High-precision Data Analysis Platform for Complex Power Networks

Abstract

The present invention discloses a high-precision data analysis platform for complex power networks. The platform includes: obtaining a topological node set and a topological node position set; collecting data upload features to obtain a set of topological node data upload feature groups; performing topological node pairing analysis on the topological node set to determine topological node pairing groups; extracting a set of group upload data; performing data processing timeliness analysis to determine a set of group data processing times; and performing collaborative analysis scheme identification to obtain a target collaborative data analysis scheme. The present invention solves the technical problems in the prior art that the processing of complex power network data lacks systematic planning, resulting in low analysis accuracy, slow overall timeliness, and poor resource collaborative utilization, and realizes the full-process refined control and efficient integrated analysis of complex power network data, improving the data analysis accuracy, processing timeliness, and resource collaborative utilization efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and more specifically, to a high-precision data analysis platform for complex power networks. Background Art

[0002] With the increasing complexity of modern power systems, especially with the introduction of smart grids and distributed energy sources, the data volume and complexity of power networks have increased significantly. The efficient operation of power networks depends on the monitoring, analysis, and processing of real-time data, especially in aspects such as the state assessment, fault diagnosis, and load forecasting of power networks. Most traditional power network monitoring systems rely on centralized data collection and processing methods. Although they can obtain data from some nodes in real time, due to the large amount and wide distribution of data, the timeliness and accuracy of information transmission and processing cannot be effectively guaranteed. In addition, due to the complex topological structure in power networks, the connection and information transmission between nodes show highly non-linear and time-varying characteristics, making it difficult for traditional data analysis methods to handle large-scale and high-dimensional data.

[0003] The existing technology has technical problems such as the lack of systematic planning for the processing of complex power network data, resulting in low analysis accuracy, slow overall timeliness, and poor resource collaborative utilization. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the above-mentioned existing technology and provide a high-precision data analysis platform for complex power networks to achieve fine control and efficient integrated analysis of the entire process of complex power network data, and improve the data analysis accuracy, processing timeliness, and resource collaborative utilization efficiency.

[0005] In view of the above problems, the present invention adopts the following technical solutions: A high-precision data analysis platform for complex power networks, comprising:

[0006] A network topology structure acquisition module, configured to acquire the network topology structure of a target power network, extract the nodes and node positions of the network topology structure, and obtain a topological node set and a topological node position set;

[0007] An uploaded feature group set acquisition module, configured to traverse the topological node set for data upload feature collection according to a preset data upload feature group, and obtain a topological node data upload feature group set;

[0008] A topological node pairing group determination module, configured to perform topological node pairing analysis on the topological node set based on the topological node data upload feature group set and the topological node position set, and determine K topological node pairing groups, where K is an integer greater than or equal to 1;

[0009] The group upload data set extraction module is used to extract the K group upload data sets of the K topological node pairing groups within a preset time window, where each group upload data corresponds to a topological node;

[0010] The group data processing time set determination module is used to traverse the K group upload data sets for data processing timeliness analysis and determine the K group data processing time sets;

[0011] The target collaborative data analysis scheme acquisition module is used to aim at reducing the overall analysis time and reducing the concurrent analysis of the pairing groups. Based on the K upload time sets of the K group upload data sets and the K group data processing time sets, it identifies the collaborative analysis scheme for the K topological node pairing groups, obtains the target collaborative data analysis scheme, and performs data analysis processing on the K group upload data sets of the K topological node pairing groups according to the target collaborative data analysis scheme.

[0012] One or more technical solutions provided by the present invention have at least the following technical effects or advantages: The present invention realizes the full-process refined control and efficient integrated analysis of complex power network data, and improves the data analysis accuracy, processing timeliness and resource collaborative utilization efficiency. Description of the Drawings

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings.

[0014] Figure 1 It is a schematic structural diagram of a high-precision data analysis platform for complex power networks provided by an embodiment of the present invention;

[0015] Figure 2 It is a schematic flow diagram of the topological node pairing group determination module in the high-precision data analysis platform for complex power networks provided by an embodiment of the present invention.

[0016] Description of the reference numerals: 10 - Network topology structure acquisition module, 20 - Upload feature group set acquisition module, 30 - Topological node pairing group determination module, 40 - Group upload data set extraction module, 50 - Group data processing time set determination module, 60 - Target collaborative data analysis scheme acquisition module. Detailed Embodiments

[0017] The present invention provides a high-precision data analysis platform for complex power networks, which is used to solve the technical problems in the prior art that the data processing of complex power networks lacks systematic planning, resulting in low analysis accuracy, slow overall timeliness, and poor resource collaborative utilization.

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0019] As Figure 1 shown, the present invention provides a high-precision data analysis platform for complex power networks, and the platform includes:

[0020] A network topology structure acquisition module 10, which is used to acquire the network topology structure of the target power network, extract the nodes and node positions of the network topology structure, and obtain a topological node set and a topological node position set.

[0021] Specifically, the network topology structure acquisition module 10 plays a key starting role in the entire data analysis process of complex power networks. First, it establishes a deep connection with various data sources and monitoring systems of the target power network, including the power grid dispatching center database, substation monitoring system, and distributed energy management system, etc. By parsing the network configuration files, communication protocol information, and real-time operation status data stored in these systems, the network topology structure of the target power network is comprehensively and accurately acquired. When extracting nodes, the module strictly follows the physical connections and logical relationships of the power network to identify key nodes such as power plants, substations, and transmission line intersections, and records them one by one. For the determination of node positions, a combination of global positioning system (GPS) data, coordinate information in geographic information system (GIS), and positioning technologies based on network communication delay measurement is used to ensure that the acquired node position information has high accuracy. Finally, all the identified nodes are formed into a topological node set, and the node position information is integrated into a topological node position set, laying a solid infrastructure for subsequent data analysis work on complex power networks.

[0022] An uploaded feature group set acquisition module 20, which is used to traverse the topological node set according to a preset data upload feature group to collect data upload features, and obtain a topological node data upload feature group set.

