Multi-source electric power information intelligent acquisition method and system

Abstract

The application discloses a kind of multi-source electric power information intelligent acquisition method and system, it is related to multi-source electric power information intelligent acquisition technical field, the implicit correlation graph between data sources is constructed by transfer entropy in the application, break through the limitation that traditional method can only handle explicit correlation, discover nonlinear causal driving relationship in electric power data;On this basis, the weighted in-out degree centrality of node, information hub degree and its change rate are calculated, to accurately identify the data source that plays a key role in system information flow;Further, dynamically determine the comprehensive acquisition priority by combining hub degree and change rate, and adaptively allocate the sampling frequency and transmission resources of each data source, avoid resource waste while ensuring the fine collection of core data;Finally, by iteratively updating the correlation graph, the system continuously tracks the evolution of electric power data interaction structure, effectively improves the intelligence, real-time and resource utilization efficiency of multi-source electric power data collection.

Description

Technical Field

[0001] This application relates to the field of intelligent acquisition technology of multi-source power information, specifically to an intelligent acquisition method and system for multi-source power information. Background Technology

[0002] In power systems, data sources are diverse, including smart meters, PMUs, SCADA systems, and weather sensors. These various data sources differ significantly in sampling frequency, time base, and communication protocols, making direct fusion and collaborative analysis of multi-source power information difficult. Traditional methods typically rely on fixed sampling periods or simple priority rules, failing to dynamically reflect the complex implicit relationships (such as causal driving structures) between data sources. Furthermore, existing technologies lack quantitative assessments of the information value and network hub status of data sources in resource allocation, easily leading to insufficient collection of critical data and excessive redundancy of secondary data, wasting communication and computing resources. Therefore, a strategy is needed that can automatically identify the correlation structure between multi-source power data and adaptively adjust sampling and transmission resources accordingly to improve the targeting of data collection and overall system efficiency.

[0003] Regarding the above-mentioned solutions, the inventors of this application have discovered that the above-mentioned technology has at least the following technical problems:

[0004] 1. Traditional methods typically rely solely on linear and symmetric correlation metrics such as correlation coefficients, failing to effectively capture the widespread nonlinear and directed causal relationships in power systems (such as the hysteresis effect of load changes on the voltage of a node), resulting in the neglect of numerous hidden causal driving paths. Furthermore, existing correlation graphs often employ undirected graph structures, lacking the ability to express information flow, making it difficult to identify key information sources and their propagation links, and thus failing to provide a theoretical basis for causal tracing and impact analysis.

[0005] 2. Traditional node importance assessment often relies on static graph topology indicators (such as degree centrality), failing to consider the directionality of information flow and differences in edge weights, and lacking dynamic tracking of the rate of change of node importance over time. Therefore, it cannot accurately locate information hub nodes at the current moment, nor can it capture the abnormal signals implied by sudden changes in hub degree, resulting in insufficient early warning capabilities for power system faults and a lack of dynamic, quantifiable protection indicators for critical nodes.

[0006] 3. Conventional data acquisition and uplink transmission resource allocation often employ fixed-frequency or uniform scheduling strategies, failing to prioritize data sources based on their dynamic importance (e.g., hub status, degree of change). This results in a significant waste of communication bandwidth and computing resources on redundant data, while high-value data sources (e.g., hub nodes, nodes experiencing drastic changes) may suffer from insufficient sampling or transmission delays. Consequently, it becomes impossible to achieve refined acquisition of high-value data and effective offloading of low-value data under limited resource conditions, making it difficult to achieve an optimal balance between resource utilization efficiency and data value density. Summary of the Invention

[0007] To address the aforementioned technical shortcomings, the purpose of this application is to provide a method and system for intelligent acquisition of multi-source power information.

[0008] To solve the above-mentioned technical problems, this application adopts the following technical solution: In the first aspect, this application provides a method for intelligent acquisition of multi-source power information, which includes the following steps: acquiring raw power information from multiple power data sources, and performing synchronization and preprocessing to obtain a standardized data sequence.

[0009] Based on standardized data sequences, the transfer entropy between various data sources is calculated, and an implicit association graph is constructed.

[0010] Based on the implicit association graph, the weighted out-degree centrality and weighted in-degree centrality of each data source node are calculated, the information hub degree of each data source node is calculated, and the rate of change of information hub degree is analyzed.

[0011] Based on information hub degree and rate of change, the comprehensive collection priority of each data source is calculated, and then the adaptive sampling frequency and transmission resource allocation of each data source are dynamically allocated.

[0012] The implicit association graph is updated based on the data obtained from the refined sampling frequency and the transmission resource allocation. The refined sampling frequency is used to perform fine-grained data acquisition on each data source.

