Measurement data dynamic partitioning method, system and equipment and storage medium

Abstract

The invention discloses a measurement data dynamic partitioning method, system and device and a storage medium, and the method comprises the steps: carrying out the abnormal measurement coarse screening of the measurement data at the current moment, and recognizing the abnormal measurement and suspicious measurement; gradually expanding the range of the area where the abnormal measurement or the suspicious measurement is located based on a preset expansion rule by taking the abnormal measurement or the suspicious measurement as a starting point; in the partition expansion process, if overlapping nodes exist between adjacent partitions, the areas where the overlapping nodes are located are merged; calculating a data correction capability index based on the weighted error entropy of the local network measurement data, and determining a minimum partition boundary range according to the index; according to the method, the dynamic change of the quality of the measured data is fully combined, the partitioning result is ensured to have dynamic adaptability, and the abnormal data can be identified and corrected in parallel in the partition, so that the abnormal data processing efficiency is remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power system dispatching automation, and particularly relates to a measurement data dynamic partitioning method, system, device and storage medium. BACKGROUND

[0002] In recent years, with the rapid development of new energy power generation technology and high-power power electronic equipment, the power supply structure and operation mode of the power system have undergone profound changes. The large-scale grid connection of renewable energy sources such as wind power and photovoltaic power has significantly enhanced the power fluctuation and randomness of the power grid, and the dynamic process has become more complex. At the same time, the continuous expansion of the power grid scale and the widespread application of intelligent sensing technology have led to an exponential increase in the amount of measurement data, bringing new opportunities and challenges for real-time monitoring and analysis of the operating state of the power grid.

[0003] To improve the accuracy and robustness of state estimation, traditional methods usually introduce an abnormal data identification link after state estimation of the power grid operation, and gradually eliminate or correct abnormal measurements through an iterative process of "estimation-detection and identification-reestimation".

[0004] However, with the expansion of the power grid scale, the number of abnormal measurements has increased significantly, and the iterative identification process has led to a sharp increase in calculation time, making the application of the above-mentioned methods in large-scale power grids face a significant computational efficiency bottleneck. SUMMARY

[0005] The purpose of the present application is to provide a measurement data dynamic partitioning method, system, device and storage medium that can improve the efficiency of abnormal measurement processing in large-scale power grids.

[0006] Technical solution: The measurement data dynamic partitioning method provided by the present application comprises the following steps:

[0007] (1) performing abnormal measurement screening on the measurement data at the current time to identify abnormal measurements or suspicious measurements;

[0008] (2) expanding the range of the region where the abnormal measurements or suspicious measurements are located based on a preset expansion rule;

[0009] (3) calculating a data correction capability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determining the minimum partition boundary range of the measurement data at the current time according to the index.

[0010] The abnormal measurement coarse screening step quickly identifies suspicious data points, effectively reduces the calculation burden of normal data, and improves the processing speed; the abnormal region range is dynamically adjusted based on the preset expansion rule, the intelligent optimization of the partition boundary is realized, and the over-correction or under-correction problem caused by the traditional fixed partition method is avoided; the data correction capability index is calculated by the weighted error entropy, the measurement data quality of the local network is scientifically evaluated, and the minimum partition boundary is accurately determined, so that the calculation resource is efficiently utilized, and the accuracy of data correction is maximally improved; the method realizes the rapid positioning and accurate correction of abnormal data, greatly improves the reliability and adaptability of the power system state estimation, and provides strong support for the safe and stable operation of the power grid.

[0011] Preferably, the step 1 of performing abnormal measurement coarse screening comprises:

[0012] (11) performing numerical range rationality check on the measurement data, and eliminating the measurement data exceeding the preset reasonable range;

[0013] (12) performing consistency check on the measurement data based on the constraint relationship between electrical quantities, and screening out the measurement data not satisfying the electrical constraint;

[0014] (13) identifying abnormal fluctuation or mutation measurement data from the remaining measurement data in combination with the rationality analysis of continuous time section power flow change;

[0015] (14) performing change trend analysis on the identified abnormal fluctuation or mutation measurement data and comparing with historical data to identify abnormal measurement or suspicious measurement.

[0016] The obviously out-of-limit data is eliminated through the numerical range check, the data violating the physical law is screened out through the electrical constraint check, the abnormal fluctuation is identified through the power flow change analysis, and the suspicious measurement is comprehensively determined in combination with the historical trend comparison, so that multi-dimensional abnormal detection is realized, invalid data is quickly excluded to reduce interference, potential abnormalities are avoided from being missed, and a high-reliability initial data basis is provided for subsequent dynamic partition.