[0023] Specifically, the working process of the upload feature group set acquisition module 20 starts from the predefined data upload feature group, which covers key indicators such as data upload speed, data upload format (such as CSV, etc.), and upload delay rate. First, it locates the set of topology nodes, and then conducts comprehensive data upload feature collection for each topology node in the set. When collecting the data upload speed, professional network monitoring tools and algorithms are used to accurately measure the amount of data per unit time when each node transfers data to the specified server or storage system, so as to evaluate its upload efficiency. For the identification of the data upload format, by analyzing the protocol header information of data transmission, file extensions, and the structured characteristics of data, etc., it accurately determines whether the format used by each node to upload data conforms to the preset standard (such as CSV format), and records relevant information. In terms of measuring the upload delay rate, with the help of high-precision clock synchronization technology, the time difference between the data sent from the node and received by the server and the theoretical transmission time is compared to calculate the upload delay rate. Through the above-mentioned detailed and comprehensive data upload feature collection for all nodes in the set of topology nodes, the characteristic information of each node is finally integrated in an orderly manner to form a set of topology node data upload feature groups, providing rich and accurate data support for the subsequent in-depth analysis of the data transmission characteristics of network nodes.

[0024] The topology node pairing group determination module 30 conducts topology node pairing analysis on the set of topology nodes based on the set of topology node data upload feature groups and the set of topology node positions, and determines K topology node pairing groups, where K is an integer greater than or equal to 1.

[0025] Specifically, the topology node pairing group determination module 30 conducts comprehensive pairing analysis on the set of topology nodes from two dimensions: data transmission characteristics and physical location distribution, based on the set of topology node data upload feature groups and the set of topology node positions. From the perspective of data transmission, by deeply analyzing the various indicators in the set of topology node data upload feature groups, such as data upload speed, upload delay rate, etc., it accurately evaluates the similarity of each node in terms of data transmission ability and efficiency. From the perspective of physical location, according to the coordinate information in the set of topology node positions, the distance between nodes is calculated to determine nodes with close physical locations. Considering these two aspects, using the pairing behavior logic in the dragonfly algorithm, the topology nodes with similar data transmission characteristics and adjacent physical locations are aggregated together. After a series of complex calculation and analysis processes, finally K topology node pairing groups are determined, where K is an integer greater than or equal to 1. The division of these groups aims to provide a more efficient and reasonable organizational structure for subsequent data processing, effectively reducing the possibility of concurrent conflicts, and improving the efficiency and accuracy of the data analysis of the entire power network.

[0026] The group upload data set extraction module 40 is used to extract the K group upload data sets of the K topological node pairing groups within a preset time window, where each group upload data corresponds to a topological node.

[0027] Specifically, the core function of the group upload data set extraction module 40 is to perform precise data extraction operations on the determined K topological node pairing groups within a preset specific time window. The module first clarifies the composition and identification information of each topological node pairing group, and then, within the set time window, establishes a stable data connection channel with the data sources associated with each group. Through real-time monitoring and data capture technologies, it collects various types of data uploaded by each topological node during this period. These data cover rich contents such as the operating state parameters of power equipment (such as voltage, current, power, etc.), monitoring data (such as temperature, pressure, etc.), and control instruction information. For each topological node pairing group, the data uploaded by its member nodes are summarized and integrated to form K independent group upload data sets, ensuring that each group upload data set completely reflects the operation and interaction of each topological node within the corresponding group within the preset time window, providing a comprehensive and accurate data basis for subsequent in-depth analysis and processing.

[0028] The group data processing time set determination module 50 is used to traverse the K group upload data sets for data processing timeliness analysis and determine the K group data processing time sets.

[0029] Specifically, the main task of the group data processing time set determination module 50 is to provide an accurate time basis for the formulation of subsequent collaborative data analysis solutions. This module first traverses each of the data sets uploaded by the K groups one by one. When traversing each group's uploaded data set, it conducts an analysis of the data processing timeliness. On the one hand, starting from the internal characteristics of the data, it analyzes factors such as the complexity of the data structure, the size of the data volume, and the standardization degree of the data format, because these factors will directly affect the data processing speed. For example, data with a complex structure and a large data volume requires more computing resources and time to process. On the other hand, it also combines the current operating state of the system, including external factors such as available computing resources (such as CPU usage rate, memory capacity, etc.), network bandwidth, and the occupancy of system resources by other tasks currently running, to comprehensively evaluate the time required to process this data. By combining the internal characteristics of the data with the external operating environment of the system, it simulates and calculates the theoretical processing time of each group's uploaded data under the current system conditions. Then, to improve the accuracy of this time evaluation, the module further retrieves the data processing time records in the historical database that are similar to the characteristics of the currently uploaded data of the group. These historical data can reflect the actual data processing duration in past similar situations and provide a valuable reference basis for the current time evaluation. Based on the comprehensive consideration of the theoretically calculated time and the historical reference time, the module finally determines the time set for each group's data processing, thereby obtaining the K group data processing time sets. The time data in these sets will serve as an important basis for subsequent determination of collaborative data analysis solutions, ensuring that the time efficiency characteristics of each group's data processing can be fully considered when formulating the solutions, and thus optimizing the time efficiency of the entire data analysis process and ensuring the efficient and orderly progress of complex power network data analysis work.

[0030] The target collaborative data analysis solution acquisition module 60 is used to identify a collaborative analysis solution for the K topology node paired groups based on the K upload time sets of the K groups' uploaded data sets and the K group data processing time sets, with the goal of reducing the overall analysis time and the concurrent analysis of paired groups, obtaining the target collaborative data analysis solution, and performing data analysis processing on the K groups' uploaded data sets of the K topology node paired groups according to the target collaborative data analysis solution.

