A distribution network carrying capacity assessment method based on power flow calculation

Abstract

The present application relates to the technical field of distribution network evaluation, and in particular to a distribution network carrying capacity evaluation method based on power flow calculation, including: data acquisition: obtaining distribution network historical operation data and real-time operation data; model construction: constructing a distribution network hybrid simulation model; model training: preprocessing historical operation data, extracting the first operation feature, and inputting it into the constructed model to obtain training results; detection: inputting the real-time operation data into the trained model to obtain detection results; carrying capacity evaluation: obtaining the constraints of the distribution network evaluation period, obtaining the training results of the same period, and recording them as input data; inputting the input data and constraints into the trained model, estimating the carrying capacity limit, inputting the detection results and input data into the model, estimating the carrying capacity; comparing the estimated carrying capacity with the carrying capacity limit, and determining whether to record or send an alarm signal. The present application has the effect of improving the accuracy of distribution network carrying capacity evaluation.

Description

Technical Field

[0001] The present application relates to the technical field of distribution network evaluation, and in particular to a distribution network carrying capacity evaluation method based on power flow calculation. Background Art

[0002] With the rapid development of the power system, the distribution network, as a key component of the power system, is crucial to ensuring power supply and meeting user needs. However, with the acceleration of urbanization and the continuous growth of power load, the distribution network is facing increasing pressure. In order to ensure the reliable operation of the distribution network, it is particularly important to accurately evaluate the carrying capacity of the distribution network.

[0003] Traditional distribution network capacity assessment methods mostly rely on empirical formulas and manual judgment. This method is not only time-consuming and labor-intensive, but also difficult to accurately reflect the actual operating status of the distribution network. In addition, due to the complexity of the distribution network structure and the diversity of operating data, traditional assessment methods are often difficult to comprehensively and accurately assess the distribution network's capacity.

[0004] In order to solve the above problems, in recent years, distribution network carrying capacity assessment methods based on power flow calculation have gradually attracted attention; power flow calculation is one of the basic tools for power system analysis, which can simulate the operating status of the power system and provide strong support for the carrying capacity assessment of the distribution network. However, most of the existing distribution network carrying capacity assessment methods based on power flow calculation only consider the steady-state operation of the distribution network, ignoring the dynamic characteristics of the distribution network, resulting in certain errors in the assessment results. Summary of the invention

[0005] In order to improve the accuracy of distribution network carrying capacity assessment, the present application provides a distribution network carrying capacity assessment method based on power flow calculation.

[0006] This application provides a distribution network carrying capacity assessment method based on power flow calculation, which adopts the following technical solutions:

[0007] A distribution network carrying capacity evaluation method based on power flow calculation includes the following steps:

[0008] Data acquisition: obtain historical and real-time operation data of the distribution network;

[0009] Model construction: A hybrid simulation model of the distribution network is built based on MATLAB and OpenDSS, which includes an upper intelligent algorithm layer, a middle power flow calculation layer, and a lower fault assessment layer. The three layers interact with each other through the COM interface of OpenDSS.

[0010] Model training: includes feature extraction step and training step;

[0011] Feature extraction: preprocess the historical operation data, and extract N first operation features from the preprocessed historical operation data;

[0012] Training: inputting N first operation characteristics into the distribution network hybrid simulation model to obtain training results;

[0013] Detection: Input the real-time operation data into the trained distribution network hybrid simulation model to obtain the detection results;

[0014] Load assessment: including constraint condition acquisition step, load estimation step, load determination step, recording step and alarm step;

[0015] Obtaining constraints: Obtain the constraints of the distribution network evaluation period, obtain the training results of the same period in the historical data, and record them as input data;

[0016] Load estimation: Based on the input data and constraints, the trained model is used to estimate the load and obtain the load capacity limit. Based on the test results and input data, the trained model is used to estimate the load capacity.

[0017] Load determination: compare the estimated load capacity with the load capacity limit to determine whether the estimated load capacity of the distribution network is less than or equal to the load capacity limit. If so, execute the recording step; otherwise, execute the alarm step;

[0018] Recording steps: Record the data and results of the bearing capacity assessment process;

[0019] Alarm step: trigger the alarm mechanism, send out an alarm signal, and output the results obtained in the detection step and the estimated carrying capacity.

[0020] By adopting the above technical solution, combining historical operation data with real-time operation data, and using the hybrid simulation model built by MATLAB and OpenDSS for training and detection, as well as the setting of fault layers, it is possible to more comprehensively simulate the actual operation status of the distribution network, integrate data analysis and model simulation methods, and combine the acquisition of constraints and the calculation of carrying capacity to improve the accuracy of carrying capacity assessment, making the assessment results closer to actual operation conditions.

[0021] The feature extraction step extracts key operating features from the preprocessed historical operating data. These features can reflect the operating laws and trends of the distribution network. These features are input into the trained model, and the model can learn the operating mode of the distribution network. Moreover, through the input of real-time operating data, the trained model can be corrected, so that it can make a more accurate estimate of future carrying capacity. This predictive capability helps to detect potential overload or failure risks in advance, and provide decision support for the operation and management of the distribution network.

[0022] This method can realize dynamic correction and evaluation of the distribution network's carrying capacity through the input of real-time operation data; as the operating status of the distribution network changes, the evaluation results will be updated accordingly to ensure the timeliness and accuracy of the evaluation; this dynamic evaluation mechanism helps to timely discover and deal with problems in the distribution network, and improve the operating efficiency and safety of the distribution network.

[0023] The entire assessment process includes multiple steps such as data acquisition, model building, model training, detection, and load assessment. Most of these steps can be automated through computer programs. This reduces the degree of human intervention and improves the efficiency and accuracy of the assessment. At the same time, the alarm step can automatically trigger the alarm mechanism and send out an alarm signal in a timely manner, which helps to quickly respond to and handle abnormal situations in the distribution network.

[0024] Through the load judgment step, the method can judge whether the estimated load capacity of the distribution network meets the requirements of the load capacity limit; if the estimated load capacity is less than or equal to the load capacity limit, it means that the current operating state of the distribution network is good and can continue to operate; if the estimated load capacity exceeds the load capacity limit, it is necessary to trigger the alarm mechanism and take corresponding measures; this judgment mechanism helps to optimize the resource allocation of the distribution network and ensure that the distribution network operates in a safe and economical state.

[0025] Optionally, an analysis step is further provided between the load determination step and the alarm step;

[0026] Analysis: including the cause acquisition step, the first judgment step, the grid connection adjustment step and the carrying capacity review step;

[0027] Reason acquisition: Based on the power grid data of the middle-level power flow calculation layer, analyze the reasons why the estimated carrying capacity exceeds the carrying capacity limit;

[0028] First judgment: determine whether the cause is an imbalance in power supply and demand. If so, perform grid connection adjustment steps;

[0029] Grid connection adjustment: Determine the number of grid connections based on the carrying capacity difference value, and connect new energy to the grid based on the number of grid connections;

[0030] Bearing capacity review: According to the data after grid connection, the real-time data is corrected and the bearing capacity is re-estimated to determine whether the estimated bearing capacity is less than or equal to the bearing capacity limit. If so, the recording step is executed, otherwise, the alarm step is executed.

[0031] By adopting the above technical solution, through the cause acquisition step, using the detailed power grid data provided by the middle-level power flow calculation layer, the specific cause leading to the estimated carrying capacity limit can be accurately analyzed. This ability to accurately locate the problem helps avoid blindly taking countermeasures and improves the efficiency and accuracy of problem solving.

[0032] In the first judgment step, by judging whether the cause is an imbalance in electricity supply and demand, the appropriate adjustment strategy can be selected intelligently; if it is judged to be an imbalance in electricity supply and demand, the grid connection adjustment step is executed, the number of grid connections is determined according to the carrying capacity difference value, and new energy is connected to the grid; this intelligent adjustment strategy can ensure the pertinence and effectiveness of the response measures, and avoid waste of resources and unnecessary adjustments.

[0033] After the grid connection adjustment, the real-time data is corrected and the load capacity is re-estimated according to the data after the grid connection through the load capacity review step; this step ensures that the effectiveness of the adjustment measures is verified, and also provides a basis for possible further adjustments in the future. This dynamic verification and adjustment mechanism makes the entire assessment method more flexible and adaptable.

[0034] Through the grid-connected adjustment steps, the number of new energy sources connected to the grid can be flexibly adjusted according to the actual needs of the power grid, thereby optimizing resource allocation; this can not only improve the power supply capacity and stability of the power grid, but also promote the full utilization of new energy, reduce dependence on traditional energy, and achieve the goal of energy conservation and emission reduction.

[0035] The introduction of the analysis step enables the entire evaluation method to promptly discover and address problems in power grid operation, thereby avoiding the accumulation and deterioration of problems. This helps to improve the efficiency and safety of power grid operation, ensure the stable operation of the power grid and the normal electricity demand of users.