[0017] Preferably, the step 2 comprises:

[0018] (21) taking the node where the abnormal measurement or the suspicious measurement is located as an initial expansion center;

[0019] (22) using a breadth-first search algorithm to gradually expand the region range from the initial expansion center to adjacent nodes;

[0020] (23) calculating the weighted error entropy of the measurement data in the current partition after each expansion;

[0021] (24) comparing the change value of the weighted error entropy before and after the expansion: if the change value exceeds the preset upper threshold, continue to expand; if the change value is lower than the preset lower threshold, stop expanding;

[0022] (25) repeating steps (22) to (24) until the expansion process of all partitions meets the stop condition.

[0023] By taking the abnormal or suspicious measurement node as the initial center, dynamically expanding the partition using breadth-first search (BFS), and combining the intelligent control of the change value of the weighted error entropy to expand the process, the range of the partition is adaptively optimized to ensure that the abnormal influence area is completely captured and the calculation efficiency is maintained. Finally, the dynamic partitioning is realized with precision and controllability.

[0024] Preferably, in the partition expansion process of step 2, if there are overlapping nodes between adjacent partitions, the area where the overlapping nodes are located is merged, including: detecting the overlapping nodes between adjacent partitions; merging adjacent partitions containing overlapping nodes to form a new unified partition; updating the partition topology relationship and boundary node information; and recalculating the weighted error entropy of the merged new partition.

[0025] By detecting and merging the overlapping nodes of adjacent partitions, redundant partitions are eliminated and the network structure is optimized to form a more reasonable unified partition. At the same time, the topology relationship and boundary information are updated to ensure data consistency, and the weighted error entropy is recalculated to evaluate the quality of the merged partition, thereby reducing the calculation complexity while improving the accuracy of anomaly analysis and the efficiency of partition management.

[0026] Preferably, the weighted error entropy formula is represented as:

[0027]

[0028] where K max is the maximum value of all measurement residuals r max , p i (r i,k ) is the probability value of the sampling interval k, and ξ i,k represents the corresponding residual value.

[0029] By introducing the maximum residual normalization and probability weighting mechanism, the error distribution characteristics of the measurement data in the partition are quantified: the interval probability is used to reflect the concentration degree of the residual distribution, and the weight is dynamically adjusted combined with the residual amplitude, so that the abnormal influence range is more sensitive captured, and objective and adaptive evaluation basis is provided for partition boundary optimization, effectively balancing local data quality and calculation efficiency.

[0030] Preferably, the determination of the minimum partition boundary range in step 3 includes:

[0031] (31) calculating the weighted error entropy gradient value of each partition boundary node;

[0032] (32) selecting a node with a gradient value change rate less than a preset threshold as a candidate boundary node;

[0033] (33) selecting a node with the minimum weighted error entropy as the final partition boundary node from the candidate boundary nodes.

[0034] By calculating the weighted error entropy gradient value of the partition boundary node, the candidate node with a slow change is screened out, and the node with the minimum entropy value is finally selected as the boundary, so as to accurately determine the minimum partition range under the premise of ensuring the data correction ability, avoid the calculation redundancy caused by excessive expansion, and ensure that the partition boundary is in a reasonable position with the weakest error influence, thereby balancing the efficiency and reliability of dynamic partition.

[0035] Preferably, the measurement data dynamic partitioning method further comprises step (4):

[0036] (41) correcting the abnormal measurement data in each partition;

[0037] (42) recalculating the state estimation results of each partition based on the corrected measurement data;

[0038] (43) integrating the state estimation results of each partition to output the whole network state estimation results.

[0039] By correcting the abnormal measurement data in each partition and recalculating the partition state estimation based on the corrected data, the whole network state estimation results are finally integrated and output, which effectively improves the data accuracy and the reliability of state estimation, and optimizes the calculation efficiency through partition parallel processing, thereby realizing accurate correction of abnormal data and efficient estimation of whole network state.

[0040] In a second aspect, the measurement data dynamic partitioning system comprises:

[0041] an abnormal measurement identification module for identifying abnormal measurement or suspicious measurement by performing coarse screening on the measurement data at the current time;

[0042] a dynamic partitioning expansion module for expanding the range of the region where the abnormal measurement or suspicious measurement is located based on a preset expansion rule;

[0043] a partition boundary determination module for calculating a data correction ability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determining the minimum partition boundary range of the measurement data at the current time according to the index.