[0031] Specifically, the core responsibility of the target collaborative data analysis solution acquisition module 60 is to determine the optimal collaborative data analysis solution to maximize the overall analysis efficiency. This module first works with the dual goals of reducing the overall analysis time and the concurrent analysis of the paired groups. The module uses a collaborative analysis solution recognizer, taking the K upload time sets of the data sets uploaded by K groups and the K data processing time sets of K groups as key input information. During the recognition process, it comprehensively considers the upload time patterns of the data of each group, analyzes the impact of different upload times on the overall analysis process. For example, processing the data of the group with an earlier upload time first may be more beneficial for shortening the overall time. At the same time, it deeply studies the data processing time characteristics of each group, including factors such as the stability of the processing duration and the correlation between different processing times. Based on these comprehensive analyses, the module determines an initial collaborative data analysis solution, clarifying the initial analysis order of the data uploaded by each group in the K paired groups of topological nodes. Then, according to this initial solution and the data sets uploaded by K groups, the initial overall analysis time and the initial concurrent analysis volume of the paired groups are accurately calculated. Then, according to the preset scientific weights, the initial overall analysis time and the initial concurrent analysis volume of the paired groups are weighted and calculated to obtain the initial solution fitness, which can comprehensively reflect the advantages and disadvantages of the initial solution in terms of overall time and concurrent analysis. Subsequently, the first fine-tuning scale is determined according to the size of the initial concurrent analysis volume of the paired groups. Based on this, the analysis order in the initial collaborative data analysis solution is randomly fine-tuned multiple times. After each fine-tuning, a new fine-tuned collaborative data analysis solution is generated, thus forming a set of fine-tuned collaborative data analysis solutions. For each solution in this set, the fitness analysis is carried out again in combination with the data sets uploaded by K groups, and the corresponding set of fine-tuned solution fitnesses is calculated. Finally, the collaborative analysis solution is recognized by carefully comparing the fitness of the initial solution and the fitnesses in the set of fine-tuned solution fitnesses. If there is a fine-tuned solution fitness in the set of fine-tuned solution fitnesses that is greater than or equal to the fitness of the initial solution, then the fine-tuned collaborative data analysis solution corresponding to the maximum fitness value is selected as the target collaborative data analysis solution; if not, the set of fine-tuning methods for obtaining this set of fine-tuned collaborative data analysis solutions is set with the number of disabled times, and the initial collaborative data analysis solution is randomly fine-tuned multiple times according to the first fine-tuning scale, and the set of fine-tuned collaborative data analysis solutions is updated according to the fine-tuning results. This process is repeated until the optimal target collaborative data analysis solution is obtained. Once the target collaborative data analysis solution is determined, the module will strictly perform efficient and accurate data analysis and processing on the data sets uploaded by K groups of the K paired groups of topological nodes according to this solution, ensuring that the data analysis work of the entire complex power grid reaches the optimal state in terms of time and resource utilization, and providing solid data support for the stable operation and optimal management of the power grid.

[0032] In a possible implementation manner, the uploading feature group set obtaining module 20 includes:

[0033] The preset data uploading feature group includes uploading speed, latency rate, encryption feature, unit data information entropy, and processing resource occupancy.

[0034] Specifically, the preset data uploading feature group covers key features in multiple aspects, providing rich information for accurately evaluating and optimizing the data uploading process. The uploading speed is a key indicator for measuring how fast data is transmitted from the topology node to the specified receiving end. A faster uploading speed means that data can reach the analysis system more quickly, reducing data transmission latency and improving the timeliness of overall data analysis. For example, in real-time monitoring of the power system, quickly uploading device operation data allows the dispatching center to promptly grasp the grid status and make more accurate decisions. The latency rate reflects the delay situation in the data uploading process relative to the ideal transmission time. A lower latency rate is crucial for power network applications with high real-time requirements, such as rapid detection and location of power faults. If the latency rate is too high, it may cause the data analysis results to lag behind the actual grid status changes, affecting the timely handling of faults and the safe and stable operation of the power grid. The encryption feature involves the encryption method and strength during data uploading. In a complex power network environment, data security is of utmost importance. Encryption can effectively protect the confidentiality and integrity of data, preventing data from being stolen or tampered with. Different encryption algorithms and key lengths will affect the efficiency and security of data uploading. Therefore, monitoring the encryption feature helps optimize the data uploading performance while ensuring security. The unit data information entropy is used to measure the amount of information contained in the data. Data with high information entropy usually means richer content and higher value, but it may also require more processing resources and time for analysis. Understanding the unit data information entropy helps allocate system resources reasonably, prioritize the processing of data with higher information entropy, and improve the pertinence and effectiveness of data analysis. The processing resource occupancy focuses on the occupancy of system computing resources (such as CPU, memory, etc.) during data uploading. Excessive resource occupancy may lead to a decline in system performance and affect the normal operation of other tasks. By monitoring the processing resource occupancy, the data uploading strategy can be adjusted in a timely manner, the system resource configuration can be optimized, and the stable and efficient operation of the entire power network data processing system can be ensured. Considering these features in the preset data uploading feature group comprehensively and deeply can comprehensively and thoroughly understand the data uploading behavior of the topology node, provide a solid foundation for subsequent data analysis, node pairing, and the formulation of collaborative analysis solutions, and contribute to improving the overall performance and reliability of the complex power network data analysis platform.

[0035] In a possible implementation manner, as Figure 2 shown, the topology node pairing group determining module 30 includes:

[0036] Based on the set of topological node data upload feature groups and the set of topological node positions, determine the comprehensive feature vector of each topological node in the topological node set, and obtain a set of topological node comprehensive feature vectors, where each topological node comprehensive feature vector corresponds to the topological node data upload feature group and the topological node position of a topological node. Taking the preset vector similarity threshold as the pairing constraint, perform aggregation analysis on the set of topological node comprehensive feature vectors to obtain K sets of aggregated topological node comprehensive feature vectors. Based on the mapping of the K sets of aggregated topological node comprehensive feature vectors, obtain the K topological node pairing groups.