[0036] Optionally, between the first determination step and the grid connection adjustment step, a grid connection evaluation step is also included;

[0037] Grid connection assessment: including model building steps, stability analysis steps, stability threshold setting steps, grid connection impact index calculation steps, grid connection impact assessment steps, impact amplitude judgment steps and grid connection decision-making steps;

[0038] Model building: Build a grid-connected simulation model based on MATLAB / Simulink or OpenDSS power simulation software;

[0039] Stability analysis: Use the built model to perform small disturbance stability analysis and large disturbance stability analysis, and output stability results;

[0040] Stability threshold setting: according to the relevant standards and specifications of power system stability, the stability threshold of each indicator is set;

[0041] Calculation of grid connection impact index: Based on the stability results, the stability thresholds of various indicators are compared, and the comprehensive stability index and grid balance index are calculated. Based on the above two indexes, the comprehensive stability index, the increase in carrying capacity, and the probability of failure or the growth trend of failures are calculated, and the grid connection feasibility index is calculated;

[0042] Grid connection impact assessment: Determine whether the feasibility index exceeds the set index threshold. If so, execute the grid connection decision step; otherwise, execute the alarm step;

[0043] Grid connection decision: Based on the grid connection impact assessment results, make grid connection execution decisions and execute grid connection adjustment steps.

[0044] By adopting the above technical solution, through the grid connection evaluation steps, using MATLAB / Simulink or OpenDSS and other power simulation software to build a grid connection simulation model, and conducting stability analysis, it is possible to scientifically evaluate the impact of new energy grid connection on grid stability. This simulation-based evaluation method can more accurately predict the operating status of the grid after grid connection, thereby improving the scientificity and accuracy of grid connection decisions.

[0045] The calculation steps of the grid-connected impact index comprehensively consider multiple factors such as the comprehensive stability index, the grid balance index, the increase in carrying capacity, the probability of failure or the failure growth trend. By calculating the grid-connected feasibility index, the comprehensive benefits of new energy grid connection can be comprehensively evaluated; this helps to optimize the grid-connected strategy and ensure that the grid after grid connection maintains good stability and security while improving its carrying capacity.

[0046] The grid-connection impact assessment step can timely discover potential risks that may exist in the grid-connection process by judging whether the feasibility index exceeds the set index threshold; once the assessment result does not meet the requirements, the alarm step is executed to issue a timely warning and take corresponding response measures to prevent the occurrence of potential risks.

[0047] The grid connection decision step makes a grid connection execution decision based on the grid connection impact assessment results, and executes the grid connection adjustment step. This flexible grid connection adjustment mechanism can flexibly adjust the number and method of grid connection of new energy sources according to the actual needs and operation of the grid, thereby improving the flexibility and adaptability of the grid.

[0048] Through scientific grid connection assessment and decision-making, it is possible to ensure that new energy is fully utilized during the grid connection process while maintaining the stability and security of the power grid. This will help promote the development and application of new energy, reduce dependence on traditional energy, and achieve optimization of energy structure and sustainable development.

[0049] The comprehensiveness and refinement of the grid connection assessment steps enable the entire assessment method to more accurately predict the grid operation status and carrying capacity changes. This helps to promptly discover and deal with problems in grid operation, enhance the reliability and stability of grid operation, and ensure the stable operation of the grid and the normal electricity demand of users.

[0050] Optionally, in the first judgment step, if the judgment reason is not an imbalance between power supply and demand, a power grid structure analysis step is performed;

[0051] Grid structure analysis: including a grid structure acquisition step, a grid structure index calculation step, a health index threshold setting step, a second judgment step, and a branch adjustment step;

[0052] Grid structure acquisition: Obtain grid structure layout information, including data on reasonableness of line layout, uniformity of transformer distribution, clarity of grid hierarchy, distribution network related data, and equipment aging;

[0053] Grid structure index calculation: according to the grid structure layout information, weights are assigned, and the grid structure index is calculated according to the product of the weights and the grid structure layout information;

[0054] Health index threshold setting: Set the health index threshold based on historical data;

[0055] Second judgment: judging whether the grid structure index is less than the health index threshold, if so, executing the branch adjustment step, otherwise, executing the alarm step;

[0056] Branch adjustment: Perform branch intervention, adjust the grid structure, and then re-execute the load estimation step.

[0057] By adopting the above technical solution, through the first judgment step, it is first quickly determined whether the power grid problem is caused by an imbalance in power supply and demand; if not, the power grid structure analysis is performed instead. This step helps to accurately identify the root cause of the power grid operation problem, avoids blind investigation, and improves the efficiency of problem solving.

[0058] The grid structure acquisition steps cover multiple dimensions such as line layout, transformer distribution, grid hierarchy clarity, distribution network correlation and equipment aging, ensuring a comprehensive assessment of the health status of the grid structure; this multi-dimensional analysis helps to discover potential problems in the grid structure and provide a basis for subsequent adjustments.

[0059] The grid structure index calculation steps quantify the complex grid structure information into an easily comparable index by assigning weights and calculating products; this method not only takes into account the differences in the importance of various factors, but also simplifies the problem-solving process and improves the accuracy and efficiency of the analysis.

[0060] The health index threshold setting and the second judgment step together constitute an intelligent early warning system that can promptly detect unhealthy conditions in the power grid structure. Once the power grid structure index is lower than the health threshold, the branch adjustment step is triggered to improve the operating status by adjusting the power grid structure. This automated adjustment mechanism helps reduce manual intervention and improve the stability and reliability of power grid operation.

[0061] After the branch is adjusted, the load estimation step is re-executed to form a closed-loop optimization process; this means that after the grid structure is adjusted, its operating status will be re-evaluated to ensure the effectiveness of the adjustment; this cyclic optimization mechanism helps to continuously improve the grid structure and enhance the overall performance of the grid.

[0062] Optionally, after the health index threshold setting step, a dynamic adjustment step is also included;

[0063] Dynamic adjustment: Obtain the complexity of the distribution network area and the degree of fluctuation in grid usage, adjust the coefficient of the health index threshold, and increase the health index threshold.

[0064] By adopting the above technical solution, the dynamic adjustment step takes into account the complexity of the distribution network area and the fluctuation of the power grid usage, both of which directly affect the health status and stability of the power grid structure; by obtaining this information and adjusting the coefficient of the health index threshold, the health index threshold is raised, making the judgment standard more in line with the actual situation, thereby enhancing the adaptability of the technical solution.

[0065] Due to the complexity of the power grid and the volatility of its usage, a fixed health index threshold may not accurately reflect the true health status of the power grid structure. The dynamic adjustment step enables the early warning system to more accurately identify potential problems in the power grid structure by adjusting the health index threshold in real time or regularly, thus avoiding false alarms or missed alarms caused by improper threshold settings.

[0066] When the grid structure index is close to but slightly higher than the health index threshold, if the actual operating state of the grid has become unstable due to complexity and volatility, the fixed threshold may cause the adjustment opportunity to be missed; the dynamic adjustment step increases the sensitivity of the early warning system by raising the threshold, which helps to take intervention measures when the grid structure problems are not serious, thereby reducing the risk of grid failures caused by misoperation.

[0067] The dynamic adjustment step also means that in areas where the grid structure is simpler and usage fluctuations are smaller, the adjustment frequency or intensity can be appropriately reduced, thereby optimizing resource allocation; when resources are limited, this optimization helps to invest more resources in areas that need more attention and adjustment, thereby improving overall operation and maintenance efficiency.

[0068] By dynamically adjusting the health index threshold, this technical solution can better cope with the uncertainties and challenges in grid operation, and improve the resilience and self-recovery ability of the grid. When faced with emergencies or abnormal fluctuations, the grid can respond more quickly, reducing the probability and scope of failure.

[0069] Optionally, between the second judgment step and the alarm step, there is also an equipment status assessment step;

[0070] Equipment status assessment: including branch calling steps, branch activation analysis steps, load-bearing capacity estimation correction steps, aging degree determination steps, aging trend prediction steps, load-bearing capacity estimation adjustment steps and load-bearing capacity re-judgment steps;

[0071] Branch call: switch and replace the spare branch to the original use line or adjust the operation mode of the existing branch;

[0072] Branch activation analysis: observe the changes in the grid operation status to indirectly determine whether key equipment in the grid is aging or has performance degradation;

[0073] Carrying capacity estimation correction: Based on the grid operation status data after the branch is activated, the carrying capacity estimation is re-estimated, the carrying capacity estimation values ​​before and after activation are compared, the impact of equipment status changes on the carrying capacity estimation is analyzed, and the carrying capacity estimation results are corrected accordingly;

[0074] Determination of aging degree: Combine the historical operation records, maintenance records and current operation status data of the equipment, and use data analysis technology to determine the aging degree of the equipment. The aging degree is determined based on a comprehensive analysis of the equipment's performance degradation curve and failure rate increase trend parameters;

[0075] Aging trend prediction: Based on the current aging degree of the equipment, historical usage and the aging rules of similar equipment, the prediction model is used to predict the future aging trend of the equipment;

[0076] Adjustment of load-bearing capacity estimate: Based on the prediction results of equipment aging degree and aging trend, further adjustment is made to the revised load-bearing capacity estimate results;

[0077] Bearing capacity re-judgment: Determine whether the estimated bearing capacity after adjustment is less than or equal to the bearing capacity limit. If so, execute the recording step, otherwise, execute the alarm step.

