Topology identification of high-voltage metering network based on correlation analysis of power change

Abstract

The invention relates to the technical field of high-voltage metering network topology identification. A traditional method has the problems of dependence on manpower, limited accuracy, poor adaptability, insufficient real-time performance and the like. According to the method, the topological structure is analyzed and recognized based on the correlation of power change, the node power data is collected, the correlation coefficient is calculated by utilizing the Pearson's correlation coefficient model and integrating multiple factors, and a basis is provided for accurate recognition. The working process comprises the steps of data acquisition, transmission, filtering, correlation coefficient calculation, clustering, topological structure optimization and the like. The case shows the processing process of the small-sized and medium-sized high-voltage metering networks. The detailed solution of the drawings presents a system deployment architecture, a method sequence and a process. The deployment architecture displays a data flow direction and a component connection relationship; according to the method sequence, all component tasks and data circulation are determined; the method process shows specific algorithm steps. The high-voltage metering network topology identification method provides an efficient and accurate solution for high-voltage metering network topology identification, and has the advantages of high accuracy, strong adaptability, good real-time performance and the like.

Description

Technical Field

[0001] This invention patent relates to the technical field of high - voltage metering network topology identification. Specifically, it is a method, component device and system for high - voltage metering network topology identification based on the correlation analysis of power changes. Background Art Existing Technologies

[0002] Traditional high - voltage metering network topology identification methods mainly include manual drawing based on equipment ledger information and analysis methods based on simple electrical quantity measurements. The method based on equipment ledger information relies on manual operation, which is not only inefficient but also prone to inaccurate topology structures due to human negligence or untimely equipment information updates. For example, during the frequent upgrading and transformation of power grid equipment, it is difficult for humans to update the ledger information in real - time, resulting in deviations between the topology map and the actual network structure.

[0003] For methods based on simple electrical quantity measurements, such as analyzing only based on single electrical quantities like voltage and current amplitudes, their accuracy is limited. In a complex high - voltage metering network with a complex network structure and diverse load changes, this single - electrical - quantity analysis method is difficult to accurately reflect the true connection relationships between nodes and is prone to misjudgment.

[0004] 2. Technological Development

[0005] With the acceleration of the intelligentization process of the power system, intelligent grid technologies have been continuously developing. Sensor technology, communication technology, and data - processing technology in high - voltage metering networks have made remarkable progress. High - precision power sensors can collect node power data more accurately, providing more accurate basic information for topology identification. The wide application of communication technologies such as the IEC61850 protocol has achieved high - speed and reliable data transmission, ensuring that data can be transmitted from the collection end to the processing center in a timely and accurate manner. In terms of data - processing technology, the development of big - data processing and analysis algorithms has made it possible to process massive metering data, creating conditions for topology identification methods based on more complex algorithms.

[0006] 3. Current Pain Points and Difficulties

[0007] In terms of accuracy: Existing technologies are difficult to accurately identify the topology structure in a complex high - voltage metering network environment. High - voltage metering networks have numerous nodes, complex connections, and frequent dynamic load changes. Traditional methods cannot effectively handle these complex situations, resulting in large errors in topology identification results and affecting the analysis and decision - making of the power system.

[0008] In terms of adaptability: Facing high - voltage metering networks of different scales and structures, traditional methods have poor adaptability. For example, when the network structure changes (such as new lines or nodes are added) or the load characteristics change, traditional methods require a large amount of manual intervention and readjustment and are difficult to automatically adapt to the dynamic changes of the network.

[0009] In terms of real-time performance: The operating state of the power system changes in real time, and the requirements for the real-time performance of topology recognition are getting higher and higher. However, the traditional methods have slow data processing speeds and cannot update the topology structure in a timely manner, failing to meet the needs of real-time monitoring and management.

[0010] 4. This technical solution

[0011] The present invention proposes a method for identifying the topology of a high-voltage metering network based on the correlation analysis of power changes. By using a power data acquisition unit to collect power, voltage, and current parameters, and a data transmission unit to achieve efficient data transmission. After preprocessing the data using a moving average filtering algorithm in the data management and analysis unit, a Pearson correlation coefficient model considering various factors (such as the accuracy of power sensors, the transmission rate of communication modules, the storage capacity of the data processing center, etc.) is used to calculate the correlation of node power changes. Then, based on the correlation matrix, the mean clustering algorithm is used to identify the topology structure, and the data accuracy is further ensured, the calculation process is monitored, and the topology structure is optimized through a preprocessing unit, an intermediate inspection unit, and a postprocessing unit. Finally, the results are output in the form of graphics and data files. This technical solution aims to overcome the deficiencies of the existing technologies, improve the accuracy, adaptability, and real-time performance of high-voltage metering network topology recognition, meet the needs of the intelligent development of modern power systems, and enhance the management and operation levels of high-voltage metering networks. Summary of the Invention

[0012] 1. Principle

[0013] The present invention identifies the topology structure of a high-voltage metering network based on the correlation analysis of power changes. In a high-voltage metering network, nodes that are closely connected to each other have a high correlation in power changes. For example, nodes on the same feeder line, their power change trends are affected by the common load changes and show similar fluctuation patterns. By collecting the node power data and using the Pearson correlation coefficient model to calculate the correlation coefficient of power changes between any two nodes, the degree of connection tightness between nodes can be reflected. When calculating the correlation coefficient, multiple factors affecting the correlation are comprehensively considered, including the weights in the original correlation coefficient calculation part (balanced by coefficients), the accuracy of power sensors (reflected by coefficients), the transmission rate of communication modules (reflected by coefficients), and the storage capacity of the data processing center (represented by coefficients). Through the comprehensive consideration of these factors, the correlation between nodes can be calculated more accurately, providing a basis for accurately identifying the topology structure.