[0037] Specifically, determine the comprehensive feature vector of each topological node based on the set of topological node data upload feature groups and the set of topological node positions. For each feature in the set of topological node data upload feature groups, such as upload speed, delay rate, encryption feature, unit data information entropy, and processing resource occupancy, corresponding weights are assigned to reflect its importance in node feature description. At the same time, the topological node location information is also quantified and taken into consideration. By fusing and calculating these weighted upload feature data with the location information, a unique comprehensive feature vector is constructed for each topological node. In this way, the comprehensive feature vectors of all topological nodes form a set of topological node comprehensive feature vectors, where each vector precisely corresponds to the complete data upload feature group of a topological node and its location information.

[0038] Next, taking the preset vector similarity threshold as the pairing constraint condition, perform aggregation analysis on the set of topological node comprehensive feature vectors. This similarity threshold is a key parameter preset according to the characteristics of the power grid and the requirements of data analysis, and is used to measure the similarity between two topological node comprehensive feature vectors. During the aggregation analysis process, calculate the similarity between each pair of vectors in the set of topological node comprehensive feature vectors one by one. When the similarity between two vectors is greater than or equal to the preset vector similarity threshold, the two corresponding topological nodes are considered to have high similarity and can be aggregated together. By continuously performing such comparison and aggregation operations, the entire set of topological node comprehensive feature vectors is finally divided into K sets of aggregated topological node comprehensive feature vectors.

[0039] Finally, in the data analysis process of a complex power grid, obtaining K paired groups of topological nodes based on the mapping of the comprehensive feature vector sets of K aggregated topological nodes is a crucial and complex process. Each comprehensive feature vector set of aggregated topological nodes is obtained through the aggregated analysis of the comprehensive feature vectors of topological nodes. The topological nodes corresponding to the vectors inside have a high degree of similarity in terms of data upload characteristics (including upload speed, delay rate, encryption characteristics, unit data information entropy, and processing resource occupancy, etc.) and topological node positions. This similarity is the basis for pairing because similar nodes are more likely to adopt similar strategies during data processing, thereby improving processing efficiency and reducing conflicts. During the mapping process, each comprehensive feature vector set of aggregated topological nodes is regarded as an independent grouping unit, and all topological nodes within the set jointly form a paired group of topological nodes. For example, if a comprehensive feature vector set of an aggregated topological node contains several topological nodes from different regions but with extremely similar data upload characteristics, then these nodes will be mapped to the same paired group of topological nodes. This means that in subsequent data processing, they will work collaboratively as a whole. Through such a mapping method, it can be ensured that the nodes within each paired group of topological nodes have high coordination, while there are certain differences between different groups, facilitating the formulation of personalized data processing strategies according to the characteristics of different groups. This mapping mechanism based on the comprehensive feature vector sets of aggregated topological nodes effectively groups the complex power grid nodes reasonably, laying a solid foundation for subsequent data processing, analysis, and optimizing the operation of the power grid, enabling the entire data analysis process to proceed more efficiently and orderly, thereby improving the overall management level and operation efficiency of the complex power grid.

[0040] In a possible implementation manner, the group data processing time set determination module 50 includes:

[0041] Traverse the K group upload data sets for feature extraction to obtain K group upload data feature sets. Using the K group upload data feature sets as indexes, retrieve the data processing times within the historical time to obtain K historical group upload data processing time clusters. Among them, each historical group upload data processing time cluster includes multiple historical group upload data processing time sets, and each historical group upload data processing time set is the multiple historical group upload data processing times of a group upload data feature within the historical time. Traverse the K historical group upload data processing time clusters for data processing time density analysis to obtain K group data processing time sets.

[0042] Specifically, first, traversing the data sets uploaded by K groups for feature extraction is a crucial starting step. In this process, for each data set uploaded by a group, data mining algorithms are used to deeply analyze the data features therein. These features cover various attributes of the data, such as the structural characteristics of the data (such as whether the data has a hierarchical structure, the distribution of data fields, etc.), the numerical distribution range of the data (maximum value, minimum value, average value, etc.), the diversity of data types (whether it contains multiple different types of data, such as numerical type, text type, etc.), and the trend of data change (stable or with obvious fluctuations), etc. Through the comprehensive extraction of these features, a unique set of features of the data uploaded by each group is formed, and a total of K such sets are obtained.

[0043] Using the K sets of features of the data uploaded by the groups as indexes, retrieve the data processing times within the historical time in the historical database. Since each set of features of the data uploaded by a group reflects the unique nature of the data of that group, using it as an index can accurately locate the data records with similar features processed in the past. For each set of features of the data uploaded by a group, search for the data processing time records that match or are similar to it in the historical database. These records form a cluster of data processing times for the historical data uploaded by the groups. Each cluster of data processing times for the historical data uploaded by the groups contains multiple sets of data processing times for the historical data uploaded by the groups. Each set of data processing times for the historical data uploaded by the groups represents the time information recorded for a data processing process with the same or similar features of the data uploaded by the group within the historical time, including the duration from the start to the completion of data processing and the time distribution of each stage during the processing.

[0044] Traverse the K clusters of data processing times for the historical data uploaded by the groups to perform data processing time density analysis to obtain K sets of data processing times for the groups. During the analysis, for each cluster of data processing times for the historical data uploaded by the groups, calculate the time density between each set of data processing times for the historical data uploaded by the groups within it. The time density can be measured in various ways, such as calculating the distribution density of time points, the frequency of time intervals, etc. Through this density analysis, the central tendency and dispersion degree of the data processing time can be discovered, thereby determining a representative data processing time for the group. Perform such analysis on all K clusters of data processing times for the historical data uploaded by the groups, and finally obtain K sets of data processing times for the groups. These sets will provide a key time reference basis for the subsequent collaborative data analysis scheme, helping to optimize the data processing process in complex power network data analysis and improve the overall analysis efficiency.