[0078] By adopting the above technical solution, through the branch call and branch activation analysis steps, it is possible to observe the changes in the power grid operation status, thereby indirectly judging whether the key equipment in the power grid is aging or performance degradation. This method can detect potential problems in time before equipment failure occurs, avoiding power grid failures caused by sudden equipment failure.

[0079] The capacity estimation correction step re-estimates the capacity based on the grid operation status data after the branch is activated, and compares the estimated capacity values ​​before and after activation, so as to analyze the impact of equipment status changes on the capacity estimation, and correct the capacity estimation results accordingly; this helps to improve the accuracy of the capacity estimation and provide a more reliable basis for the dispatch and operation of the grid.

[0080] The aging degree determination step combines the equipment's historical operation records, maintenance records, and current operating status data, and uses data analysis technology to determine the equipment's aging degree. This method takes into account multiple factors, such as the equipment's performance degradation curve and failure rate increase trend parameters, and can more scientifically evaluate the equipment's aging status, providing more accurate guidance for equipment maintenance and replacement.

[0081] The aging trend prediction step uses a prediction model to predict the future aging trend of the equipment based on the current aging degree of the equipment, historical usage, and aging patterns of similar equipment. This helps to plan equipment maintenance and replacement plans in advance, avoid power grid failures caused by equipment aging, and improve the stability and reliability of the power grid.

[0082] Based on the predicted results of the equipment aging degree and aging trend, the capacity estimation adjustment step further adjusts the revised capacity estimation results; this helps to more accurately evaluate the grid's carrying capacity, provide more refined management for the grid's dispatch and operation, and ensure that the grid operates in a safe and reliable state.

[0083] Optionally, a data updating step is further provided after the aging degree determining step;

[0084] Data update: associate the estimated equipment aging degree with the real-time data and perform data simulation to form simulated data, divide the simulated data into m second operating characteristics, and count the second operating characteristics into the first operating characteristics and input them into the training step.

[0085] By adopting the above technical solution, the data update step associates the estimated equipment aging degree with the real-time data, and performs data simulation to form simulated data; this step ensures that the equipment aging assessment does not only rely on historical data and static analysis, but can reflect the current status of the equipment in real time; through the generation of simulated data, a more comprehensive understanding of the operating performance of the equipment at different aging degrees can be achieved.

[0086] The simulation data is divided into m second operating features, which are counted into the original first operating features and then input into the training step. This approach expands the feature set and provides a richer data foundation for machine learning or predictive models. By continuously introducing new operating features, the model can learn more about the relationship between equipment aging and performance changes, thereby improving the accuracy of prediction and evaluation.

[0087] The data update step actually constitutes a continuous learning mechanism. With the continuous introduction of new data and continuous training of the model, the equipment aging assessment model can gradually adapt to changes in grid operating conditions and improve its ability to respond to new situations and new problems. This iterative update method helps to maintain the timeliness and accuracy of the model and ensure the effectiveness of equipment status assessment.

[0088] Through real-time data updates and feature expansion, the equipment status assessment model can provide more accurate and comprehensive equipment status information; this means better decision support for power grid managers; based on real-time equipment status information, managers can formulate maintenance plans and adjust operation strategies more scientifically, thereby improving the overall performance and reliability of the power grid.

[0089] Through continuous equipment status assessment and model optimization, the power grid can better cope with the challenges brought by equipment aging; timely detection and taking measures before equipment failure occurs will help reduce power grid failures caused by equipment failure and enhance the resilience and self-recovery capabilities of the power grid.

[0090] Optionally, a data conversion step is provided between the detection step and the constraint condition acquisition step;

[0091] The data conversion step includes: a modeling step, a prediction step, a second detection step, a feedback step, a trend acquisition step, a third judgment step and a fallback step;

[0092] Modeling: Establish a Bayesian network model, where the first-level nodes of the Bayesian network model are historical operation data, and the directed edges of the Bayesian network model are the conversion probabilities between various historical operation data;

[0093] Prediction: Input the real-time operation data into the Bayesian network model to obtain the operation data at the next moment, which is recorded as the estimated operation data;

[0094] Second detection: input the estimated operation data into the trained distribution network hybrid simulation model, obtain the detection result at the next moment, record it as the auxiliary detection result, and perform the feedback step;

[0095] Feedback: Adjust the lower fault assessment layer in the hybrid simulation model of the distribution network based on real-time data and estimated operating data;

[0096] Trend acquisition: based on the test results obtained in the test steps and the auxiliary test results, obtain the change trend of the test results;

[0097] Third judgment: judging whether the change trend is in line with expectations, if so, executing the constraint condition acquisition step; if not, executing the fallback step;

[0098] Rollback: Roll back adjustments made in the feedback step.

[0099] By adopting the above technical solution, the Bayesian network model established through the modeling step can use the conversion probability relationship between historical operation data to intelligently predict the real-time operation data and obtain the estimated operation data at the next moment. This prediction capability provides a more accurate data basis for subsequent load assessment and helps to discover potential operation problems in advance.

[0100] The second detection step inputs the estimated operation data into the trained distribution network hybrid simulation model to obtain the auxiliary detection results at the next moment; this step makes full use of the advantages of the simulation model in simulating complex power grid operation scenarios, and combined with the estimated operation data, it can further improve the accuracy and reliability of the simulation.

[0101] The feedback step adjusts the lower fault assessment layer in the distribution network hybrid simulation model according to real-time data and estimated operating data, realizing the adaptive learning of the model; this adjustment mechanism helps the model better adapt to changes in the grid operating status and improves the accuracy of the assessment.

[0102] The trend acquisition step obtains the changing trend of the detection results based on the detection results and auxiliary detection results; the third judgment step decides whether to execute the load assessment step based on whether this changing trend is in line with expectations; this trend analysis and judgment mechanism helps to timely discover abnormal changes in the operating status of the power grid and provide a basis for subsequent decision-making.

[0103] The rollback step rolls back the adjustments made in the feedback step, ensuring that when the trend does not meet expectations, the evaluation process can stably roll back to the previous state, avoiding deviations in subsequent evaluation results due to incorrect adjustments; this rollback mechanism enhances the robustness and stability of the evaluation process.

[0104] Optionally, before the prediction step, a data adjustment step is also included;

[0105] Data adjustment: including aging collection steps and model optimization steps;

[0106] Aging collection: Obtain estimated aging data based on the future aging trend of the equipment in the aging trend step;

[0107] Model optimization: Add secondary nodes and associated nodes to the Bayesian network model, the secondary nodes are estimated aging data, the associated nodes are historical fault data, and directed edges from the primary nodes to the secondary nodes, directed edges between the primary nodes and the associated nodes, and directed edges between the secondary nodes and the associated nodes are established to obtain an optimized Bayesian network model;

[0108] In the prediction step, the real-time operation data and simulation data are input into the optimized Bayesian network model together to obtain the operation data at the next moment as the new estimated operation data.

[0109] By adopting the above technical solution, the estimated aging data obtained through the aging collection step can reflect the future aging status of the equipment. This data is introduced into the Bayesian network model as a secondary node. This introduction method enables the model to more comprehensively consider the impact of equipment aging on the operating status of the power grid when predicting the operating data at the next moment, thereby improving the accuracy and reliability of the prediction.

[0110] The model optimization step adds secondary nodes and associated nodes to the Bayesian network model, and establishes corresponding directed edges to obtain the optimized Bayesian network model; this optimization method enables the model to better adapt to changes in the operating status of the power grid, especially the impact of the important factor of equipment aging. By continuously optimizing the model structure, the model's sensitivity and response speed to changes in the operating status of the power grid can be improved.

[0111] In the prediction step, the real-time operation data and simulation data are input into the optimized Bayesian network model to obtain the operation data at the next moment. This data fusion method can make full use of the advantages of real-time data and simulation data to improve the accuracy and comprehensiveness of the prediction results. At the same time, by continuously introducing new data, the prediction ability of the model can be further improved.

[0112] The combination of data adjustment and prediction steps provides grid managers with more accurate and comprehensive grid operation status prediction information. Based on this prediction information, managers can make more scientific decisions such as maintenance plans and adjustment of operation strategies, thereby improving the overall performance and reliability of the grid.

[0113] By considering the impact of equipment aging trends on the operating status of the power grid and optimizing the Bayesian network model for prediction, power grid managers can better identify potential operating risks; this improvement in risk management capabilities helps to promptly discover and address safety hazards in the power grid and avoid or reduce power grid failures caused by equipment aging.

[0114] In summary, the present application includes at least one of the following beneficial technical effects:

[0115] 1. By combining historical operation data and real-time operation data, and using the hybrid simulation model built by MATLAB and OpenDSS for training and testing, as well as the setting of fault layers, the actual operation status of the distribution network can be simulated more comprehensively. The comprehensive data analysis and model simulation methods, combined with the acquisition of constraints and the calculation of carrying capacity, improve the accuracy of carrying capacity assessment, making the assessment results closer to the actual operation situation; the feature extraction step extracts key operating features from the pre-processed historical operation data. These features can reflect the operating laws and trends of the distribution network; these features are input into the trained model, and the model can learn the operation mode of the distribution network. Through the input of real-time operation data, the trained model can be corrected, so that it can make a more accurate estimate of future carrying capacity; this prediction capability helps to discover potential overload or fault risks in advance and provide decision support for the operation and management of the distribution network.

