A method and system for collaborative management of distributed power grid terminals

Abstract

The present invention discloses a method and system for collaborative management of distributed terminals in a power grid, and relates to the technical field of power grid information processing, including: obtaining terminal nodes of distributed terminals in the power grid; performing data collection on the terminal nodes to obtain terminal monitoring data sets; obtaining an isolated terminal anti-risk assessment model, inputting the terminal monitoring data sets into the isolated terminal anti-risk assessment model for anti-risk assessment, and outputting anti-risk performance indicators; training an inter-class balanced classifier, inputting the anti-risk performance indicators into the collaborative terminal anti-risk assessment model to perform balanced division of the terminal nodes, and obtaining balanced division results that meet balanced constraint conditions, wherein the balanced division results are N inter-class collaborative groups; and performing inter-class collaborative management of distributed terminals in the power grid according to the N inter-class collaborative groups. The present invention solves the technical problems of insufficient control accuracy and low management efficiency in the prior art, and achieves the technical effect of improving the operation efficiency and stability of the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid information processing, and in particular to a method and system for collaborative management of distributed power grid terminals. Background Art

[0002] With the continuous development of the power industry and the accelerated advancement of smart grid construction, distributed terminals in the power grid play a vital role in the power system. These terminals are not only responsible for data collection, transmission and monitoring, but are also directly related to the reliability, safety and economy of power supply. However, with the increasing complexity of the power grid structure, the widespread access to distributed energy and the diversification of user needs, the management and control of distributed terminals in the power grid face huge challenges.

[0003] Traditional power grid distributed terminal management methods mainly rely on centralized control mode, which can basically meet the needs when the power grid is small and the structure is relatively simple. However, in the current power grid environment, due to the sharp increase in the number of devices, the complexity of network structure and the diversification of energy types, the centralized control mode has gradually exposed problems such as insufficient control accuracy and low management efficiency. Summary of the invention

[0004] The present application provides a method and system for collaborative management of distributed terminals in a power grid, which are used to solve the technical problems of insufficient control accuracy and low management efficiency in the prior art.

[0005] In view of the above problems, the present application provides a method and system for collaborative management of distributed terminals in a power grid.

[0006] In a first aspect of the present application, a method for coordinated management of distributed terminals in a power grid is provided, the method comprising:

[0007] Acquire multiple terminal nodes of a distributed terminal of a power grid; perform data collection on the multiple terminal nodes to acquire multiple terminal monitoring data sets, wherein the multiple terminal nodes correspond to the multiple terminal monitoring data sets; acquire an isolated-terminal anti-risk assessment model, input the multiple terminal monitoring data sets into the isolated-terminal anti-risk assessment model for anti-risk assessment, and output multiple anti-risk performance indicators; train an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, input the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division on the multiple terminal nodes, and obtain a balanced division result that satisfies a balanced constraint condition, wherein the balanced division result is N inter-class collaborative groups, each inter-class collaborative group includes at least two terminal nodes; perform inter-class collaborative management on the distributed terminals of the power grid according to the N inter-class collaborative groups.

[0008] A second aspect of the present application provides a power grid distributed terminal collaborative management system, the system comprising:

[0009] A terminal node acquisition module, wherein the terminal node acquisition module acquires multiple terminal nodes of a power grid distributed terminal; a data acquisition module, wherein the data acquisition module performs data acquisition on the multiple terminal nodes to acquire multiple terminal monitoring data sets, wherein the multiple terminal nodes correspond to the multiple terminal monitoring data sets; an anti-risk assessment module, wherein the anti-risk assessment module acquires an isolated-terminal anti-risk assessment model, inputs the multiple terminal monitoring data sets into the isolated-terminal anti-risk assessment model for anti-risk assessment, and outputs multiple anti-risk performance indicators; a balanced partitioning module, wherein the balanced partitioning module trains an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, inputs the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced partitioning on the multiple terminal nodes, and obtains a balanced partitioning result that satisfies a balanced constraint condition, wherein the balanced partitioning result is N inter-class collaborative groups, and each inter-class collaborative group includes at least two terminal nodes; an inter-class collaborative management module, wherein the inter-class collaborative management module performs inter-class collaborative management on the power grid distributed terminals according to the N inter-class collaborative groups.