[0044] Preferably, the abnormal measurement identification module is specifically used to check the reasonableness of the numerical range of the measurement data and remove measurement data that exceeds the preset reasonable range; to perform consistency verification on the measurement data based on the constraint relationship between electrical quantities and filter out measurement data that does not meet the electrical constraints; to identify abnormal fluctuation or sudden change measurement data from the remaining measurement data by combining the reasonableness analysis of the continuous time section power flow change; to perform trend analysis on the identified abnormal fluctuation or sudden change measurement data and compare it with historical data to identify abnormal or suspicious measurements.

[0045] Preferably, the dynamic partition expansion module is specifically used to take the node where the abnormal or suspicious measurement is located as the initial expansion center; adopt a breadth-first search algorithm to gradually expand the region range from the initial expansion center to adjacent nodes; after each expansion, calculate the weighted error entropy of the measurement data in the current partition; compare the change value of the weighted error entropy before and after the expansion: if the change value exceeds a preset upper threshold, then continue the expansion; if the change value is lower than a preset lower threshold, then stop the expansion; repeat the above steps until the expansion process of all partitions meets the stopping condition.

[0046] Preferably, the dynamic partition expansion module is specifically used to merge the regions where the overlapping nodes are located if there are overlapping nodes between adjacent partitions during the partition expansion process, including: detecting overlapping nodes between adjacent partitions; merging adjacent partitions containing overlapping nodes to form a new unified partition; updating the partition topology and boundary node information; and recalculating the weighted error entropy of the merged new partition.

[0047] Preferably, the dynamic partitioning extension module is specifically used to express the weighted error entropy formula as follows:

[0048]

[0049] Among them, K max r is the maximum value of all measurement residuals max p i (r i,k Let ξ be the probability value obtained by sampling within the interval k. i,k This represents the corresponding residual value.

[0050] Preferably, the partition boundary determination module is specifically used to calculate the weighted error entropy gradient value of each partition boundary node; select nodes whose gradient value change rate is less than a preset threshold as candidate boundary nodes; and select the node with the smallest weighted error entropy among the candidate boundary nodes as the final partition boundary node.

[0051] Preferably, the partition boundary determination module is specifically used to correct abnormal measurement data in each partition; recalculate the state estimation results of each partition based on the corrected measurement data; integrate the state estimation results of each partition, and output the state estimation results of the entire network.

[0052] Thirdly, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed by the dynamic partitioning method for measurement data.

[0053] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned dynamic partitioning method for measurement data.

[0054] Beneficial Effects: Compared with existing technologies, this invention has the following significant advantages: 1. It fully integrates the dynamic changes in measurement data quality, ensuring the dynamic adaptability of the partitioning results. Within a partition, the identification and correction of abnormal data can be carried out in parallel, thereby significantly improving the efficiency of abnormal data processing; 2. By establishing a dynamic partitioning model based on the abnormal measurement center, the identification and correction of abnormal data can be carried out in parallel within the partition, effectively solving the efficiency bottleneck of the traditional serial processing method, reducing the time consumption of large-scale power grid abnormal data processing, and significantly improving processing efficiency; 3. By adopting a partitioning boundary determination method based on weighted error entropy gradient, the optimal partitioning boundary is intelligently selected by quantitatively analyzing the contribution of boundary nodes to data correction, ensuring both the high correlation of data within the partition and avoiding the waste of computing resources caused by excessive partitioning; 4. It innovatively integrates the links of anomaly identification, dynamic partitioning, data correction, and state estimation to form a complete closed-loop processing system, improving processing efficiency while ensuring the accuracy and reliability of the final state estimation results. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0056] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0057] like Figure 1 As shown, the present invention provides a method for dynamic partitioning of measurement data, comprising the following steps:

[0058] S1. Perform a rough screening of abnormal measurements on the measurement data at the current moment to identify abnormal or suspicious measurements.

[0059] S2. Starting from the abnormal or suspicious measurement, gradually expand the range of its area based on the preset expansion rules.

[0060] S3, calculate a data correction capability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determine a minimum partition boundary range of the measurement data at the current time according to the index.

[0061] The embodiments of the present application will be described in detail below.