[0116] 2. By considering the impact of equipment aging trends on the operating status of the power grid and optimizing the Bayesian network model for prediction, power grid managers can better identify potential operating risks; this improvement in risk management capabilities helps to promptly discover and address safety hazards in the power grid and avoid or reduce power grid failures caused by equipment aging. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] Figure 1 is a flow chart of the evaluation method in Example 1 of the present application;

[0118] Figure 2 is a sub-step flow chart of the analysis step in Example 2 of the present application;

[0119] Figure 3 is a sub-step flow chart of the grid connection assessment step in Example 2 of the present application;

[0120] Figure 4 It is a sub-step flow chart of the device status evaluation step in Example 2 of the present application;

[0121] Figure 5 It is a sub-step flowchart of the data adjustment step in Example 2 of the present application. DETAILED DESCRIPTION

[0122] The following combination Figures 1 to 5 This application is described in further detail.

[0123] This embodiment discloses a distribution network carrying capacity evaluation method based on power flow calculation.

[0124] Example 1: Reference Figure 1 ,The distribution network carrying capacity evaluation method based on power flow calculation includes the following steps:

[0125] Data acquisition: obtain historical and real-time operation data of the distribution network;

[0126] Specifically, the Supervisory Control and Data Acquisition (SCADA) system is used to obtain the historical and real-time operation data of the distribution network. The historical operation data includes past load, voltage, current and other parameters as well as existing fault data, while the real-time operation data includes the current load, voltage, current and other parameters as well as existing equipment parameters. The data should be exported from the SCADA system regularly and stored in the database for subsequent analysis and processing.

[0127] Model construction: A hybrid simulation model of the distribution network is built based on MATLAB and OpenDSS, which includes an upper intelligent algorithm layer, a middle power flow calculation layer, and a lower fault assessment layer. The three layers interact with each other through the COM interface of OpenDSS.

[0128] Specifically, MATLAB is used for programming and simulation. MATLAB has powerful numerical calculation and simulation capabilities and is suitable for building complex distribution network models. OpenDSS is an open source distribution system simulation software used to build a detailed model of the distribution network. Through the COM interface of OpenDSS, MATLAB can interact with OpenDSS for data to achieve hybrid simulation. A hybrid simulation model is built that includes an upper intelligent algorithm layer, a middle power flow calculation layer, and a lower fault assessment layer. The upper intelligent algorithm layer is used to handle advanced analysis tasks such as optimization and decision support; the middle power flow calculation layer is used to perform power flow calculations on the distribution network; and the lower fault assessment layer is used to assess the fault conditions of the distribution network.

[0129] Model training: includes feature extraction step and training step;

[0130] Feature extraction: preprocess the historical operation data, and extract N first operation features from the preprocessed historical operation data;

[0131] Training: inputting N first operation characteristics into the distribution network hybrid simulation model to obtain training results;

[0132] Specifically, use programming languages ​​such as Python to perform data preprocessing, such as data cleaning and data transformation; extract n first operating features from the preprocessed historical operating data, which may be load peak, voltage fluctuation range, current average, etc.; input the extracted features into a hybrid simulation model constructed by MATLAB and OpenDSS; and obtain the training results of the model through simulation operation, which can be used for subsequent detection and load evaluation.

[0133] Detection: Input the real-time operation data into the trained distribution network hybrid simulation model to obtain the detection results;

[0134] Specifically, real-time operating data is input into the trained hybrid simulation model; the model performs analysis and calculation based on the real-time data and outputs detection results, which reflect the current status of the distribution network.

[0135] Load assessment: including constraint condition acquisition step, load estimation step, load determination step, recording step and alarm step;

[0136] Obtaining constraints: Obtain the constraints of the distribution network evaluation period, obtain the training results of the same period in the historical data, and record them as input data;

[0137] Load estimation: Based on the input data and constraints, the trained distribution network hybrid simulation model is used to estimate the load and obtain the load capacity limit. Based on the test results and input data, the trained model is used to estimate the load capacity.

[0138] Specifically, input data: this includes historical operating data, real-time operating data, and constraints, etc.; model parameters: these parameters are learned by the model during the training process and are used to describe the characteristics and behavior of the distribution network; output results: that is, the estimated capacity, which is calculated by the model based on the input data and constraints.

[0139] For example, a simplified load assessment formula may be as follows: ;

[0140] is the estimated carrying capacity, is the input data, including historical operation data, real-time operation data, etc.; is the model parameter; is the constraint condition, such as load limit, voltage limit, current limit, etc.; the model parameter can be set according to the actual situation of the power grid, and the constraint condition can be obtained according to the power grid architecture and power grid setting rules;

[0141] F represents a function. In this embodiment, a linear regression model is selected as the load assessment function. The updated load assessment formula is: ;

[0142] in , is the intercept term,

[0143] , , …, is the input data ( ) coefficient.

[0144] , , …, are the model parameters ( ) coefficient.

[0145] , , …, is a constraint condition ( ) coefficient.

[0146] , , …, It is a specific input data item.

[0147] , , …, is a specific model parameter item.

[0148] , , …, It is a specific constraint item; each input item (data, parameter, constraint) has a corresponding coefficient, which is learned through the model training process and describes the influence of each input item on the estimated carrying capacity.

[0149] For example, suppose we have a simple distribution network model, and we have obtained historical operation data and real-time operation data from the database. In addition, we have also set some constraints, such as maximum load limit, voltage fluctuation range, etc.

[0150] Input data: Historical operation data: including past load, voltage, current and other parameters;

[0151] Real-time operation data: load, voltage, current and other parameters at the current moment;

[0152] Constraints: Maximum load does not exceed 1000kW, voltage fluctuation range is within ±5%;

[0153] Model training: Use historical operation data to train the model and learn the characteristics and behaviors of the distribution network;

[0154] Load estimation: Input real-time operation data and constraints into the trained model;

[0155] The model performs calculations based on the input data and constraints to produce an estimated carrying capacity.

[0156] For example, the model may calculate that under the current real-time operating data and constraints, the estimated carrying capacity of the distribution network is 800kW, which means that the distribution network can operate safely and stably without exceeding this load.

[0157] Load determination: compare the estimated load capacity with the load capacity limit to determine whether the estimated load capacity of the distribution network is less than or equal to the load capacity limit. If so, execute the recording step; otherwise, execute the alarm step;

[0158] Recording steps: Record the data and results of the bearing capacity assessment process;

[0159] Alarm step: trigger the alarm mechanism, send out an alarm signal, and output the results obtained in the detection step and the estimated carrying capacity.

[0160] Specifically, the estimated bearing capacity is compared with the bearing capacity limit; if the estimated bearing capacity is less than or equal to the bearing capacity limit, the recording step is executed; if the estimated bearing capacity is greater than the bearing capacity limit, the alarm step is executed; the data and results of the bearing capacity assessment process are recorded using a database or file system; these data can be used for subsequent analysis and decision support.

[0161] If the estimated carrying capacity exceeds the carrying capacity limit, the alarm mechanism is triggered; the alarm signal can be sent through email, text message, sound, etc.; at the same time, the results obtained from the detection steps and the estimated carrying capacity are output so that the operation and maintenance personnel can take timely measures.

[0162] The load capacity limit is obtained through a series of analyses and calculations based on the physical characteristics, operating parameters and safety standards of the distribution network. Specifically, the ways to obtain the load capacity limit include: Statistical analysis based on historical data: By analyzing the historical operating data of the distribution network, key parameters such as the average load, maximum load, and load fluctuation range of the distribution network can be obtained. These parameters can be used as a reference for formulating the load capacity limit.

[0163] Calculation based on physical models: Using the physical model of the distribution network (such as power flow model, short-circuit current model, etc.), combined with the topological structure and equipment parameters of the distribution network, the bearing capacity limit of the distribution network under different working conditions can be calculated.

[0164] Setting based on safety standards: The load capacity limit of the distribution network can be set according to relevant national and industry safety standards, such as load limit, voltage fluctuation range, current limit, etc. These standards take into account the safety and stability requirements of the distribution network.

[0165] This embodiment is preferably obtained based on statistical analysis of historical data, and after the bearing capacity value is obtained, it is lowered by 5%.

[0166] For example, historical operation data:

[0167] Load peak (past month): 800kW, 850kW, 900kW, …, 1000kW (30 days of data in total);

[0168] Voltage fluctuation range (past month): ±3%, ±2.5%, ±2%, ..., ±1% (30 days of data in total);

[0169] Other parameters (such as current, power factor, etc.) also have 30 days of historical data;

[0170] Real-time operation data: current load: 950kW; current voltage fluctuation range: ±2%; other real-time parameters (such as current current, current power factor, etc.);

[0171] Constraints: Maximum load limit: 1000kW, after a 5% reduction, the load capacity limit is 950kW; Voltage fluctuation range limit: ±5%

[0172] Other constraints (such as current limit, power factor limit, etc.);

[0173] Model training: Use MATLAB and OpenDSS to build a hybrid simulation model; extract features such as load peak and voltage fluctuation range from preprocessed historical operating data and input these features into the model; the model training process involves multiple iterations to minimize the prediction error by adjusting the model parameters.