[0010] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] The present application obtains multiple terminal nodes of a distributed terminal of a power grid; performs data collection on multiple terminal nodes to obtain multiple terminal monitoring data sets, wherein multiple terminal nodes correspond to multiple terminal monitoring data sets; obtains an isolated-terminal anti-risk assessment model, inputs multiple terminal monitoring data sets into the isolated-terminal anti-risk assessment model for anti-risk assessment, and outputs multiple anti-risk performance indicators; trains an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, and inputs multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division of multiple terminal nodes, and obtains a balanced division result that satisfies the balanced constraint condition, wherein the balanced division result is N inter-class collaborative groups, each inter-class collaborative group includes at least two terminal nodes; and performs inter-class collaborative management of distributed terminals of the power grid according to the N inter-class collaborative groups. The present invention solves the technical problems of insufficient control accuracy and low management efficiency in the prior art, and achieves the technical effect of improving the operation efficiency and stability of the power grid through real-time monitoring, data analysis, collaborative control and other means. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 A schematic flow chart of a method for collaborative management of distributed terminals in a power grid provided in an embodiment of the present application.

[0014] Figure 2 A schematic diagram of the structure of a power grid distributed terminal collaborative management system provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: terminal node acquisition module 11, data collection module 12, risk resistance assessment module 13, balanced division module 14, inter-class collaborative management module 15. DETAILED DESCRIPTION

[0016] The present application provides a method and system for collaborative management of distributed terminals in a power grid, aiming to solve the technical problems of insufficient control accuracy and low management efficiency in the prior art, and achieves the technical effect of improving the operating efficiency and stability of the power grid through real-time monitoring, data analysis, collaborative control and other means.

[0017] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0018] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0019] Embodiment 1

[0020] like Figure 1 As shown, the present application provides a method for coordinated management of distributed power grid terminals, the method comprising:

[0021] Step S100: Acquire multiple terminal nodes of a distributed terminal of a power grid.

[0022] In the embodiment of the present application, the power grid management system is used to scan the entire power grid to identify all terminal devices connected to the power grid. These devices are discovered using network protocols such as SNMP and Modbus.

[0023] Through the above process, multiple terminal nodes of the power grid distributed terminal are obtained.

[0024] Step S200: collecting data from the plurality of terminal nodes to obtain a plurality of terminal monitoring data sets, wherein the plurality of terminal nodes correspond to the plurality of terminal monitoring data sets.

[0025] In the embodiment of the present application, the data type to be collected is first determined, such as voltage, current, power, temperature, etc. Then, data is collected according to the set sampling frequency through pre-installed sensors and smart meters and other devices. For example, a smart meter records voltage, current and power data once a minute.

[0026] Clean the collected raw data to remove noise, outliers, etc. to ensure the accuracy and reliability of the data. Assign a unique identifier to each terminal node, such as a device ID, and associate the collected data with the identifier.

[0027] Through the above process, multiple terminal monitoring data sets are obtained, wherein each terminal node corresponds to its own terminal monitoring data set through a device ID.

[0028] Step S300: obtaining an isolated-terminal anti-risk assessment model, inputting the plurality of terminal monitoring data sets into the isolated-terminal anti-risk assessment model to perform anti-risk assessment, and outputting a plurality of anti-risk performance indicators.

[0029] In an embodiment of the present application, an isolated-terminal anti-risk assessment model based on deep learning is first obtained. The model is already trained and is used to assess the anti-risk capability of power grid terminals.

[0030] Next, data preprocessing is performed on multiple terminal monitoring data sets, including data cleaning, format conversion, outlier processing, etc., to ensure the quality and accuracy of the data to meet the needs of model operation.

[0031] The preprocessed data is input into the isolated-terminal anti-risk assessment model, which performs anti-risk assessment on the input data set and calculates the anti-risk performance index of each terminal node. Multiple anti-risk performance indicators are obtained through model calculation, including the fault recovery time, fault frequency, communication stability and maintenance response speed of each terminal node.

