Method and system for evaluating peak bearing capacity of power distribution network

Abstract

The invention provides a power distribution network peak bearing capacity evaluation method and system, and the method comprises the following steps: reading an evaluation request, and analyzing the evaluation request to generate a bearing capacity evaluation parameter set; the method specifically comprises the following steps: performing time dimension multi-scale analysis on new energy output and load demand change; performing spatial dimension analysis on the spatial topological structure of the power distribution network; calculating an evaluation value of each node in the power distribution network based on the evaluation parameter set; selecting an optimal target node according to the evaluation value; dynamically allocating an evaluation task to the optimal target node; and the optimal target node obtains the evaluation task, then performs task decomposition, executes task evaluation and outputs an evaluation result. Generating a power distribution network peak bearing capacity threshold according to an evaluation result; and comparing the real-time operation data with the peak bearing capacity threshold value of the power distribution network to generate early warning information. Based on the method, the invention also provides a power distribution network peak bearing capacity evaluation system. According to the invention, refined evaluation and real-time early warning of the bearing capacity of the power distribution network are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network carrying capacity assessment, and in particular relates to a method and system for assessing the peak carrying capacity of a distribution network. Background Art

[0002] In recent years, with the explosive growth of distributed renewable energy grid-connected installations and the widespread adoption of new energy vehicles, distribution network capacity in some regions has reached saturation, making it difficult for upgrades to keep pace with energy development. Consequently, the issue of distribution network capacity and its improvement has become increasingly prominent. Accurately assessing the peak capacity of distribution networks is crucial for ensuring safe and stable grid operation, rationally planning grid upgrades, and promoting efficient energy utilization.

[0003] The method and system for constructing a static load model and applying it to power flow calculations, disclosed in Chinese patent CN118074136A, rely on historical load data or fixed topology models. These methods are unable to dynamically respond to real-time changes such as the integration of distributed power sources, fluctuations in renewable energy output (such as intermittent photovoltaic / wind power generation), and sudden load changes. Furthermore, in practical applications, the static model often produces short-term peak load capacity prediction errors exceeding 25%. A low-voltage substation topology identification method and system, disclosed in Chinese patent CN116581867A, uses traditional early warning systems that rely on fixed threshold triggers and lack dynamic predictions based on probabilistic risk. Field data shows that in the event of a sudden load surge, the average response time of the existing system exceeds 15 minutes, far exceeding the second-level dispatching requirements. Therefore, existing distribution network peak load capacity assessment technologies assess load capacity based on the total power delivered. However, with the addition of clean energy sources such as wind and photovoltaic power, the power supply system has become more diverse. Furthermore, the volatility of renewable energy generation naturally conflicts with the steady-state output characteristics of traditional power sources. Existing technologies lack a multi-timescale (seconds to hours) source-grid-load coordinated optimization mechanism. Summary of the Invention

[0004] To address the aforementioned technical issues, this paper proposes a method and system for evaluating the peak carrying capacity of distribution networks. By leveraging a multi-time and space-scale model, it accurately analyzes the dynamic interactions between the source, grid, and load across different time dimensions, avoiding the risk of misjudgment or missed detection. Furthermore, an intelligent matching algorithm assigns tasks to nodes with sufficient resources and superior historical performance, reducing computational latency and ensuring timely warnings within seconds.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A method for evaluating the peak carrying capacity of a distribution network comprises the following steps: Reading an assessment request, where the assessment request is a distribution network peak carrying capacity assessment request issued when new energy output or load demand changes; The evaluation request is parsed to generate a set of carrying capacity evaluation parameters. The evaluation request is specifically: performing a multi-scale analysis of the changes in renewable energy output and load demand in the time dimension; and performing a spatial dimension analysis of the spatial topology of the distribution network. Calculating an evaluation value for each node in the distribution network based on an evaluation parameter set; selecting an optimal target node based on the evaluation value; and dynamically assigning evaluation tasks to the optimal target node; After obtaining the evaluation task, the optimal target node decomposes the task, performs task evaluation, and outputs the evaluation results.

[0006] Furthermore, the method also includes: generating a distribution network peak carrying capacity threshold value based on the evaluation result; and comparing the real-time operation data with the distribution network peak carrying capacity threshold value to generate early warning information.

