A power distribution network evaluation method and system

Abstract

The application provides a power distribution network evaluation method and system, and relates to the technical field of power system operation control. The method comprises the following steps: obtaining an extreme event index of a power distribution network in a current time period according to a measurement data stream, and judging whether the power distribution network has an extreme event according to the extreme event index; classifying the measurement data stream through a data quality grading mechanism, and then constructing an observation weight matrix of the extreme event; supplementing missing data according to the observation weight matrix, and dynamically updating topology information of the power distribution network in combination with the supplemented missing data, and obtaining a real-time network model corresponding to the extreme event according to the topology information; and obtaining a state evaluation result of the power distribution network in the current time period according to the real-time network model and the extreme event index. The application improves the evaluation accuracy of the power distribution network under extreme conditions.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, and more specifically, to a distribution network assessment method and system. Background Technology

[0002] Assessing the status of the distribution network is a crucial step in ensuring the safe and stable operation of the power grid. Existing technologies mainly rely on steady-state measurement data provided by monitoring devices such as smart meters. By using algorithms such as weighted least squares, robust estimation, or Kalman filtering, the operating status of the distribution network, such as node voltage and branch power, can be perceived and assessed in real time. This can adequately meet the needs of distribution network assessment under normal operating conditions and provide important decision-making basis for the operation and maintenance of the power grid.

[0003] In related technologies, under extreme conditions such as typhoons, torrential rains and floods, or extreme low temperatures, monitoring equipment may generate a large amount of missing or abnormal data due to malfunctions or data transmission interruptions. Furthermore, when the topology of the distribution network undergoes drastic changes, such as large-scale line switching or rapid integration and disengagement of distributed energy resources, data acquisition and transmission can be affected, resulting in significant data loss and severely impacting the accuracy and reliability of distribution network assessments. Summary of the Invention

[0004] The problem addressed by this invention is how to improve the accuracy of power distribution network assessment under extreme conditions.

[0005] To address the above problems, this invention provides a power distribution network assessment method and system.

[0006] In a first aspect, the present invention provides a power distribution network assessment method, comprising:

[0007] Acquire the measurement data stream of the operating status of the distribution network in the current time period;

[0008] Based on the measured data stream, the extreme event index of the distribution network in the current time period is obtained, and based on the extreme event index, it is determined whether there is an extreme event in the distribution network.

[0009] If so, the measurement data stream is classified through a data quality grading mechanism to obtain complete data segments and missing data segments in the measurement data stream;

[0010] Based on the data in the complete data segment, construct the observation weight matrix of the extreme event;

[0011] The missing data segment is obtained by performing a weighted tensor completion operation on the missing data segment based on the observation weight matrix.

[0012] Based on the complete data segment and the missing data, the topology information of the distribution network is dynamically updated, and based on the topology information, the real-time network model corresponding to the extreme event is obtained;

[0013] Based on the real-time network model and the extreme event index, the estimated values ​​of node voltage and branch power of the distribution network are obtained through variable structure factor graph reasoning.

[0014] Based on the estimated values ​​of the node voltage and the branch power, the state assessment result of the distribution network in the current time period is obtained.

[0015] Optionally, obtaining the extreme event index of the distribution network in the current time period based on the measurement data stream includes:

[0016] Feature extraction is performed on the measurement data stream to obtain meteorological disturbance feature quantities and topological change feature quantities;

[0017] The meteorological disturbance intensity index is obtained by comparing the meteorological disturbance characteristic quantity with the preset meteorological threshold.

[0018] The topology change feature quantity is compared with a preset topology threshold to obtain a topology change rate index;

[0019] The extreme event index is generated by weighted fusion of the meteorological disturbance intensity index and the topological change rate index.

[0020] Optionally, determining whether an extreme event exists in the distribution network based on the extreme event index includes:

[0021] The extreme event index is compared with a preset threshold range for extreme events;

[0022] If the extreme event index is within any of the threshold ranges, then the preset extreme event corresponding to the threshold range is taken as the extreme event of the distribution network.

[0023] If the extreme event index is not within the threshold range of any of the preset extreme events, then the power distribution network is determined to be in normal operation.

[0024] Optionally, the step of classifying the measurement data stream through a data quality grading mechanism to obtain complete data segments and missing data segments in the measurement data stream includes:

[0025] The data quality level of each measurement point is obtained based on the data integrity, timestamp continuity, and numerical reasonableness of each measurement point in the measurement data stream;

[0026] Based on the data quality level, determine whether data exists in the measurement point; wherein, when data exists in the measurement point, the measurement point is regarded as a complete data point; when no data exists in the measurement point, the measurement point is regarded as a complete data point.

[0027] Based on the complete data points and the missing data points, the complete data segments and the missing data segments in the measurement data stream are determined respectively.

[0028] Optionally, constructing the observation weight matrix of the extreme event based on the data in the complete data segment includes:

[0029] Based on the numerical confidence, temporal confidence, and spatial correlation of each measurement point in the complete data segment, a first weighted component negatively correlated with the extreme event index is determined.

[0030] Based on the data quality level of each measurement point in the complete data segment, a second weight component that is positively correlated with the data quality level is determined.

[0031] The first weight component and the second weight component are fused to obtain the comprehensive weight of each measurement point;

[0032] The comprehensive weights are arranged according to the time dimension, node dimension, and data type dimension to obtain the observation weight matrix.

[0033] Optionally, the step of performing weighted tensor completion operation on the missing data segment based on the observation weight matrix to obtain the missing data of the missing data segment includes:

[0034] Using time, nodes, and data types as three-dimensional coordinates, and mapping the complete data segment and the missing data segment to an initial tensor, wherein the initial tensor is set to zero at the missing position of the missing data;

[0035] A weighted mask is applied to the initial tensor based on the observation weight matrix to obtain a weighted tensor completion objective function;

[0036] The global optimal solution of the tensor completion objective function is obtained by solving the convex optimization iterative algorithm to obtain the completed tensor;

[0037] The missing data of the missing data segment is the value corresponding to the missing position in the completed tensor.

[0038] Optionally, the step of dynamically updating the topology information of the distribution network based on the complete data segment and the missing data, and obtaining the real-time network model corresponding to the extreme event based on the topology information, includes:

[0039] The complete data segment and the missing data are fused to obtain the fused measurement data stream.

[0040] Data is extracted from the fused measurement data stream to generate a real-time topology of the current power grid connection relationship of the distribution network;

[0041] The real-time topology is compared with the pre-stored baseline topology, and the line parameters, distributed power source access points, and load models of the distribution network are synchronously corrected based on the comparison results to obtain the real-time network model corresponding to the extreme event.

[0042] Optionally, the step of obtaining the estimated values ​​of node voltage and branch power of the distribution network through variable structure factor graph inference based on the real-time network model and the extreme event index includes:

[0043] Based on the node-branch relationship of the real-time network model, an initial factor graph containing node variables, measurement factors, pseudo-measurement factors, and event-topology coupling factors is generated.