[0032] Step S400: training an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, inputting the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division on the multiple terminal nodes, and obtaining a balanced division result that satisfies the balanced constraint conditions, wherein the balanced division result is N inter-class collaborative groups, and each inter-class collaborative group includes at least two terminal nodes.

[0033] In the embodiment of the present application, when training the inter-class balanced classifier, the K-means clustering model is selected, and the number of clusters N is set to the desired number of cooperative groups. The K-means model is trained using the standardized anti-risk performance index, and the cluster center and the number of iterations are adjusted to improve the classification accuracy.

[0034] The inter-class balance classifier is embedded with a collaborative-terminal anti-risk assessment model, which is integrated by training multiple temporal network layers. The multiple temporal network layers are respectively trained by long short-term memory network model, GRU model and recurrent neural network model through historical traffic monitoring samples.

[0035] The trained collaborative-terminal anti-risk assessment model is embedded in the K-means classifier, and the anti-risk performance index data is input into the classifier. The K-means model generates preliminary collaborative groups. The collaborative-terminal anti-risk assessment model evaluates the anti-risk performance of each collaborative group and calculates the average index and variance of each group. If the initial division does not meet the equilibrium condition, the genetic algorithm is used to adjust the collaborative group and iterate the optimization until the condition is met. The final balanced division result is output, including N inter-class collaborative groups, each of which contains at least two terminal nodes.

[0036] Step S500: performing inter-class collaborative management on the power grid distributed terminals according to the N inter-class collaborative groups.

[0037] In the embodiment of the present application, when performing inter-class collaborative management, the types of resources that can be shared are determined, such as power output, backup power supply, maintenance resources, etc. Set at load peak, sudden failure, equipment maintenance. Use the DRMS ​​system to dynamically monitor and adjust shared resources.

[0038] Ensure that high-risk endpoints provide support to low-risk endpoints during high load or failure situations. Use DRMS ​​to monitor and adjust shared resources in real time.

[0039] Use SolarWinds, Nagios, etc. to collect real-time operation status, load conditions, and fault information. Dynamically adjust the resource allocation of each terminal node, such as power output, communication bandwidth, etc., based on real-time data and preset rules. Use automated deployment tools such as Ansible and Puppet to automatically adjust resource configuration based on monitoring data and deployment strategies.

[0040] Through the above process, inter-class collaborative management of distributed terminals in the power grid is completed.

[0041] Furthermore, step S300 in the method provided in the application embodiment also includes:

[0042] Perform anti-risk factor analysis on the multiple terminal monitoring data sets to obtain multiple anti-risk factors, wherein the multiple anti-risk factors include fault recovery time, fault occurrence frequency, communication stability, and maintenance response speed; generate an anti-risk factor sample set based on the multiple anti-risk factors; train the isolated-terminal anti-risk assessment model according to the anti-risk factor sample set and identification information that identifies the anti-risk performance, and connect the output end of the isolated-terminal anti-risk assessment model to the input end of the inter-class balanced classifier.

[0043] In an embodiment of the present application, data cleaning is performed on multiple terminal monitoring data sets, and then anti-risk factor analysis is performed. Anti-risk factors include fault recovery time, fault frequency, communication stability, and maintenance response speed. The fault recovery time is the time required for each terminal node to recover from a faulty state to a normal state. The fault frequency is the number of times a terminal node fails within a specific time period, such as the number of times a fault occurs within a month. Communication stability evaluates whether the communication connection between the terminal and other systems or devices is stable, and is expressed as a percentage of packet loss rate or delay. The maintenance response speed is the response time for each terminal node failure.

[0044] The acquired anti-risk factors are combined with the corresponding terminal instances to form an anti-risk factor dataset. Each data point contains the terminal's identification information and the specific value of each anti-risk factor.

[0045] Next, identification information to identify the risk resistance performance is added to each terminal or each risk resistance factor data point in the risk resistance factor sample set. These identification information are divided into risk levels 1, 2, and 3. Among them, level 1 has the highest risk and level 3 has the lowest risk. The identification is determined by technical experts.