[0007] Furthermore, based on the evaluation results, the process of generating the distribution network peak carrying capacity threshold includes: Construct a nonparametric kernel density estimate based on the evaluation results; ; in, For the point The probability density estimate at ; The sample size of the evaluation result data; is the bandwidth parameter; is the kernel function; For the historical assessment data points; ; in, is the peak carrying capacity threshold of the distribution network; is the complex power divergence; is the probability of system risk; Represents the upper limit of risk tolerance.

[0008] Furthermore, before performing multi-scale analysis of changes in new energy output and load demand in the time dimension, it also includes: setting a time granularity; the time granularity includes a second-level time granularity for analyzing instantaneous fluctuations in new energy output and a minute-level time granularity for analyzing sudden changes in load demand.

[0009] Furthermore, the process of parsing the assessment request and generating the carrying capacity assessment parameter set includes: Hilbert-Huang transform is used to extract the instantaneous fluctuation IMF component to obtain the time dimension subset Constructing ARIMA-GARCH joint model to predict load mutation and obtain spatial dimension subsets; The time dimension subset and space dimension subset are used to construct the evaluation parameter set of coupled new energy fluctuation characteristics.

[0010] Furthermore, the process of calculating the evaluation value of each node in the distribution network based on the evaluation parameter set includes: calculating the evaluation value of each node in the distribution network based on the evaluation parameter set and the real-time operating status, resource availability, and historical performance data of the target node in the distribution network; the evaluation value is: ; in, is the evaluation value, is the first dynamic adjustment coefficient, is the second dynamic adjustment coefficient, is the third dynamic adjustment coefficient; is the deviation rate between the current load of the target node and the historical average load; The percentage of assessment tasks that were not completed on time within 24 hours; The number of failures of the target node in the past 24 hours.

[0011] Furthermore, after dynamically allocating the evaluation task to the optimal target node, it also includes: associating the evaluation task, target node location information, time granularity and spatial granularity and storing them in the evaluation data storage device; specifically: storing the second-level data in slices according to timestamps, and storing the spatial topology data in a graph structure.

[0012] Furthermore, the method further comprises: When the optimal target node is evaluating a task, if the local resources of the optimal target node are insufficient, the evaluation task is split into multiple subtasks and assigned to other target nodes for processing; When other target nodes process subtasks, they exchange gradient information in an encrypted manner; After the subtasks are processed, the subtask execution results are aggregated to obtain the final evaluation results.

[0013] The present invention also proposes a distribution network peak carrying capacity evaluation system, which includes a reading module, an analysis module, a calculation module and an evaluation module; The reading module is used to read an evaluation request, which is a distribution network peak carrying capacity evaluation request issued when the new energy output or load demand changes; The parsing module is used to parse the evaluation request to generate a carrying capacity evaluation parameter set; the parsing evaluation request specifically includes: performing a time-dimensional multi-scale analysis of the changes in the output of new energy and load demand; and performing a spatial-dimensional analysis of the spatial topology of the distribution network; The calculation module is used to calculate the evaluation value of each node in the distribution network based on the evaluation parameter set; select the optimal target node according to the evaluation value; and dynamically assign the evaluation task to the optimal target node; The evaluation module is used to decompose the task and perform task evaluation after the optimal target node obtains the evaluation task and outputs the evaluation result.

[0014] Furthermore, it also includes an early warning module; The early warning module is used to generate a distribution network peak carrying capacity threshold according to the evaluation results; and compare the real-time operation data with the distribution network peak carrying capacity threshold to generate early warning information.