[0013] Preferably, the step of acquiring the raw power information from multiple power data sources and synchronizing and preprocessing it to obtain a standardized data sequence includes: S201, acquiring the raw power information from multiple power data sources, wherein the multiple power data sources include smart meter data sources, phasor measurement unit (PMU) data sources, SCADA data sources, and environmental sensor data sources.

[0014] S202, timestamp alignment and missing value filling are performed on the original power information from each data source to generate a multi-source power data set after synchronous preprocessing.

[0015] S203, standardize the data sequences of each data source in the multi-source power data set after synchronous preprocessing to obtain a standardized data sequence.

[0016] Preferably, the step of calculating the transfer entropy between data sources based on the standardized data sequence and constructing an implicit association graph includes: S301, calculating the transfer entropy between the data sources based on the standardized data sequence. The data source for the first The transfer entropy of each data source.

[0017] S302, based on the transfer entropy between each data source, normalizes each transfer entropy to obtain the normalized directed edge weights.

[0018] S303 uses each data source as a vertex, retains directed edges with weights greater than a preset association strength threshold, and constructs an implicit association graph.

[0019] Preferably, the step of calculating the transfer entropy between data sources based on the standardized data sequence and constructing an implicit association graph includes: S301, calculating the transfer entropy between the data sources based on the standardized data sequence. The data source for the first The transfer entropy of each data source.

[0020] S302, based on the transfer entropy between each data source, normalizes each transfer entropy to obtain the normalized directed edge weights.

[0021] S303 uses each data source as a vertex, retains directed edges with weights greater than a preset association strength threshold, and constructs an implicit association graph.

[0022] Preferably, the step of calculating the weighted out-degree centrality and weighted in-degree centrality of each data source node based on the implicit association graph, calculating the information hub degree of each data source node, and analyzing the rate of change of the information hub degree includes: S501, calculating the weighted out-degree centrality and weighted in-degree centrality of each data source node based on the edge weight matrix of the implicit association graph.

[0023] S502, based on the weighted out-degree centrality and the weighted in-degree centrality, calculate the information hub degree of each data source node.

[0024] S503, construct a sliding time window sequence, calculate the rate of change of the information hub degree of each data source node within multiple consecutive time windows, and the rate of change is used to characterize the trend of the evolution of the status of the data source node in the implicit association graph.

[0025] Preferably, the step of calculating the information hub degree of each data source node based on the weighted out-degree centrality and the weighted in-degree centrality includes: the expression for the information hub degree is:

[0026]

[0027] in, For the first Weighted out-degree centrality of each data source node For the first Weighted in-degree centrality of each data source node, This is a preset non-zero small value used to prevent division by zero errors when the weighted in-degree centrality is zero.

[0028] Preferably, the step of calculating the comprehensive acquisition priority of each data source based on the information hub degree and the rate of change, and then dynamically allocating the adaptive sampling frequency and transmission resource allocation of each data source, includes: S701, calculating the comprehensive acquisition priority of each data source based on the information hub degree and the rate of change, wherein the comprehensive acquisition priority is the product of the information hub degree and the evolution gain factor, and the evolution gain factor is calculated based on the rate of change.

[0029] S702, based on the comprehensive acquisition priority, determine the adaptive sampling frequency of each data source, wherein the adaptive sampling frequency is the frequency value after adjusting the base sampling frequency according to the degree to which the comprehensive acquisition priority deviates from the average priority of all data sources.

[0030] S703, based on the comprehensive acquisition priority, determine the transmission resource allocation for each data source, wherein the transmission resource allocation is the total system transmission resource after allocating the resources according to the ratio of the comprehensive acquisition priority of each data source to the total priority.

[0031] Preferably, the specific calculation process for the transmission resource allocation is as follows:

[0032]

[0033] in, For the total system transmission resource budget, For the first The overall collection priority of each data source This is the sum of the collection priorities for all data sources. This represents the total number of data sources.

[0034] Preferably, the step of performing refined acquisition on each data source based on the adaptive sampling frequency and the transmission resource allocation, and updating the implicit correlation graph based on the data obtained from the refined acquisition, includes: S901, adjusting the acquisition time interval of each data source according to the adaptive sampling frequency, allocating a transmission channel and buffer space for each data source according to the transmission resource allocation, performing refined acquisition, and obtaining refined acquisition data for each data source.

[0035] S902, based on the refined collected data, update the edge weight matrix of the implicit association graph to obtain the updated implicit association graph.

[0036] S903, based on the updated implicit association graph, cyclically execute the steps of calculating the transfer entropy, calculating the information hub degree, calculating the comprehensive acquisition priority, and refining the acquisition, so as to realize the online evolution of the implicit association graph and the adaptive dynamic adjustment of the acquisition strategy.

