load curve data or charging load curve data of each typical power supply area within a specified time period. The area types include residential areas, commercial areas, industrial areas and highway service areas. Among them, electricity load curve data is obtained for residential areas and commercial areas, and charging load curve data is obtained for industrial areas and highway service areas.

[0018] When selecting typical power supply areas for residential areas, factors to consider include the service life of distribution transformers and the capacity of distribution lines; when selecting typical power supply areas for commercial areas, factors to consider include load density, peak-valley fluctuations, power supply reliability, and potential expansion needs; when selecting typical power supply areas for industrial areas, factors to consider include the proportion of charging load in the total load of the park and the availability of data from the energy management platform; when selecting typical power supply areas for highway service areas, factors to consider include the service area's construction and service life, the proportion of fast-charging or supercharging facilities for electric vehicles, and the commissioning time of distribution transformers.

[0019] By refining the classification of historical data by season and typical day type, and calculating the average value of multiple sampling periods, the data used for subsequent analysis can fully cover the time characteristics of different load scenarios, providing high-quality basic data support for dynamic simultaneity analysis.

[0020] Clearly distinguishing the specific considerations for selecting typical power supply areas in four categories—residential areas, commercial areas, industrial areas, and highway service areas—provides a clear and operable standard for the data selection process, which helps ensure the consistency and comparability of analysis results under different scenarios.

[0021] As a preferred approach, based on historical charging load data and base load data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor is analyzed. The obtained data covers distribution transformer capacity, location attributes, distribution transformer commissioning year, number of electricity users, number of connected charging facilities, charging facility capacity, charging load and base load information.

[0022] The analysis includes the following procedures:

[0023] The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under different numbers of charging piles were statistically analyzed, and the impact of the number of charging facilities on the simultaneity rate in typical scenarios in different regions was analyzed.

[0024] The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under the condition of connecting charging facilities with different rated power are statistically analyzed to analyze the impact of the charging facility power on the simultaneity rate.

[0025] The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different regional attributes were statistically analyzed to examine the impact of distribution transformer location attributes on the simultaneity rate.

[0026] The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different commissioning times were statistically analyzed to examine the impact of the commissioning year of the distribution transformer on the simultaneity rate.

[0027] The acquired dataset encompasses multi-dimensional information such as distribution transformer capacity, location attributes, commissioning year, number of users, number and capacity of connected charging facilities, charging load, and base load, providing a comprehensive data foundation for simultaneity rate analysis and ensuring the reliability and coverage of the analysis results. By statistically analyzing the temporal and probability density distributions of the simultaneity rate from several dimensions—number of charging facilities, facility power, location attributes, and commissioning year—the intrinsic correlation between various influencing factors and the simultaneity rate can be systematically revealed, providing multi-dimensional feature basis for the prediction model.

[0028] Preferably, the conditional variational autoencoder model uses historical charging simultaneity rate data as time-series input and distribution transformer characteristic information as conditional input. The encoder compresses the historical charging simultaneity rate data into a latent space representation, and then the decoder reconstructs the simultaneity rate curve from the latent space. The distribution transformer characteristic information includes the distribution transformer charging facility capacity, distribution transformer user load or number of users, and the year the distribution transformer was put into operation. The process of generating the dynamic simultaneity rate prediction result of the distribution transformer charging includes: sampling in the latent space, decoding the sampling result using the trained decoder, and generating and outputting the dynamic simultaneity rate prediction curve of the distribution transformer charging.

[0029] As a preferred method, based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and combined with the dynamic simultaneity coefficient, the available capacity of the charging facilities of the distribution transformer is determined, including the following process:

[0030] Based on the preset safe operating capacity threshold of the distribution transformer, the base load and charging load are subtracted to obtain the available charging margin for each typical day and time period.

[0031] The dynamic simultaneity coefficient is determined based on the ratio of charging load to total rated charging power in each time period.

[0032] The available capacity for each time period on each typical day is determined based on the ratio of the available charging margin to the dynamic simultaneity coefficient.

[0033] Take the minimum available capacity for each time period within each typical day, and then take the minimum available capacity for each typical day as the available capacity for the charging facilities of the distribution transformer.

[0034] The calculation expressions for the above parameters are as follows:

[0035] ,

[0036] ,

[0037] ,

[0038] ,

[0039] in, This represents the available charging margin for a typical day during the t-th period. For a typical day s, the dynamic simultaneity coefficient for the t-th time period is ; This represents the available capacity for a typical day during the t-th time period; The charging capacity of the distribution transformer can be opened; This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day. The rated charging power is the total daily amount.

[0040] Based on the preset safe operating capacity threshold, the available capacity is calculated time-by-time using a dynamic simultaneity coefficient, and the minimum value of multiple time periods and multiple typical days is taken as the final result, making the assessment result conservative and safe, which is conducive to the safe and stable operation of the distribution network.