[0015] The effects provided in the summary of the invention are only the effects of the embodiments, not all the effects of the invention. One of the above technical solutions has the following advantages or beneficial effects: The present invention proposes a method and system for evaluating the peak carrying capacity of a distribution network. The method includes the following steps: reading an evaluation request, which is a distribution network peak carrying capacity evaluation request issued when the output of new energy or load demand changes; parsing the evaluation request to generate a carrying capacity evaluation parameter set; parsing the evaluation request specifically includes: performing a time dimension multi-scale analysis of the changes in the output of new energy and load demand; and performing a spatial dimension analysis of the spatial topology of the distribution network; calculating the evaluation value of each node in the distribution network based on the evaluation parameter set; selecting the optimal target node according to the evaluation value; dynamically allocating the evaluation task to the optimal target node; after the optimal target node obtains the evaluation task, it performs task decomposition, executes task evaluation, and outputs the evaluation result. Generate a distribution network peak carrying capacity threshold based on the evaluation result; compare the real-time operating data with the distribution network peak carrying capacity threshold to generate early warning information. Based on a method for evaluating the peak carrying capacity of a distribution network, a system for evaluating the peak carrying capacity of a distribution network is also proposed. The present invention can accurately analyze the dynamic interaction between source, network and load in different time dimensions, avoiding the risk of misjudgment or missed judgment. The intelligent matching algorithm allocates tasks to nodes with sufficient resources and better historical performance, reducing computing delays and ensuring the timeliness of second-level warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of a method for evaluating the peak carrying capacity of a distribution network proposed in Example 1 of the present invention; Figure 2 This is a schematic diagram of a distribution network peak carrying capacity assessment system proposed in Example 2 of the present invention. DETAILED DESCRIPTION

[0017] In order to clearly illustrate the technical features of this solution, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings. The disclosure below provides many different embodiments or examples for realizing different structures of the present invention. In order to simplify the disclosure of the present invention, the components and settings of specific examples are described below. In addition, the present invention may repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplicity and clarity and does not itself indicate the relationship between the various embodiments and / or settings discussed. It should be noted that the components illustrated in the accompanying drawings are not necessarily drawn to scale. The present invention omits descriptions of well-known components and processing technologies and processes to avoid unnecessary limitations on the present invention.

[0018] Example 1 Embodiment 1 of the present invention proposes a method for evaluating the peak carrying capacity of a distribution network, which is used to solve the technical problem of misalignment between the output of renewable energy and load demand in terms of temporal and spatial distribution. Figure 1 This is a flow chart of a method for evaluating the peak carrying capacity of a distribution network proposed in Example 1 of the present invention; In step S100, an evaluation request is read, where the evaluation request is a distribution network peak carrying capacity evaluation request issued when the output of new energy or load demand changes; The distribution network monitoring system receives evaluation requests from the distribution network monitoring system through an API interface or message queue. The distribution network monitoring system issues an evaluation request when it detects a significant change in renewable energy output (such as photovoltaic and wind power) or load demand. The evaluation request includes information on the spatiotemporal distribution of renewable energy output and load demand.

[0019] In step S110, the evaluation request is parsed to generate a set of carrying capacity evaluation parameters; the evaluation request is specifically parsed to: perform a time-dimensional multi-scale analysis of the changes in renewable energy output and load demand; and perform a spatial-dimensional analysis of the spatial topology of the distribution network; The present invention first sets different time granularities according to the type of renewable energy output (such as instantaneous fluctuations in photovoltaic output), and considers the spatial topology of the distribution network to generate an evaluation parameter set containing time and space dimensions.

[0020] For photovoltaic output, a time granularity of seconds is used to analyze instantaneous fluctuations in output data; for load demand, a time granularity of minutes is used to analyze its sudden changes. Furthermore, the spatial topology of the distribution network is considered to generate an evaluation parameter set that includes both temporal and spatial dimensions.

[0021] The process of parsing the assessment request and generating the carrying capacity assessment parameter set includes: using Hilbert-Huang transform to extract the instantaneous fluctuation IMF component to obtain the time dimension subset; constructing the ARIMA-GARCH joint model to predict load mutations to obtain the space dimension subset; using the time dimension subset and the space dimension subset to construct the assessment parameter set coupled with the new energy fluctuation characteristics.

[0022] When parsing an evaluation request, the system first sets a time granularity, such as seconds or minutes, and then allocates time-dependent parameters to subsets of the time dimension based on the granularity. Time granularity includes seconds and minutes, with seconds accounting for PV fluctuations and minutes accounting for sudden load changes. Simultaneously, a spatial granularity is set based on the distribution network topology, and spatial-dependent parameters are allocated to subsets of the spatial dimension. By constructing an evaluation parameter set that couples the fluctuation characteristics of renewable energy sources, the system can comprehensively characterize the dynamic coupling relationship between renewable energy output fluctuations and sudden load changes.