[0037] In a second aspect, this application provides a system for intelligent acquisition of multi-source power information, comprising: preferably, a multi-source power data synchronization and preprocessing module, which acquires raw power information from multiple power data sources, and performs synchronization and preprocessing to obtain a standardized data sequence.

[0038] The transfer entropy association graph construction module calculates the transfer entropy between various data sources based on standardized data sequences and constructs an implicit association graph.

[0039] The information hub degree and rate of change analysis module, based on the implicit association graph, calculates the weighted out-degree centrality and weighted in-degree centrality of each data source node, calculates the information hub degree of each data source node, and analyzes the rate of change of the information hub degree.

[0040] The adaptive sampling and resource allocation module calculates the comprehensive acquisition priority of each data source based on information hub degree and change rate, and then dynamically allocates the adaptive sampling frequency and transmission resource allocation of each data source.

[0041] The refined data acquisition and map iterative update module performs refined data acquisition on each data source based on the adaptive sampling frequency and the transmission resource allocation, and updates the implicit association map based on the data obtained from the refined data acquisition.

[0042] The beneficial effects of this application are as follows:

[0043] 1. This application provides a multi-source power information intelligent acquisition method and system. By constructing an implicit correlation graph between data sources through transfer entropy, it overcomes the limitation of traditional methods that can only handle explicit correlations and discovers nonlinear causal driving relationships in power data. On this basis, it calculates the weighted in-degree centrality and information hub degree of nodes and their rate of change, thereby accurately identifying data sources that play a key role in the flow of system information. Furthermore, it dynamically determines the comprehensive acquisition priority by combining hub degree and rate of change, and adaptively allocates the sampling frequency and transmission resources of each data source, avoiding resource waste while ensuring the refined acquisition of core data. Finally, by iteratively updating the correlation graph, the system continuously tracks the evolution of the power data interaction structure, effectively improving the intelligence, real-time performance, and resource utilization efficiency of multi-source power data acquisition.

[0044] 2. This application breaks through the limitation that traditional correlation coefficients can only describe linear and symmetric relationships. It adopts transfer entropy to quantify the nonlinear and directed causal dependence strength between data sources and constructs an implicit correlation graph based on this. It can discover hidden causal driving paths in the power system (such as the hysteresis effect of load changes on the voltage of a certain node), providing a theoretical basis for identifying key information sources and propagation links, and making up for the shortcomings of traditional undirected graphs in expressing the direction of information flow.

[0045] 3. This application defines weighted out-degree or in-degree centrality on the implicit association graph, and then calculates the information hub degree (combining the influence of out-degree or in-degree) and dynamically tracks its rate of change, quantifying the core position and evolution trend of each data source in the information network; it can not only locate the information hub node (such as the master control data source) at the current moment, but also capture abnormal signals of hub degree change, providing quantifiable dynamic indicators for early warning of power system faults and protection of critical nodes.

[0046] 4. This application integrates the static importance of information hub degree with the dynamic sensitivity of change rate to establish a comprehensive collection priority, and adjusts the sampling frequency and uplink transmission resource allocation of each data source online accordingly; under the condition of limited communication bandwidth and computing resources, it prioritizes the fine collection of high-priority data sources (such as hub nodes and nodes with drastic changes), while reducing the sampling burden of redundant data, and achieving the optimal balance between resource utilization efficiency and data value density. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating the implementation steps of the method described in this application.

[0049] Figure 2 This is a schematic diagram of the system structure connection of this application. Detailed Implementation

[0050] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0051] Please see Figure 1As shown, this application provides a method for intelligent acquisition of multi-source power information in the first aspect, including: acquiring raw power information from multiple power data sources, and synchronizing and preprocessing it to obtain a standardized data sequence;

[0052] In a specific example, the step of acquiring raw power information from multiple power data sources and synchronizing and preprocessing it to obtain a standardized data sequence includes: S201, acquiring raw power information from multiple power data sources, wherein the multiple power data sources include smart meter data sources, phasor measurement unit (PMU) data sources, SCADA data sources, and environmental sensor data sources.

[0053] S202, timestamp alignment and missing value filling are performed on the original power information from each data source to generate a multi-source power data set after synchronous preprocessing.

[0054] S203, standardize the data sequences of each data source in the multi-source power data set after synchronous preprocessing to obtain a standardized data sequence.

[0055] It should be noted that the smart meter data source is used to provide power consumption data and voltage and current measurement data on the user side; the phasor measurement unit (PMU) data source is used to provide synchronous phasor measurement data of transmission lines and substation nodes; the SCADA data source is used to provide telemetry and teleindication data of the power grid dispatch automation system; and the environmental sensor data source is used to provide environmental parameter data such as temperature and humidity.