[0044] Based on the extreme event index, determine the adaptive weights of the event-topology coupling factor;

[0045] Based on the adaptive weights, the node variables and factor nodes in the initial factor graph are adjusted to obtain the variable structure factor graph corresponding to the real-time topology.

[0046] Confidence propagation inference is performed on the variable structure factor graph to obtain estimates of the node voltage and branch power of the distribution network.

[0047] Optionally, obtaining the state assessment result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power includes:

[0048] Based on the preset voltage safety range and branch power current carrying limit, the estimated value is scanned for exceeding the limit to generate node voltage deviation index and branch power exceeding the limit index.

[0049] The node voltage deviation index and the branch power over-limit index are mapped to the risk quantification model to obtain the comprehensive risk score of the distribution network in the current time period.

[0050] The status assessment result is obtained based on the comprehensive risk score.

[0051] Secondly, the present invention provides a power distribution network assessment system, comprising:

[0052] The data monitoring unit is used to acquire the measurement data stream of the operating status of the distribution network in the current time period;

[0053] An extreme event judgment unit is used to obtain the extreme event index of the distribution network in the current time period based on the measurement data stream, and to determine whether there is an extreme event in the distribution network based on the extreme event index.

[0054] A data classification unit is used to classify the measurement data stream through a data quality grading mechanism if the condition is met, to obtain complete data segments and missing data segments in the measurement data stream.

[0055] A matrix construction unit is used to construct the observation weight matrix of the extreme event based on the data in the complete data segment;

[0056] The completion operation unit is used to perform a weighted tensor completion operation on the missing data segment according to the observation weight matrix to obtain the missing data of the missing data segment;

[0057] The model building unit is used to dynamically update the topology information of the distribution network based on the complete data segment and the missing data, and to obtain the real-time network model corresponding to the extreme event based on the topology information.

[0058] The estimation unit is used to obtain estimated values ​​of node voltage and branch power of the distribution network through variable structure factor graph reasoning based on the real-time network model and the extreme event index.

[0059] The evaluation unit is used to obtain the state evaluation result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power.

[0060] The distribution network assessment method and system of this invention obtains extreme event indicators of the distribution network in the current time period based on the measurement data stream, thereby achieving accurate monitoring and rapid response to the distribution network's operating status. When extreme events such as extreme weather or drastic changes in topology occur, traditional methods often struggle to accurately assess the distribution network status due to significant data gaps or anomalies. Therefore, it is necessary to analyze the measurement data stream to obtain extreme event indicators and determine whether extreme events exist in the distribution network. This invention, through an active monitoring and judgment mechanism, avoids the assessment delays or failures caused by data problems under extreme conditions in traditional methods. Furthermore, the combination of a data quality grading mechanism and weighted tensor completion operations effectively solves the problem of data gaps and anomalies under extreme events, further improving the accuracy of the assessment. Specifically, when the existence of an extreme event is determined, the measurement data stream is classified through the data quality grading mechanism to distinguish between complete and missing data segments. Then, an observation weight matrix for extreme events is constructed based on the complete data segments, and this matrix is ​​used to perform weighted tensor completion operations on the missing data segments to obtain estimated values ​​for the missing data. This invention not only fully utilizes the information from existing complete data but also considers data quality and correlation through weighting, making the completed data more reliable. Compared with traditional data interpolation or simple estimation methods, weighted tensor completion operations can more accurately restore missing data, reduce the impact of data errors on the evaluation results, and thus improve the accuracy of the evaluation. The efficiency and accuracy of the evaluation are improved by dynamically updating the distribution network topology information and constructing a real-time network model. Specifically, when extreme events occur, the topology of the distribution network may change drastically, such as line switching, rapid access to or disconnection of distributed energy sources, etc. Traditional methods often struggle to capture these changes in a timely manner, leading to discrepancies between the evaluation model and the actual operating state, resulting in distorted evaluation results. This invention, however, dynamically updates the distribution network topology information based on the complete data segments and the completed missing data, and accordingly obtains a real-time network model corresponding to the extreme events. This ensures that the evaluation model accurately reflects the actual operating state of the distribution network under extreme conditions, providing strong support for subsequent accurate evaluation. Meanwhile, the construction of a real-time network model improves evaluation efficiency and avoids repeated corrections and iterations caused by model update lags. This invention uses a variable structure factor graph to derive estimates of node voltages and branch power in a distribution network based on a real-time network model and extreme event indicators. The variable structure factor graph can be flexibly adjusted according to the real-time topology and operating status of the distribution network, thus more accurately capturing the changing relationship between node voltages and branch power. Compared with traditional fixed-model evaluation methods, variable structure factor graph inference can respond more quickly to changes brought about by extreme events, reduce computational redundancy in the inference process, and improve evaluation efficiency. Furthermore, by comprehensively considering the real-time network model and extreme event indicators, the estimated values ​​are closer to the actual operating state, thereby improving the accuracy of the evaluation.In summary, this invention significantly improves the accuracy of distribution network assessment under extreme conditions by actively monitoring extreme event indicators, data quality grading and weighted tensor completion, dynamic topology updating and real-time network model construction, variable structure factor graph reasoning, and generating detailed and accurate state assessment results. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating a power distribution network assessment method according to an embodiment of the present invention;

[0062] Figure 2 This is a schematic diagram of the distribution network assessment system in another embodiment of the present invention. Detailed Implementation

[0063] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0064] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0065] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0066] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0067] The names of the messages or information exchanged between the multiple devices in the embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of these messages or information.

[0068] To address the problems existing in the aforementioned related technologies, this embodiment provides a power distribution network assessment method and system.

[0069] Combination Figure 1 As shown in the figure, an embodiment of the present invention provides a power distribution network assessment method, including:

[0070] Acquire the measurement data stream of the operating status of the distribution network in the current time period.

[0071] Specifically, acquiring the measurement data stream of the distribution network's operational status in the current time period is fundamental to the entire evaluation method. This step involves real-time acquisition of the distribution network's operational status data through multiple monitoring devices, such as smart meters, μPMUs, and SCADA systems. This operational status data includes, but is not limited to, node voltage, branch current, power flow, and switch status. By aggregating this data, a continuous measurement data stream is formed, providing the raw input for subsequent analysis and processing. This process ensures the real-time nature and completeness of the data, providing a solid data foundation for subsequent extreme event detection and status assessment.

[0072] Based on the measured data stream, the extreme event index of the distribution network in the current time period is obtained, and based on the extreme event index, it is determined whether there is an extreme event in the distribution network.

[0073] Specifically, the first step is to perform in-depth analysis of the measurement data to identify whether the distribution network is in an extreme operating state. In this embodiment of the invention, statistical analysis, signal processing, and other techniques are used to extract features from the data and compare them with features under normal operating conditions. By setting reasonable thresholds or pattern matching rules, it is possible to determine whether the current state meets the definition of an extreme event.

[0074] If so, the measurement data stream is classified through a data quality grading mechanism to obtain complete data segments and missing data segments in the measurement data stream.