[0046] According to the anti-risk factor sample set and the identification information that identifies the anti-risk performance, the isolated-terminal anti-risk assessment model is trained. When training the isolated-terminal anti-risk assessment model, a suitable machine learning model is selected according to the data characteristics. In this application, a random forest model is used for training. The anti-risk factor sample set is divided into a training set and a test set, and the ratio of the training set to the test set is 7:3. Use the training set to train the model and adjust the model parameters to improve the evaluation accuracy. Use the test set to verify the model performance and calculate evaluation indicators such as accuracy, precision, recall, and F1-score to ensure the generalization ability of the model. The training of the isolated-terminal anti-risk assessment model is completed through the above process.

[0047] Finally, data pipeline tools such as Apache Kafka, RabbitMQ, and Apache Nifi are used to build a data transmission pipeline to connect the output of the isolated-terminal anti-risk assessment model with the input of the inter-class balance classifier.

[0048] Further, the multiple anti-risk performance indicators are input into the collaborative-terminal anti-risk assessment model to perform balanced division on the multiple terminal nodes to obtain a balanced division result that satisfies the balanced constraint condition. The method further includes:

[0049] The multiple terminal nodes are initialized and classified by using the inter-class balanced classifier to obtain multiple initialized collaborative groups; the multiple anti-risk performance indicators are input into the collaborative-terminal anti-risk assessment model, and the collaborative-terminal anti-risk assessment model performs collaborative anti-risk performance assessment on the multiple initialized collaborative groups, and outputs multiple collaborative anti-risk performance indicators corresponding to the multiple initialized collaborative groups; the inter-class balanced classifier determines whether the multiple collaborative anti-risk performance indicators meet the balanced constraint conditions, and if the balanced constraint conditions are not met, the multiple initialized collaborative groups are optimized until the balanced constraint conditions are met, and a balanced division result is output.

[0050] In an embodiment of the present application, the anti-risk performance index data output by the isolated-terminal anti-risk assessment model in the previous step is input into the inter-class balance classifier, and initial classification is performed according to preset classification standards, such as risk level, to obtain multiple initialization collaborative groups.

[0051] The collaborative group data obtained by initialization classification is input into the collaborative-terminal anti-risk evaluation model. The collaborative-terminal anti-risk evaluation model evaluates the anti-risk performance of each collaborative group and obtains the collaborative anti-risk performance index corresponding to each collaborative group.

[0052] Next, the inter-class equilibrium classifier determines whether the current division meets the requirements based on the preset equilibrium constraints, such as the variance of the risk resistance performance index of each collaborative group is less than a specific threshold. Calculate the variance of the risk resistance performance index of each collaborative group and compare it with the preset equilibrium constraints. If all collaborative groups meet the equilibrium constraints, output the current equilibrium division result. If some collaborative groups do not meet the equilibrium constraints, perform optimization adjustments. Use optimization algorithms such as genetic algorithms, simulated annealing, particle swarm optimization, etc. to adjust the collaborative groups that do not meet the conditions. Reclassify and evaluate the adjusted risk resistance performance indicators to generate new collaborative groups. Iterate the optimization and evaluation until all collaborative groups meet the equilibrium constraints. Finally, output the equilibrium division result that meets the equilibrium constraints.

[0053] Furthermore, the method further comprises:

[0054] The inter-class balance classifier determines whether the multiple collaborative risk resistance performance indicators meet the balance constraint conditions, and the balance constraint conditions include a first constraint condition and a second constraint condition; wherein, the first constraint condition is that the collaborative risk resistance performance indicators corresponding to each of the N inter-class collaborative groups are greater than or equal to a preset collaborative risk resistance performance indicator; the second constraint condition is that the variance of the inter-group risk resistance performance indicators of each of the N inter-class collaborative groups is less than the expected variance.

[0055] In the embodiment of the present application, the equilibrium constraint condition includes a first constraint condition and a second constraint condition.

[0056] The first constraint condition is that the collaborative anti-risk performance index corresponding to each inter-class collaborative group in the N inter-class collaborative groups is greater than or equal to the preset collaborative anti-risk performance index. For example, the preset threshold of the fault recovery time is 0.2 hours, the fault frequency is 2 times per month, the communication stability is 0.9, and the maintenance response speed is 1 hour. Only when all the anti-risk performance indicators of the collaborative group are greater than or equal to the preset collaborative anti-risk performance indicators, the first constraint condition is considered to be met.