[0023] In step S120, an evaluation value of each node in the distribution network is calculated based on the evaluation parameter set; an optimal target node is selected according to the evaluation value; and an evaluation task is dynamically assigned to the optimal target node; An improved Q-learning algorithm is used to dynamically assign evaluation tasks to the optimal target node based on the target node's real-time operating status, resource availability, and historical performance data. Notification messages containing time and space granularity information are sent to the target node in real time. The evaluation value is: ; in, is the evaluation value, is the first dynamic adjustment coefficient, is the second dynamic adjustment coefficient, is the third dynamic adjustment coefficient; is the deviation rate between the current load of the target node and the historical average load; The percentage of assessment tasks that were not completed on time within 24 hours; The number of failures of the target node in the past 24 hours.

[0024] When assigning evaluation tasks, the system uses an improved Q-learning algorithm to calculate the evaluation value of each node based on the target node's real-time CPU usage, memory occupancy, and other operating status information, as well as the node's current computing resource availability, combined with performance data such as historical task processing speed and accuracy. The node with the highest evaluation value is selected as the optimal target node, and the evaluation task is assigned to it to ensure efficient task processing. Suppose there are three target nodes, A, B, and C. Node A currently has a low load, sufficient resources, and good historical performance; Node B has a high load and limited resources; and Node C has experienced a failure in the past 24 hours. Based on the improved Q-learning algorithm, the system will prioritize assigning the evaluation task to Node A because it has the best evaluation conditions.

[0025] In step S130, the evaluation task, target node location information, time granularity, and spatial granularity are associated and stored in the evaluation data storage device. Specifically, a hybrid engine of column storage and time slicing is used to store second-level data in slicing by timestamp, and spatial topology data is stored in a graph structure. Second-level data is sharded by timestamp and stored in Apache Parquet columnar storage; Spatial topology data is stored in the Neo4j database in the form of a graph structure; A spatiotemporal joint index relationship network is established to achieve second-level data retrieval through dual indexing of timestamp and node ID.

[0026] The system uses the Apache Parquet columnar storage format to accelerate data access. Furthermore, the Neo4j database is used for graph storage, supporting complex relational queries and analysis. Furthermore, by constructing a spatiotemporal joint index relationship network, it enables rapid data retrieval and association analysis, providing strong support for the rapid processing of evaluation tasks.

[0027] The joint indexing of columnar storage and graph databases further improves the efficiency of spatiotemporal data queries and supports real-time analysis of trillions of data.

[0028] In step S140 , the optimal target node obtains the evaluation task, decomposes the task, performs task evaluation, and outputs the evaluation result.

[0029] The target node obtains the evaluation task from the storage device according to the notification message, decomposes the task in parallel, processes it using computing resources, and outputs the evaluation result to the storage device; When notifying the target node, a notification message is sent containing time granularity (seconds and minutes) and spatial granularity (specific to the target distribution network area). Upon receiving the notification message, the target node retrieves the corresponding assessment task from the assessment data storage device based on the time and spatial granularity information in the message. This ensures that the target node can accurately and quickly obtain the assessment task it needs to process.

[0030] The notification message includes time granularity and space granularity information. The target node obtains the corresponding evaluation task from the evaluation data storage device according to the time granularity and space granularity information and the location information of the target node.

[0031] When the optimal target node is performing task evaluation, if the local resources of the optimal target node are insufficient, the evaluation task is split into multiple subtasks and assigned to other target nodes for processing; when other target nodes process subtasks, they exchange gradient information using Paillier homomorphic encryption.

[0032] After subtasks are processed, their execution results are aggregated to produce the final evaluation result. Through task splitting, encrypted exchange of gradient information, and model parameter aggregation, multi-node collaborative processing of evaluation tasks is achieved. When the target node processes an evaluation task and local resources are insufficient, the system splits the task into multiple subtasks and assigns them to multiple nodes for processing. During processing, each node exchanges gradient information in an encrypted manner to ensure data security. After processing is complete, the system aggregates the model parameters of each node to produce the final evaluation result. This fully utilizes the computing resources of multiple nodes and improves the processing efficiency of evaluation tasks.