[0045] In a possible implementation manner, the module 50 for determining the set of data processing times for the groups includes:

[0046] Extract the first historical group's uploaded data processing time set from the K historical group's uploaded data processing time clusters; calculate the time mean of the first historical group's uploaded data processing time set to obtain the first historical group's uploaded data processing time mean, and use the first historical group's uploaded data processing time mean as the density analysis center to construct the first initial neighborhood according to a preset density analysis radius; based on the preset density analysis radius, perform neighborhood edge outward diffusion on the first initial neighborhood to obtain the first diffusion neighborhood; determine whether the neighborhood density of the first initial neighborhood is less than the neighborhood density of the first diffusion neighborhood. If so, perform neighborhood edge outward diffusion on the first diffusion neighborhood according to the preset density analysis radius until the preset diffusion times are satisfied to obtain the first target neighborhood; calculate the mean of the multiple first historical group's uploaded data processing times within the first target neighborhood to obtain the first group's data processing time; perform data processing time density analysis on the K historical group's uploaded data processing time clusters to obtain the K group's data processing time sets.

[0047] Specifically, in the process of complex power network data analysis, extracting the first historical group's uploaded data processing time set from the K historical group's uploaded data processing time clusters is an important basic operation. The K historical group's uploaded data processing time clusters are retrieved from a vast amount of historical data by using the K group's uploaded data feature sets as indexes. Each historical group's uploaded data processing time cluster contains rich historical information, which is a set of time data recorded when processing data for groups with similar data upload characteristics in different past periods. When extracting the first historical group's uploaded data processing time set, it is first necessary to determine the basis or order of selection. This may be based on various factors, such as time sequence (selecting the earliest or latest historical data), data integrity (selecting the historical data with the most complete records), or relevance to the current system state (selecting the historical data of the period most similar to the current power network operating conditions), etc. If extracted according to the time sequence, then the set of data processing times that are the earliest recorded on the time axis will be selected from the first historical group's uploaded data processing time cluster and defined as the first historical group's uploaded data processing time set. Each element in this set represents the time spent processing the data uploaded by the group at a specific historical moment. These time data include the entire duration from when the data starts to enter the processing process to the final completion of processing, as well as the time consumption of each possible processing sub-stage. By accurately extracting the first historical group's uploaded data processing time set, it provides an initial data sample for subsequent operations such as calculating the time mean, constructing neighborhoods, and performing density analysis. It is a crucial starting step in the process of determining the entire group's data processing time set, and is of great significance for finally obtaining an accurate group's data processing time set and further optimizing the complex power network data analysis scheme.

[0048] Calculate the time mean for the set of data processing times uploaded by the first historical group that has been extracted. Each data in the set represents the time spent by the group on data processing at a certain historical moment. Sum up these time data and then divide by the number of data in the set. The result obtained is the mean data processing time for the first historical group to upload data. This mean value can comprehensively reflect the average time consumption level of the historical group in the past data processing process and is an important benchmark value for subsequent analysis. Using the calculated mean data processing time for the first historical group to upload data as the density analysis center, start constructing the first initial neighborhood. The preset density analysis radius is a key parameter preset before data analysis according to the characteristics of power network data and analysis requirements. It determines the size of the neighborhood range. In the time dimension, centered on the time mean, extend the time range determined by the preset density analysis radius forward and backward. This range constitutes the first initial neighborhood. For example, if the preset density analysis radius is 5 minutes and the time mean is exactly 10 o'clock, then the first initial neighborhood may be the range of all possible time data related to the group's data processing within the time period from 9:55 to 10:05 (the actual situation will be determined according to the specific distribution and time scale of the data). The purpose of constructing this initial neighborhood is to initially delimit the data range close to the average time level, so as to further explore the distribution density of data processing time within this local range, lay a foundation for subsequent neighborhood diffusion and density analysis, and thus more accurately determine the set of group data processing times, providing a reliable time-related basis for the data analysis of the entire complex power network.

[0049] Based on the preset density analysis radius, expand the first initial neighborhood outward from the neighborhood edge to obtain the first diffusion neighborhood. The preset density analysis radius is a key parameter preset according to the characteristics of power network data and analysis requirements. It determines the step size of neighborhood expansion. In the time dimension, starting from the boundary of the first initial neighborhood, expand outward according to the preset density analysis radius, and determine the newly included time data range as the first diffusion neighborhood. For example, if the initial neighborhood is a range that fluctuates 5 minutes above and below the time mean, and the preset density analysis radius is 2 minutes, then the first diffusion neighborhood will expand to a range that fluctuates 7 minutes above and below the time mean, including all relevant time data within this new range.

[0050] Next, it is determined whether the neighborhood density of the first initial neighborhood is less than the neighborhood density of the first diffusion neighborhood. There are various methods for calculating the neighborhood density. For example, the ratio of the number of time data points within the neighborhood to the size of the neighborhood range can be calculated. If the density of the first initial neighborhood is less than that of the first diffusion neighborhood, it indicates that there is more information related to the data processing time within a larger range, and further exploration is required. Therefore, continue to expand outward from the neighborhood edge of the first diffusion neighborhood according to the preset density analysis radius, and repeat this judgment and diffusion process until the preset number of diffusion times is satisfied. The setting of the preset number of diffusion times is to prevent the analysis range from becoming too large due to excessive diffusion and losing pertinence, while also ensuring that the potential laws in the data can be fully mined. Through such an iterative process, the first target neighborhood is finally obtained. This neighborhood has a relatively high time data density under the condition of satisfying the diffusion condition limitations.

[0051] Then, calculate the mean value of the data processing times uploaded by multiple first historical groups within the first target neighborhood to obtain the data processing time of the first group. Within the first target neighborhood, perform a summation operation on all the data processing times uploaded by the first historical groups, and then divide by the number of time data to obtain a mean value that can more accurately reflect the typical time consumption of data processing by this group under such data characteristics and historical circumstances. This data processing time of the first group will be used as an important reference value for subsequent analysis.

[0052] Finally, repeat the above entire process for the data processing time clusters uploaded by K historical groups, that is, sequentially extract the corresponding historical group uploaded data processing time sets from each historical group uploaded data processing time cluster, perform operations such as mean value calculation, neighborhood construction, and diffusion analysis. Finally, obtain K sets of data processing times for the groups. These sets will provide a comprehensive and accurate time reference basis for the data analysis of the entire complex power network, contribute to formulating a more optimized collaborative data analysis plan, improve data processing efficiency, and ensure the stable operation and effective management of the power network.