[0041] The distribution network open capacity assessment system considering the access of large-scale charging facilities as described in this invention includes:

[0042] Data selection module: used to acquire historical charging load data and base load data of distribution transformers on multiple typical days;

[0043] Charging Simultaneity Analysis Module: Based on the acquired data, this module analyzes the correlation between the charging simultaneity rate of the distribution transformer and at least one influencing factor, including the number of charging facilities, the power of the charging facilities, the location attributes of the distribution transformer, and the year the distribution transformer was put into operation.

[0044] Dynamic charging simultaneity rate prediction module: It is used to generate dynamic charging simultaneity rate prediction results for distribution transformers under specific scenarios by taking historical charging simultaneity rate data and distribution transformer characteristic information as input and using a conditional variational autoencoder model.

[0045] Openable capacity calculation module: Based on the dynamic simultaneity rate prediction results of the distribution transformer charging, it calculates the available charging margin of the distribution transformer in each time period, and determines the openable capacity of the charging facilities of the distribution transformer by combining the dynamic simultaneity coefficient.

[0046] Preferably, the acquisition of historical charging load data and base load data of the distribution transformer on multiple typical days includes:

[0047] The operation data of the distribution transformer in the analyzed area within a preset historical period are obtained. According to the season, the data of three typical days in each season, namely the weekdays, weekends and statutory holidays with the largest electricity consumption, the largest electricity load, the largest charging volume and the largest charging load, are selected. The average value of the data of the three typical days is calculated for the corresponding multiple sampling periods to obtain the four load types of the three typical days in each season.

[0048] Select typical power supply areas corresponding to different area types, and obtain the electricity load curve data or charging load curve data of each typical power supply area within a specified time period. The area types include residential areas, commercial areas, industrial areas and highway service areas. Among them, electricity load curve data is obtained for residential areas and commercial areas, and charging load curve data is obtained for industrial areas and highway service areas.

[0049] When selecting typical power supply areas for residential areas, factors to consider include the service life of distribution transformers and the capacity of distribution lines; when selecting typical power supply areas for commercial areas, factors to consider include load density, peak-valley fluctuations, power supply reliability, and potential expansion needs; when selecting typical power supply areas for industrial areas, factors to consider include the proportion of charging load in the total load of the park and the availability of data from the energy management platform; when selecting typical power supply areas for highway service areas, factors to consider include the service area's construction and service life, the proportion of fast-charging or supercharging facilities for electric vehicles, and the commissioning time of distribution transformers.

[0050] As a preferred approach, based on historical charging load data and base load data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor is analyzed. The obtained data covers distribution transformer capacity, location attributes, distribution transformer commissioning year, number of electricity users, number of connected charging facilities, charging facility capacity, charging load and base load information.

[0051] The analysis includes the following procedures:

[0052] The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under different numbers of charging piles were statistically analyzed, and the impact of the number of charging facilities on the simultaneity rate in typical scenarios in different regions was analyzed.

[0053] The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under the condition of connecting charging facilities with different rated power are statistically analyzed to analyze the impact of the charging facility power on the simultaneity rate.

[0054] The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different regional attributes were statistically analyzed to examine the impact of distribution transformer location attributes on the simultaneity rate.

[0055] The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different commissioning times were statistically analyzed to examine the impact of the commissioning year of the distribution transformer on the simultaneity rate.

[0056] Preferably, the conditional variational autoencoder model uses historical charging simultaneity rate data as time-series input and distribution transformer characteristic information as conditional input. The encoder compresses the historical charging simultaneity rate data into a latent space representation, and then the decoder reconstructs the simultaneity rate curve from the latent space. The distribution transformer characteristic information includes the distribution transformer charging facility capacity, distribution transformer user load or number of users, and the year the distribution transformer was put into operation. The process of generating the dynamic simultaneity rate prediction result of the distribution transformer charging includes: sampling in the latent space, decoding the sampling result using the trained decoder, and generating and outputting the dynamic simultaneity rate prediction curve of the distribution transformer charging.

[0057] As a preferred method, based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and combined with the dynamic simultaneity coefficient, the available capacity of the charging facilities of the distribution transformer is determined, including the following process:

[0058] Based on the preset safe operating capacity threshold of the distribution transformer, the base load and charging load are subtracted to obtain the available charging margin for each typical day and time period.

[0059] The dynamic simultaneity coefficient is determined based on the ratio of charging load to total rated charging power in each time period.

[0060] The available capacity for each time period on each typical day is determined based on the ratio of the available charging margin to the dynamic simultaneity coefficient.

[0061] Take the minimum available capacity for each time period within each typical day, and then take the minimum available capacity for each typical day as the available capacity for the charging facilities of the distribution transformer.