[0075] Specifically, data reliability becomes paramount once extreme events are detected. Therefore, a data quality grading mechanism is needed to help identify which data is reliable and which may be missing or abnormal due to equipment failure or transmission problems. Based on this, each data point undergoes meticulous verification, such as determining its completeness, timestamp continuity, and numerical reasonableness. This ensures that subsequent data processing is based only on reliable data segments. Inappropriate data is also identified as missing and thus classified as a missing data segment, providing a basis for completing the missing data. Furthermore, if no extreme events are detected, it is determined that no extreme events exist in the distribution network.

[0076] Based on the data in the complete data segment, construct the observation weight matrix of the extreme event.

[0077] Specifically, to effectively complete missing data, further analysis of the existing complete data is needed to construct a weight matrix that reflects the importance and relevance of the data. This involves modeling the spatial and temporal characteristics of the data to determine the relative importance of each data point in describing the distribution network status. The construction of the weight matrix needs to comprehensively consider the location, time, and relationship with other data points of the data, thereby providing a scientific basis for subsequent tensor completion.

[0078] The missing data segment is obtained by performing a weighted tensor completion operation on the missing data segment based on the observation weight matrix.

[0079] Specifically, the missing data is completed using the constructed observation weight matrix. In this embodiment of the invention, a weighted tensor completion operation can be employed. Through mathematical models and algorithms, the missing data values ​​are estimated based on known data and their weights. This allows for the accurate reconstruction of missing information while considering data quality and relevance, making the entire dataset more complete and reliable. In a preferred embodiment of the invention, a weighted low-rank approximation model can be used. An initial tensor is first constructed based on the complete data, and missing data is treated as zero and assigned a lower weight. Then, the completed tensor is obtained step by step through iterative solving using alternating least squares, thereby estimating the missing node voltage and branch power data, providing complete data support for the state assessment of the distribution network.

[0080] Based on the complete data segment and the missing data, the topology information of the distribution network is dynamically updated, and based on the topology information, the real-time network model corresponding to the extreme event is obtained.

[0081] Specifically, the structure of the distribution network may change during extreme events, such as line switching or equipment failure. Therefore, it is necessary to update the distribution network topology information based on the latest data. This requires real-time monitoring of network structure changes and adjustments to the network model accordingly. Dynamically updating the topology information ensures that the network model accurately reflects the current physical structure and operating status of the distribution network, providing an accurate basis for subsequent analysis.

[0082] Based on the real-time network model and the extreme event index, the estimated values ​​of node voltage and branch power of the distribution network are obtained through variable structure factor graph reasoning.

[0083] Specifically, by combining real-time network models and extreme event indicators, variable structure factor graph inference is used to estimate key parameters of the distribution network. In other words, the structure and parameters of the inference model need to be flexibly adjusted according to the current network structure and operating status. Through such dynamic inference, even under complex and uncertain conditions caused by extreme events, key parameters such as node voltage and branch power can be estimated more accurately.

[0084] Based on the estimated values ​​of the node voltage and the branch power, the state assessment result of the distribution network in the current time period is obtained.

[0085] Specifically, based on estimated parameters such as node voltage and branch power, a comprehensive analysis is conducted to assess the current state of the distribution network. First, each parameter needs to be compared with preset safety standards to identify potential risk points and problem areas. Through a comprehensive evaluation of these parameters, a complete status report is generated, providing decision support in extreme situations and enabling staff to receive accurate assessments of the distribution network, thereby ensuring its safe and stable operation.

[0086] The distribution network assessment method of this invention obtains extreme event indicators of the distribution network in the current time period based on the measurement data stream, thereby achieving accurate monitoring and rapid response to the distribution network's operating status. When extreme events such as extreme weather or drastic changes in topology occur, traditional methods often struggle to accurately assess the distribution network status due to significant data gaps or anomalies. Therefore, this invention analyzes the measurement data stream to obtain extreme event indicators and determines whether extreme events exist in the distribution network. Through an active monitoring and judgment mechanism, it avoids the assessment delays or failures caused by data problems under extreme conditions, as is common in traditional methods. Furthermore, the combination of a data quality grading mechanism and weighted tensor completion operations effectively solves the problem of data gaps and anomalies under extreme events, further improving the accuracy of the assessment. Specifically, when the existence of an extreme event is determined, the measurement data stream is classified through the data quality grading mechanism to distinguish between complete and missing data segments. Then, an observation weight matrix for extreme events is constructed based on the complete data segments, and this matrix is ​​used to perform weighted tensor completion operations on the missing data segments to obtain estimated values ​​for the missing data. This invention not only fully utilizes the information from existing complete data but also considers data quality and correlation through weighting, making the completed data more reliable. Compared with traditional data interpolation or simple estimation methods, weighted tensor completion operations can more accurately restore missing data, reduce the impact of data errors on the evaluation results, and thus improve the accuracy of the evaluation. The efficiency and accuracy of the evaluation are improved by dynamically updating the distribution network topology information and constructing a real-time network model. Specifically, when extreme events occur, the topology of the distribution network may change drastically, such as line switching, rapid access to or disconnection of distributed energy sources, etc. Traditional methods often struggle to capture these changes in a timely manner, leading to discrepancies between the evaluation model and the actual operating state, resulting in distorted evaluation results. This invention, however, dynamically updates the distribution network topology information based on the complete data segments and the completed missing data, and accordingly obtains a real-time network model corresponding to the extreme events. This ensures that the evaluation model accurately reflects the actual operating state of the distribution network under extreme conditions, providing strong support for subsequent accurate evaluation. Meanwhile, the construction of a real-time network model improves evaluation efficiency and avoids repeated corrections and iterations caused by model update lags. This invention uses a variable structure factor graph to derive estimates of node voltages and branch power in a distribution network based on a real-time network model and extreme event indicators. The variable structure factor graph can be flexibly adjusted according to the real-time topology and operating status of the distribution network, thus more accurately capturing the changing relationship between node voltages and branch power. Compared with traditional fixed-model evaluation methods, variable structure factor graph inference can respond more quickly to changes brought about by extreme events, reduce computational redundancy in the inference process, and improve evaluation efficiency. Furthermore, by comprehensively considering the real-time network model and extreme event indicators, the estimated values ​​are closer to the actual operating state, thereby improving the accuracy of the evaluation.In summary, this invention significantly improves the accuracy of distribution network assessment under extreme conditions by actively monitoring extreme event indicators, data quality grading and weighted tensor completion, dynamic topology updating and real-time network model construction, variable structure factor graph reasoning, and generating detailed and accurate state assessment results.

[0087] Optionally, obtaining the extreme event index of the distribution network in the current time period based on the measurement data stream includes:

[0088] Feature extraction is performed on the measurement data stream to obtain meteorological disturbance feature quantities and topological change feature quantities;

[0089] The meteorological disturbance intensity index is obtained by comparing the meteorological disturbance characteristic quantity with the preset meteorological threshold.

[0090] The topology change feature quantity is compared with a preset topology threshold to obtain a topology change rate index;

[0091] The extreme event index is generated by weighted fusion of the meteorological disturbance intensity index and the topological change rate index.