[0062] S1: coarse screening of abnormal measurement for the measurement data at the current time, and identifying abnormal measurement or suspicious measurement;

[0063] In the embodiments of the present application, coarse screening of abnormal measurement and identification of abnormal measurement or suspicious measurement include:

[0064] Abnormal measurement is preliminarily screened through rationality check of measurement value range, constraint relationship check between different electrical measurements, and rationality check of continuous time section flow change. Specifically, for different types of electrical measurements (such as voltage, current, active power, and reactive power), a reasonable numerical range threshold is preset. For example, the normal range of voltage measurement can be set to 90% to 110% of the rated voltage, and the normal range of current measurement can be determined according to the current carrying capacity of the line or equipment. If the measurement value exceeds the preset reasonable range, it is determined as abnormal measurement or suspicious measurement. In addition, for some special working conditions (such as during equipment switching or fault), the range threshold can be dynamically adjusted to adapt to different operating states.

[0065] In addition to the numerical range and the constraint relationship, the change trend of the measurement data and the comparison with the historical data can also be considered to further improve the accuracy of the discrimination. Specifically, it includes: trend analysis: using a sliding window or a time series analysis method (such as moving average, exponential smoothing) to model the short-term trend of the measurement data, if the current measurement significantly deviates from the predicted trend, it is determined as abnormal; historical data comparison: comparing the current measurement value with the historical data at the same period (such as data under the same load level and similar operation mode), if the deviation exceeds the reasonable range, it is marked as suspicious measurement; machine learning assistance: for complex scenarios, an abnormal detection model (such as Isolation Forest, One-Class SVM) based on historical data can be trained to identify potential abnormalities through unsupervised learning method.

[0066] S2: taking abnormal measurement or suspicious measurement as the starting point, gradually expanding the range of the region around the center;

[0067] In the embodiments of the present application, taking abnormal measurement or suspicious measurement as the starting point, gradually expanding the range of the region around the center, includes:

[0068] From each abnormal / suspected measurement node, a breadth-first search (BFS) algorithm is used to gradually expand the area range and obtain the boundary of each partition to be expanded; specifically including: first checking the adjacent nodes (such as the nearest busbar, line or transformer node) directly connected to the abnormal node; if the measurement data of the first layer nodes still shows abnormality or uncertainty, then continue to expand to the second layer, third layer nodes until the possible affected area is covered; during the expansion process, if the measurement data of a certain layer node is normal, then stop the expansion in this direction to avoid unnecessary calculation overhead; during the expansion process, record the boundary of each partition to be expanded (i.e. the outermost layer of the current search) in order to subsequently perform error entropy evaluation.

[0069] The weighted error entropy of the target measurement after expansion of each partition is calculated. If the weighted error entropy after expansion of the partition still significantly rises or falls, it indicates that the measurement data of the expanded area is helpful to improve the correction ability of the overall data. When the weighted error entropy tends to be stable, it indicates that the partition expansion has reached the effect limit, and the partition process ends at this time.

[0070] S3: During the partition expansion process, if there are overlapping nodes between adjacent partitions, then these overlapping areas should be merged;

[0071] In the embodiment of the present application, if there are overlapping nodes between adjacent partitions, then these overlapping areas should be merged, including the following steps:

[0072] During the partition expansion process, the boundary node set of each partition is monitored in real time. If a node belongs to the boundary or internal node set of two or more partitions to be expanded, then it is determined to be an overlapping node;

[0073] In combination with the power grid topology, the electrical connection relationship of the overlapping node is analyzed, for example: if the overlapping node is a key busbar or a tie switch, it may indicate that multiple abnormal areas exist strong coupling and need to be merged first; if the overlapping node is only an ordinary line node, then the consistency of its measurement data needs to be further evaluated;

[0074] When the overlapping node is detected, whether to trigger merging can be determined according to the following rules: data consistency check: if the state (normal / abnormal) of the measurement data of the overlapping node in different partitions is consistent, then directly merge; error entropy evaluation: calculate the weighted error entropy (WEE) of the partition before and after merging, if the WEE significantly decreases after merging, then perform merging; otherwise, keep the independent partition;

[0075] By merging the overlapping areas, repeated calculation is avoided, the processing of redundant information is reduced, and at the same time the continuity of the partition and the integrity of the data are ensured, thereby improving the rationality and effectiveness of the partition division.

[0076] S4: Use the weighted error entropy of local network measurement data as a data correction capability index to determine the minimum partition boundary range;

[0077] In this embodiment of the invention, the calculation of the weighted error entropy is implemented as follows:

[0078] The weighted error entropy formula is expressed as follows:

[0079]

[0080] Among them, K max r is the maximum value of all measurement residuals max p i (r i,k Let ξ be the probability value obtained by sampling within the interval k. i,k This represents the corresponding residual value.