[0174] Load estimation: The real-time operating data is input into the trained model. The model calculates the estimated load capacity based on the input data and constraints. In this example, the model estimates the load capacity to be 980kW.

[0175] Comparison process: The estimated carrying capacity (980kW) is compared with the carrying capacity limit (950kW); since the estimated carrying capacity is greater than the carrying capacity limit, the alarm mechanism needs to be triggered.

[0176] If the estimated bearing capacity is greater than the bearing capacity limit, the data and results of the bearing capacity assessment process shall be recorded, including real-time operating data, estimated bearing capacity, bearing capacity limit, etc.

[0177] Embodiment 1 of the present application constructs a lower fault layer through a simulation model, which is convenient for analyzing distribution network faults and truly reflecting the accuracy of data. It also relies on the estimated calculation of the carrying capacity to analyze the carrying capacity and reduce the actual demand for excessive carrying capacity, which does not match the distribution network and causes distribution network faults.

[0178] Example 2: Reference Figure 2 , the difference between this embodiment and embodiment 1 is that an analysis step is provided between the load determination step and the alarm step;

[0179] Analysis: including the cause acquisition step, the first judgment step, the grid connection adjustment step and the carrying capacity review step;

[0180] Reason acquisition: Based on the power grid data of the middle-level power flow calculation layer, analyze the reasons why the estimated carrying capacity exceeds the carrying capacity limit;

[0181] Specifically, this step can be implemented by adding a cause analysis layer in the above-mentioned distribution network hybrid simulation model, and after inputting the data calculated by the middle-level power flow calculation layer, the cause can be obtained through simulation; or with the help of professional power grid data analysis tools, such as power system simulation software (such as Simulink, PSCAD, ETAP, etc. in MATLAB) or distribution network management system (DMS); first, the power grid data provided by the middle-level power flow calculation layer is imported into the analysis tool; then, the power flow calculation module or fault diagnosis module in the analysis tool is used to conduct an in-depth analysis of the power grid data to find out the specific cause of the estimated carrying capacity limit;

[0182] First judgment: determine whether the cause is an imbalance in power supply and demand. If so, perform grid connection adjustment steps;

[0183] Specifically, in the first judgment step, it is necessary to judge whether the cause of the carrying capacity limit is the imbalance of power supply and demand based on the analysis results of the cause acquisition step. This is achieved by comparing the supply and demand data of the power grid. If the supply and demand data differ greatly and the imbalance of supply and demand persists in multiple time periods, it can be judged that there is an imbalance of power supply and demand;

[0184] For example, by comparing the supply and demand data during peak hours in the region, it was found that the imbalance between supply and demand does exist, and the insufficient output of renewable energy power generation is the main reason for the imbalance.

[0185] Grid connection adjustment: Determine the number of grid connections based on the carrying capacity difference value, and connect new energy to the grid based on the number of grid connections;

[0186] Specifically, in the grid connection adjustment step, the number of grid connections needs to be determined based on the carrying capacity difference value (i.e. the difference between the estimated carrying capacity and the actual demand); this requires considering the output characteristics of new energy power generation, the acceptance capacity of the power grid, the economic feasibility of grid connection and the number of distribution networks that can actually be connected to the grid, and then, based on the determined number of grid connections, the new energy grid connection operation is carried out. This can be achieved by negotiating with new energy power generation companies, signing grid connection agreements, adjusting power grid dispatching strategies, etc.

[0187] Carrying capacity review: Based on the data after grid connection, the real-time data is corrected, the carrying capacity is re-estimated, and then the carrying capacity judgment step is performed.

[0188] Specifically, in the carrying capacity review step, the real-time data needs to be corrected according to the data after grid connection, and the carrying capacity needs to be re-estimated; this can be achieved by recalculating the flow, updating the grid model, etc.; then, the re-estimated carrying capacity is compared with the carrying capacity limit to determine whether the grid is still in a safe operating state.

[0189] The model has clear hierarchy and complementary functions: the upper intelligent algorithm layer is responsible for advanced analysis and decision-making, such as optimization algorithms and prediction models; the middle power flow calculation layer performs core power flow calculations and simulates the operation status of the power grid, which is the basis for carrying capacity assessment; the lower fault assessment layer simulates and evaluates power grid faults and provides power grid performance data under fault conditions.

[0190] Data interaction and collaborative work: Through the COM interface of OpenDSS, the three-layer model can achieve efficient data interaction and collaborative work, ensuring the accuracy and real-time nature of the evaluation results.

[0191] Through feature extraction and model training, the operation rules and characteristics of the power grid are learned from historical operation data; the real-time operation data is input into the trained model to make real-time estimation of the carrying capacity; the carrying capacity limit is calculated according to the constraints of the distribution network evaluation period (such as load demand, power grid structure, etc.) combined with the training results of the same period in the historical data; in the carrying judgment step, if the estimated carrying capacity exceeds the carrying capacity limit, the reason for the excessive carrying capacity is accurately located through the power grid data of the middle-level flow calculation layer; if the reason is determined to be an imbalance in power supply and demand, the number of grid connections is determined according to the carrying capacity difference value, and new energy grid connection adjustments are made to balance the supply and demand of the power grid; after the grid connection adjustment, the real-time acquired data is corrected and the carrying capacity is re-estimated to ensure that the power grid operates within a safe and economical range; through continuous carrying capacity review and optimization adjustment, dynamic management and continuous optimization of the power grid carrying capacity are achieved.

[0192] The setting and collaborative work of the three-layer simulation model can comprehensively and accurately simulate the operating status of the power grid and improve the accuracy of the carrying capacity assessment; the comprehensive evaluation based on historical data and real-time data, as well as the calculation of the carrying capacity limit considering the constraints, provides a scientific basis for power grid dispatch and operation decisions; through targeted grid connection adjustments and carrying capacity reviews, it can quickly respond to changes in power grid supply and demand and improve power grid operation efficiency and stability; the entire evaluation process realizes data-driven and intelligent management, which improves the automation level and response speed of power grid management.

[0193] Reference Figure 3 , in other embodiments, between the first determination step and the grid connection adjustment step, a grid connection evaluation step is also included;

[0194] Grid connection assessment: including model building steps, stability analysis steps, stability threshold setting steps, grid connection impact index calculation steps, grid connection impact assessment steps, impact amplitude judgment steps and grid connection decision-making steps;

[0195] Model building: Build a grid-connected simulation model based on MATLAB / Simulink or OpenDSS power simulation software;

[0196] Specifically, the tool selection is: based on MATLAB / Simulink or OpenDSS power simulation software; according to the actual structure and parameters of the grid-connected system, build the corresponding grid-connected simulation model in MATLAB / Simulink or OpenDSS; set the parameters of the model, including the parameters of power supply, load, line, transformer and other components; verify the model to ensure the accuracy and reliability of the model. Among them, the setting of parameters is based on the collection of relevant information of the actual power grid, including power grid structure, equipment parameters, operation data, etc., and the list of parameters that need to be set is sorted out according to the information; according to the actual situation of the power grid and the simulation requirements, the reasonable value range of each parameter is determined; for uncertain parameters, relevant standards, specifications or empirical values ​​can be referred to for estimation; the value of each parameter is set in the simulation software according to the parameter list and value range; pay attention to maintaining consistency and coordination between parameters; verify the set parameters to ensure that there are no omissions or errors; make necessary adjustments and optimizations based on the verification results, and save the set parameter values ​​for use in subsequent simulations; record the process and results of parameter setting for subsequent verification and analysis.

[0197] Stability analysis: Use the built model to perform small disturbance stability analysis and large disturbance stability analysis, and output stability results;

[0198] Specifically, the constructed grid-connected simulation model is used to carry out small disturbance stability analysis and large disturbance stability analysis.

[0199] Small disturbance stability analysis: By applying small disturbances to the system and observing the changes in the system state variables, we can determine whether the system is stable.

[0200] Large disturbance stability analysis: By applying large disturbances to the system, such as load mutations, line failures, etc., observe whether the system can recover to a stable state.

[0201] Output stability results, including time response curve of system state variables, oscillation frequency, damping ratio and other indicators.

[0202] Stability threshold setting: according to the relevant standards and specifications of power system stability, the stability threshold of each indicator is set;

[0203] Specifically, stability thresholds of various indicators are set according to relevant standards and specifications for power system stability; these thresholds may include voltage fluctuation range, frequency deviation range, power oscillation amplitude, etc.; when setting the thresholds, the actual operating conditions and safety margin of the system need to be considered.

[0204] Calculation of grid connection impact index: Based on the stability results, the stability thresholds of various indicators are compared, and the comprehensive stability index and grid balance index are calculated. Based on the above two indexes, the comprehensive stability index, the increase in carrying capacity, and the probability of failure or the growth trend of failures are calculated, and the grid connection feasibility index is calculated;

[0205] Specifically, the comprehensive stability index is obtained by weighted summing of multiple stability indicators (such as voltage stability, frequency stability, power angle stability, etc.) or using other composite function forms. The calculation of each stability indicator is based on the system's state variables (such as voltage, current, power, etc.) and time response characteristics.