[0092] Specifically, firstly, meteorological disturbance features and topology change features are extracted from the measurement data stream through feature extraction. Meteorological disturbance features include meteorological data such as wind speed and rainfall, while topology change features include changes in the physical structure of the distribution network, such as changes in switch states and line switching. These features directly reflect the changes in the external environment and internal structure faced by the distribution network. Next, the extracted features are compared with preset meteorological and topology thresholds to obtain meteorological disturbance intensity and topology change rate indices. The preset thresholds can be set based on the analysis of historical data of the distribution network and the understanding of the characteristics of extreme events. The comparison quantifies the impact of current meteorological disturbances and topology changes on the distribution network. Finally, the meteorological disturbance intensity and topology change rate indices are combined using a weighted fusion method to generate a comprehensive extreme event index. The weighted fusion process needs to consider the importance of the two indices, and the weights are usually determined based on historical data and expert experience. This method can comprehensively reflect the influence of multiple factors in a single index, providing a scientific basis for subsequent judgment on whether extreme events exist in the distribution network.

[0093] In a preferred embodiment of the invention, the distribution network is equipped with various monitoring devices capable of collecting data such as wind speed, rainfall, and switch status in real time. Meteorological thresholds (e.g., wind speed exceeding 30 m / s) and topological thresholds (e.g., switch switching frequency exceeding 5 times / hour) are preset. During a strong typhoon, the monitoring devices detect wind speeds reaching 35 m / s, a surge in rainfall, and multiple switch switching events occurring in the distribution network due to the typhoon. After extracting these features, they are compared with preset thresholds to obtain meteorological disturbance intensity and topological change rate indices, respectively. Based on historical data and expert experience, meteorological disturbances are given a greater weight (e.g., 0.7), while topological changes are given a smaller weight (e.g., 0.3). Through weighted fusion, an extreme event index is generated and compared with a preset extreme event threshold to determine if the distribution network is in an extreme event state. This determination triggers subsequent data processing and status assessment procedures, enabling maintenance personnel to take timely measures to ensure the safe and stable operation of the distribution network.

[0094] In this embodiment of the invention, precise monitoring and rapid response of the distribution network under extreme conditions are achieved through feature extraction and weighted fusion methods. First, feature extraction technology can extract key information from massive amounts of measurement data, reducing data processing complexity and improving processing efficiency. Second, by comparing with preset thresholds, the impact of meteorological disturbances and topological changes can be quantified, making the assessment process more objective and operable. Finally, the weighted fusion method integrates different types of feature quantities into a comprehensive index, which not only improves the accuracy of the judgment but also enhances the adaptability and robustness of the method.

[0095] Optionally, determining whether an extreme event exists in the distribution network based on the extreme event index includes:

[0096] The extreme event index is compared with a preset threshold range for extreme events;

[0097] If the extreme event index is within any of the threshold ranges, then the preset extreme event corresponding to the threshold range is taken as the extreme event of the distribution network.

[0098] If the extreme event index is not within the threshold range of any of the preset extreme events, then the power distribution network is determined to be in normal operation.

[0099] Specifically, the extreme event index is first compared with preset threshold ranges. These threshold ranges are pre-set based on historical data and expert experience, reflecting the impact of different types of extreme events on the distribution network. If the extreme event index falls within a certain threshold range, it indicates that the current state of the distribution network conforms to the characteristics of the extreme event corresponding to that threshold range. This method, by setting multiple threshold ranges, can classify and identify extreme events of different degrees, thereby providing more detailed information for subsequent response measures. If the extreme event index is not within any preset threshold range, the distribution network is determined to be in normal operation. The threshold comparison-based judgment method is simple, direct, easy to implement, and can quickly provide judgment results, making it suitable for application in real-time monitoring systems. In a preferred embodiment of the invention, the first threshold range (0-0.3) corresponds to minor extreme events, the second threshold range (0.3-0.6) corresponds to moderate extreme events, and the third threshold range (0.6-1.0) corresponds to severe extreme events. During a strong typhoon, the monitoring equipment detected an extreme event index of 0.75 for the distribution network. The system compares the indicator with the preset threshold range and finds that it is within the third threshold range (0.6-1.0), therefore it judges that the distribution network is in a severe extreme event state.

[0100] In this embodiment of the invention, by setting threshold ranges to determine whether extreme events exist in the distribution network, a rapid and accurate assessment of the distribution network's operating status is achieved. This allows for timely identification of whether the distribution network is in an extreme event state, providing a timely trigger signal for subsequent response measures. By setting multiple threshold ranges, not only can the existence of extreme events be identified, but the severity of the events can also be categorized, helping maintenance personnel to take appropriate countermeasures based on the severity of the events.

[0101] Optionally, the step of classifying the measurement data stream through a data quality grading mechanism to obtain complete data segments and missing data segments in the measurement data stream includes:

[0102] The data quality level of each measurement point is obtained based on the data integrity, timestamp continuity, and numerical reasonableness of each measurement point in the measurement data stream;

[0103] Based on the data quality level, determine whether data exists in the measurement point; wherein, when data exists in the measurement point, the measurement point is regarded as a complete data point; when no data exists in the measurement point, the measurement point is regarded as a complete data point.

[0104] Based on the complete data points and the missing data points, the complete data segments and the missing data segments in the measurement data stream are determined respectively.

[0105] Specifically, the data quality level of each measurement point in the measurement data stream is determined based on its data integrity, timestamp continuity, and numerical reasonableness. Data integrity primarily checks for missing or erroneous data; timestamp continuity ensures the data is continuous and uninterrupted over time; and numerical reasonableness verifies whether the data is within the expected physical range, such as whether the voltage is within the normal range or the current exceeds the equipment capacity. Based on a comprehensive evaluation of these dimensions, each measurement point is assigned a data quality level, reflecting the data's reliability and usability. Next, based on the data quality level, it is determined whether valid data exists in the measurement point. If data exists at the measurement point and the data quality level meets the preset criteria, the measurement point is considered to contain complete data; otherwise, it is considered missing data. By traversing all measurement points, the measurement data stream can be divided into complete data segments and missing data segments. For example, consecutive complete data points form a complete data segment, and consecutive missing data points form a missing data segment. This process not only identifies problems in the data but also provides a clear classification basis for subsequent data processing, ensuring the accuracy and reliability of subsequent analysis. For example, if a power distribution network operates during heavy rain, some smart meters may experience data transmission interruptions due to water ingress, while other devices operate normally. A data quality grading mechanism is used to analyze the measurement data stream. For each measurement point, its data integrity, timestamp continuity, and numerical reasonableness are checked. For instance, if a smart meter's data is missing for a period, its timestamps are discontinuous, and its values ​​exceed the normal range, it is rated as low-quality data. These low-quality data points are marked as missing data points, and consecutive missing data points are clustered into missing data segments. Simultaneously, another set of measurement points with complete data, continuous timestamps, and reasonable values ​​are identified, marked as complete data points, and clustered into complete data segments. This successfully distinguishes which data can be used for subsequent analysis and which data needs to be supplemented or corrected, thus ensuring the accuracy and reliability of the status assessment.