[0081] S5: First, abnormal data in each partition is processed in a targeted manner. After identifying the type of abnormality through intelligent algorithms, different correction methods are adopted: obvious erroneous data is smoothed and verified using technology, hidden errors are corrected by comparing with historical data, and random noise is filtered. Quality checks are performed at each step to ensure that the corrected data is reasonable and reliable.

[0082] Next, the state of each partition is recalculated based on the corrected data. The most suitable calculation method is selected according to the characteristics of the partition, and parallel computing is used to improve efficiency. Finally, the calculation results of all partitions are intelligently integrated, with a focus on coordinating the consistency of boundary data. The final output is a complete network state estimation result. The whole process achieves efficient and accurate data processing, which not only ensures the calculation speed but also improves the reliability of the results.

[0083] Furthermore, based on a similar inventive concept, embodiments of the present invention also provide a dynamic measurement data partitioning system corresponding to the aforementioned dynamic measurement data partitioning method, comprising:

[0084] The abnormal measurement identification module is used to perform a rough screening of abnormal measurements on the measurement data at the current moment, and to identify abnormal or suspicious measurements.

[0085] The dynamic partitioning expansion module is used to gradually expand the range of the area where the abnormal or suspicious measurement is located, based on a preset expansion rule.

[0086] The partition boundary determination module is used to calculate the data correction capability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determine the minimum partition boundary range of the measurement data at the current moment according to the index.

[0087] Furthermore, the abnormal measurement identification module is specifically used to check the reasonableness of the numerical range of the measurement data and remove measurement data that exceeds the preset reasonable range; to perform consistency verification on the measurement data based on the constraint relationship between electrical quantities and filter out measurement data that does not meet the electrical constraints; to identify abnormal fluctuation or sudden change measurement data from the remaining measurement data by combining the reasonableness analysis of the continuous time section power flow change; to perform trend analysis on the identified abnormal fluctuation or sudden change measurement data and compare it with historical data to identify abnormal or suspicious measurements.

[0088] Furthermore, the dynamic partition expansion module is specifically used to take the node where the abnormal or suspicious measurement is located as the initial expansion center; employing a breadth-first search algorithm, it gradually expands the region range from the initial expansion center to adjacent nodes; after each expansion, it calculates the weighted error entropy of the measurement data within the current partition; it compares the change in weighted error entropy before and after expansion: if the change exceeds a preset upper threshold, it continues to expand; if the change is lower than a preset lower threshold, it stops expanding; it repeats the above steps until the expansion process of all partitions meets the stopping condition.

[0089] Furthermore, the dynamic partition expansion module is specifically used to merge the regions where the overlapping nodes are located if there are overlapping nodes between adjacent partitions during the partition expansion process. This includes: detecting overlapping nodes between adjacent partitions; merging adjacent partitions containing overlapping nodes to form a new unified partition; updating the partition topology and boundary node information; and recalculating the weighted error entropy for the merged new partition.

[0090] Furthermore, the dynamic partitioning extension module is specifically used to express the weighted error entropy formula as follows:

[0091]

[0092] Among them, K max r is the maximum value of all measurement residuals max p i (r i,k Let ξ be the probability value obtained by sampling within the interval k. i,k This represents the corresponding residual value.

[0093] Furthermore, the partition boundary determination module is specifically used to calculate the weighted error entropy gradient value of each partition boundary node; select nodes whose gradient value change rate is less than a preset threshold as candidate boundary nodes; and select the node with the smallest weighted error entropy among the candidate boundary nodes as the final partition boundary node.

[0094] Further, the partition boundary determination module is specifically configured to correct the abnormal measurement data in each partition; recalculate the state estimation result of each partition based on the corrected measurement data; integrate the state estimation result of each partition to output the whole network state estimation result.

[0095] The application further discloses an electronic device.

[0096] Specifically, the electronic device can be a computer device such as a desktop computer, a notebook computer, a palm computer and a cloud server. The computer device can include but is not limited to a processor and a memory. The processor and the memory can be connected through a bus or other means. The processor can be a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, graphics processing units (GPU), embedded neural network processing units (NPU) or other dedicated deep learning coprocessors, discrete gates or transistor logic devices, discrete hardware components, etc. chips, or combinations of the above chips.