[0062] The calculation expressions for the above parameters are as follows:

[0063] ,

[0064] ,

[0065] ,

[0066] ,

[0067] in, This represents the available charging margin for a typical day during the t-th period. For a typical day s, the dynamic simultaneity coefficient for the t-th time period is ; This represents the available capacity for a typical day during the t-th time period; The charging capacity of the distribution transformer can be opened; This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day. The rated charging power is the total daily amount.

[0068] Beneficial Effects: By analyzing the dynamic influencing factors of charging simultaneity rate and introducing a simultaneity rate prediction model based on conditional variational autoencoders, the shortcomings of existing static simultaneity rate assessment methods are overcome. This model can accurately reflect the actual carrying capacity bottleneck of distribution transformers under different time periods and regional characteristics. The system-calculated available capacity of distribution transformer charging facilities can provide a basis for operation scheduling and the application for installation of equipment such as distribution transformers and charging piles, providing effective support for the safe and stable operation of distribution networks under the large-scale development of electric vehicles. Attached Figure Description

[0069] Figure 1 This is a schematic diagram of the framework of the distribution network open capacity assessment method and system of the present invention;

[0070] Figure 2 A flowchart illustrating the typical daily classification and historical data selection method;

[0071] Figure 3 This is a schematic diagram of the simultaneous rate curve generation model based on a conditional variational autoencoder.

[0072] Figure 4 A schematic diagram illustrating the time distribution and probability density distribution of the simultaneous charging rate for electric vehicles, wherein... Figure 4 (a) Comparison under different numbers of charging facilities. Figure 4 (b) Comparison of power levels for different charging facilities. Figure 4 (c) Comparison under different regional attributes Figure 4 (d) Comparison under different transformer commissioning times;

[0073] Figure 5 A diagram showing the comparison between the predicted and actual values ​​of the simultaneous charging rate curve of distribution transformers under different scenarios;

[0074] Figure 6 A schematic diagram of the calculation process for the open capacity of distribution transformers. Detailed Implementation

[0075] like Figure 1 As shown, the method for assessing the open capacity of a distribution network considering the access of large-scale charging facilities, as described in this invention, includes:

[0076] Obtain historical charging load data and base load data of the distribution transformer on multiple typical days.

[0077] In practice, in order to calculate the dynamic simultaneity rate of distribution transformer charging, it is first necessary to collect and preprocess historical charging data.

[0078] Specifically, it is divided into typical daily classification and historical data selection, and typical power supply area data selection:

[0079] Typical daily classification and historical data selection: Please refer to Figure 2 Historical operating data of distribution transformers in the analyzed area over the past year were obtained and divided into four seasons. For each season, data on three typical days were selected: weekdays, weekends, and statutory holidays with the highest electricity consumption, highest electricity load, highest charging volume, and highest charging load. For the selected typical day data, the average value of 96 sampling periods was calculated with a sampling resolution of 15 minutes to obtain the four load types of data for the three typical days in each of the four seasons: highest electricity consumption, highest electricity load, highest charging volume, and highest charging load.

[0080] Typical power supply area data selection: To differentiate the load demand of different types of areas and form a differentiated analysis of diverse scenario characteristics, it is necessary to select typical power supply areas corresponding to each area type and obtain their electricity load curve data and charging load curve data. Area types include residential areas, commercial areas, industrial areas, and highway service areas.

[0081] For residential areas, when selecting typical power supply areas, key characteristics such as the long service life of distribution transformers and insufficient distribution line capacity, which are common in old communities, are taken into consideration. Load curve data of residential areas in four seasons are obtained to reflect the electricity load characteristics of residential life.

[0082] For commercial areas, when selecting typical power supply areas, key considerations include high load density, significant peak-valley fluctuations, strict power supply reliability requirements, and potential expansion needs. The annual load curve data of the commercial area is obtained to reflect the high frequency of power load fluctuations in the commercial area.

[0083] For industrial zones, when selecting typical power supply areas, key considerations include the core proportion of charging load to the total load of the zone and the unified management of data by the zone's energy platform. The annual charging load curve data of the industrial zone is obtained to reflect the charging load characteristics driven by the industrial zone's production plan.

[0084] For highway service areas, when selecting typical power supply areas, key considerations include the service area's construction and service life, the proportion of fast charging and supercharging facilities for electric vehicles, and the commissioning time of distribution transformers. The annual charging load curve data of highway service areas is obtained to reflect the high-frequency fluctuation of charging load characteristics of highway service areas.

[0085] Based on the acquired data, the correlation between the simultaneous charging rate of distribution transformers and the number of charging facilities, the power of charging facilities, the location attributes of distribution transformers, and the year of commissioning of distribution transformers was analyzed.