[0056] It should be noted that the original power information from each data source is timestamped and filled with missing values. Specifically, the data from each data source is aligned according to a unified time base. For data gaps caused by communication delays or equipment failures, interpolation or forward padding is used to fill in the gaps, generating a continuous, synchronously preprocessed multi-source power data set.

[0057] It should be noted that the standardization process for the synchronously preprocessed data sequences of each data source in the multi-source power data set is specifically as follows: the mean and standard deviation of the synchronously preprocessed data sequences of each data source are calculated; based on the mean and the standard deviation, the data values ​​in the data sequences are standardized so that data from data sources with different dimensions can be analyzed and compared on the same scale.

[0058] It should be noted that the standardization process is as follows: subtract the mean of the data sequence from each data value in the synchronously preprocessed data sequence of the data source to obtain the difference; divide the difference by the standard deviation of the data sequence to obtain the standardized data value.

[0059] Based on standardized data sequences, the transfer entropy between various data sources is calculated, and an implicit association graph is constructed.

[0060] In a specific example, the step of calculating the transfer entropy between data sources based on the standardized data sequence and constructing an implicit association graph includes: S301, calculating the transfer entropy between the data sources based on the standardized data sequence. The data source for the first The transfer entropy of each data source.

[0061] S302, based on the transfer entropy between each data source, normalizes each transfer entropy to obtain the normalized directed edge weights.

[0062] S303 uses each data source as a vertex, retains directed edges with weights greater than a preset association strength threshold, and constructs an implicit association graph.

[0063] It should be noted that the normalization process for the transfer entropy is as follows: obtain the maximum and minimum values ​​of the transfer entropy among all data sources, subtract the minimum value from each transfer entropy and divide by the difference between the maximum and minimum values ​​to obtain the normalized directed edge weights, so that the values ​​of all directed edge weights are between 0 and 1.

[0064] The directed edge weights are further normalized, and the specific calculation process is as follows:

[0065]

[0066] in For the first The data source for the first Normalized directed edge weights for each data source For the first The data source for the first The transfer entropy of a data source To be the minimum value of transitive entropy, This represents the maximum value in the propagation entropy.

[0067] It should be noted that the implicit association graph is constructed as follows: each data source is used as the vertex of the graph, the directed edges whose weights are greater than a preset association strength threshold are used as the set of directed edges of the graph, and the edge weight matrix of the graph is constructed using the weights of the directed edges to obtain the implicit association graph.

[0068] In a specific instance, the calculation of the first The data source for the first The transfer entropy of each data source includes: calculated using the formula The result is the The data source for the first Transmission entropy of each data source ,in For the first The data source within the time window Standardized data values ​​at each moment For the first The data source within the time window Standardized data values ​​at each moment For the first The data source within the time window Standardized data values ​​at each moment Represents the probability density function. This represents the conditional probability density function.

[0069] It should be noted that the calculation of the first... The data source for the first The transit entropy of a data source is used to characterize the data source. For data source The direction and intensity of information transmission, specifically: based on the first The standardized data value of the first data source at the current moment, the first... The standardized data value of the first data source at the current moment and the first... The standardized data value of the data source at the next time step is used to calculate the first data point. The data source for the first The transfer entropy of the first data source, to quantify the first Does the data source contain information that can be used to predict the first...? Additional information about the future state of each data source.

[0070] This application breaks through the limitation that traditional correlation coefficients can only describe linear and symmetric relationships. It adopts transfer entropy to quantify the nonlinear and directed causal dependence strength between data sources and constructs an implicit correlation graph based on this. It can discover hidden causal driving paths in the power system (such as the hysteresis effect of load changes on the voltage of a node), providing a theoretical basis for identifying key information sources and propagation links, and making up for the inability of traditional undirected graphs to express the direction of information flow.

[0071] Based on the implicit association graph, the weighted out-degree centrality and weighted in-degree centrality of each data source node are calculated, the information hub degree of each data source node is calculated, and the rate of change of information hub degree is analyzed.

[0072] In a specific example, the step of calculating the weighted out-degree centrality and weighted in-degree centrality of each data source node based on the implicit association graph, calculating the information hub degree of each data source node, and analyzing the rate of change of the information hub degree includes: S501, calculating the weighted out-degree centrality and weighted in-degree centrality of each data source node based on the edge weight matrix of the implicit association graph.

[0073] S502, based on the weighted out-degree centrality and the weighted in-degree centrality, calculate the information hub degree of each data source node.