[0097] The memory, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and modules. The processor executes various functions and data processing of the processor by running the non-transitory software programs, instructions and modules stored in the memory. The memory can include a program storage area and a data storage area, wherein the program storage area can store application programs required by the control unit and at least one function; the data storage area can store data created by the processor, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory. In some embodiments, the memory can optionally include a memory remotely arranged relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0098] The application further discloses a computer readable storage medium.

[0099] Specifically, the computer readable storage medium is configured to store a computer program, and the computer program is configured to instruct a processor to implement the method in the above method embodiments when executed.

[0100] Those skilled in the art can understand that all or part of the processes in the above method embodiments of the present application can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. When the program is executed, the program can include the processes of the above method embodiments. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), etc. The storage medium can also include a combination of the above types of memories.

Claims

1. A method for dynamically partitioning measurement data, characterized in that, The method comprises the following steps: (1) performing abnormal measurement rough screening on the measurement data at the current moment to identify abnormal measurement or suspicious measurement; (2) taking the abnormal measurement or suspicious measurement as a starting point, and gradually expanding the range of the region based on a preset expansion rule; (3) calculating a data correction capability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determining the minimum partition boundary range of the measurement data at the current moment according to the index.

2. The metrology data dynamic partitioning method of claim 1, wherein, Step 1 of performing abnormal measurement rough screening comprises: (11) performing numerical range rationality checking on the measurement data, and eliminating measurement data beyond the preset reasonable range; (12) performing consistency checking on the measurement data based on the constraint relationship between electrical quantities, and screening out measurement data that does not satisfy the electrical constraint; (13) identifying abnormal fluctuation or mutation measurement data from the remaining measurement data in combination with the rationality analysis of continuous time section power flow change; (14) performing change trend analysis on the identified abnormal fluctuation or mutation measurement data and comparing with historical data to identify abnormal measurement or suspicious measurement.

3. The method of dynamic partitioning of metrology data according to claim 1, wherein, Step 2 comprises: (21) taking the node where the abnormal measurement or suspicious measurement is located as an initial expansion center; (22) using a breadth-first search algorithm to gradually expand the region range from the initial expansion center to adjacent nodes; (23) calculating the weighted error entropy of the measurement data in the current partition after each expansion; (24) comparing the change value of the weighted error entropy before and after expansion: if the change value exceeds a preset upper threshold, continue to expand; if the change value is lower than a preset lower threshold, stop expanding; (25) repeating steps (22) to (24) until the expansion process of all partitions meets the stop condition.

4. The method of dynamic partitioning of metrology data according to claim 1, wherein, In the partition expansion process of step 2, further comprising: if there are overlapping nodes between adjacent partitions, performing merging processing on the region where the overlapping nodes are located, including: detecting the overlapping nodes between adjacent partitions; merging adjacent partitions containing overlapping nodes to form a new unified partition; updating the partition topology relationship and boundary node information; recalculating the weighted error entropy of the merged new partition.

5. The method of dynamic partitioning of metrology data according to claim 3 or 4, wherein, The weighted error entropy formula is represented as: where K max is the maximum value of all measurement residuals r max , p i (r i,k ) is the probability value of sampling the interval k, and ξ i,k represents the corresponding residual value.

6. The method of dynamic partitioning of metrology data according to claim 1, wherein, Step 3 of determining the minimum partition boundary range comprises: (31) calculating the weighted error entropy gradient value of each partition boundary node; (32) selecting nodes with a gradient value change rate less than a preset threshold as candidate boundary nodes; (33) selecting the node with the smallest weighted error entropy as the final partition boundary node from the candidate boundary nodes.

7. The method of dynamic partitioning of metrology data according to claim 1, wherein, Further comprising step (4): (41) performing correction processing on the abnormal measurement data in each partition; (42) recalculating the state estimation results of each partition based on the corrected measurement data; (43) integrating the state estimation results of each partition to output the whole network state estimation result.

8. A metrology data dynamic partitioning system, comprising: Comprise: an abnormal measurement identification module, configured to perform abnormal measurement rough screening on the measurement data at the current moment to identify abnormal measurement or suspicious measurement; a dynamic partition expansion module, configured to take the abnormal measurement or suspicious measurement as a starting point, and gradually expand the range of the region based on a preset expansion rule; A partition boundary determination module is configured to calculate a data correction capability index based on the range of the region and the weighted error entropy of the corresponding local network measurement data, and determine the minimum partition boundary range of the measurement data at the current time according to the index.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement the measurement data dynamic partitioning method of any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the measurement data dynamic partitioning method of any one of claims 1 to 7.

Images