[0086] In practical implementation, based on the charging curves of transformer substations and electric vehicle user data in typical power supply areas, a dataset is constructed that includes information such as transformer capacity, location attributes, transformer commissioning year, number of power users, number of connected charging facilities, charging facility capacity, charging load, and base load, providing data support for simultaneous rate analysis.

[0087] Analyzing relationships specifically includes the following processes:

[0088] Based on the dataset, the temporal distribution and probability density distribution of the simultaneous charging rate of electric vehicles were statistically analyzed under different numbers of charging piles connected to the distribution transformer (taking 10-15, 15-20, and 30-40 charging piles as examples). Please refer to [link / reference needed]. Figure 4 (a). By analyzing the curve trend in the figure, we study the impact of the number of charging facilities on the simultaneous charging rate in typical scenarios in different regions.

[0089] Based on the dataset, the temporal distribution and probability density distribution of electric vehicle charging simultaneity rates were statistically analyzed when distribution transformers were connected to charging facilities with different rated power (taking 6-10kW, 10-15kW, and 15-25kW charging piles as examples). Please refer to [link to relevant documentation]. Figure 4 (b) By analyzing the curve trend in the figure, we study the impact of the increase in charging facility power on the charging simultaneity rate.

[0090] Based on the dataset, the temporal distribution and probability density distribution of simultaneous electric vehicle charging rates under different regional attributes (taking urban areas, county towns, and rural areas as examples) are statistically analyzed. Please refer to [link to relevant documentation]. Figure 4 (c). By analyzing the curve trends in the figure, we study the impact of the regional characteristics of the distribution transformer on the charging simultaneity rate.

[0091] Based on the dataset, the temporal distribution and probability density distribution of the simultaneous charging rate of electric vehicles were statistically analyzed under different commissioning periods (taking the commissioning periods of 1997-2007, 2007-2017, and 2017-2025 as examples). Please refer to [link to relevant documentation]. Figure 4 (d). By analyzing the curve trend in the figure, we study the intrinsic relationship between the commissioning time of distribution transformers and the charging simultaneity rate in residential areas, commercial areas, industrial areas and highway service areas.

[0092] Using historical charging simultaneity rate data and distribution transformer characteristic information as input, a conditional variational autoencoder model is used to generate dynamic simultaneity rate prediction results for distribution transformer charging under specific scenarios.

[0093] Please see Figure 3 In its implementation, the model adopts a conditional variational autoencoder architecture, which is divided into two stages: the training process and the curve generation process.

[0094] During training, the historical time-series data used to train the encoder-decoder network is historical charging simultaneity rate data. The input transformer characteristic information includes three categories: transformer charging facility capacity, transformer user load / number of users, and transformer commissioning year. The encoder and decoder adopt a symmetrical structure; the encoder compresses the historical simultaneity rate data into a latent space representation, and the decoder reconstructs the simultaneity rate curve from the latent space representation. After training, a dynamic simultaneity rate curve generation model for transformer charging based on a conditional variational autoencoder is obtained.

[0095] The process of generating the dynamic charging simultaneity rate curve of the distribution transformer includes: latent space sampling step, decoder decoding step, and finally generating and outputting the dynamic charging simultaneity rate prediction curve and envelope interval of the distribution transformer.

[0096] To verify the effectiveness of the model, Figure 5 The comparison between the predicted dynamic simultaneous charging rate curves of distribution transformers and the actual values ​​is shown for two typical scenarios. Scenario 1 corresponds to a rural area with 133 users and 10 charging facilities; Scenario 2 corresponds to an urban area with 337 users and 23 charging facilities. Figure 5 It is evident that the predicted charging dynamic simultaneity rate curve can effectively fit the actual historical charging dynamic simultaneity rate, verifying the effectiveness of the distribution transformer charging dynamic simultaneity rate curve generation model based on conditional variational autoencoder.

[0097] Based on the prediction results of the dynamic simultaneous charging rate of the distribution transformer, the available capacity of the distribution transformer is calculated.

[0098] Please see Figure 6 In practice, 80% of the rated capacity of the distribution transformer is used as the preset safe operating capacity threshold, and the fact that the distribution transformer is not overloaded at any time is used as the criterion for judging the safe carrying capacity of the charging facility. The dynamic simultaneity rate of typical days in each scenario is used to calculate the open capacity of the distribution transformer in each time period.

[0099] Calculate the available charging margin. The available charging margin is obtained by subtracting the base load and charging load from the preset safe operating capacity threshold. The calculation expression is:

[0100] ,

[0101] in: This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day.

[0102] Calculate the dynamic simultaneity factor. The dynamic simultaneity factor is obtained by dividing the charging load of each time period by the total rated charging power under the corresponding typical day. The calculation expression is as follows:

[0103] ,

[0104] in: The total rated charging power for a typical day is the sum of the rated power of all charging piles. This is the dynamic simultaneity coefficient.