[0014] 2. Implementation

[0015] Component composition and connection implementation

[0016] The power data acquisition unit is responsible for collecting parameters such as power, voltage, and current, and the data it collects is output and connected to the input of the data transmission unit. The sensors in the power data acquisition unit can monitor the electrical quantity information of the node in real time.

[0017] The data transmission unit includes a data sending sub-module, a data receiving sub-module, a protocol conversion sub-module, and a transmission control sub-module. The data sending sub-module packs and encodes the collected data. The protocol conversion sub-module converts the data format output by the sensor into a format compliant with the IEC61850 protocol, and then through the monitoring and management of the transmission control sub-module, the data is transmitted to the input of the data management and analysis unit. The data receiving sub-module is responsible for receiving the feedback information or control instructions from the data processing center.

[0018] The data management and analysis unit includes a data receiving module, a data preprocessing module, a correlation analysis module, and a topology identification module. The data receiving module receives the transmitted data. The data preprocessing module uses the moving average filtering algorithm to clean the data (remove outliers) and smooth the data (eliminate high-frequency noise). The processed data is calculated for correlation by the correlation analysis module using the Pearson correlation coefficient model to generate a correlation matrix. Finally, the topology identification module clusters the nodes according to the correlation matrix through the mean clustering algorithm to obtain the clustering result and initially identify the topology structure.

[0019] In addition, it also includes a preprocessing unit (including a data verification module and a data format conversion module), an intermediate inspection unit (including a data integrity inspection module and a correlation calculation monitoring module), and a postprocessing unit (including a topology structure optimization module and a result output module). The output of the preprocessing unit is connected to the input of the original data preprocessing module of the data processing center to perform preliminary verification and format conversion on the collected data. The input of the intermediate inspection unit is connected to the output of the correlation analysis module in the data management and analysis unit to monitor the correlation analysis result in real time. The input of the postprocessing unit is connected to the output of the original topology identification module in the data management and analysis unit to optimize the identified topology structure and output it.

[0020] 3. Algorithm Step Implementation

[0021] First, the power sensor collects the node power data ( is the node number, is the time).

[0022] Next, with the help of the communication module, the collected data is transmitted to the data processing center according to the IEC61850 protocol. After receiving the data, the data processing center uses the moving average filtering algorithm to filter the data and obtain .

[0023] Then, for any two nodes and , calculate the Pearson correlation coefficient model correlation coefficient of their power changes . The calculation formula is as follows:

[0024] ,

[0025] where is used to balance the weight of the original correlation coefficient calculation part in the whole, , reflects the influence of the power sensor accuracy on the correlation calculation, , reflects the influence of the communication module transmission rate on the correlation, , represents the influence of the data processing center storage capacity on the correlation, is the time series length, and are the means of the power change sequences of nodes and respectively, is the power sensor accuracy (the value range is ), is the communication rate of the communication module (unit: Mbps), is the data processing center storage capacity (unit: GB).

[0026] Finally, based on the calculated correlation matrix , use the mean clustering algorithm to cluster the nodes and obtain the clustering result , and then identify the topological structure of the high-voltage metering network. The topological structure optimization module in the post-processing unit will also optimize the topological structure according to the minimum spanning tree algorithm, remove unreasonable connection relationships, and obtain the optimized topological structure , and then the result output module will output it in the form of graphics and data files.

[0027] 4. Beneficial effects

[0028] Improve accuracy: By comprehensively considering various factors to calculate the correlation coefficient of the node power changes, it can more accurately reflect the true connection relationship between nodes, thereby improving the accuracy of topological recognition. Compared with traditional methods, the present invention can more accurately identify the topological structure in a complex high-voltage metering network environment and reduce misjudgment.

[0029] Enhanced adaptability: The correlation analysis method based on power change does not depend on specific network structures and load characteristics, and can adapt to high-voltage metering networks of different scales and structures. Whether it is a change in the network structure (such as adding or deleting lines and nodes) or a change in load characteristics, the present invention can quickly and accurately identify the topological structure by recalculating the correlation coefficient and performing cluster analysis, without the need for a large amount of manual intervention.

[0030] Improved real-time performance: The data acquisition, transmission, and processing link of the present invention is reasonably designed, enabling fast data acquisition, efficient transmission, and timely processing. In the case of real-time changes in the operating state of the power system, the topological structure can be updated in a timely manner to meet the requirements of real-time monitoring and management, which helps to improve the operating efficiency and reliability of the power system.