[0206] Comprehensive stability index = a*voltage stability index + b*frequency stability index + c*power angle stability index + d*other stability index (a, b, c, d are weight coefficients), among which, multiple other stability indicators can be set and the weights are determined according to their importance.

[0207] Grid balance index: The grid balance index is used to evaluate the supply and demand balance of the grid after grid connection. It is calculated based on factors such as the power flow, load level, and power generation output of the grid.

[0208] Grid balance index = (actual power generation output - actual load demand) / total system capacity.

[0209] Increase in carrying capacity: Increase in carrying capacity is used to quantify the degree of improvement in the grid-connected plan on the grid's carrying capacity; Increase in carrying capacity = (stability index after grid-connection - stability index before grid-connection) / stability index before grid-connection.

[0210] Failure probability or failure growth trend: These indicators are used to evaluate the impact of the grid connection scheme on the risk of grid failure. The failure probability is calculated based on historical failure data, system operating status, equipment reliability and other factors. The failure growth trend may be predicted by analyzing the changes in system stability over time. The specific calculation formula may involve probability statistics, reliability analysis, fault tree analysis and other methods. This embodiment can be formed based on the results of the fault layer output in the hybrid simulation model of the distribution network and analyzed according to the time series.

[0211] Grid connection feasibility index: The grid connection feasibility index is calculated by combining all the above indicators and is used to evaluate the feasibility of the grid connection plan; its specific calculation formula involves weighted summation, fuzzy comprehensive evaluation, hierarchical analysis method and other methods.

[0212] This embodiment adopts the following calculation method: Grid connection feasibility index = f (stability comprehensive index, power grid balance index, carrying capacity improvement range, fault probability / fault growth trend) (where f is an evaluation function, and this embodiment adopts a weighted method for calculation, i.e., f (S, B, C, P) = w1*S + w2*B + w3*C-w4*P), w1, w2, w3, and w4 are weight coefficients corresponding to the qualitative comprehensive index, power grid balance index, carrying capacity improvement range, and fault probability / fault growth trend, respectively, and S, B, C, and P represent the qualitative comprehensive index, power grid balance index, carrying capacity improvement range, and fault probability / fault growth trend, respectively. The weight coefficient can be determined by expert scoring; or the weight coefficient can be determined according to the information entropy of each parameter. The smaller the information entropy, the greater the amount of information provided by the parameter, and the larger the weight coefficient should be.

[0213] Grid connection impact assessment: Determine whether the feasibility index exceeds the set index threshold. If so, execute the grid connection decision step; otherwise, execute the alarm step;

[0214] Grid connection decision: Based on the grid connection impact assessment results, make grid connection execution decisions and execute grid connection adjustment steps.

[0215] Specifically, determine whether the grid-connected feasibility index exceeds the set index threshold; if the feasibility index exceeds the threshold, the grid-connected scheme is considered feasible and the grid-connected decision-making step is executed; otherwise, the grid-connected scheme is considered to have risks and the alarm step is executed; based on the grid-connected impact assessment results, a grid-connected execution decision is made; if the assessment results show that the grid-connected scheme is feasible and the impact is within an acceptable range, the grid-connected adjustment step is executed to prepare for grid implementation; if the assessment results show that the grid-connected scheme has unacceptable risks or impacts, the grid-connected scheme needs to be readjusted or other measures need to be taken.

[0216] Reference Figure 3 In other embodiments, in the first determination step, if the determination reason is not an imbalance in power supply and demand, a power grid structure analysis step is performed;

[0217] Grid structure analysis: including a grid structure acquisition step, a grid structure index calculation step, a health index threshold setting step, a second judgment step, and a branch adjustment step;

[0218] Grid structure acquisition: Obtain grid structure layout information, including data on reasonableness of line layout, uniformity of transformer distribution, clarity of grid hierarchy, distribution network related data, and equipment aging;

[0219] Grid structure index calculation: according to the grid structure layout information, weights are assigned, and the grid structure index is calculated according to the product of the weights and the grid structure layout information;

[0220] Health index threshold setting: Set the health index threshold based on historical data;

[0221] Second judgment: judging whether the grid structure index is less than the health index threshold, if so, executing the branch adjustment step, otherwise, executing the alarm step;

[0222] Branch adjustment: Perform branch intervention, adjust the grid structure, and then re-execute the load estimation step.

[0223] Specifically, the real-time structural layout information of the power grid is obtained through professional power grid monitoring and management systems (such as EMS, DMS, etc.); sensors, remote monitoring equipment and other means are used to collect relevant data such as line layout, transformer distribution, power grid level, distribution network association and equipment aging in real time; the collected data is preprocessed, including data cleaning, format conversion, etc., to ensure the accuracy and consistency of the data; according to different aspects of the power grid structure layout information (such as the rationality of the line layout, the uniformity of the transformer distribution, etc.), corresponding weights are assigned, and the weight assignment can be determined based on expert experience, historical data and the actual needs of power grid operation; the power grid structure index is calculated based on the product of the weight and the power grid structure layout information, and the calculation formula can be expressed as: power grid structure index = Σ (weight i * power grid structure layout information i); the calculated power grid structure index is standardized for comparison and analysis with other indicators; according to historical data, the distribution of the power grid structure index is statistically analyzed to determine the reasonable range of the health index; the threshold of the health index is set in combination with the actual power grid operation and expert experience. When the grid structure index is lower than the threshold, it is considered that there are potential risks or problems in the grid structure; the calculated grid structure index is compared with the health index threshold; if the grid structure index is less than the health index threshold, the branch adjustment step is executed; otherwise, the alarm step is executed to remind relevant personnel to pay attention to the potential risks or problems of the grid structure; based on the results of the grid structure analysis, the location of the branch that needs to be adjusted and the adjustment plan are determined; the remote control system or on-site operation means are used to intervene and adjust the branches of the grid; after the adjustment is completed, the load estimation step is re-executed to evaluate the load capacity and operating status of the grid.

[0224] Reference Figure 3 , in other embodiments, after the health index threshold setting step, a dynamic adjustment step is also included;

[0225] Dynamic adjustment: Obtain the complexity of the distribution network area and the degree of fluctuation in grid usage, adjust the coefficient of the health index threshold, and increase the health index threshold.

[0226] Specifically, the GIS system is used to obtain information such as the geographical distribution, line density, and equipment type of the distribution network area; combined with the power grid planning software, complex factors such as the topological structure, power supply reliability, and load distribution of the distribution network area are analyzed; based on the analysis results, the complexity of the distribution network area is quantitatively evaluated to obtain a value or level that reflects the complexity; through the power grid monitoring and management system, the load data, voltage data, current data, etc. of the power grid are obtained in real time; the distribution network hybrid simulation model is used to analyze historical load data and predict load fluctuations in the future; based on the load forecast results and real-time load data, the degree of fluctuation in power grid use is calculated to obtain a value or indicator that reflects the size of the fluctuation.

[0227] Based on the complexity of the distribution network area and the degree of fluctuation in power grid usage, an adjustment coefficient is set. The coefficient can be in the form of a linear function, a nonlinear function, or a lookup table, and different adjustment values ​​are output according to the complexity and fluctuation.

[0228] Set a baseline health index threshold (such as the threshold set based on historical data in the previous step); calculate the adjustment coefficient based on the quantitative assessment results of the complexity of the distribution network area and the degree of fluctuation in grid usage; multiply the adjustment coefficient by the baseline health index threshold to obtain the adjusted health index threshold, and use the calculated adjusted health index threshold as the new threshold for subsequent grid structure analysis and judgment.

[0229] For example, suppose a regional power grid is dynamically adjusting the health index threshold:

[0230] The GIS system is used to obtain information such as the geographical distribution and line density of the distribution network area in the region.

[0231] Through the analysis of power grid planning software, it was found that the topological structure of the distribution network area in the region is complex, the power supply reliability is low, and the load distribution is uneven; after quantitative evaluation, the value of the complexity of the distribution network area is 0.8 (assuming the full score is 1).

[0232] Obtain the degree of fluctuation in grid usage: obtain real-time load data through the grid monitoring and management system; analyze historical load data and predict that the load will fluctuate greatly in the next month; calculate the value of the grid usage fluctuation to be 0.7 (assuming the full score is 1); set the benchmark health index threshold to 80; set the adjustment coefficient to 1.2 based on the complexity of the distribution network area and the degree of fluctuation in grid usage (the coefficient is calculated based on the quantitative evaluation results of the complexity and fluctuation and the preset adjustment rules); multiply the adjustment coefficient by the benchmark health index threshold to obtain an adjusted health index threshold of 96; use the adjusted health index threshold of 96 as the new threshold for subsequent grid structure analysis and judgment.