[0105] Calculate the available capacity for each time period. The available capacity for each time period is obtained by dividing the available charging margin by the corresponding dynamic simultaneity coefficient. The calculation expression is:

[0106] ,

[0107] in, This represents the available capacity during the tt period of a typical day, in kW; the meanings of the other parameters are the same as before.

[0108] Determine the final available capacity. First, for a single typical day, take the minimum value among the 96 available capacity values ​​within that typical day. Then, for the minimum values ​​corresponding to each of the three types of typical days—weekdays, weekends, and public holidays—take the final minimum value as the final result of the available charging capacity for the distribution transformer. The calculation expression is:

[0109] ,

[0110] in, The charging capacity of the distribution transformer can be opened.

[0111] This invention also provides a distribution network open capacity assessment system considering the access of large-scale charging facilities, comprising:

[0112] Data selection module: used to acquire historical charging load data and base load data of distribution transformers on multiple typical days;

[0113] Specifically, it is divided into typical daily classification and historical data selection, and typical power supply area data selection:

[0114] Typical daily classification and historical data selection: Please refer to Figure 2 Historical operating data of distribution transformers in the analyzed area over the past year were obtained and divided into four seasons. For each season, data on three typical days were selected: weekdays, weekends, and statutory holidays with the highest electricity consumption, highest electricity load, highest charging volume, and highest charging load. For the selected typical day data, the average value of 96 sampling periods was calculated with a sampling resolution of 15 minutes to obtain the four load types of data for the three typical days in each of the four seasons: highest electricity consumption, highest electricity load, highest charging volume, and highest charging load.

[0115] Typical power supply area data selection: To differentiate the load demand of different types of areas and form a differentiated analysis of diverse scenario characteristics, it is necessary to select typical power supply areas corresponding to each area type and obtain their electricity load curve data and charging load curve data. Area types include residential areas, commercial areas, industrial areas, and highway service areas.

[0116] For residential areas, when selecting typical power supply areas, key characteristics such as the long service life of distribution transformers and insufficient distribution line capacity, which are common in old communities, are taken into consideration. Load curve data of residential areas in four seasons are obtained to reflect the electricity load characteristics of residential life.

[0117] For commercial areas, when selecting typical power supply areas, key considerations include high load density, significant peak-valley fluctuations, strict power supply reliability requirements, and potential expansion needs. The annual load curve data of the commercial area is obtained to reflect the high frequency of power load fluctuations in the commercial area.

[0118] For industrial zones, when selecting typical power supply areas, key considerations include the core proportion of charging load to the total load of the zone and the unified management of data by the zone's energy platform. The annual charging load curve data of the industrial zone is obtained to reflect the charging load characteristics driven by the industrial zone's production plan.

[0119] For highway service areas, when selecting typical power supply areas, key considerations include the service area's construction and service life, the proportion of fast charging and supercharging facilities for electric vehicles, and the commissioning time of distribution transformers. The annual charging load curve data of highway service areas is obtained to reflect the high-frequency fluctuation of charging load characteristics of highway service areas.

[0120] Based on the acquired data, the correlation between the simultaneous charging rate of distribution transformers and the number of charging facilities, the power of charging facilities, the location attributes of distribution transformers, and the year of commissioning of distribution transformers was analyzed.

[0121] In practical implementation, based on the charging curves of transformer substations and electric vehicle user data in typical power supply areas, a dataset is constructed that includes information such as transformer capacity, location attributes, transformer commissioning year, number of power users, number of connected charging facilities, charging facility capacity, charging load, and base load, providing data support for simultaneous rate analysis.

[0122] Charging Simultaneity Analysis Module: Based on the acquired data, this module analyzes the correlation between the charging simultaneity rate of the distribution transformer and at least one influencing factor, including the number of charging facilities, the power of the charging facilities, the location attributes of the distribution transformer, and the year the distribution transformer was put into operation.

[0123] Analyzing relationships specifically includes the following processes:

[0124] Based on the dataset, the temporal distribution and probability density distribution of the simultaneous charging rate of electric vehicles were statistically analyzed under different numbers of charging piles connected to the distribution transformer (taking 10-15, 15-20, and 30-40 charging piles as examples). Please refer to [link / reference needed]. Figure 4 (a). By analyzing the curve trend in the figure, we study the impact of the number of charging facilities on the simultaneous charging rate in typical scenarios in different regions.

[0125] Based on the dataset, the temporal distribution and probability density distribution of electric vehicle charging simultaneity rates were statistically analyzed when distribution transformers were connected to charging facilities with different rated power (taking 6-10kW, 10-15kW, and 15-25kW charging piles as examples). Please refer to [link to relevant documentation]. Figure 4 (b) By analyzing the curve trend in the figure, we study the impact of the increase in charging facility power on the charging simultaneity rate.