[0031] Optimized data processing process: The pre-processing unit ensures the accuracy and format consistency of the acquired data. The intermediate inspection unit monitors the correlation calculation results in real time. The post-processing unit optimizes the topological structure and outputs the results. The entire data processing process is more perfect, improving the quality and efficiency of data processing, and helping to enhance the management level of the high-voltage metering network.

[0032] Description of the drawings

[0033] Figure 1 Deployment schematic diagram of topological identification of high-voltage metering network based on correlation analysis of power change

[0034] Figure 2 Schematic diagram of the sequence of the topological identification method of the high-voltage metering network based on correlation analysis of power change

[0035] Figure 3 Schematic diagram of the process of the topological identification method of the high-voltage metering network based on correlation analysis of power change

[0036] Figure 4 Schematic diagram of the grouping of the topological 5-node network

[0037] Figure 5 Schematic diagram of the topological 15-node network

[0038] Figure 6 Schematic diagram of the sequence diagram of the power data acquisition unit

[0039] Figure 7 Composition diagram of the data management and analysis unit

[0040] Figure 8 Relationship diagram of the power data transmission unit

[0041] Figure 9 Component relationship diagram of the pre-intermediate-post processing process

[0042] Figure 10 Component relationship diagram of the high-voltage metering network topological identification components

[0043] Figure 11 Component Relationship Diagram of High-Voltage Metering Network Topology Identification Device

[0044] Figure 12 Component Relationship Diagram of High-Voltage Metering Network Topology Identification System Specific Embodiments

[0045] Embodiment 1, in combination with Technical Solution 1, Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 The detailed explanation is as follows:

[0046] Among them, Case 1 and Case 2 are both specific cases, each containing 5 and 15 nodes respectively. Generally, it can be extended to a scale from hundreds to thousands and tens of thousands, only limited by the physical performance limit, and can meet the requirements in terms of principle. Specifically as follows:

[0047] 1. Case 1

[0048] Suppose there is a small high-voltage metering network with 5 nodes (numbered 1 - 5). The power data acquisition unit collects the power data of each node within a certain period as follows (unit: kW):

[0049] Node number t = 1 s t = 2 s t = 3 s t = 4 s t = 5 s 1 50 52 48 51 49 2 48 50 46 49 47 3 60 62 58 61 59 4 55 57 53 56 54 5 70 72 68 71 69

[0050] The working process is as follows:

[0051] 1. Data acquisition: The power sensor collects the power data of the above nodes.

[0052] 2. Data transmission: The data transmission unit transmits the data to the data processing center.

[0053] 3. Data filtering: The data processing center filters the data using the moving average filtering algorithm (assuming the filtering window is 3). Taking Node 1 as an example, the calculation process is as follows:

[0054]

[0055]

[0056]

[0057] 4. Calculate the correlation coefficient:

[0058] Assume the power sensor accuracy (then ), communication rate (then ), data processing center storage capacity (then ), .

[0059] Calculate the correlation coefficient r12 between node 1 and node 2:

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066] Calculate the correlation coefficients between other nodes in the same way to obtain the correlation matrix R.

[0067] 5. Node clustering: Based on the correlation matrix, use the K-means clustering algorithm (assuming ) to cluster the nodes to obtain the clustering result . For example, nodes 1, 2, and 4 may be clustered into one class, and nodes 3 and 5 may be clustered into another class to initially identify the topological structure.

[0068] 6. Topological structure optimization: The topological structure optimization module in the post-processing unit optimizes the topological structure according to the minimum spanning tree algorithm, removes unreasonable connection relationships, and obtains the optimized topological structure .

[0069] 7. Result output: The final result output module will output in the form of graphics and data files as Figure 4 shown.

[0070] 2. Case 2

[0071] Consider a medium-sized high-voltage metering network with 15 nodes. At a certain moment t, after the power data of each node is collected, it undergoes data transmission and preprocessing. Assume that nodes 6, 7, and 8 are nodes on the same line, and their power change trends are similar.

[0072] When calculating the correlation coefficient, according to the above formula, under the conditions of specific power sensor accuracy, communication module transmission rate, and data processing center storage capacity (assuming power sensor accuracy , communication module communication rate , data processing center storage capacity ).

[0073] Take the correlation coefficient r67 between computing nodes 6 and 7 as an example:

[0074] 1. First, calculate the mean value:

[0075]

[0076]

[0077] 2. Then, calculate the numerator part:

[0078]

[0079] 3. Calculate the denominator part:

[0080]

[0081]

[0082] 4. Calculate the correlation coefficient:

[0083]

[0084] Since , for the coefficient , .

[0085]

[0086] Calculating the correlation coefficients between node 6 and node 8, and between node 7 and node 8 in the same way also results in relatively high values. During the clustering process, they are very likely to be clustered into one category, which is consistent with their connection relationship on the same line in the actual high-voltage metering network. Through the correlation calculation and clustering analysis of all the network nodes, the topological structure of the high-voltage metering network is finally identified, and after being optimized and output by the post-processing unit, as Figure 5 shown.