[0074] S503, construct a sliding time window sequence, calculate the rate of change of the information hub degree of each data source node within multiple consecutive time windows, and the rate of change is used to characterize the trend of the evolution of the status of the data source node in the implicit association graph.

[0075] It should be noted that weighted out-degree centrality represents the sum of the weights of directed edges from the given data source node to all other data source nodes, while weighted in-degree centrality represents the sum of the weights of directed edges from all other data source nodes to the given data source node. The weighted out-degree centrality of each data source node is calculated specifically by: for the edge weight matrix of the implicit association graph... Summing the weights of all directed edges in the first row yields the result. The weighted out-degree centrality of the nth data source node; the larger this value, the wider the influence of that data source on the information of other data sources. The weighted in-degree centrality of each data source node is calculated specifically by: [calculating the weighted in-degree centrality of the nth node in the implicit association graph's edge weight matrix]. Summing the weights of all directed edges in the list yields the result of the first... The weighted in-degree centrality of each data source node; the larger the value, the more deeply the data source is influenced by other data sources.

[0076] Furthermore, the expressions for weighted out-degree centrality and weighted in-degree centrality are:

[0077]

[0078]

[0079] in, For from the first The data source points to the first The weights of directed edges from each data source. For from the first The data source points to the first The weights of directed edges from each data source. This represents the total number of data sources.

[0080] It should be noted that constructing a sliding time window sequence and calculating the rate of change of the information hub degree of each data source node within multiple consecutive time windows is specifically as follows: subtracting the information hub degree of the previous time window from the information hub degree of the current time window to obtain the difference; dividing the difference by the sum of the information hub degree of the previous time window and a preset non-zero small amount to obtain the rate of change; a positive rate of change indicates that the position of the data source node in the implicit association graph is rising, and a negative rate indicates that the position is falling.

[0081] Furthermore, the expression for the rate of change is:

[0082]

[0083] in, For the first The current data source node is the _th data source node. Information hub capacity within a time window For the first The data source node in the previous... Information hub capacity within a time window This is a preset non-zero small quantity.

[0084] In a specific example, the calculation of the information hub degree of each data source node based on the weighted out-degree centrality and the weighted in-degree centrality includes: the expression for the information hub degree is:

[0085]

[0086] in, For the first Weighted out-degree centrality of each data source node For the first Weighted in-degree centrality of each data source node, This is a preset non-zero small value used to prevent division by zero errors when the weighted in-degree centrality is zero.

[0087] It should be noted that the information hub degree is the ratio between the weighted out-degree centrality and the weighted in-degree centrality. The information hub degree of each data source node is calculated as follows: the weighted out-degree centrality is used as the numerator, and the sum of the weighted in-degree centrality and a preset non-zero small quantity is used as the denominator, and the ratio of the two is calculated. An information hub degree greater than 1 indicates that the data source node is a net sender of information, less than 1 indicates that it is a net receiver of information, and equal to 1 indicates that sending and receiving are balanced.

[0088] This application defines weighted out-degree or in-degree centrality on the implicit association graph, and then calculates the information hub degree (combining the influence of out-degree or in-degree) and dynamically tracks its rate of change, quantifying the core position and evolution trend of each data source in the information network; it can not only locate the information hub node (such as the master control data source) at the current moment, but also capture abnormal signals of hub degree mutation, providing quantifiable dynamic indicators for early warning of power system faults and protection of critical nodes.

[0089] Based on information hub degree and rate of change, the comprehensive collection priority of each data source is calculated, and then the adaptive sampling frequency and transmission resource allocation of each data source are dynamically allocated.

[0090] In a specific example, the step of calculating the comprehensive acquisition priority of each data source based on the information hub degree and the rate of change, and then dynamically allocating the adaptive sampling frequency and transmission resource allocation of each data source, includes: S701, calculating the comprehensive acquisition priority of each data source based on the information hub degree and the rate of change, wherein the comprehensive acquisition priority is the product of the information hub degree and the evolution gain factor, and the evolution gain factor is calculated based on the rate of change.

[0091] S702, based on the comprehensive acquisition priority, determine the adaptive sampling frequency of each data source, wherein the adaptive sampling frequency is the frequency value after adjusting the base sampling frequency according to the degree to which the comprehensive acquisition priority deviates from the average priority of all data sources.

[0092] S703, based on the comprehensive acquisition priority, determine the transmission resource allocation for each data source, wherein the transmission resource allocation is the total system transmission resource after allocating the resources according to the ratio of the comprehensive acquisition priority of each data source to the total priority.