[0053] In a possible implementation manner, the target collaborative data analysis scheme acquisition module 60 includes:

[0054] Using a collaborative analysis scheme recognizer to perform collaborative analysis on the set of data processing times of the K groups and the set of the K upload times, to obtain an initial collaborative data analysis scheme, where the initial collaborative data analysis scheme is the analysis order of the uploaded data of each group in the initially analyzed set of uploaded data of the K groups; determining an initial overall analysis time and an initial concurrent analysis volume of paired groups based on the initial collaborative data analysis scheme and the set of uploaded data of the K groups; analyzing the initial overall analysis time and the initial concurrent analysis volume of paired groups to obtain an initial scheme fitness; determining a first fine-tuning scale according to the magnitude of the initial concurrent analysis volume of paired groups, and performing multiple random fine-tunings on the analysis order in the initial collaborative data analysis scheme based on the first fine-tuning scale to obtain a set of fine-tuned collaborative data analysis schemes; traversing the set of fine-tuned collaborative data analysis schemes and the set of uploaded data of the K groups to perform fitness analysis to obtain a set of fine-tuned scheme fitnesses; performing collaborative analysis scheme recognition according to the fitness magnitudes of the initial scheme fitness and the set of fine-tuned scheme fitnesses to obtain the target collaborative data analysis scheme.

[0055] Specifically, use a collaborative analysis scheme recognizer to perform collaborative analysis on the set of data processing times of K groups and the set of K upload times to determine an initial collaborative data analysis scheme. The collaborative analysis scheme recognizer comprehensively considers the time data in the set of data processing times of each group and the time information in the set of upload times. For the set of data processing times, it analyzes factors such as the average time required for each group to process data and the fluctuation range of time to evaluate the difficulty and time requirements of data processing for different groups. At the same time, the data in the set of upload times can reflect the order in which data arrives at the analysis system and the timeliness of data transmission. By weighing these factors, the recognizer determines the analysis order of the uploaded data of each group in the initially analyzed set of uploaded data of the K groups, forming an initial collaborative data analysis scheme. For example, if a group has a short data processing time and an early upload time, it may be given priority for analysis.

[0056] Next, determine the initial overall analysis time and the initial concurrent analysis volume of paired groups based on the initial collaborative data analysis scheme and the set of uploaded data of the K groups. When determining the initial overall analysis time, calculate the sum of the data processing time and upload time of each group in turn according to the analysis order set in the initial collaborative data analysis scheme, and at the same time consider the possible data dependencies and transmission waiting times between different groups, so as to obtain the total time required from the start of processing the data of the first group to the completion of processing the data of the last group, that is, the initial overall analysis time. For the initial concurrent analysis volume of paired groups, it is necessary to analyze how many paired groups' data are in a concurrent processing state at the same time under the initial collaborative data analysis scheme, and this quantity reflects the parallel processing degree of the system under the initial scheme.

[0057] Then, analyze the initial overall analysis time and the initial concurrent analysis volume of the paired groups to obtain the fitness of the initial solution. Generally, according to the specific goals and requirements of power network data analysis, corresponding weights are set for the overall analysis time and the concurrent analysis volume of the paired groups, and the two are combined into an initial solution fitness index through weighted calculation. For example, if the system pays more attention to quickly completing the overall analysis, the weight of the overall analysis time may be higher; if it is desired to improve the parallel processing ability of the system, the weight of the concurrent analysis volume of the paired groups will increase accordingly.

[0058] After that, determine the first fine-tuning scale according to the size of the initial concurrent analysis volume of the paired groups. If the initial concurrent analysis volume of the paired groups is large, it indicates that the system already has a high degree of parallel processing under the initial solution. At this time, the first fine-tuning scale may be relatively small to avoid system performance degradation caused by excessive adjustment; conversely, if the initial concurrent analysis volume of the paired groups is small, a larger fine-tuning scale can be adopted to try to improve the parallel processing ability. Based on the first fine-tuning scale, perform multiple random fine-tuning on the analysis order in the initial collaborative data analysis solution. Each fine-tuning will generate a new analysis order, thus obtaining a set of fine-tuned collaborative data analysis solutions.

[0059] Start traversing each solution in the set of fine-tuning collaborative data analysis solutions. For each fine-tuning solution, process the data in the data set uploaded by K groups in the order set for analysis. During the processing, closely monitor the flow and processing of the data, and accurately calculate the processing time of each group's data and the overall analysis time of the entire system under this solution. This requires considering factors such as the transfer delay between various processing links and the impact of the processing sequence of different groups' data on the overall time. At the same time, analyze the concurrent analysis situation of paired groups under this fine-tuning solution, determine how many paired groups' data are in the concurrent processing state at the same time, and count the concurrent analysis volume. By comprehensively considering these two key indicators, the overall analysis time and the concurrent analysis volume of paired groups, calculate according to the pre-set fitness calculation model. This fitness calculation model will assign corresponding weights to the overall analysis time and the concurrent analysis volume according to the actual requirements and goals of complex power network data analysis. For example, if quickly completing the overall analysis is crucial for real-time monitoring of the power network, the weight of the overall analysis time may be relatively high; if emphasizing the full utilization of system resources and parallel processing capabilities, the weight of the concurrent analysis volume of paired groups will increase accordingly. Obtain the fitness value of each fine-tuning solution through weighted calculation, record these fitness values one by one, and finally form a set of fine-tuning solution fitness values. This set comprehensively reflects the comprehensive performance of each fine-tuning collaborative data analysis solution in terms of the overall analysis time and the concurrent analysis volume of paired groups, providing an important basis for subsequent comparison and selection of the optimal solution. Through in-depth analysis of the set of fine-tuning solution fitness values, the advantages and disadvantages of each solution can be clearly understood, so as to screen out the solution most suitable for complex power network data analysis, ensure the efficient and stable operation of the system, and provide accurate and timely data support for the management and decision-making of the power network.