[0106] In this embodiment of the invention, by meticulously classifying each measurement point in the measurement data stream, the targeted and effective nature of data processing is ensured. Complete data segments can be directly used for subsequent analysis and modeling, while missing data segments can be specifically completed or corrected. This not only improves the efficiency of data processing but also enhances the robustness of the entire evaluation system. By identifying and classifying data quality issues, misjudgments caused by data anomalies are reduced, thereby improving the accuracy and reliability of state assessment.

[0107] Optionally, constructing the observation weight matrix of the extreme event based on the data in the complete data segment includes:

[0108] Based on the numerical confidence, temporal confidence, and spatial correlation of each measurement point in the complete data segment, a first weighted component negatively correlated with the extreme event index is determined.

[0109] Based on the data quality level of each measurement point in the complete data segment, a second weight component that is positively correlated with the data quality level is determined.

[0110] The first weight component and the second weight component are fused to obtain the comprehensive weight of each measurement point;

[0111] The comprehensive weights are arranged according to the time dimension, node dimension, and data type dimension to obtain the observation weight matrix.

[0112] Specifically, firstly, based on the numerical reliability, temporal confidence, and spatial correlation of each measurement point in the complete data segment, a first weighted component negatively correlated with the extreme event index is determined. Numerical reliability reflects the physical reasonableness of the data, temporal confidence ensures the continuity of the data over time, and spatial correlation considers the correlation of the data in geographical distribution. These factors collectively affect the reliability of the data under extreme events; therefore, the first weighted component is negatively correlated with the extreme event index, meaning that the more severe the extreme event, the lower the data reliability may be, and the lower the weight will be accordingly. Secondly, based on the data quality level of each measurement point, a second weighted component positively correlated with the data quality level is determined. The data quality level is obtained through multi-dimensional evaluation of data integrity, timestamp continuity, and numerical reasonableness, and can directly reflect the usability and accuracy of the data. Therefore, the second weighted component is positively correlated with the data quality level; the higher the data quality, the greater the weight. Finally, the first and second weighted components are merged to obtain the comprehensive weight of each measurement point. This fusion process is usually achieved through weighted summation or other mathematical methods, aiming to comprehensively consider the impact of extreme events and the reliability of the data itself. The comprehensive weight not only reflects the reliability of data under extreme events but also its importance in the overall dataset. The comprehensive weights are arranged according to the time dimension, node dimension, and data type dimension to form the observation weight matrix. This matrix provides a scientific basis for subsequent weighted tensor completion operations, ensuring the scientific nature and accuracy of the completion process. In a preferred embodiment of the invention, it is assumed that the distribution network operates during a typhoon, and some monitoring devices experience data transmission interruptions or anomalies due to strong winds and heavy rain. Complete and missing data segments are identified through a data quality grading mechanism. For each measurement point in a complete data segment, its numerical reliability, temporal confidence, and spatial correlation are further analyzed. For example, the data from a certain smart meter is numerically reasonable, has continuous timestamps, and exhibits strong spatial correlation with the data from other adjacent meters; therefore, its first weight component is high. Simultaneously, the data quality level of this smart meter is also high, thus its second weight component is also high. These two weight components are fused to obtain the comprehensive weight of the measurement point, and a high weight value is assigned to the corresponding position in the observation weight matrix. For measurement points with lower data quality or those significantly affected by typhoons, their overall weight will be reduced accordingly. This approach allows for more accurate identification and utilization of high-quality data, providing a reliable basis for subsequent weighted tensor completion calculations and state assessments.

[0113] In this embodiment of the invention, by comprehensively considering numerical confidence, temporal confidence, spatial correlation, and data quality level, the observation weight matrix can more accurately reflect the reliability and importance of each measurement point under extreme events. This weight matrix not only reduces the impact of data anomalies caused by extreme events on the evaluation results but also improves the confidence of the supplementary data, thereby enhancing the robustness and accuracy of the entire evaluation system. By rationally allocating weights, high-quality data plays a greater role in the supplementation and evaluation process, further enhancing the confidence of the evaluation results.

[0114] Optionally, the step of performing weighted tensor completion operation on the missing data segment based on the observation weight matrix to obtain the missing data of the missing data segment includes:

[0115] Using time, nodes, and data types as three-dimensional coordinates, and mapping the complete data segment and the missing data segment to an initial tensor, wherein the initial tensor is set to zero at the missing position of the missing data;

[0116] A weighted mask is applied to the initial tensor based on the observation weight matrix to obtain a weighted tensor completion objective function;

[0117] The global optimal solution of the tensor completion objective function is obtained by solving the convex optimization iterative algorithm to obtain the completed tensor;

[0118] The missing data of the missing data segment is the value corresponding to the missing position in the completed tensor.

[0119] Specifically, time, nodes, and data types are used as three-dimensional coordinates to map complete and missing data segments into an initial tensor. In this tensor, the positions of missing data are set to zero to clearly distinguish complete and missing data in subsequent calculations. Next, a weighted mask is applied to the initial tensor based on the observation weight matrix to generate a weighted objective function. This objective function aims to minimize the difference between the completed data and the known data, while maximizing data integrity and consistency. The global optimal solution of this objective function can be found using a convex optimization iterative algorithm, yielding the completed tensor. Finally, the estimated value of the missing data is obtained based on the values ​​at the missing positions in the completed tensor. In a preferred embodiment of the invention, time, nodes, and data types are used as three-dimensional coordinates to construct the initial tensor. ,in It refers to the number of time points. It is the number of nodes. It refers to the number of data types.

[0120] For locations with missing data, set the corresponding position in the initial tensor to zero. That is:

[0121] ;

[0122] in, It is in time ,node Data types The observed value at that location.

[0123] Based on the observation weight matrix For the initial tensor Applying a weighted mask yields the weighted tensor-complete objective function:

[0124] ;

[0125] in, It is the complete tensor to be solved. This represents element-wise multiplication. It is the Frobenius norm. It is the nuclear norm. It is the regularization parameter.

[0126] The objective function described above is solved using a convex optimization iterative algorithm (such as the Alternating Direction Multiplier Method, ADMM) to obtain the completed tensor. The specific iterative steps are as follows:

[0127] initialization , Lagrange multipliers .

[0128] Iterative updates:

[0129] ;

[0130] ;

[0131] ;

[0132] in, It is a soft threshold operator. It is a threshold parameter. : No. The completed tensor estimate of the next iteration. It is the observed initial tensor, where the missing positions are zero. It is the first The Lagrange multipliers in the next iteration are used for balancing. and , This is the observation weight matrix, representing the confidence level of each observation. It is element-wise multiplication, used to multiply the weight matrix. Applied to tensors This highlights the discrepancy between the observed values ​​and the current estimates. It is the first The auxiliary variable is updated in the next iteration. It is a soft thresholding operator used to threshold tensors, promoting sparse solutions. Threshold parameter. The size of the regularization term is controlled, thereby affecting the sparsity of the solution. It is the current completed tensor estimate minus the Lagrange multiplier, used as the input for the soft threshold operator. It is the first The Lagrange multiplier update value in the next iteration. It represents the difference between the current completed tensor and the auxiliary variable, used to update the Lagrange multipliers to adjust the balance between the two in subsequent iterations. It is in time ,node Data types Estimates of missing data.