[0126] Based on the dataset, the temporal distribution and probability density distribution of simultaneous electric vehicle charging rates under different regional attributes (taking urban areas, county towns, and rural areas as examples) are statistically analyzed. Please refer to [link to relevant documentation]. Figure 4 (c). By analyzing the curve trends in the figure, we study the impact of the regional characteristics of the distribution transformer on the charging simultaneity rate.

[0127] Based on the dataset, the temporal distribution and probability density distribution of the simultaneous charging rate of electric vehicles were statistically analyzed under different commissioning periods (taking the commissioning periods of 1997-2007, 2007-2017, and 2017-2025 as examples). Please refer to [link to relevant documentation]. Figure 4 (d). By analyzing the curve trend in the figure, we study the intrinsic relationship between the commissioning time of distribution transformers and the charging simultaneity rate in residential areas, commercial areas, industrial areas and highway service areas.

[0128] Dynamic charging simultaneity rate prediction module: It is used to generate dynamic charging simultaneity rate prediction results for distribution transformers under specific scenarios by taking historical charging simultaneity rate data and distribution transformer characteristic information as input and using a conditional variational autoencoder model.

[0129] During training, the historical time-series data used to train the encoder-decoder network is historical charging simultaneity rate data. The input transformer characteristic information includes three categories: transformer charging facility capacity, transformer user load / number of users, and transformer commissioning year. The encoder and decoder adopt a symmetrical structure; the encoder compresses the historical simultaneity rate data into a latent space representation, and the decoder reconstructs the simultaneity rate curve from the latent space representation. After training, a dynamic simultaneity rate curve generation model for transformer charging based on a conditional variational autoencoder is obtained.

[0130] The process of generating the dynamic charging simultaneity rate curve of the distribution transformer includes: latent space sampling step, decoder decoding step, and finally generating and outputting the dynamic charging simultaneity rate prediction curve and envelope interval of the distribution transformer.

[0131] Openable capacity calculation module: Based on the dynamic simultaneity rate prediction results of the distribution transformer charging, it calculates the available charging margin of the distribution transformer in each time period, and determines the openable capacity of the charging facilities of the distribution transformer by combining the dynamic simultaneity coefficient.

[0132] The specific process is as follows: 80% of the rated capacity of the distribution transformer is used as the preset safe operating capacity threshold, and the fact that the distribution transformer is not overloaded at any time is used as the criterion for judging the safe carrying capacity of the charging facility. The dynamic simultaneity rate of typical days under various scenarios is used to calculate the open capacity of the distribution transformer in each time period.

[0133] Calculate the available charging margin. The available charging margin is obtained by subtracting the base load and charging load from the preset safe operating capacity threshold. The calculation expression is:

[0134] ,

[0135] in: This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day.

[0136] Calculate the dynamic simultaneity factor. The dynamic simultaneity factor is obtained by dividing the charging load of each time period by the total rated charging power under the corresponding typical day. The calculation expression is as follows:

[0137] ,

[0138] in: The total rated charging power for a typical day is the sum of the rated power of all charging piles. This is the dynamic simultaneity coefficient.

[0139] Calculate the available capacity for each time period. The available capacity for each time period is obtained by dividing the available charging margin by the corresponding dynamic simultaneity coefficient. The calculation expression is:

[0140] ,

[0141] in, This represents the available capacity during the tt period of a typical day, in kW; the meanings of the other parameters are the same as before.

[0142] Determine the final available capacity. First, for a single typical day, take the minimum value among the 96 available capacity values ​​within that typical day. Then, for the minimum values ​​corresponding to each of the three types of typical days—weekdays, weekends, and public holidays—take the final minimum value as the final result of the available charging capacity for the distribution transformer. The calculation expression is:

[0143] ,

[0144] in, The charging capacity of the distribution transformer can be opened.

[0145] By combining the above four modules, the available capacity of distribution transformer charging facilities is calculated according to the aforementioned method for assessing the available capacity of the distribution network. Based on the available capacity of distribution transformer charging facilities obtained by this system, it can accurately reflect the actual capacity bottleneck of distribution transformers under different time periods and regional characteristics, taking into account the differences in the characteristics of diverse scenarios and the shortcomings of distribution capacity under specific scenarios, thus providing strong support for the safe and stable operation of the distribution network under the large-scale development of electric vehicles.

Claims

1. A method for assessing the open capacity of a distribution network considering the access of large-scale charging facilities, characterized in that, include: Acquire historical charging load data and base load data of the distribution transformer on multiple typical days; Based on historical charging load data and basic load data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor is analyzed. The influencing factors include the number of charging facilities, the power of charging facilities, the location attributes of distribution transformers, and the year of commissioning of distribution transformers. Using historical charging simultaneity rate data and distribution transformer characteristic information as input, a conditional variational autoencoder model is used to generate dynamic simultaneity rate prediction results for distribution transformer charging under specific scenarios. Based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and the available capacity of the charging facilities of the distribution transformer is determined by combining the dynamic simultaneity coefficient.