[0093] It should be noted that the calculation of the comprehensive collection priority of each data source is as follows: the information hub degree is multiplied by the evolution gain factor, where the evolution gain factor is 1 plus the product of the preset evolution trend gain coefficient and the rate of change. This formula ensures that the data source that is currently an information hub and whose status is rising receives the highest comprehensive collection priority, while the priority of hub nodes whose status is declining is reduced, thereby freeing up resources for emerging new hub nodes.

[0094] Furthermore, the expression for the overall acquisition priority is:

[0095]

[0096] in, For the first The information hub status of each data source For the first The rate of change in the information hub degree of each data source This is the preset evolution trend gain coefficient.

[0097] It should be noted that determining the adaptive sampling frequency for each data source specifically involves: calculating the average of the comprehensive acquisition priorities of all data sources; for each data source, subtracting the average from its comprehensive acquisition priority to obtain the difference; dividing the difference by the average to obtain the amplitude modulation ratio; multiplying the base sampling frequency by 1 and summing it with the amplitude modulation ratio to obtain the adaptive sampling frequency for that data source; for data sources with a comprehensive acquisition priority higher than the average, the adaptive sampling frequency is higher than the base sampling frequency, and for data sources with a priority lower than the average, the adaptive sampling frequency is lower than the base sampling frequency.

[0098] Furthermore, the expression for the adaptive sampling frequency is:

[0099]

[0100] in, The preset base sampling frequency, For the first The overall collection priority of each data source The average priority of all data sources.

[0101] In a specific instance, the calculation process for the allocated transmission resources is as follows:

[0102]

[0103] in, For the total system transmission resource budget, For the first The overall collection priority of each data source This is the sum of the collection priorities for all data sources. This represents the total number of data sources.

[0104] It should be noted that determining the transmission resource allocation for each data source involves: summing the overall collection priorities of all data sources to obtain a total priority; for each data source, dividing its overall collection priority by the total priority to obtain the resource allocation ratio for that data source; and multiplying the total system transmission resource budget by the resource allocation ratio to obtain the transmission resource allocation for that data source. This allocation method ensures that data sources with higher overall collection priorities receive more transmission bandwidth and cache space.

[0105] This application integrates the static importance of information hub degree with the dynamic sensitivity of change rate to establish a comprehensive collection priority, and adjusts the sampling frequency and uplink transmission resource allocation of each data source online accordingly. Under the condition of limited communication bandwidth and computing resources, it prioritizes the fine collection of high-priority data sources (such as hub nodes and nodes with drastic changes), while reducing the sampling burden of redundant data, and achieves the optimal balance between resource utilization efficiency and data value density.

[0106] The implicit association graph is updated based on the data obtained from the refined sampling frequency and the transmission resource allocation. The refined sampling frequency is used to perform fine-grained data acquisition on each data source.

[0107] In a specific example, the step of performing refined collection on each data source based on the adaptive sampling frequency and the transmission resource allocation, and updating the implicit correlation graph based on the data obtained from the refined collection, includes: S901, adjusting the collection time interval of each data source according to the adaptive sampling frequency, allocating a transmission channel and buffer space for each data source according to the transmission resource allocation, performing refined collection, and obtaining refined collection data for each data source.

[0108] S902, based on the refined collected data, update the edge weight matrix of the implicit association graph to obtain the updated implicit association graph.

[0109] S903, based on the updated implicit association graph, cyclically execute the steps of calculating the transfer entropy, calculating the information hub degree, calculating the comprehensive acquisition priority, and refining the acquisition, so as to realize the online evolution of the implicit association graph and the adaptive dynamic adjustment of the acquisition strategy.

[0110] It should be noted that the fine-grained acquisition process involves: distributing the adaptive sampling frequency to the acquisition units of each data source, and the acquisition units sampling data according to the adjusted time interval; allocating corresponding communication bandwidth and local cache space to each data source according to the transmission resource allocation; and after acquisition, obtaining high-resolution data containing timestamps and measurement values ​​to obtain fine-grained acquisition data for each data source.

[0111] It should be noted that updating the implicit correlation graph specifically involves: using the newly acquired refined data as input to a new standardized data sequence, recalculating the transfer entropy between any two data sources, recalculating the directed edge weights, and updating the edge weight matrix of the implicit correlation graph so that the implicit correlation graph can reflect the latest correlation between the data sources.

[0112] It should be noted that the above steps are executed cyclically as follows: using the updated implicit association graph as input for the new round of graph structure analysis, the weighted out-degree centrality and weighted in-degree centrality of each data source node are recalculated, the information hub degree and its rate of change are recalculated, the comprehensive acquisition priority is recalculated, the adaptive sampling frequency and transmission resource allocation are reallocated, and refined acquisition is re-executed, forming a closed-loop evolution mechanism of "acquisition-graph construction-structure analysis-resource allocation-reacquisition". This enables the acquisition strategy to continuously and adaptively adjust to the dynamic changes in the association between data sources without manual intervention.