[0133] Based on the completed tensor The values ​​at the missing locations are used to obtain an estimate of the missing data. For each missing location... The estimated value is:

[0134] ;

[0135] in, It is to complete the tensor The value at the corresponding position is used as an estimate of the missing data.

[0136] In this embodiment of the invention, by utilizing a three-dimensional tensor structure, the potential relationships between time, nodes, and data types can be fully leveraged, thereby improving the quality of the completed data. The introduction of an observation weight matrix ensures that high-quality data plays a greater role in the completion process, reducing the impact of low-quality data or noise. Furthermore, the use of a convex optimization iterative algorithm guarantees the stability and convergence of the solution process, making the completion results more reliable. This not only improves the accuracy of distribution network condition assessment but also enhances its adaptability and robustness under extreme events.

[0137] Optionally, the step of dynamically updating the topology information of the distribution network based on the complete data segment and the missing data, and obtaining the real-time network model corresponding to the extreme event based on the topology information, includes:

[0138] The complete data segment and the missing data are fused to obtain the fused measurement data stream.

[0139] Data is extracted from the fused measurement data stream to generate a real-time topology of the current power grid connection relationship of the distribution network;

[0140] The real-time topology is compared with the pre-stored baseline topology, and the line parameters, distributed power source access points, and load models of the distribution network are synchronously corrected based on the comparison results to obtain the real-time network model corresponding to the extreme event.

[0141] Specifically, firstly, by fusing complete data segments and supplemented missing data, a complete measurement data stream is obtained, ensuring data continuity and integrity and providing a reliable data foundation for subsequent topology updates. Next, data extraction is performed on the fused measurement data stream to generate the real-time topology of the distribution network. Generating the real-time topology requires extracting key information such as switch states and protection signals from the data; this information reflects the connection relationships and structural changes of the distribution network under extreme events. Finally, the generated real-time topology is compared with a pre-stored baseline topology to identify newly added, disconnected, or reconstructed branches and nodes. Based on these differences, the line parameters, distributed power source access points, and load models of the distribution network are simultaneously corrected, ultimately obtaining a real-time network model corresponding to the extreme event. This process not only captures changes in the distribution network structure in a timely manner but also ensures the accuracy and real-time performance of the network model, providing a reliable foundation for subsequent state assessment. In a preferred embodiment of the invention, during a strong typhoon, some lines of a distribution network experienced faults due to excessive wind force, leading to changes in the topology. The topology information is dynamically updated and a real-time network model is constructed through the following steps: The complete data segment and the padded missing data are merged to obtain the merged measurement data stream. Assume the complete data segment is... The missing data segment after completion is The fused measurement data stream is as follows:

[0142] ;

[0143] Key information such as switch status and protection signals are extracted from the fused measurement data stream to generate a real-time topology. The real-time topology can be represented as an adjacency matrix. ,in It is the number of nodes. Represents a node and nodes There are connections between them. This indicates no connection.

[0144] Real-time topology With the pre-stored baseline topology Perform a difference comparison to identify newly added, disconnected, or reconstructed branches and nodes. The comparison formula is:

[0145] ;

[0146] in, This represents the XOR operation. Represents a node and nodes The connection state between them has changed.

[0147] Based on the discrepancies, the line parameters, distributed generation connection points, and load models of the distribution network are adjusted synchronously. For example, if a branch is disconnected in the real-time topology, the admittance of that branch is set to zero, and the corresponding node parameters are updated.

[0148] Based on the corrected topology information and parameters, a real-time network model corresponding to the extreme events is constructed. The real-time network model can be represented as a graph structure. ,in It is a set of nodes. It is a collection of branch roads, each branch road It has corresponding parameters (such as resistance, reactance, etc.).

[0149] In this embodiment of the invention, by dynamically updating the topology information of the distribution network and constructing a real-time network model, the accuracy and efficiency of distribution network status assessment can be significantly improved. The generation and updating of the real-time topology ensures that the assessment model accurately reflects the actual operating state of the distribution network under extreme events, avoiding assessment errors caused by model lag. By comparing the differences with the baseline topology, changes in the distribution network structure can be quickly identified, and line parameters, distributed generation access points, and load models can be corrected simultaneously, making the assessment model closer to actual operating conditions.

[0150] Optionally, the step of obtaining the estimated values ​​of node voltage and branch power of the distribution network through variable structure factor graph inference based on the real-time network model and the extreme event index includes:

[0151] Based on the node-branch relationship of the real-time network model, an initial factor graph containing node variables, measurement factors, pseudo-measurement factors, and event-topology coupling factors is generated.

[0152] Based on the extreme event index, determine the adaptive weights of the event-topology coupling factor;

[0153] Based on the adaptive weights, the node variables and factor nodes in the initial factor graph are adjusted to obtain the variable structure factor graph corresponding to the real-time topology.

[0154] Confidence propagation inference is performed on the variable structure factor graph to obtain estimates of the node voltage and branch power of the distribution network.

[0155] Specifically, an initial factor graph is generated based on the node-branch relationships of the real-time network model. This initial factor graph includes node variables, measurement factors, pseudo-measurement factors, and event-topology coupling factors. Node variables represent state quantities such as voltage at each node in the distribution network; measurement factors reflect observed values ​​of node voltage and branch power based on actual measurement data; pseudo-measurement factors are used to supplement insufficient measurement data and are generated from historical or predicted data; and event-topology coupling factors capture the impact of extreme events on the distribution network topology and operating status. Next, adaptive weights for the event-topology coupling factors are determined based on extreme event indices. These weights reflect the severity of the impact of extreme events on the distribution network; the higher the extreme event index, the greater the impact of the event-topology coupling factors on the inference process. Then, the node variables and factor nodes in the initial factor graph are adjusted according to the adaptive weights to obtain a variable-structure factor graph corresponding to the real-time topology. By adjusting the connection relationships and weights of node variables and factor nodes, the factor graph can adapt to changes in the topology and operating status of the distribution network under extreme events. Finally, belief propagation inference is performed on the variable structure factor graph. Belief propagation inference is a distributed inference algorithm based on a probabilistic graphical model. By passing messages in the factor graph, it gradually converges to the estimates of node voltage and branch power.