[0087] The following is a detailed explanation Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 :

[0088] Figure 1 : Topological Identification Deployment of High-Voltage Metering Network Based on Correlation Analysis of Power Changes

[0089] This figure shows the deployment architecture of the entire high-voltage metering network topology identification system. From left to right, it is mainly divided into three parts: the power data acquisition end (field devices), the data transmission network, and the data processing center (server). The power data acquisition unit is located at the power data acquisition end and is responsible for collecting parameters such as power, voltage, and current of the nodes. The collected data is transmitted to the data processing center through the data transmission network via the data transmission unit. The data transmission unit ensures the stable and efficient transmission of data, including the packing and encoding of the data sending sub-module, the protocol conversion sub-module to convert the data format into a format compliant with the IEC61850 protocol, and the monitoring and management of the transmission control sub-module. The data management and analysis unit in the data processing center processes and analyzes the received data. Each part works together to achieve the topology identification function. This deployment architecture clearly presents the data flow from acquisition to processing, reflecting the connection relationship and functional division of each component of the system at the physical level.

[0090] Figure 2 : Sequence of the high-voltage metering network topology identification method based on correlation analysis of power changes

[0091] This figure describes the interaction sequence between the user and each component of the system and the data processing flow in the form of a sequence diagram. The user first starts data acquisition, triggering the power data acquisition unit to perform the acquisition operation. The collected data is sent by the power data acquisition unit to the data transmission unit, and the data transmission unit processes and transmits the data, transmitting the processed data to the data management and analysis unit. After receiving the data, the data management and analysis unit performs a series of data processing and analysis operations, and finally provides the topology structure identification result to the user. This figure intuitively shows the operation sequence of the entire method, clarifies the tasks of each component at different stages and the data flow process, and helps to understand the dynamic operation mechanism of the system.

[0092] Figure 3 : Schematic diagram of the high-voltage metering network topology identification method based on correlation analysis of power changes

[0093] This flow chart details the specific algorithm steps for high-voltage metering network topology identification. Starting from collecting node power data by the power sensor, then transmitting the data to the data processing center through the communication module, and the data processing center performs preprocessing. Then, calculate the Pearson correlation coefficient model correlation coefficient of power changes for any two nodes, and then use the mean clustering algorithm for clustering, obtain the clustering result and then identify the topology structure. The entire process has clear steps and rigorous logic, which is a high-level summary of the algorithm implementation process and helps to deeply understand the core algorithm process of the technical solution.

[0094] Figure 4 、 Figure 5 is the schematic diagram of the node topology; Figure 4It is a schematic diagram of packet division of a topological 5-node network. There are two types of nodes in the figure. The first type of nodes includes nodes 1 and 2, and the second type of nodes includes nodes 3, 4, and 5. Node 1 and Node 2 are respectively connected to Node 3, and Node 3 is connected to Node 4 and Node 5.

[0095] Figure 5 It is a schematic diagram of a topological 15-node network. There are also two types of nodes in the figure. The first type of nodes includes 6, 7, 8, and the second type of nodes includes 1, 2, 3, 4, 5, 10, 9, 11, 12, 13, 14, 15. The connection relationships between the nodes are as follows: Node 1 is connected to Node 3, Node 2 is connected to Node 4, Node 3 is connected to Node 5, Node 4 is connected to Node 9, Node 5 is connected to Node 10, Node 9 is connected to Node 14, Node 10 is connected to Node 15, Node 14 is connected to Node 11, Node 15 is connected to Node 12, Node 11 is connected to Node 13, Node 6 is connected to Node 3, Node 7 is connected to Node 4, Node 8 is connected to Node 5. These connection relationships form a relatively complex network topology structure.

[0096] Embodiment 2, in combination with Technical Solution 2, Figure 6 is explained in detail as follows:

[0097] Claim 2 states that the power data acquisition unit includes acquiring power, voltage, and current parameters. It can be seen from the sequence diagram schematic of the power data acquisition unit in Figure 6 its working process and data flow direction.

[0098] Data acquisition startup: The user first starts the data acquisition operation, and this action triggers the power data acquisition unit to start working, thus initiating the entire data acquisition process.

[0099] Acquiring parameter information: The sensors in the power data acquisition unit continuously monitor the electrical quantity information of the nodes. As described in Embodiment 2, the electrical quantity information acquired here includes power, voltage, and current parameters. For example, in an actual high-voltage metering network, the sensors will accurately obtain the power, voltage, and current values of each node at different times, providing a comprehensive data basis for subsequent analysis.

[0100] Data transmission preparation: The acquired parameter data such as power, voltage, and current are preliminarily processed in the power data acquisition unit and then prepared to be transmitted to the data transmission unit. During this process, the accuracy and integrity of the data are ensured to prepare for further processing in the data management and analysis unit.