[0113] Please see Figure 2 As shown, this application provides a system for intelligent acquisition of multi-source power information in a second aspect.

[0114] The system 100 of the intelligent acquisition method for multi-source power information described in this invention can be installed in an electronic device. Depending on the functions implemented, the system 100 may include a multi-source power data synchronization and preprocessing module 101, a transfer entropy correlation map construction module 102, an information hub degree and rate of change analysis module 103, an adaptive sampling and resource allocation module 104, and a refined acquisition and map iterative update module 105. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0115] In this embodiment, the functions of each module / unit are as follows: The multi-source power data synchronization and preprocessing module acquires the original power information from multiple power data sources, performs synchronization and preprocessing, and obtains a standardized data sequence.

[0116] The transfer entropy association graph construction module calculates the transfer entropy between various data sources based on standardized data sequences and constructs an implicit association graph.

[0117] The information hub degree and rate of change analysis module, based on the implicit association graph, calculates the weighted out-degree centrality and weighted in-degree centrality of each data source node, calculates the information hub degree of each data source node, and analyzes the rate of change of the information hub degree.

[0118] The adaptive sampling and resource allocation module calculates the comprehensive acquisition priority of each data source based on information hub degree and change rate, and then dynamically allocates the adaptive sampling frequency and transmission resource allocation of each data source.

[0119] The refined data acquisition and map iterative update module performs refined data acquisition on each data source based on the adaptive sampling frequency and the transmission resource allocation, and updates the implicit association map based on the data obtained from the refined data acquisition.

[0120] This application provides a multi-source power information intelligent acquisition method and system. By constructing an implicit correlation graph between data sources through transfer entropy, it overcomes the limitation of traditional methods that can only handle explicit correlations and discovers nonlinear causal driving relationships in power data. Based on this, it calculates the weighted in-degree centrality and information hub degree of nodes and their rate of change, thereby accurately identifying data sources that play a key role in the flow of system information. Furthermore, it dynamically determines the comprehensive acquisition priority by combining hub degree and rate of change, and adaptively allocates the sampling frequency and transmission resources of each data source, avoiding resource waste while ensuring the refined acquisition of core data. Finally, by iteratively updating the correlation graph, the system continuously tracks the evolution of the power data interaction structure, effectively improving the intelligence, real-time performance, and resource utilization efficiency of multi-source power data acquisition.

[0121] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0122] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0123] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0124] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0125] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent acquisition of multi-source power information, characterized in that, include: The raw power information from multiple power data sources is acquired, synchronized, and preprocessed to obtain a standardized data sequence. Based on standardized data sequences, the transfer entropy between various data sources is calculated, and an implicit correlation graph is constructed. Based on the implicit association graph, the weighted out-degree centrality and weighted in-degree centrality of each data source node are calculated, the information hub degree of each data source node is calculated, and the rate of change of information hub degree is analyzed. Based on information hub degree and change rate, the comprehensive collection priority of each data source is calculated, and then the adaptive sampling frequency and transmission resource allocation of each data source are dynamically allocated. The implicit association graph is updated based on the data obtained from the refined sampling frequency and the transmission resource allocation. The refined sampling frequency is used to perform fine-grained data acquisition on each data source.

2. The intelligent acquisition method for multi-source power information according to claim 1, characterized in that, The process of acquiring raw power information from multiple power data sources, synchronizing and preprocessing it to obtain a standardized data sequence includes: S201, Obtain raw power information from multiple power data sources, including smart meter data sources, phasor measurement unit (PMU) data sources, SCADA data sources, and environmental sensor data sources; S202, timestamp alignment and missing value filling are performed on the original power information from each data source to generate a multi-source power data set after synchronous preprocessing; S203, standardize the data sequences of each data source in the multi-source power data set after synchronous preprocessing to obtain a standardized data sequence.

3. The intelligent acquisition method for multi-source power information according to claim 1, characterized in that, The process of calculating the transfer entropy between data sources based on standardized data sequences and constructing an implicit correlation graph includes: S301, based on the standardized data sequence, calculate the... The data source for the first The transfer entropy of a data source; S302, based on the transfer entropy between each data source, normalize each transfer entropy to obtain the normalized directed edge weights; S303 uses each data source as a vertex, retains directed edges with weights greater than a preset association strength threshold, and constructs an implicit association graph.