[0156] In a preferred embodiment of the invention, during a flood disaster caused by a rainstorm, some lines of the distribution network were submerged, leading to changes in the topology and anomalies in node voltage and branch power. First, an initial factor graph is generated based on the node-branch relationships of the real-time network model. The real-time network model reflects the current connection relationships and parameters of the distribution network. The initial factor graph constructed based on this model includes the voltage variables of each node, the power variables of each branch, and the corresponding measurement factors, pseudo-measurement factors, and event-topology coupling factors. Next, the calculated extreme event index reflects the degree of impact of the rainstorm and flood on the distribution network. Based on this index, a large adaptive weight is assigned to the event-topology coupling factor, highlighting its importance in inference. Subsequently, the initial factor graph is adjusted according to the adaptive weights, changing the connection strength between node variables and the event-topology coupling factor to form a variable structure factor graph. Finally, a belief propagation inference algorithm is used to perform iterative calculations on the variable structure factor graph. Initially, the node voltage message is initialized based on prior probabilities and measurement data; as the iteration progresses, the message is continuously updated and transmitted. After multiple iterations, estimated values ​​of voltage at each node and power in each branch were obtained, presenting the actual operating status of the power distribution network under rainstorm and flood disasters, providing a key basis for subsequent fault diagnosis and recovery strategy formulation.

[0157] In this embodiment of the invention, by constructing a variable structure factor graph that includes event-topology coupling factors, the impact of extreme events on the distribution network can be fully considered, making the state estimation closer to the actual operating conditions. The generation of the initial factor graph provides a basic framework for inference, while the introduction of adaptive weights enhances the sensitivity to extreme events, enabling the inference process to dynamically respond to extreme events of different intensities. The formation of the variable structure factor graph further improves the adaptability of the model, enabling it to accurately reflect the real-time topology and operating status of the distribution network.

[0158] Optionally, obtaining the state assessment result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power includes:

[0159] Based on the preset voltage safety range and branch power current carrying limit, the estimated value is scanned for exceeding the limit to generate node voltage deviation index and branch power exceeding the limit index.

[0160] The node voltage deviation index and the branch power over-limit index are mapped to the risk quantification model to obtain the comprehensive risk score of the distribution network in the current time period.

[0161] The status assessment result is obtained based on the comprehensive risk score.

[0162] Specifically, firstly, based on preset voltage safety ranges and branch power current carrying limits, the estimated values ​​are scanned for exceeding limits. By comparing the voltage estimate of each node with the preset voltage safety range, a node voltage deviation index is calculated; simultaneously, the power estimate of each branch is compared with its power current carrying limit to calculate a branch power exceeding limit index. This step accurately identifies nodes and branches in the distribution network where voltage and power anomalies occur, providing a quantitative basis for subsequent risk assessment. Next, the node voltage deviation index and branch power exceeding limit index are mapped into a risk quantification model. This model comprehensively considers the impact of voltage deviation and power exceeding limits on the safe operation of the distribution network. Through specific mapping relationships and weight allocations, a comprehensive risk score for the distribution network in the current time period is calculated. The mapping process typically involves normalization and weighted summation of different indicators to ensure that the comprehensive risk score fully reflects the overall operational risk of the distribution network. Finally, based on the magnitude of the comprehensive risk score, the state assessment result of the distribution network is determined. Different risk level ranges are typically set, such as low risk, medium risk, and high risk. The risk level is determined by the range in which the comprehensive risk score falls, thus providing operation and maintenance personnel with an intuitive and clear basis for decision-making.

[0163] In a preferred embodiment of the present invention, it is assumed that after a strong wind, the estimated node voltages and branch power values ​​of a distribution network have been derived through a variable structure factor diagram. The distribution network has 100 nodes and 80 branches. First, based on the preset voltage safety range (0.9 pu-1.1 pu) and the branch power current carrying limit (based on the branch thermal limit power), the estimated values ​​of all node voltages and branch power are scanned for exceeding limits. The scan reveals that the voltage estimates of 5 nodes exceed the voltage safety range, of which 3 nodes have low voltages (below 0.9 pu) and 2 nodes have high voltages (above 1.1 pu); simultaneously, the power estimates of 3 branches exceed 80% of the branch power current carrying limit. For these exceeding-limit situations, node voltage deviation indices and branch power exceeding-limit indices are calculated. The voltage deviation index can be calculated using the following formula:

[0164] ;

[0165] Assume the total calculated voltage deviation index is 0.12. The branch power over-limit index can be calculated using the following formula:

[0166] ;

[0167] Where m is the total number of branches, and we assume the sum of the calculated branch power exceedance indices is 0.08. Next, we map these two indices to the risk quantification model. The risk quantification model can be set as follows:

[0168] ;

[0169] in, and These are the weights of the voltage deviation index and the branch power over-limit index, respectively. Assuming... , Substituting the values ​​into the calculation, we get:

[0170] .

[0171] Based on the preset risk level ranges (assuming low risk: 0-0.1, medium risk: 0.1-0.3, and high risk: 0.3-1.0), the comprehensive risk score of 0.128 falls within the medium risk range. Therefore, the distribution network's condition assessment result for the current time period is medium risk. This result suggests that operation and maintenance personnel need to further inspect and handle nodes and branches that exceed limits to prevent potential faults and ensure the safe and stable operation of the distribution network.

[0172] In this embodiment of the invention, over-limit scanning can promptly detect voltage deviations and power over-limit issues in the distribution network, providing early warning for the prevention and handling of potential faults. Mapping over-limit indicators to a risk quantification model allows different types of indicators to be integrated within a unified risk assessment framework, avoiding the one-sidedness that may result from evaluating a single indicator.

[0173] Combination Figure 2 As shown in the figure, an embodiment of the present invention provides a power distribution network assessment system, comprising:

[0174] The data monitoring unit is used to acquire the measurement data stream of the operating status of the distribution network in the current time period;

[0175] An extreme event judgment unit is used to obtain the extreme event index of the distribution network in the current time period based on the measurement data stream, and to determine whether there is an extreme event in the distribution network based on the extreme event index.

[0176] A data classification unit is used to classify the measurement data stream through a data quality grading mechanism if the condition is met, to obtain complete data segments and missing data segments in the measurement data stream.

[0177] A matrix construction unit is used to construct the observation weight matrix of the extreme event based on the data in the complete data segment;

[0178] The completion operation unit is used to perform a weighted tensor completion operation on the missing data segment according to the observation weight matrix to obtain the missing data of the missing data segment;

[0179] The model building unit is used to dynamically update the topology information of the distribution network based on the complete data segment and the missing data, and to obtain the real-time network model corresponding to the extreme event based on the topology information.

[0180] The estimation unit is used to obtain estimated values ​​of node voltage and branch power of the distribution network through variable structure factor graph reasoning based on the real-time network model and the extreme event index.

[0181] The evaluation unit is used to obtain the state evaluation result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power.

[0182] The distribution network assessment system of the present invention has the same advantages over the prior art as the aforementioned distribution network assessment method, and will not be repeated here.