[0101] Data Transmission and Subsequent Processing: The power data acquisition unit outputs and connects the data to the input of the data transmission unit, and the data transmission unit then transmits the data to the data management and analysis unit. In the data management and analysis unit, the data reception module receives the data, the data preprocessing module uses the moving average filtering algorithm to clean and smooth the data, the correlation analysis module calculates the correlation using the Pearson correlation coefficient model, and the topology identification module clusters the nodes through the mean clustering algorithm based on the correlation matrix, and finally identifies the topology structure.

[0102] Through the combination of Embodiment 2 and Figure 6, the functions, working processes of the power data acquisition unit in the entire high-voltage metering network topology identification system and its cooperation with other units can be clearly understood, ensuring the accuracy and effectiveness starting from the data acquisition source.

[0103] Embodiment 3, combined with Technical Solution 3, Figure 7 is explained in detail as follows:

[0104] Embodiment 3 makes detailed regulations on the data management and analysis unit, and Figure 7 visually presents its composition structure. The combination of the two can deeply analyze the key significance of this unit in the high-voltage metering network topology identification. The data reception module, as the starting point of the data management and analysis process, is responsible for receiving the power data transmitted from the communication module. In the high-voltage metering network, the power data of each node is collected by the power data acquisition unit and then sent by the communication module. This module ensures the accurate reception of the data, laying the foundation for subsequent processing. Then, the data preprocessing module performs data cleaning (removing outliers) and smoothing processing (using a filtering algorithm to eliminate high-frequency noise) on the received data. Since the actual collected data is easily interfered, such as outliers generated by equipment failures and high-frequency noise introduced by electromagnetic interference, after being processed by a specific algorithm (such as the moving average filtering algorithm), the data becomes more stable and reliable, facilitating subsequent correlation analysis and improving the accuracy of topology identification. Subsequently, the correlation analysis module calculates the correlation of power changes using the Pearson correlation coefficient model based on the preprocessed data and generates a correlation matrix. Since the power changes of closely connected nodes in the high-voltage metering network are related, and the power change trends of nodes on the same feeder are similar due to the influence of common loads, this module clearly reflects the degree of connection tightness between nodes by accurately calculating the coefficients between nodes, providing a key basis for topology identification. Finally, the topology identification module clusters the nodes using the K-means clustering algorithm based on the correlation matrix to identify the topology structure. It classifies the nodes according to the degree of node correlation, and nodes with high correlation are clustered into one category. For example, through this clustering analysis of a medium-sized high-voltage metering network, the node connection relationship can be clarified, and the network topology structure can be outlined, providing important support for the management and operation of the high-voltage metering network. The cooperation of each module ensures the efficient and accurate operation of the entire data management and analysis unit.

[0105] Embodiment 4, combined with Technical Solution 4,Figure 8 The detailed explanation is as follows:

[0106] Embodiment 4 details the composition of the power data transmission unit, including a data sending sub-module, a data receiving sub-module, a protocol conversion sub-module, and a transmission control sub-module. And Figure 8 clearly shows the connection relationships and cooperation processes among these sub-modules. The data sending sub-module packs and encodes the parameter data such as power, voltage, and current collected by the power sensor, organizes and converts the discrete data into a format suitable for transmission according to rules. Just like packing and labeling scattered items, it ensures that the data enters the transmission link accurately and completely, providing a reliable data source for subsequent processing. The protocol conversion sub-module, as a "translator", converts the non-IEC61850 protocol format output by the power sensor into this protocol format, enabling it to communicate compatibly with the data processing center, avoiding transmission errors caused by format incompatibility, and ensuring a smooth link. The transmission control sub-module monitors and manages during data transmission, real-time monitors the status, responds to situations such as network congestion and signal interference, tracks the sending and receiving of data packets, and takes measures such as retransmission or speed adjustment when abnormalities are found, ensuring that the data reaches the processing center in a timely and complete manner, meeting the real-time requirements, and providing timely and accurate data for topology recognition. The data receiving sub-module is responsible for receiving the feedback information or control instructions from the processing center, realizing effective interaction between the two. If the processing center issues an instruction when it discovers data problems, this module ensures that the instruction is correctly received and executed, guaranteeing the collaborative work of the system. Through the combination of Embodiment 4 and Figure 8 it can be seen that the sub-modules of the power data transmission unit work together to ensure the efficient, accurate, and stable transmission of data between the acquisition end and the processing center, providing a solid guarantee for topology recognition.