[0060] Finally, identify the collaborative data analysis solution based on the fitness of the initial solution and the set of fine-tuning solution fitness values to obtain the target collaborative data analysis solution. If there is a fine-tuning solution fitness value in the set of fine-tuning solution fitness values that is greater than or equal to the fitness of the initial solution, then select the fine-tuning collaborative data analysis solution with the largest fitness value as the target collaborative data analysis solution; if not, it may be necessary to adjust the fine-tuning scale or adopt other optimization strategies, continue to perform fine-tuning and update the set of fine-tuning solution fitness values, and repeat this process until the optimal target collaborative data analysis solution is found. This target collaborative data analysis solution will be able to maximize the analysis efficiency and optimize resource utilization on the premise of meeting the requirements of complex power network data analysis, providing strong support for the safe and stable operation and effective management of the power network.

[0061] In a possible implementation manner, the target collaborative data analysis solution acquisition module 60 includes:

[0062] Determine whether there is a fitness value of the fine-tuning scheme in the fitness set of the fine-tuning scheme that is greater than or equal to the fitness of the initial scheme. If so, use the fine-tuning collaborative data analysis scheme corresponding to the maximum value in the fitness set of the fine-tuning scheme as the target collaborative data analysis scheme; if not, set the number of disabled times for the fine-tuning method set that obtains the fine-tuning collaborative data analysis scheme set, and continue to perform multiple random fine-tuning on the initial collaborative data analysis scheme according to the first fine-tuning scale, and update the fine-tuning collaborative data analysis scheme set according to the fine-tuning result, where the number of disabled times is the number of times the fine-tuning method is prohibited from being used.

[0063] Specifically, conduct a comprehensive inspection of the fitness set of the fine-tuning scheme to determine whether there is a fitness value of the fine-tuning scheme that is greater than or equal to the fitness of the initial scheme. This comparison process is crucial because it directly determines the subsequent operation direction. If such a scheme is found in the fitness set of the fine-tuning scheme, it means that through the fine-tuning operation, a scheme that performs better or at least not worse than the initial scheme in terms of the overall analysis time and the concurrent analysis volume of the paired group has been found. At this time, directly select the fine-tuning collaborative data analysis scheme with the largest fitness value in the fitness set of the fine-tuning scheme as the target collaborative data analysis scheme. This scheme is considered to be the one that can best meet the complex power network data analysis requirements under the current analysis conditions. It can ensure the efficient operation of the system while maximizing resource utilization and improving analysis efficiency.

[0064] However, if after judgment, there is no fitness value of the fine-tuning scheme in the fitness set of the fine-tuning scheme that is greater than or equal to the fitness of the initial scheme, it indicates that the current fine-tuning operation has not produced a better scheme. In this case, it is necessary to adjust and optimize the fine-tuning process. Set the number of disabled times for the fine-tuning method set that obtains the fine-tuning collaborative data analysis scheme set to avoid repeated use of ineffective fine-tuning methods and prevent getting stuck in a local optimum. Then, continue to perform multiple random fine-tuning on the initial collaborative data analysis scheme according to the first fine-tuning scale. After each fine-tuning, update the fine-tuning collaborative data analysis scheme set according to the new fine-tuning result, incorporate the newly generated fine-tuning scheme into the set, recalculate its fitness value, and update relevant information such as the number of disabled times. By continuously repeating this process, gradually explore a better solution space until a target collaborative data analysis scheme that meets the conditions is found. This iterative optimization method ensures that the finally determined target collaborative data analysis scheme has high reliability and effectiveness in a complex and changing power network data environment, and can provide strong decision-making support for aspects such as the stable operation, fault prediction, and resource scheduling of the power network.

[0065] In a possible implementation manner, the target collaborative data analysis scheme acquisition module 60 includes:

[0066] Perform weighted calculation on the initial overall analysis time and the initial concurrent analysis volume of paired groups according to the preset weights to obtain the initial scheme fitness.

[0067] Specifically, the initial overall analysis time reflects the total duration from the start of processing the data uploaded by the first group to the completion of the data processing of the last group. This time indicator is crucial for the timeliness and real-time nature of power network data analysis. Because a longer overall analysis time may lead to a lag in decision-making and an inability to respond promptly to the dynamic changes in the power network. For example, in the power fault detection scenario, if the overall analysis time is too long, it may delay the timing of fault location and repair, affecting the stable operation of the power system. The initial concurrent analysis volume of paired groups reflects the number of paired groups that can simultaneously perform data processing at the same moment under the initial collaborative data analysis scheme. A higher concurrent analysis volume means that the system can make more full use of computing resources, improve the parallelism of data processing, and thus speed up the overall analysis progress to a certain extent. However, too high a concurrent analysis volume may also bring problems such as resource competition, and a suitable balance point needs to be found in the trade-off. The preset weights are set in advance according to the characteristics of complex power networks, the key points of data analysis, and actual application requirements. If the power network has extremely high requirements for real-time performance, such as in the emergency control scenario of the power system, a higher weight will be assigned to the initial overall analysis time to ensure that the scheme can quickly give analysis results; while in the case where computing resources are relatively sufficient and the overall efficiency improvement of the system is emphasized, the weight of the initial concurrent analysis volume of paired groups will be appropriately increased. By multiplying the initial overall analysis time by its corresponding preset weight and adding the product of the initial concurrent analysis volume of paired groups and its preset weight, the initial scheme fitness can be obtained. This fitness value comprehensively considers the influence of two key factors and can intuitively reflect the performance of the initial collaborative data analysis scheme in terms of overall performance. The higher the initial scheme fitness, the better the scheme is in balancing the overall analysis time and the concurrent analysis volume, and the more likely it is to become the final target collaborative data analysis scheme; on the contrary, the scheme needs to be further optimized and adjusted, and its fitness can be improved through subsequent fine-tuning and other operations. This weighted calculation method provides a scientific and quantitative evaluation method for the selection and optimization of complex power network data analysis schemes, which helps to improve the effectiveness and reliability of the entire data analysis process.

[0068] It should be noted that the above sequence of embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. And the above specific embodiments of this specification have been described. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0069] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

[0070] This specification and the drawings are merely exemplary descriptions of the present invention and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present invention. Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention is intended to include these changes and modifications.