[0183] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for evaluating power distribution networks, characterized in that, include: Acquire the measurement data stream of the operating status of the distribution network in the current time period; Based on the measured data stream, the extreme event index of the distribution network in the current time period is obtained, and based on the extreme event index, it is determined whether there is an extreme event in the distribution network. If so, the measurement data stream is classified using a data quality grading mechanism to obtain complete data segments and missing data segments within the measurement data stream. Specifically, this includes: determining the data quality level of each measurement point based on its data completeness, timestamp continuity, and numerical reasonableness; determining whether data exists at the measurement point based on the data quality level; wherein, when data exists at the measurement point, the measurement point is considered a complete data point; when no data exists at the measurement point, the measurement point is considered a missing data point; and determining the complete data segments and missing data segments in the measurement data stream based on the complete data points and the missing data points, respectively. Based on the data in the complete data segment, an observation weight matrix for the extreme event is constructed. Specifically, this includes: determining a first weight component negatively correlated with the extreme event index based on the numerical confidence, temporal confidence, and spatial correlation of each measurement point in the complete data segment; determining a second weight component positively correlated with the data quality level of each measurement point in the complete data segment; fusing the first weight component and the second weight component to obtain a comprehensive weight for each measurement point; and arranging the comprehensive weights according to the time dimension, node dimension, and data type dimension to obtain the observation weight matrix. The missing data segment is filled using a weighted tensor based on the observation weight matrix to obtain the missing data. Specifically, this includes: using time, node, and data type as three-dimensional coordinates, and mapping the complete data segment and the missing data segment to an initial tensor, where the initial tensor is set to zero at the missing data position; applying a weighted mask to the initial tensor based on the observation weight matrix to obtain a weighted tensor completion objective function; solving the global optimal solution of the tensor completion objective function using a convex optimization iterative algorithm to obtain the completed tensor; and using the values ​​corresponding to the missing positions in the completed tensor as the missing data of the missing data segment. Based on the complete data segment and the missing data, the topology information of the distribution network is dynamically updated, and based on the topology information, the real-time network model corresponding to the extreme event is obtained; Based on the real-time network model and the extreme event index, the estimated values ​​of node voltage and branch power of the distribution network are obtained through variable structure factor graph reasoning. Based on the estimated values ​​of the node voltage and the branch power, the state assessment result of the distribution network in the current time period is obtained.

2. The distribution network assessment method according to claim 1, characterized in that, The step of obtaining the extreme event index of the distribution network in the current time period based on the measured data stream includes: Feature extraction is performed on the measurement data stream to obtain meteorological disturbance feature quantities and topological change feature quantities; The meteorological disturbance intensity index is obtained by comparing the meteorological disturbance characteristic quantity with the preset meteorological threshold. The topology change feature is compared with a preset topology threshold to obtain a topology change rate index; The extreme event index is generated by weighted fusion of the meteorological disturbance intensity index and the topological change rate index.

3. The distribution network assessment method according to claim 1, characterized in that, The step of determining whether an extreme event exists in the distribution network based on the extreme event index includes: The extreme event index is compared with a preset threshold range for extreme events; If the extreme event index is within any of the threshold ranges, then the preset extreme event corresponding to the threshold range is taken as the extreme event of the distribution network. If the extreme event index is not within the threshold range of any of the preset extreme events, then the power distribution network is determined to be in normal operation.

4. The distribution network assessment method according to claim 1, characterized in that, The step of dynamically updating the topology information of the distribution network based on the complete data segment and the missing data, and obtaining the real-time network model corresponding to the extreme event based on the topology information, includes: The complete data segment and the missing data are fused to obtain the fused measurement data stream. Data is extracted from the fused measurement data stream to generate a real-time topology of the current power grid connection relationship of the distribution network; The real-time topology is compared with the pre-stored baseline topology, and the line parameters, distributed power source access points, and load models of the distribution network are synchronously corrected based on the comparison results to obtain the real-time network model corresponding to the extreme event.

5. The distribution network assessment method according to claim 4, characterized in that, The step of obtaining estimates of node voltages and branch power of the distribution network through variable structure factor graph reasoning based on the real-time network model and the extreme event index includes: Based on the node-branch relationship of the real-time network model, an initial factor graph containing node variables, measurement factors, pseudo-measurement factors, and event-topology coupling factors is generated. Based on the extreme event index, determine the adaptive weights of the event-topology coupling factor; Based on the adaptive weights, the node variables and factor nodes in the initial factor graph are adjusted to obtain the variable structure factor graph corresponding to the real-time topology. Confidence propagation inference is performed on the variable structure factor graph to obtain estimates of the node voltage and branch power of the distribution network.

6. The distribution network assessment method according to claim 4, characterized in that, The step of obtaining the state assessment result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power includes: Based on the preset voltage safety range and branch power current carrying limit, the estimated value is scanned for exceeding the limit to generate node voltage deviation index and branch power exceeding the limit index. The node voltage deviation index and the branch power over-limit index are mapped to the risk quantification model to obtain the comprehensive risk score of the distribution network in the current time period. The status assessment result is obtained based on the comprehensive risk score.

7. A power distribution network assessment system, characterized in that, include: The data monitoring unit is used to acquire the measurement data stream of the operating status of the distribution network in the current time period; An extreme event judgment unit is used to obtain the extreme event index of the distribution network in the current time period based on the measurement data stream, and to determine whether there is an extreme event in the distribution network based on the extreme event index. A data classification unit, if applicable, classifies the measurement data stream using a data quality grading mechanism to obtain complete data segments and missing data segments within the measurement data stream. Specifically, this includes: determining the data quality level of each measurement point based on its data completeness, timestamp continuity, and numerical reasonableness; determining whether data exists at the measurement point based on the data quality level; wherein, when data exists at the measurement point, the measurement point is considered a complete data point; when no data exists at the measurement point, the measurement point is considered a missing data point; and determining the complete data segment and the missing data segment in the measurement data stream based on the complete data point and the missing data point, respectively. A matrix construction unit is used to construct an observation weight matrix for the extreme event based on the data in the complete data segment. Specifically, this includes: determining a first weight component negatively correlated with the extreme event index based on the numerical confidence, temporal confidence, and spatial correlation of each measurement point in the complete data segment; determining a second weight component positively correlated with the data quality level of each measurement point in the complete data segment; fusing the first weight component and the second weight component to obtain a comprehensive weight for each measurement point; and arranging the comprehensive weights according to the time dimension, node dimension, and data type dimension to obtain the observation weight matrix. The completion operation unit is used to perform weighted tensor completion operation on the missing data segment according to the observation weight matrix to obtain the missing data of the missing data segment; specifically, it includes: using time, node, and data type as three-dimensional coordinates, and mapping the complete data segment and the missing data segment to an initial tensor, wherein the initial tensor is set to zero at the missing position of the missing data; applying a weighted mask to the initial tensor according to the observation weight matrix to obtain a weighted tensor completion objective function; solving the global optimal solution of the tensor completion objective function through a convex optimization iterative algorithm to obtain the completed tensor; and using the value corresponding to the missing position in the completed tensor as the missing data of the missing data segment; The model building unit is used to dynamically update the topology information of the distribution network based on the complete data segment and the missing data, and to obtain the real-time network model corresponding to the extreme event based on the topology information. The estimation unit is used to obtain estimated values ​​of node voltage and branch power of the distribution network through variable structure factor graph reasoning based on the real-time network model and the extreme event index. The evaluation unit is used to obtain the state evaluation result of the distribution network in the current time period based on the estimated values ​​of the node voltage and the branch power.

Images