[0107] Embodiment 5, combined with Technical Solution 5, Figure 8 The detailed explanation is as follows:

[0108] The adjustment coefficient in the Pearson correlation coefficient model specified in Embodiment 5 (value range, balancing the calculation weights of the original correlation coefficient and introducing the influence of other factors), (related to the accuracy of the power sensor, the higher the accuracy, the greater the impact on the correlation calculation), (related to the communication rate of the communication module, the faster the rate, the greater the impact), (related to the storage capacity of the data processing center, the larger the capacity, the greater the impact), combined with Figure 9 showing the component relationships of the pre-processing unit, intermediate inspection unit, post-processing unit, and data management and analysis unit, profoundly affects the accuracy and reliability of the high-voltage metering network topology identification. The data verification module of the pre-processing unit verifies the collected power data. After correct verification, it is converted into a standard format by the format conversion module to ensure the accuracy and consistency of the data, laying a foundation for subsequent calculations. The data integrity check module of the intermediate inspection unit monitors the correlation analysis data. If it is complete, the correlation calculation monitoring module continues to monitor the calculation. If it is incomplete, it triggers a re-sampling or processing mechanism to ensure the accuracy and reliability of the calculation. The topology structure optimization module of the post-processing unit optimizes the preliminarily identified topology structure using the minimum spanning tree algorithm, removing unreasonable connections. The result output module then outputs the optimized topology structure in graphical and file forms for easy viewing and analysis. In short, the combination of the two clearly presents the coefficient influence and the cooperation of each unit to ensure the quality of the entire topology identification process.

[0109] Embodiment 6, combined with Technical Solution 6, Figure 9 is explained in detail as follows:

[0110] Embodiment 6 is further enriched on the basis of the original high-voltage metering network topology identification method, introducing a pre-processing unit, an intermediate inspection unit, and a post-processing unit. Figure 9 clearly presents the relationships between the components of the high-voltage metering network topology identification. The combination of the two comprehensively improves the quality and efficiency of the topology identification.

[0111] The pre-processing unit includes a data verification module and a data format conversion module, and its output is connected to the input of the original data pre-processing module in the data processing center. After data collection, the data verification module carefully checks the accuracy of the collected power data, such as through checksum calculation. If the verification is correct, the data format conversion module converts the data into a standard format suitable for subsequent processing, ensuring that the data can smoothly enter the data processing center for further analysis, effectively avoiding subsequent problems caused by inconsistent data formats or data errors, and laying a solid data foundation for the entire topology identification process.

[0112] The intermediate inspection unit includes a data integrity inspection module and a correlation calculation monitoring module, and its input is connected to the output of the correlation analysis module in the data management and analysis unit. During the correlation analysis process, the data integrity inspection module checks in real time whether the data is complete. Once it is found that the data is missing or incomplete, it can trigger the data re-acquisition or processing mechanism in time to ensure the integrity of the analysis data. At the same time, the correlation calculation monitoring module closely monitors the correlation calculation process, monitors the calculation results in real time, ensures the accuracy of the calculation, and provides a reliable basis for the accurate identification of the topological structure.

[0113] The post-processing unit includes a topological structure optimization module and a result output module, and its input is connected to the output of the original topological identification module in the data management and analysis unit. After the topological identification module initially identifies the topological structure, the topological structure optimization module uses an optimization algorithm (such as the minimum spanning tree algorithm) to optimize it, removes unreasonable connection relationships, and obtains a more accurate and concise topological structure. Subsequently, the result output module outputs the optimized topological structure in the form of graphics and data files, which is convenient for relevant personnel to intuitively view and deeply analyze the topological structure of the high-voltage metering network, and provides strong support for the operation management and decision-making of the power system.

[0114] By closely combining Embodiment 6 with Figure 9, it can be clearly seen the positions and collaborative effects of each processing unit in the entire topological identification system. From the front-end verification of data to the intermediate monitoring and then to the back-end optimization output, each link cooperates with each other, greatly improving the accuracy, reliability and practicability of the topological identification of the high-voltage metering network.

[0115] Embodiment 7, combined with Technical Solution 7, Figure 9 is explained in detail as follows:

[0116] Embodiment 7 clarifies that the topological structure optimization module in the post-processing unit optimizes the topological structure according to the minimum spanning tree algorithm. Figure 9 It shows the component relationship of the high-voltage metering network topological identification device, and the combination of the two is of great significance for improving the topological identification effect. In the high-voltage metering network, the initially identified topological structure often has redundant or unreasonable connections, such as false connections caused by temporary faults or measurement errors. The minimum spanning tree algorithm adopted by the topological structure optimization module can, according to the node connection weights (related to factors such as the correlation with power changes), remove unnecessary connections while ensuring the network connectivity, and obtain the optimized topological structure , making the topological structure more concise and accurate, which is conducive to work such as power system analysis. From Figure 9 it can be seen that in the entire high-voltage metering network topological identification device, the power data acquisition unit collects data, which is transmitted to the data management and analysis unit through the transmission unit for processing and analysis to obtain the initial topological structure. The topological structure optimization module is closely connected to the topological identification module and receives And after optimization, it is transmitted to the result output module. The preprocessing unit ensures data quality, the intermediate inspection unit monitors the calculation results, and each component works in coordination. The topology optimization module plays a crucial role in improving the quality of the final topology structure, jointly contributing to the accurate identification and optimized output of the high-voltage metering network topology, and strongly supporting the efficient operation and management of the power system.