[0233] Reference Figure 4 In other embodiments, between the second determination step and the alarm step, there is also a device status evaluation step;

[0234] Equipment status assessment: including branch calling steps, branch activation analysis steps, load-bearing capacity estimation correction steps, aging degree determination steps, aging trend prediction steps, load-bearing capacity estimation adjustment steps and load-bearing capacity re-judgment steps;

[0235] Branch call: switch and replace the spare branch to the original use line or adjust the operation mode of the existing branch;

[0236] Specifically, in the power grid automation control system, identify and locate the backup branch that needs to be called or the existing branch that needs to be adjusted; send instructions through the remote control terminal to switch the backup branch to the working state, or adjust the operation mode of the existing branch (such as adjusting current, voltage and other parameters); monitor and record the changes in the power grid operation status during the branch switching or adjustment process.

[0237] Branch activation analysis: observe the changes in the grid operation status to indirectly determine whether key equipment in the grid is aging or has performance degradation;

[0238] Specifically, the power grid monitoring and analysis software is used to obtain real-time power grid operation status data, including voltage, current, power factor, etc.; the changing trend of the power grid operation status, such as voltage fluctuations and current changes, is observed through the data visualization platform; the changes in the power grid operation status are analyzed to indirectly determine whether key equipment in the power grid is aging or has degraded performance, such as increased equipment heat generation and decreased power factor.

[0239] Carrying capacity estimation correction: Based on the grid operation status data after the branch is activated, the carrying capacity estimation is re-estimated, the carrying capacity estimation values ​​before and after activation are compared, the impact of equipment status changes on the carrying capacity estimation is analyzed, and the carrying capacity estimation results are corrected accordingly;

[0240] Specifically, after the branch is activated, the capacity estimation step is re-executed, and the capacity estimation is re-performed based on the new grid operation status data; the capacity estimation values ​​before and after activation are compared, and the impact of equipment status changes on the capacity estimation is analyzed; based on the analysis results, the capacity estimation results are corrected to make them closer to the actual grid operation status.

[0241] Determination of aging degree: Combine the historical operation records, maintenance records and current operation status data of the equipment, and use data analysis technology to determine the aging degree of the equipment. The aging degree is determined based on a comprehensive analysis of the equipment's performance degradation curve and failure rate increase trend parameters;

[0242] Specifically, historical data such as the equipment's operation records and maintenance records are obtained from the equipment history database. Data analysis and mining software is used to process and analyze the equipment's historical data to extract key information such as the equipment's performance degradation curve and failure rate increase trend. Comprehensively analyze the equipment's performance degradation curve and failure rate increase trend to determine the degree of equipment aging.

[0243] Aging trend prediction: Based on the current aging degree of the equipment, historical usage and the aging rules of similar equipment, the prediction model is used to predict the future aging trend of the equipment;

[0244] Specifically, a prediction model is constructed based on the current aging degree of the equipment, historical usage, and aging patterns of similar equipment. The prediction model is used to input data such as the current aging degree of the equipment, historical usage, and other data to predict aging trends. The prediction results are output, including key information such as the future aging degree of the equipment, expected lifespan, and so on.

[0245] Adjustment of load-bearing capacity estimate: Based on the prediction results of equipment aging degree and aging trend, the revised load-bearing capacity estimate results are further adjusted.

[0246] Specifically, based on the prediction results of equipment aging degree and aging trend, the revised carrying capacity estimation results are further adjusted; considering the impact of equipment aging on grid operation, such as decreased equipment capacity and reduced reliability, the relevant parameters in the carrying capacity estimation results are adjusted; the adjusted carrying capacity estimation results are output to provide a reference for grid operation decisions.

[0247] Reference Figure 4 In other embodiments, a data updating step is further provided after the aging degree determining step;

[0248] Data update: associate the estimated equipment aging degree with the real-time data and perform data simulation to form simulated data, divide the simulated data into m second operating characteristics, and count the second operating characteristics into the first operating characteristics and input them into the training step.

[0249] Specifically, operating data is collected through sensors installed on the equipment, and the collected data is processed and analyzed using data analysis software to estimate the degree of equipment aging. This involves comprehensive consideration of factors such as the equipment's performance parameters, operating time, and failure history.

[0250] Correlate the estimated equipment aging with real-time data. This can be achieved through data correlation software that can identify and match related data sets; simulate the correlated data using data simulation software to form simulated data, which may involve simulation of the equipment's operating state, simulation of failure modes, etc., and divide the simulated data into m second operating features. This can be achieved through data analysis techniques, such as feature extraction, feature selection, etc.; count the second operating features into the first operating features. This involves adding new feature data to the original feature data set for use in subsequent training steps; train the machine learning model using a training data set containing the first operating features (including the newly counted second operating features); adjust the model's parameters through iterative optimization algorithms (such as gradient descent) to improve its prediction accuracy of the equipment status; during the training process, techniques such as cross-validation can be used to evaluate the performance of the model and make necessary adjustments.

[0251] Reference Figure 5 In other embodiments, a data conversion step is provided between the detection step and the constraint condition acquisition step;

[0252] The data conversion step includes: a modeling step, a prediction step, a second detection step, a feedback step, a trend acquisition step, a third judgment step and a fallback step;

[0253] Modeling: Establish a Bayesian network model, where the first-level nodes of the Bayesian network model are historical operation data, and the directed edges of the Bayesian network model are the conversion probabilities between various historical operation data;

[0254] Prediction: Input the real-time operation data into the Bayesian network model to obtain the operation data at the next moment, which is recorded as the estimated operation data;

[0255] Second detection: input the estimated operation data into the trained distribution network hybrid simulation model, obtain the detection result at the next moment, record it as the auxiliary detection result, and perform the feedback step;

[0256] Feedback: Adjust the lower fault assessment layer in the hybrid simulation model of the distribution network based on real-time data and estimated operating data;

[0257] Trend acquisition: based on the test results obtained in the test steps and the auxiliary test results, obtain the change trend of the test results;

[0258] Third judgment: judging whether the change trend is in line with expectations, if so, executing the constraint condition acquisition step; if not, executing the fallback step;

[0259] Rollback: Roll back adjustments made in the feedback step.

[0260] Specifically, historical operation data is collected, including various performance parameters, operation status, fault records, etc. of the equipment; Bayesian network modeling software is used to build a Bayesian network model based on the logical relationship of historical data and the conversion probability between data. The first-level node is the historical operation data, and the directed edge represents the conversion probability between data; the model is trained and verified to ensure that it can accurately reflect the historical operation status and conversion relationship of the equipment; the real-time operation data is input into the trained Bayesian network model, and the Bayesian reasoning engine is used to calculate the operation data at the next moment, that is, the estimated operation data, based on the real-time data and the conversion probability in the model; the estimated operation data is output for use in subsequent steps.

[0261] The estimated operation data is input into the trained distribution network hybrid simulation model. The simulation model performs simulation operation according to the estimated operation data to obtain the detection result at the next moment, that is, the auxiliary detection result; the auxiliary detection result is output for feedback and adjustment of subsequent steps.

[0262] Compare real-time data and estimated operating data, analyze the differences and causes between the two, and adjust the lower fault assessment layer in the distribution network hybrid simulation model based on the analysis results to improve the accuracy and reliability of the model. Record the adjustment process so that it can be rolled back when necessary.

[0263] Collect the test results obtained in the test step and the auxiliary test results obtained in the second test step; use data analysis software to perform time series analysis on the test results, obtain the change trend of the test results, and output the change trend for use in subsequent steps.

[0264] According to the preset expected change trend, determine whether the change trend of the detection result meets expectations. If it meets expectations, execute the constraint condition acquisition step; if it does not meet expectations, execute the fallback step.

[0265] According to the adjustment process recorded in the feedback step, the hybrid simulation model of the distribution network is rolled back and restored to the state before the adjustment so as to perform prediction and detection again.

[0266] Reference Figure 5 , in other embodiments, before the prediction step, a data adjustment step is also included;

[0267] Data adjustment: including aging collection steps and model optimization steps;

[0268] Aging collection: Obtain estimated aging data based on the future aging trend of the equipment in the aging trend step;

[0269] Model optimization: Add secondary nodes and associated nodes to the Bayesian network model, the secondary nodes are estimated aging data, the associated nodes are historical fault data, and directed edges from the primary nodes to the secondary nodes, directed edges between the primary nodes and the associated nodes, and directed edges between the secondary nodes and the associated nodes are established to obtain an optimized Bayesian network model;

[0270] In the prediction step, the real-time operation data and simulation data are input into the optimized Bayesian network model together to obtain the operation data at the next moment as the new estimated operation data.

[0271] Specifically, first, it is necessary to collect the historical operation data of the equipment, which should include the performance parameters, operating time, maintenance records, etc. of the equipment; use data analysis software to process and analyze the historical data to identify the aging trend of the equipment; this involves operations such as data cleaning, conversion, and aggregation, as well as the use of statistical methods or machine learning algorithms to identify patterns and trends in the data; based on the identified aging trends, predict the aging of the equipment in the future and generate estimated aging data; these data will be used in subsequent model optimization steps. In the existing Bayesian network model, secondary nodes and associated nodes are added. The secondary nodes represent the estimated aging data, and the associated nodes represent the historical fault data; according to the logical relationship and data dependency of the nodes, the primary nodes are respectively established to point to the aging data. The directed edges of the secondary nodes, the directed edges of the primary nodes and the associated nodes, and the directed edges of the secondary nodes and the associated nodes, the directed edges represent the conversion probabilities between the historical operation data, and these conversion probabilities reflect the causal relationship or correlation between the historical operation data; the optimized model is trained and verified using the Bayesian network modeling software to ensure that the model can accurately reflect the dependency and probability distribution between the data; the real-time operation data and the simulation data are input into the optimized Bayesian network model together, the real-time operation data can come from the equipment's sensors or monitoring system, and the simulation data can be generated based on historical data or expert experience; the Bayesian network inference engine is used to infer and calculate the model to obtain the operation data at the next moment as the new estimated operation data.