[0057] The second constraint is that the variance of the inter-group anti-risk performance index of each inter-class collaborative group in the N inter-class collaborative groups is less than the expected variance. When judging whether the second constraint is met, the variance of each anti-risk performance index of all collaborative groups is calculated. For example, a statistical function is used to calculate the variance of the fault recovery time of each collaborative group. The calculated variance is compared with the preset expected variance. For example, the expected variance of the fault recovery time is set to 0.01, and the expected variance of the communication stability is set to 0.005. If the variance of the anti-risk performance indicators of all collaborative groups is less than the expected variance, the second constraint is met.

[0058] Furthermore, the method further comprises:

[0059] The inter-class balanced classifier also includes a third constraint condition, which includes a preset discreteness; obtaining the distribution state of the risk resistance performance indicators of each inter-class collaborative group; performing a discreteness analysis on the distribution state of the risk resistance performance indicators to obtain the discreteness corresponding to each inter-class collaborative group; the inter-class balanced classifier performs discrete constraints on the distribution state of the indicators within each inter-class collaborative group according to the preset discreteness.

[0060] In the embodiment of the present application, the inter-class balanced classifier also includes a third constraint condition. The third constraint condition requires that the terminal nodes within each inter-class collaborative group have a certain degree of dispersion in the risk resistance performance index, that is, there are both terminal nodes with high risk resistance performance and terminal nodes with low risk resistance performance in the group, but the dispersion should not exceed the preset dispersion threshold.

[0061] Obtain the risk resistance performance index data of each inter-class collaborative group from the collaborative-terminal risk resistance assessment model. These data include indicators such as the fault recovery time, fault frequency, communication stability, and maintenance response speed of each terminal node. Use data analysis tools such as Pandas or R to calculate the distribution status of the risk resistance performance index of each collaborative group, including the maximum, minimum, mean, median, standard deviation, and other statistics of each indicator.

[0062] Conduct a dispersion analysis on the distribution status of the risk resistance performance indicators of each collaborative group. Commonly used dispersion indicators include range, variance and standard deviation. Calculate the dispersion of each risk resistance performance indicator of each collaborative group.

[0063] Next, the inter-class balance classifier makes a discrete constraint judgment on the distribution state of the indicators within each collaborative group according to the preset discreteness. The preset discreteness is set by technical experts according to actual needs. For example, the discreteness threshold of the fault recovery time is set to 0.05 hours, and the discreteness threshold of the fault occurrence frequency is set to 1 time / month.

[0064] If the dispersion of the risk resistance performance indicators of all collaborative groups meets the preset dispersion threshold, it is considered that the current division result meets the third constraint condition.

[0065] If the discreteness of any collaborative group does not meet the preset threshold, the collaborative group needs to be adjusted. Use optimization algorithms such as genetic algorithms, simulated annealing, particle swarm optimization, etc. to adjust the collaborative groups that do not meet the discreteness constraints. When the discreteness is too small, transfer some terminal nodes from the current collaborative group to other collaborative groups to increase the performance difference of the nodes within the current collaborative group. When the discreteness is too large, adjust the configuration of the members within the group while keeping the total number of collaborative groups unchanged so that the performance indicator difference of the nodes within each collaborative group is moderate. Repeat adjustments and evaluations until the discreteness of all collaborative groups meets the preset threshold and the risk resistance performance indicators within the group are reasonably distributed. Complete the discrete constraints on the distribution status of the indicators within each inter-class collaborative group.

[0066] Further, the collaborative-terminal anti-risk assessment model is trained, and the method further includes:

[0067] Train multiple timing network layers, the multiple timing network layers include continuous flow peak timing period, intermittent flow peak timing period and smooth flow timing period; collect continuous flow collaborative monitoring samples, intermittent flow collaborative monitoring samples and smooth flow collaborative monitoring samples; train the multiple timing network layers according to the continuous flow collaborative monitoring samples, the intermittent flow collaborative monitoring samples and the smooth flow collaborative monitoring samples, and construct a collaborative-terminal anti-risk assessment model.