[0033] Suppose target node A is running low on resources when processing an evaluation task. It splits the task into a local computation subtask and a global model update subtask. It then uses Paillier homomorphic encryption to exchange gradient information with other nodes. Finally, the system aggregates the model parameters of each node to generate a global evaluation result.

[0034] In step S150, a distribution network peak carrying capacity threshold is generated according to the evaluation result; and the real-time operation data is compared with the distribution network peak carrying capacity threshold to generate early warning information.

[0035] The target node determines in real time whether it has sufficient resources to handle the evaluation task: If resources are insufficient, the assessment task will be intelligently placed in the task queue and automatically processed when resources are available; If resources are sufficient, the evaluation task is carefully decomposed and processed in parallel using the powerful computing resources of the target nodes.

[0036] Based on the time granularity and space granularity information, the evaluation results are efficiently and accurately extracted from the evaluation data storage device, and combined with the preset evaluation rules and thresholds to generate targeted early warning information.

[0037] Based on the evaluation results, the process of generating the distribution network peak carrying capacity threshold includes: Construct a nonparametric kernel density estimate based on the evaluation results; ; in, For the point The probability density estimate at ; The sample size of the evaluation result data; is the bandwidth parameter; is the kernel function; For the historical assessment data points; ; in, is the peak carrying capacity threshold of the distribution network; is the complex power divergence; is the probability of system risk; Represents the upper limit of risk tolerance.

[0038] When generating thresholds, a probability distribution is calculated based on historical data and real-time fluctuations in renewable energy output. For example, the system can calculate the fluctuation range of renewable energy output over a period of time and combine it with real-time data to predict future output trends. Based on this information, the system dynamically adjusts the thresholds, ensuring that when real-time data exceeds the threshold, warning information can be triggered more accurately.

[0039] Embodiment 1 of the present invention proposes a method for evaluating the peak carrying capacity of a distribution network, which can accurately analyze the dynamic interaction between source, grid, and load in different time dimensions, avoid the risk of misjudgment or missed judgment, and the intelligent matching algorithm assigns tasks to nodes with sufficient resources and better historical performance, reducing computing delays and ensuring the timeliness of second-level warnings.

[0040] Example 1 of the present invention proposes a method for evaluating the peak carrying capacity of a distribution network. It uses task allocation to improve the load balancing of computing nodes, thereby greatly shortening the task response delay, and dynamically adjusts the threshold in conjunction with the probabilistic risk model. The false alarm rate and missed alarm rate are both reduced, thereby improving the resilience of the power grid.

[0041] Example 2 Based on the method for evaluating the peak carrying capacity of a distribution network proposed in Example 1 of the present invention, Example 2 of the present invention further proposes a system for evaluating the peak carrying capacity of a distribution network. Figure 2This is a schematic diagram of a distribution network peak carrying capacity assessment system proposed in Example 2 of the present invention; the system includes: a reading module, an analysis module, a calculation module, and an assessment module; The reading module is used to read an evaluation request, wherein the evaluation request is a distribution network peak carrying capacity evaluation request issued when the output of new energy or load demand changes; The parsing module is used to parse the evaluation request to generate a set of carrying capacity evaluation parameters. The parsing evaluation request specifically includes: performing a multi-scale analysis of the changes in renewable energy output and load demand in the time dimension; and performing a spatial dimension analysis of the spatial topology of the distribution network. The calculation module is used to calculate the evaluation value of each node in the distribution network based on the evaluation parameter set; select the optimal target node according to the evaluation value; and dynamically assign the evaluation task to the optimal target node; The evaluation module is used to obtain the evaluation task at the optimal target node, perform task decomposition, execute task evaluation, and output the evaluation results.

[0042] In the reading module of the present application, a time granularity is set; the time granularity includes a second-level time granularity for analyzing instantaneous fluctuations in new energy output and a minute-level time granularity for analyzing sudden changes in load demand.

[0043] In the parsing module, the process of parsing the assessment request and generating the carrying capacity assessment parameter set includes: Hilbert-Huang transform is used to extract the instantaneous fluctuation IMF component to obtain the time dimension subset Constructing ARIMA-GARCH joint model to predict load mutation and obtain spatial dimension subsets; The time dimension subset and space dimension subset are used to construct the evaluation parameter set of coupled new energy fluctuation characteristics.