[0117] Embodiment 8, in combination with Technical Solution 8, Figure 10 is elaborated in detail as follows:

[0118] The realization of the high-voltage metering network topology recognition function based on specific components. The components involved in Embodiment 8 are key parts in the high-voltage metering network topology recognition system, which includes at least one unit that can execute the high-voltage metering network topology recognition method for correlation analysis based on power changes. From Figure 10 it can be seen the core position of this component in the system, and it works closely and in coordination with other components. At the data acquisition end, its power data acquisition unit uses sensors to collect node power data , providing raw data for subsequent processes. These data, in cooperation with the data transmission unit, are transmitted to the data processing center after operations such as packaging and encoding. In the data processing center, it works in coordination with the data management and analysis unit. For example, the data preprocessing module processes the data it collects using a moving average filtering algorithm, and the correlation analysis module calculates the correlation coefficient considering component-related factors, and the topology recognition module also relies on it when clustering based on the correlation matrix to obtain . At the same time, it cooperates with the preprocessing unit to verify and convert the format of the collected data; the intermediate inspection unit monitors the results of its correlation analysis; the post-processing unit optimizes and outputs the provided by it. The effective operation of this component improves the accuracy of the system in identifying the topology structure, enhances adaptability because its function does not depend on specific structures and load characteristics, and improves real-time performance through efficient cooperation, contributing to the improvement of the operation efficiency, reliability of the power system and the management level of the high-voltage metering network.

[0119] Embodiment 9, in combination with Technical Solution 9, Figure 11 is elaborated in detail as follows:

[0120] A device using the components as described in Embodiment 8, characterized in that the device includes at least one component as defined in Embodiment 8. These components cooperate with each other to jointly execute the high-voltage metering network topology recognition method for correlation analysis based on power changes to achieve the effective recognition of the high-voltage metering network topology structure.

[0121] Specifically, the components in the device work together in the following way: The power data acquisition unit is responsible for collecting parameters such as power, voltage, and current, and the data it collects is transmitted via the data transmission unit. The data sending sub-module in the data transmission unit packs and encodes the collected data, the protocol conversion sub-module converts the data format into a format compliant with the IEC61850 protocol, and the transmission control sub-module monitors and manages the transmission process to ensure the complete and timely transmission of data to the data management and analysis unit. After receiving the data, the data preprocessing module in the data management and analysis unit uses the moving average filtering algorithm to clean and smooth the data, the correlation analysis module uses the Pearson correlation coefficient model to calculate the correlation of power changes and generates a correlation matrix, and the topology identification module clusters the nodes using the mean clustering algorithm based on the correlation matrix to obtain the clustering result, and then preliminarily identifies the topology of the high-voltage metering network.

[0122] In addition, the preprocessing unit (including the data verification module and the data format conversion module) in the device performs preliminary verification and format conversion on the collected data to ensure data accuracy and format consistency; the intermediate inspection unit (including the data integrity inspection module and the correlation calculation monitoring module) monitors the correlation analysis results in real time; the post-processing unit (including the topology structure optimization module and the result output module) optimizes the identified topology structure and outputs it. Through such a collaborative working mechanism, the entire device can accurately and efficiently identify the topology of the high-voltage metering network, adapt to different network environments and operation requirements, and provide strong support for the stable operation and management of the power system.

[0123] Embodiment 10, in combination with Technical Solution 10, Figure 12 is explained in detail as follows:

[0124] Embodiment 10 relates to a high-voltage metering network topology identification system, which includes at least one device using the device as in Embodiment 9, and these devices are connected to each other and work together to form a complete high-voltage metering network topology identification system. In this system, each device further includes at least one unit using the method as in Embodiment 7, and these units can execute the high-voltage metering network topology identification method based on the correlation analysis of power changes, collect parameters such as power through the power data acquisition unit, transmit data with the help of the data transmission unit, and process and analyze data using the data management and analysis unit, including filtering the data using the moving average filtering algorithm, calculating the correlation coefficient of the Pearson correlation coefficient model of power changes, and using The mean clustering algorithm identifies a series of operations such as topological structures, and further ensures data accuracy, monitors the calculation process, and optimizes the topological structure through a preprocessing unit, an intermediate inspection unit, and a postprocessing unit. Finally, the topological structure recognition results are output in the form of graphics and data files, realizing the accurate recognition of the topological structure of the high-voltage metering network to meet the needs of the intelligent development of modern power systems and improve the management and operation level of the high-voltage metering network.

Claims

1. A method for identifying high-voltage metering network topology based on correlation analysis of power changes, characterized in that: include: Power data acquisition unit, data management and analysis unit, data transmission unit; Parts Contact: The output of the power data acquisition unit is connected to the input of the data transmission unit, and the output of the data transmission unit is connected to the input of the data management and analysis unit, forming a link for data acquisition, transmission and processing. Algorithm steps: Step SA01: Power sensor collects node power data , is the node number, For time; Step SA02: With the help of the communication module, the collected data is transmitted to the data processing center according to the IEC61850 protocol; Step SA03: After receiving the data, the data processing center performs preprocessing and applies the moving average filtering algorithm. Filter the data to get ; Step SA04: For any two nodes and , calculate the Pearson correlation coefficient model correlation coefficient of its power change , the calculation formula is: in , , , The adjustment factor Used to balance the weight of the original correlation coefficient calculation part in the whole. Reflects the impact of power sensor accuracy on correlation calculations, Reflects the impact of the communication module transmission rate on the correlation, represents the effect of the storage capacity of the data processing center on the correlation, is the length of the time series, and Node and The mean of the power variation series, is the power sensor accuracy, the value range here is 0.5%-100%, The communication rate of the communication module, in Mbps. is the storage capacity of the data processing center, in GB. , , is a constant and can be adjusted according to actual conditions to optimize the influence of the coefficient on the correlation calculation; Step SA05: Based on the calculated correlation matrix ,use Mean Clustering Algorithm Cluster the nodes and obtain the clustering results , and then identify the topology of the high-voltage metering network.