4. The intelligent acquisition method for multi-source power information according to claim 3, characterized in that, The process of calculating the transfer entropy between data sources based on standardized data sequences and constructing an implicit correlation graph includes: S301, based on the standardized data sequence, calculate the... The data source for the first The transfer entropy of a data source; S302, based on the transfer entropy between each data source, normalize each transfer entropy to obtain the normalized directed edge weights; S303 uses each data source as a vertex, retains directed edges with weights greater than a preset association strength threshold, and constructs an implicit association graph.

5. The intelligent acquisition method for multi-source power information according to claim 1, characterized in that, The implicit association graph is used to calculate the weighted out-degree centrality and weighted in-degree centrality of each data source node, calculate the information hub degree of each data source node, and analyze the rate of change of information hub degree, including: S501, based on the edge weight matrix of the implicit association graph, calculates the weighted out-degree centrality and weighted in-degree centrality of each data source node; S502, Calculate the information hub degree of each data source node based on the weighted out-degree centrality and the weighted in-degree centrality; S503, construct a sliding time window sequence, calculate the rate of change of the information hub degree of each data source node within multiple consecutive time windows, and the rate of change is used to characterize the trend of the evolution of the status of the data source node in the implicit association graph.

6. The intelligent acquisition method for multi-source power information according to claim 5, characterized in that, The calculation of the information hub degree of each data source node based on the weighted out-degree centrality and the weighted in-degree centrality includes: The expression for information hub degree is: ； in, For the first Weighted out-degree centrality of each data source node For the first Weighted in-degree centrality of each data source node, This is a preset non-zero small value used to prevent division by zero errors when the weighted in-degree centrality is zero.

7. The intelligent acquisition method for multi-source power information according to claim 1, characterized in that, The process of calculating the comprehensive acquisition priority of each data source based on information hub degree and rate of change, and then dynamically allocating the adaptive sampling frequency and transmission resource allocation of each data source, includes: S701, based on the information hub degree and the rate of change, calculate the comprehensive collection priority of each data source, wherein the comprehensive collection priority is the product of the information hub degree and the evolution gain factor, and the evolution gain factor is calculated based on the rate of change; S702, based on the comprehensive acquisition priority, determine the adaptive sampling frequency of each data source, wherein the adaptive sampling frequency is the frequency value after adjusting the base sampling frequency according to the degree to which the comprehensive acquisition priority deviates from the average priority of all data sources; S703, based on the comprehensive acquisition priority, determine the transmission resource allocation for each data source, wherein the transmission resource allocation is the total system transmission resource after allocating the resources according to the ratio of the comprehensive acquisition priority of each data source to the total priority.

8. The intelligent acquisition method for multi-source power information according to claim 7, characterized in that, The specific calculation process for the allocated transmission resources is as follows: ； in, For the total system transmission resource budget, For the first The overall collection priority of each data source This is the sum of the collection priorities for all data sources. This represents the total number of data sources.

9. The intelligent acquisition method for multi-source power information according to claim 1, characterized in that, The process of performing refined data collection on each data source based on adaptive sampling frequency and transmission resource allocation, and updating the implicit correlation graph based on the data obtained from the refined data collection, includes: S901, adjust the acquisition time interval of each data source according to the adaptive sampling frequency, allocate transmission channels and buffer space for each data source according to the transmission resource allocation amount, perform fine-grained acquisition, and obtain fine-grained acquisition data of each data source; S902, Based on the refined collected data, update the edge weight matrix of the implicit association graph to obtain the updated implicit association graph; S903, based on the updated implicit association graph, cyclically execute the steps of calculating the transfer entropy, calculating the information hub degree, calculating the comprehensive acquisition priority, and refining the acquisition, so as to realize the online evolution of the implicit association graph and the adaptive dynamic adjustment of the acquisition strategy.

10. A system for implementing the intelligent acquisition method for multi-source power information according to any one of claims 1-9, characterized in that, include: The multi-source power data synchronization and preprocessing module acquires raw power information from multiple power data sources, performs synchronization and preprocessing, and obtains a standardized data sequence. The transfer entropy association graph construction module calculates the transfer entropy between various data sources based on standardized data sequences and constructs an implicit association graph. The information hub degree and rate of change analysis module, based on the implicit association graph, calculates the weighted out-degree centrality and weighted in-degree centrality of each data source node, calculates the information hub degree of each data source node, and analyzes the rate of change of information hub degree. The adaptive sampling and resource allocation module calculates the comprehensive acquisition priority of each data source based on information hub degree and rate of change, and then dynamically allocates the adaptive sampling frequency and transmission resource allocation of each data source. The refined data acquisition and map iterative update module performs refined data acquisition on each data source based on the adaptive sampling frequency and the transmission resource allocation, and updates the implicit association map based on the data obtained from the refined data acquisition.

Images