[0044] In the calculation module, the process of calculating the evaluation value of each node in the distribution network based on the evaluation parameter set includes: calculating the evaluation value of each node in the distribution network based on the evaluation parameter set according to the real-time operating status, resource availability and historical performance data of the target node in the distribution network; the evaluation value is: ; in, is the evaluation value, is the first dynamic adjustment coefficient, is the second dynamic adjustment coefficient, is the third dynamic adjustment coefficient; is the deviation rate between the current load of the target node and the historical average load; The percentage of assessment tasks that were not completed on time within 24 hours; The number of failures of the target node in the past 24 hours.

[0045] In the evaluation module, the evaluation task, target node location information, time granularity and spatial granularity are associated and stored in the evaluation data storage device; specifically: second-level data is stored in shards according to timestamps, and spatial topology data is stored in a graph structure.

[0046] When the optimal target node is evaluating a task, if the local resources of the optimal target node are insufficient, the evaluation task is split into multiple subtasks and assigned to other target nodes for processing; When other target nodes process subtasks, they exchange gradient information in an encrypted manner; After the subtasks are processed, the subtask execution results are aggregated to obtain the final evaluation results.

[0047] The system also includes an early warning module; generates a distribution network peak carrying capacity threshold based on the evaluation results; and compares real-time operating data with the distribution network peak carrying capacity threshold to generate early warning information.

[0048] Based on the evaluation results, the process of generating the distribution network peak carrying capacity threshold includes: Construct a nonparametric kernel density estimate based on the evaluation results; ; in, For the point The probability density estimate at ; The sample size of the evaluation result data; is the bandwidth parameter; is the kernel function; For the historical assessment data points; ; in, is the peak carrying capacity threshold of the distribution network; is the complex power divergence; is the probability of system risk; Represents the upper limit of risk tolerance.

[0049] When generating thresholds, a probability distribution is calculated based on historical data and real-time fluctuations in renewable energy output. For example, the system can calculate the fluctuation range of renewable energy output over a period of time and combine it with real-time data to predict future output trends. Based on this information, the system dynamically adjusts the thresholds, ensuring that when real-time data exceeds the threshold, warning information can be triggered more accurately.

[0050] Embodiment 2 of the present invention proposes a distribution network peak carrying capacity assessment system, which can accurately analyze the dynamic interaction between source, grid and load in different time dimensions, avoid the risk of misjudgment or missed judgment, and the intelligent matching algorithm assigns tasks to nodes with sufficient resources and better historical performance, reducing computing delays and ensuring the timeliness of second-level warnings.

[0051] Example 2 of the present invention proposes a distribution network peak carrying capacity assessment system, which uses task allocation to improve the load balance of computing nodes, thereby greatly shortening the task response delay, and dynamically adjusts the threshold in conjunction with the probabilistic risk model, reducing the false alarm rate and missed alarm rate, thereby improving the resilience of the power grid.

[0052] The description of the relevant parts of the distribution network peak carrying capacity assessment system provided in Example 2 of the present application can be found in the detailed description of the corresponding parts of the distribution network peak carrying capacity assessment method provided in Example 1 of the present application, and will not be repeated here.

[0053] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device comprising a series of elements are inherent to the elements. In the absence of further restrictions, the elements limited by the statement "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device comprising the elements. In addition, the above-mentioned technical solutions provided in the embodiments of the present application are not described in detail in accordance with the corresponding technical solutions in the prior art to achieve the same principle, so as to avoid excessive elaboration.

[0054] Although the above description is of specific embodiments of the present invention in conjunction with the accompanying drawings, it does not limit the scope of protection of the present invention. For those skilled in the art, other different forms of modifications or variations can be made based on the above description. It is not necessary and impossible to list all embodiments here. Based on the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without expending creative effort are still within the scope of protection of the present invention.

Claims

1. A method for evaluating the peak carrying capacity of a distribution network, characterized in that: The following steps are involved: Reading an assessment request, where the assessment request is a distribution network peak carrying capacity assessment request issued when new energy output or load demand changes; The evaluation request is parsed to generate a set of carrying capacity evaluation parameters. The evaluation request is specifically: performing a multi-scale analysis of the changes in renewable energy output and load demand in the time dimension; and performing a spatial dimension analysis of the spatial topology of the distribution network. Calculate the evaluation value of each node in the distribution network based on the evaluation parameter set; Selecting an optimal target node according to the evaluation value; dynamically allocating evaluation tasks to the optimal target node; After obtaining the evaluation task, the optimal target node decomposes the task, performs task evaluation, and outputs the evaluation results.

2. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: The method further includes: generating a distribution network peak carrying capacity threshold value according to the evaluation result; and comparing the real-time operation data with the distribution network peak carrying capacity threshold value to generate early warning information.

3. The method for evaluating the peak carrying capacity of a distribution network according to claim 2, wherein: Based on the evaluation results, the process of generating the distribution network peak carrying capacity threshold includes: Construct a nonparametric kernel density estimate based on the evaluation results; ； in, For the point The probability density estimate at ; The sample size of the evaluation result data; is the bandwidth parameter; is the kernel function; For the historical assessment data points; ； in, is the peak carrying capacity threshold of the distribution network; is the complex power divergence; is the probability of system risk; Represents the upper limit of risk tolerance.

4. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: Before performing multi-scale analysis of changes in new energy output and load demand in the time dimension, the method also includes: setting a time granularity; the time granularity includes a second-level time granularity for analyzing instantaneous fluctuations in new energy output and a minute-level time granularity for analyzing sudden changes in load demand.

5. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: The process of parsing the assessment request and generating the carrying capacity assessment parameter set includes: Hilbert-Huang transform is used to extract the instantaneous fluctuation IMF component to obtain the time dimension subset Constructing ARIMA-GARCH joint model to predict load mutation and obtain spatial dimension subsets; The time dimension subset and space dimension subset are used to construct the evaluation parameter set of coupled new energy fluctuation characteristics.

6. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: The process of calculating the evaluation value of each node in the distribution network based on the evaluation parameter set includes: calculating the evaluation value of each node in the distribution network based on the evaluation parameter set according to the real-time operating status, resource availability and historical performance data of the target node in the distribution network; the evaluation value is: ； in, is the evaluation value, is the first dynamic adjustment coefficient, is the second dynamic adjustment coefficient, is the third dynamic adjustment coefficient; is the deviation rate between the current load of the target node and the historical average load; The percentage of assessment tasks that were not completed on time within 24 hours; The number of failures of the target node in the past 24 hours.

7. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: After dynamically allocating the evaluation task to the optimal target node, it also includes: associating the evaluation task, target node location information, time granularity and spatial granularity and storing them in the evaluation data storage device; specifically: storing the second-level data in slices according to timestamps, and storing the spatial topology data in a graph structure.

8. The method for evaluating the peak carrying capacity of a distribution network according to claim 1, wherein: The method further comprises: When the optimal target node is evaluating a task, if the local resources of the optimal target node are insufficient, the evaluation task is split into multiple subtasks and assigned to other target nodes for processing; When other target nodes process subtasks, they exchange gradient information in an encrypted manner; After the subtasks are processed, the subtask execution results are aggregated to obtain the final evaluation results.

9. A distribution network peak carrying capacity assessment system, characterized in that: It includes reading module, parsing module, calculation module and evaluation module; The reading module is used to read an evaluation request, which is a distribution network peak carrying capacity evaluation request issued when the new energy output or load demand changes; The parsing module is used to parse the evaluation request to generate a carrying capacity evaluation parameter set; the parsing evaluation request specifically includes: performing a time-dimensional multi-scale analysis of the changes in the output of new energy and load demand; and performing a spatial-dimensional analysis of the spatial topology of the distribution network; The calculation module is used to calculate the evaluation value of each node in the distribution network based on the evaluation parameter set; Selecting an optimal target node according to the evaluation value; dynamically allocating evaluation tasks to the optimal target node; The evaluation module is used to decompose the task and perform task evaluation after the optimal target node obtains the evaluation task and outputs the evaluation result.

10. The distribution network peak carrying capacity evaluation system according to claim 9, characterized in that: It also includes an early warning module; The early warning module is used to generate a distribution network peak carrying capacity threshold according to the evaluation results; and compare the real-time operation data with the distribution network peak carrying capacity threshold to generate early warning information.

Images