Claims

1. A high-precision data analysis system for complex power networks, characterized in that Including: A network topology structure acquisition module, configured to acquire the network topology structure of a target power network, extract the nodes and node positions of the network topology structure, and obtain a topological node set and a topological node position set; An upload feature group set acquisition module, configured to collect data upload features by traversing the topological node set according to a preset data upload feature group, and obtain a topological node data upload feature group set; A topological node pairing group determination module, configured to perform topological node pairing analysis on the topological node set based on the topological node data upload feature group set and the topological node position set, and determine K topological node pairing groups, where K is an integer greater than or equal to 1; A group upload data set extraction module, configured to extract K group upload data sets of the K topological node pairing groups within a preset time window, where each group upload data corresponds to a topological node; A group data processing time set determination module, configured to perform data processing timeliness analysis by traversing the K group upload data sets, and determine K group data processing time sets; A target collaborative data analysis scheme acquisition module, configured to identify a collaborative analysis scheme for the K topological node pairing groups based on the K upload time sets of the K group upload data sets and the K group data processing time sets with the goal of reducing the overall analysis time and reducing the concurrent analysis of the pairing groups, obtain a target collaborative data analysis scheme, and perform data analysis processing on the K group upload data sets of the K topological node pairing groups according to the target collaborative data analysis scheme; The topological node pairing group determination module specifically includes: Based on the topological node data upload feature group set and the topological node position set, determine the comprehensive feature vector of each topological node in the topological node set, and obtain a topological node comprehensive feature vector set, where each topological node comprehensive feature vector corresponds to the topological node data upload feature group and the topological node position of a topological node; Perform aggregation analysis on the topological node comprehensive feature vector set with a preset vector similarity threshold as the pairing constraint, and obtain K aggregated topological node comprehensive feature vector sets; Map and obtain the K topological node pairing groups based on the K aggregated topological node comprehensive feature vector sets.

2. The system according to claim 1, wherein The preset data upload feature group includes upload speed, delay rate, encryption feature, unit data information entropy, and processing resource occupancy.

3. The system according to claim 1, characterized in that, The group data processing time set determination module specifically includes: Traverse the K group upload data sets to perform feature extraction, and obtain K group upload data feature sets; Use the K group upload data feature sets as indexes to retrieve the data processing times within the historical time, and obtain K historical group upload data processing time clusters, where each historical group upload data processing time cluster includes multiple historical group upload data processing time sets, and each historical group upload data processing time set is the multiple historical group upload data processing times of a group upload data feature within the historical time; Traverse the data processing time clusters uploaded by the K historical groups to perform data processing time density analysis, and obtain the data processing time sets of the K groups.

4. The system according to claim 3, characterized in that, The group data processing time set determination module is used for: Extract the first historical group's uploaded data processing time set from the K historical groups' uploaded data processing time clusters; Calculate the time mean of the first historical group's uploaded data processing time set, obtain the first historical group's uploaded data processing time mean, and use the first historical group's uploaded data processing time mean as the density analysis center to construct the first initial neighborhood according to the preset density analysis radius; Based on the preset density analysis radius, perform neighborhood edge outward diffusion on the first initial neighborhood to obtain the first diffusion neighborhood; Judge whether the neighborhood density of the first initial neighborhood is less than the neighborhood density of the first diffusion neighborhood. If so, perform neighborhood edge outward diffusion on the first diffusion neighborhood according to the preset density analysis radius until the preset diffusion times are satisfied to obtain the first target neighborhood; Calculate the mean of the data processing times of multiple first historical groups uploaded within the first target neighborhood to obtain the first group's data processing time; Perform data processing time density analysis on the K historical groups' uploaded data processing time clusters to obtain the data processing time sets of the K groups.

5. The system according to claim 1, wherein The target collaborative data analysis scheme acquisition module is specifically used for: Use the collaborative analysis scheme recognizer to perform collaborative analysis on the data processing time sets of the K groups and the K upload time sets to obtain the initial collaborative data analysis scheme, where the initial collaborative data analysis scheme is the analysis order of the uploaded data of each group in the initially analyzed K groups' uploaded data sets; Based on the initial collaborative data analysis scheme and the K groups' uploaded data sets, determine the initial overall analysis time and the initial concurrent analysis volume of the paired groups; Analyze the initial overall analysis time and the initial concurrent analysis volume of the paired groups to obtain the initial scheme fitness; Determine the first fine-tuning scale according to the magnitude of the initial concurrent analysis volume of the paired groups, and perform multiple random fine-tunings on the analysis order in the initial collaborative data analysis scheme based on the first fine-tuning scale to obtain the fine-tuned collaborative data analysis scheme set; Traverse the fine-tuned collaborative data analysis scheme set and the K groups' uploaded data sets to perform fitness analysis to obtain the fine-tuned scheme fitness set; Perform collaborative analysis scheme recognition according to the fitness magnitudes of the initial scheme fitness and the fine-tuned scheme fitness set to obtain the target collaborative data analysis scheme.

6. The system according to claim 5, wherein The target collaborative data analysis scheme acquisition module is used for: Judge whether there is a fine-tuned scheme fitness in the fine-tuned scheme fitness set that is greater than or equal to the initial scheme fitness. If so, use the fine-tuned collaborative data analysis scheme corresponding to the maximum value of the fine-tuned scheme fitness set as the target collaborative data analysis scheme; If not, set the disable times for the set of fine-tuning methods for obtaining the set of fine-tuned collaborative data analysis solutions, and continue to perform multiple random fine-tuning on the initial collaborative data analysis solution according to the first fine-tuning scale, and update the set of fine-tuned collaborative data analysis solutions according to the fine-tuning results, where the disable times is the number of times the fine-tuning method is prohibited from being used.

7. The system according to claim 5, wherein Perform weighted calculation on the initial overall analysis time and the initial concurrent analysis volume of the paired groups according to the preset weights to obtain the initial solution fitness.

Images