2. The method according to claim 1, characterized in that: The power data acquisition unit also includes collecting power, voltage, and current data parameters.

3. The method according to claim 2, characterized in that in, The data management and analysis unit includes a data receiving module, a data preprocessing module, a correlation analysis module and a topology recognition module; The data receiving module receives the power data transmitted from the communication module; The data preprocessing module cleans the received data, removes outliers, and uses a smoothing filter algorithm to eliminate high-frequency noise, making the data more suitable for subsequent analysis; , The correlation analysis module uses the Pearson correlation coefficient model to calculate the correlation of power changes and generate a correlation matrix; The topology identification module clusters the nodes based on the correlation matrix and uses the K-means clustering algorithm to identify the topological structure of the high-voltage metering network.

4. The method according to claim 3, characterized in that in, The power data transmission unit includes a data transmission submodule, a data receiving submodule, a protocol conversion submodule and a transmission control submodule; The data transmission submodule packages and encodes the data collected by the power sensor for transmission; The data receiving submodule receives feedback information or control instructions from the data processing center; The protocol conversion submodule is responsible for converting the data format output by the power sensor into a format that complies with the IEC61850 protocol to achieve compatible communication with the data processing center; The transmission control submodule monitors and manages the data transmission process to ensure the integrity and timeliness of data transmission.

5. The method according to claim 4, characterized in that in, In the Pearson correlation coefficient model, the , , , ,include: Said The value range is ,This coefficient is mainly used to balance the weight of the original correlation coefficient calculation part in the whole, ensuring that the correlation of the power change itself dominates the calculation, while reasonably introducing the influence of other factors; Said The value of is consistent with the power sensor accuracy Related, its value function is ,in ; Said The value of the communication module communication rate Related, its value function is ,in ; Said The value of and the storage capacity of the data processing center Related, its value function is ,in , this coefficient represents the impact of the storage capacity of the data processing center on the correlation.

6. The method according to claim 5, characterized in that Also includes, among which Parts list: The pre-processing unit includes a data verification module and a data format conversion module, the intermediate inspection unit includes a data integrity inspection module and a correlation calculation monitoring module, and the post-processing unit includes a topology structure optimization module and a result output module; Component relationship: The output of the pre-processing unit is connected to the input of the original data pre-processing module of the data processing center, forming a connection link between data acquisition and pre-processing. After the data collected by the power sensor is initially checked and format converted by the pre-processing unit, it smoothly enters the data processing center for further processing to ensure the accuracy and format consistency of the data; The input of the intermediate inspection unit is connected to the output of the correlation analysis module in the data processing center to monitor the correlation analysis results in real time; The input of the post-processing unit is connected to the output of the original topology recognition module in the data processing center to optimize and output the recognized topology structure; step: Step SS01: The data verification module in the pre-processing unit verifies the collected power data Perform verification. If the verification is correct, the data is converted into a standard format through the data format conversion module. , if the checksum is wrong, it is marked as wrong data; Step SS02: The data integrity check module in the intermediate check unit checks whether the data after the correlation analysis is complete. If it is complete, the correlation calculation monitoring module continues to monitor the correlation calculation results. If it is incomplete, the data re-collection or processing mechanism is triggered; Step SS03: The topology optimization module in the post-processing unit optimizes the identified topology Optimize and use optimization algorithms Remove unreasonable connection relationships to obtain an optimized topological structure , then the result output module will Output in the form of graphics and data files.

7. The method according to claim 6, characterized in that It also includes that the topology structure optimization module in the post-processing unit optimizes the topology structure according to the minimum spanning tree algorithm to remove unnecessary connection relationships.

8. A component using the method according to claim 7, characterized in that: The method comprises at least one unit using the method according to claim 7, and the unit is capable of executing the method for identifying the topology of the high-voltage metering network based on the correlation analysis of power changes.

9. A device using the component according to claim 8, characterized in that: The method comprises at least one use component according to claim 8, which is capable of executing the high-voltage metering network topology identification method based on power variation correlation analysis.

10. A system using the device as claimed in claim 9, characterized in that: It comprises at least one device as claimed in claim 9, which is capable of executing the high-voltage metering network topology identification method based on power change correlation analysis, and the devices are interconnected and work together to form a complete high-voltage metering network topology identification system.

Images