2. The method according to claim 1, characterized in that, The acquisition of historical charging load data and base load data of the distribution transformer on multiple typical days includes: The operation data of the distribution transformer in the analyzed area within a preset historical period are obtained. According to the season, the data of three typical days in each season, namely the weekdays, weekends and statutory holidays with the largest electricity consumption, the largest electricity load, the largest charging volume and the largest charging load, are selected. The average value of the data of the three typical days is calculated for the corresponding multiple sampling periods to obtain the four load types of the three typical days in each season. Select typical power supply areas corresponding to different area types, and obtain the electricity load curve data or charging load curve data of each typical power supply area within a specified time period. The area types include residential areas, commercial areas, industrial areas and highway service areas. Among them, electricity load curve data is obtained for residential areas and commercial areas, and charging load curve data is obtained for industrial areas and highway service areas. When selecting typical power supply areas for residential areas, factors to consider include the service life of distribution transformers and the capacity of distribution lines; when selecting typical power supply areas for commercial areas, factors to consider include load density, peak-valley fluctuations, power supply reliability, and potential expansion needs; when selecting typical power supply areas for industrial areas, factors to consider include the proportion of charging load in the total load of the park and the availability of data from the energy management platform; when selecting typical power supply areas for highway service areas, factors to consider include the service area's construction and service life, the proportion of fast-charging or supercharging facilities for electric vehicles, and the commissioning time of distribution transformers.

3. The method according to claim 1, characterized in that, Based on historical charging load data and basic load data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor is analyzed. The data obtained covers distribution transformer capacity, location attributes, distribution transformer commissioning year, number of power users, number of connected charging facilities, charging facility capacity, charging load and basic load information. The analysis includes the following procedures: The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under different numbers of charging piles were statistically analyzed, and the impact of the number of charging facilities on the simultaneity rate in typical scenarios in different regions was analyzed. The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under the condition of connecting charging facilities with different rated power are statistically analyzed to analyze the impact of the charging facility power on the simultaneity rate. The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different regional attributes were statistically analyzed to examine the impact of distribution transformer location attributes on the simultaneity rate. The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different commissioning times were statistically analyzed to examine the impact of the commissioning year of the distribution transformer on the simultaneity rate.

4. The method according to claim 1, characterized in that: The conditional variational autoencoder model takes historical charging simultaneity rate data as time-series input and distribution transformer characteristic information as conditional input. The encoder compresses the historical charging simultaneity rate data into a latent space representation, and then the decoder reconstructs the simultaneity rate curve from the latent space. The distribution transformer characteristic information includes the distribution transformer charging facility capacity, distribution transformer user load or number of users, and the year the distribution transformer was put into operation. The process of generating the dynamic simultaneous rate prediction result of distribution transformer charging includes: sampling in the latent space, decoding the sampling result using the trained decoder, and generating and outputting the dynamic simultaneous rate prediction curve of distribution transformer charging.

5. The method according to claim 1, characterized in that, Based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and combined with the dynamic simultaneity coefficient, the available charging capacity of the distribution transformer is determined, including the following process: Based on the preset safe operating capacity threshold of the distribution transformer, the base load and charging load are subtracted to obtain the available charging margin for each typical day and time period. The dynamic simultaneity coefficient is determined based on the ratio of charging load to total rated charging power in each time period. The available capacity for each time period on each typical day is determined based on the ratio of the available charging margin to the dynamic simultaneity coefficient. Take the minimum available capacity for each time period within each typical day, and then take the minimum available capacity for each typical day as the available capacity for the charging facilities of the distribution transformer. The calculation expressions for the above parameters are as follows: ， ， ， ， in, This represents the available charging margin for a typical day during the t-th period. For a typical day s, the dynamic simultaneity coefficient for the t-th time period is ; This represents the available capacity for a typical day during the t-th time period. The charging capacity of the distribution transformer can be opened; This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day. The rated charging power is the total daily amount.

6. A distribution network open capacity assessment system, characterized in that, include: Data selection module: used to acquire historical charging load data and base load data of distribution transformers on multiple typical days; Charging Simultaneity Analysis Module: Based on the acquired data, this module analyzes the correlation between the charging simultaneity rate of the distribution transformer and at least one influencing factor, including the number of charging facilities, the power of the charging facilities, the location attributes of the distribution transformer, and the year the distribution transformer was put into operation. Dynamic charging simultaneity rate prediction module: It is used to generate dynamic charging simultaneity rate prediction results for distribution transformers under specific scenarios by taking historical charging simultaneity rate data and distribution transformer characteristic information as input and using a conditional variational autoencoder model. Openable capacity calculation module: Based on the dynamic simultaneity rate prediction results of the distribution transformer charging, it calculates the available charging margin of the distribution transformer in each time period, and determines the openable capacity of the charging facilities of the distribution transformer by combining the dynamic simultaneity coefficient.