[0272] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A distribution network carrying capacity assessment method based on power flow calculation, characterized in that: The following steps are involved: Data acquisition: obtain historical and real-time operation data of the distribution network; Model construction: A hybrid simulation model of the distribution network is built based on MATLAB and OpenDSS, which includes an upper intelligent algorithm layer, a middle power flow calculation layer, and a lower fault assessment layer. The three layers interact with each other through the COM interface of OpenDSS. Model training: Includes feature extraction step and training step; Feature extraction: preprocess the historical operation data, and extract N first operation features from the preprocessed historical operation data; Training: inputting N first operation characteristics into the distribution network hybrid simulation model to obtain training results; Detection: Input the real-time operation data into the trained distribution network hybrid simulation model to obtain the detection results; Load assessment: including constraint condition acquisition step, load estimation step, load determination step, recording step and alarm step; Obtaining constraints: Obtain the constraints of the distribution network evaluation period, obtain the training results of the same period in the historical data, and record them as input data; Load estimation: Based on the input data and constraints, the trained model is used to estimate the load and obtain the load capacity limit. Based on the test results and input data, the trained model is used to estimate the load capacity. Load determination: compare the estimated load capacity with the load capacity limit to determine whether the estimated load capacity of the distribution network is less than or equal to the load capacity limit. If so, execute the recording step; otherwise, execute the alarm step; Recording steps: Record the data and results of the bearing capacity assessment process; Alarm step: trigger the alarm mechanism, send out an alarm signal, and output the results obtained in the detection step and the estimated carrying capacity.

2. The distribution network carrying capacity evaluation method based on power flow calculation according to claim 1 is characterized in that: An analysis step is also provided between the load determination step and the alarm step; Analysis: including the cause acquisition step, the first judgment step, the grid connection adjustment step and the carrying capacity review step; Reason acquisition: Based on the power grid data of the middle-level power flow calculation layer, analyze the reasons why the estimated carrying capacity exceeds the carrying capacity limit; First judgment: determine whether the cause is an imbalance in power supply and demand. If so, perform grid connection adjustment steps; Grid connection adjustment: Determine the number of grid connections based on the carrying capacity difference value, and connect new energy to the grid based on the number of grid connections; Bearing capacity review: According to the data after grid connection, the real-time data is corrected and the bearing capacity is re-estimated to determine whether the estimated bearing capacity is less than or equal to the bearing capacity limit. If so, the recording step is executed, otherwise, the alarm step is executed.

3. The distribution network carrying capacity evaluation method based on power flow calculation according to claim 2 is characterized in that: Between the first judgment step and the grid connection adjustment step, a grid connection evaluation step is also included; Grid connection assessment: including model building steps, stability analysis steps, stability threshold setting steps, grid connection impact index calculation steps, grid connection impact assessment steps, impact amplitude judgment steps and grid connection decision-making steps; Model building: Build a grid-connected simulation model based on MATLAB / Simulink or OpenDSS power simulation software; Stability analysis: Use the built model to perform small disturbance stability analysis and large disturbance stability analysis, and output stability results; Stability threshold setting: according to the relevant standards and specifications of power system stability, the stability threshold of each indicator is set; Calculation of grid connection impact index: Based on the stability results, the stability thresholds of various indicators are compared, and the comprehensive stability index and grid balance index are calculated. Based on the above two indexes, the comprehensive stability index, the increase in carrying capacity, and the probability of failure or the growth trend of failures are calculated, and the grid connection feasibility index is calculated; Grid connection impact assessment: Determine whether the feasibility index exceeds the set index threshold. If so, execute the grid connection decision step; otherwise, execute the alarm step; Grid connection decision: Based on the grid connection impact assessment results, make grid connection execution decisions and execute grid connection adjustment steps.

4. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 2, characterized in that: In the first determination step, if the determination reason is not an imbalance in power supply and demand, a power grid structure analysis step is performed; Grid structure analysis: including a grid structure acquisition step, a grid structure index calculation step, a health index threshold setting step, a second judgment step, and a branch adjustment step; Grid structure acquisition: Obtain grid structure layout information, including data on reasonableness of line layout, uniformity of transformer distribution, clarity of grid hierarchy, distribution network related data, and equipment aging; Grid structure index calculation: according to the grid structure layout information, weights are assigned, and the grid structure index is calculated according to the product of the weights and the grid structure layout information; Health index threshold setting: Set the health index threshold based on historical data; Second judgment: judging whether the grid structure index is less than the health index threshold, if so, executing the branch adjustment step, otherwise, executing the alarm step; Branch adjustment: Perform branch intervention, adjust the grid structure, and then re-execute the load estimation step.

5. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 4 is characterized in that: After the health index threshold setting step, a dynamic adjustment step is also included; Dynamic adjustment: Obtain the complexity of the distribution network area and the degree of fluctuation in grid usage, adjust the coefficient of the health index threshold, and increase the health index threshold.

6. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 4, characterized in that: Between the second judgment step and the alarm step, there is also an equipment status evaluation step; Equipment status assessment: including branch calling steps, branch activation analysis steps, load-bearing capacity estimation correction steps, aging degree determination steps, aging trend prediction steps, load-bearing capacity estimation adjustment steps and load-bearing capacity re-judgment steps; Branch call: switch and replace the spare branch to the original use line or adjust the operation mode of the existing branch; Branch activation analysis: observe the changes in the grid operation status to indirectly determine whether key equipment in the grid is aging or has performance degradation; Carrying capacity estimation correction: Based on the grid operation status data after the branch is activated, the carrying capacity estimation is re-estimated, the carrying capacity estimation values ​​before and after activation are compared, the impact of equipment status changes on the carrying capacity estimation is analyzed, and the carrying capacity estimation results are corrected accordingly; Determination of aging degree: Combine the historical operation records, maintenance records and current operation status data of the equipment, and use data analysis technology to determine the aging degree of the equipment. The aging degree is determined based on a comprehensive analysis of the equipment's performance degradation curve and failure rate increase trend parameters; Aging trend prediction: Based on the current aging degree of the equipment, historical usage and the aging rules of similar equipment, the prediction model is used to predict the future aging trend of the equipment; Adjustment of load-bearing capacity estimate: Based on the prediction results of equipment aging degree and aging trend, further adjustment is made to the revised load-bearing capacity estimate results; Bearing capacity re-judgment: Determine whether the estimated bearing capacity after adjustment is less than or equal to the bearing capacity limit. If so, execute the recording step, otherwise, execute the alarm step.

7. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 6, characterized in that: A data updating step is also provided after the aging degree determining step; Data update: associate the estimated equipment aging degree with the real-time data and perform data simulation to form simulated data, divide the simulated data into m second operating characteristics, and count the second operating characteristics into the first operating characteristics and input them into the training step.

8. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 2, characterized in that: A data conversion step is also provided between the detection step and the constraint condition acquisition step; The data conversion step includes: a modeling step, a prediction step, a second detection step, a feedback step, a trend acquisition step, a third judgment step and a fallback step; Modeling: Establish a Bayesian network model, where the first-level nodes of the Bayesian network model are historical operation data, and the directed edges of the Bayesian network model are the conversion probabilities between various historical operation data; Prediction: Input the real-time operation data into the Bayesian network model to obtain the operation data at the next moment, which is recorded as the estimated operation data; Second detection: input the estimated operation data into the trained distribution network hybrid simulation model, obtain the detection result at the next moment, record it as the auxiliary detection result, and perform the feedback step; Feedback: Adjust the lower fault assessment layer in the hybrid simulation model of the distribution network based on real-time data and estimated operating data; Trend acquisition: based on the test results obtained in the test steps and the auxiliary test results, obtain the change trend of the test results; Third judgment: judging whether the change trend is in line with expectations, if so, executing the constraint condition acquisition step; if not, executing the fallback step; Rollback: Roll back adjustments made in the feedback step.

9. The method for evaluating the carrying capacity of a distribution network based on power flow calculation according to claim 8, characterized in that: Before the prediction step, a data adjustment step is also included; Data adjustment: including aging collection steps and model optimization steps; Aging collection: Obtain estimated aging data based on the future aging trend of the equipment in the aging trend step; Model optimization: Add secondary nodes and associated nodes to the Bayesian network model, the secondary nodes are estimated aging data, the associated nodes are historical fault data, and directed edges from the primary nodes to the secondary nodes, directed edges between the primary nodes and the associated nodes, and directed edges between the secondary nodes and the associated nodes are established to obtain an optimized Bayesian network model; In the prediction step, the real-time operation data and simulation data are input into the optimized Bayesian network model together to obtain the operation data at the next moment as the new estimated operation data.

Images