7. The system according to claim 6, characterized in that, The acquisition of historical charging load data and base load data of the distribution transformer on multiple typical days includes: The operation data of the distribution transformer in the analyzed area within a preset historical period are obtained. According to the season, the data of three typical days in each season, namely the weekdays, weekends and statutory holidays with the largest electricity consumption, the largest electricity load, the largest charging volume and the largest charging load, are selected. The average value of the data of the three typical days is calculated for the corresponding multiple sampling periods to obtain the four load types of the three typical days in each season. Select typical power supply areas corresponding to different area types, and obtain the electricity load curve data or charging load curve data of each typical power supply area within a specified time period. The area types include residential areas, commercial areas, industrial areas and highway service areas. Among them, electricity load curve data is obtained for residential areas and commercial areas, and charging load curve data is obtained for industrial areas and highway service areas. When selecting typical power supply areas for residential areas, factors to consider include the service life of distribution transformers and the capacity of distribution lines; when selecting typical power supply areas for commercial areas, factors to consider include load density, peak-valley fluctuations, power supply reliability, and potential expansion needs; when selecting typical power supply areas for industrial areas, factors to consider include the proportion of charging load in the total load of the park and the availability of data from the energy management platform; when selecting typical power supply areas for highway service areas, factors to consider include the service area's construction and service life, the proportion of fast-charging or supercharging facilities for electric vehicles, and the commissioning time of distribution transformers.

8. The system according to claim 6, characterized in that, Based on historical charging load data and basic load data, the correlation between the simultaneous charging rate of distribution transformers and at least one influencing factor is analyzed. The data obtained covers distribution transformer capacity, location attributes, distribution transformer commissioning year, number of power users, number of connected charging facilities, charging facility capacity, charging load and basic load information. The analysis includes the following procedures: The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under different numbers of charging piles were statistically analyzed, and the impact of the number of charging facilities on the simultaneity rate in typical scenarios in different regions was analyzed. The time distribution and probability density distribution of the charging simultaneity rate of the distribution transformer under the condition of connecting charging facilities with different rated power are statistically analyzed to analyze the impact of the charging facility power on the simultaneity rate. The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different regional attributes were statistically analyzed to examine the impact of distribution transformer location attributes on the simultaneity rate. The temporal distribution and probability density distribution of charging simultaneity rate of distribution transformers under different commissioning times were statistically analyzed to examine the impact of the commissioning year of the distribution transformer on the simultaneity rate.

9. The system according to claim 6, characterized in that: The conditional variational autoencoder model takes historical charging simultaneity rate data as time-series input and distribution transformer characteristic information as conditional input. The encoder compresses the historical charging simultaneity rate data into a latent space representation, and then the decoder reconstructs the simultaneity rate curve from the latent space. The distribution transformer characteristic information includes the distribution transformer charging facility capacity, distribution transformer user load or number of users, and the year the distribution transformer was put into operation. The process of generating the dynamic simultaneous rate prediction result of distribution transformer charging includes: sampling in the latent space, decoding the sampling result using the trained decoder, and generating and outputting the dynamic simultaneous rate prediction curve of distribution transformer charging.

10. The method according to claim 6, characterized in that, Based on the prediction results of the dynamic simultaneity rate of distribution transformer charging, the available charging margin of the distribution transformer in each time period is calculated, and combined with the dynamic simultaneity coefficient, the available charging capacity of the distribution transformer is determined, including the following process: Based on the preset safe operating capacity threshold of the distribution transformer, the base load and charging load are subtracted to obtain the available charging margin for each typical day and time period. The dynamic simultaneity coefficient is determined based on the ratio of charging load to total rated charging power in each time period. The available capacity for each time period on each typical day is determined based on the ratio of the available charging margin to the dynamic simultaneity coefficient. Take the minimum available capacity for each time period within each typical day, and then take the minimum available capacity for each typical day as the available capacity for the charging facilities of the distribution transformer. The calculation expressions for the above parameters are as follows: ， ， ， ， in, This represents the available charging margin for a typical day during the t-th period. For a typical day s, the dynamic simultaneity coefficient for the t-th time period is ; This represents the available capacity for a typical day during the t-th time period. The charging capacity of the distribution transformer can be opened; This refers to the rated power of the distribution transformer. This represents the base load for a typical day during period t. This represents the charging load during the t-th period of a typical day. The rated charging power is the total daily amount.

Images