[0068] In an embodiment of the present application, when training multiple timing network layers, continuous flow collaborative monitoring samples, intermittent flow collaborative monitoring samples and smooth flow collaborative monitoring samples are collected.

[0069] Sustained traffic refers to situations where the terminal node maintains high load for a long time, such as during the peak power consumption period in summer. Through smart meters, sensors and SCADA systems, data of power grid terminal nodes during sustained high load periods, including voltage, current, power, etc., are collected to obtain continuous traffic collaborative monitoring samples.

[0070] Intermittent traffic refers to situations where the load of the terminal node fluctuates dramatically in a short period of time, such as large equipment that is started intermittently. Collect data from the power grid terminal nodes during periods of drastic load fluctuations, record the frequency and amplitude of load changes, and obtain intermittent traffic collaborative monitoring samples.

[0071] Smooth flow refers to the situation where the load of the terminal node is relatively stable, such as daily electricity consumption. Collect data of the power grid terminal nodes during the stable load period, record the load stability, voltage and current changes, etc., and obtain smooth flow collaborative monitoring samples.

[0072] When training multiple time series network layers, use the long short-term memory network model to process continuous high-load data. Use the GRU model to process data with severe load fluctuations. Use the recurrent neural network model to process stable load data. Next, use the continuous traffic collaborative monitoring samples to train the long short-term memory network model, use the intermittent traffic collaborative monitoring samples to train the GRU model, and use the smooth traffic collaborative monitoring samples to train the recurrent neural network model. After training, multiple time series network layers are obtained.

[0073] Finally, the multiple time series network layers obtained through training are integrated to complete the construction of the collaborative-terminal anti-risk assessment model.

[0074] Furthermore, step S500 in the method provided in the application embodiment also includes:

[0075] Determine a first inter-class collaborative group among the N inter-class collaborative groups, wherein the first inter-class collaborative group includes K terminal nodes; identify communication configuration parameters of the K terminal nodes, generate communication protocols corresponding to the K terminal nodes according to the communication configuration parameters, and perform inter-class collaborative management of the K terminal nodes according to the communication protocol.

[0076] In the embodiment of the present application, the initial classification results of N inter-class cooperative groups are obtained from the output of the inter-class balanced classifier. The cooperative group with the best anti-risk performance index is selected as the first inter-class cooperative group. The first inter-class cooperative group includes K terminal nodes.

[0077] Use the existing network management system and terminal equipment management system to collect the communication configuration parameters of each terminal node, including communication protocol type, bandwidth, delay, network topology, etc. According to the communication configuration parameters of K terminal nodes, use the existing communication protocol generation tools, such as network configuration management software, to generate the corresponding communication protocol. The protocol content includes the specific method of communication between nodes, the format and frequency of data transmission, error handling mechanism, etc. Next, use network simulation tools such as OPNET and NS2 to verify the generated communication protocol to ensure its feasibility in practical applications.

[0078] Finally, inter-class collaborative management is performed on the K terminal nodes according to the communication protocol. Inter-class collaborative management includes that when the high-risk terminal has sufficient resources, it provides the low-risk terminal with necessary resource support, such as computing power, storage space, network bandwidth, etc. Through resource sharing, the resource load of the entire collaborative group is balanced to improve the overall efficiency. When a low-risk terminal faces a resource bottleneck, resources are allocated to the low-risk terminal from the high-risk terminal or other resource pools.

[0079] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects:

[0080] The present application obtains multiple terminal nodes of a distributed terminal of a power grid; performs data collection on multiple terminal nodes to obtain multiple terminal monitoring data sets, wherein multiple terminal nodes correspond to multiple terminal monitoring data sets; obtains an isolated-terminal anti-risk assessment model, inputs multiple terminal monitoring data sets into the isolated-terminal anti-risk assessment model for anti-risk assessment, and outputs multiple anti-risk performance indicators; trains an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, and inputs multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division of multiple terminal nodes, and obtains a balanced division result that satisfies the balanced constraint condition, wherein the balanced division result is N inter-class collaborative groups, each inter-class collaborative group includes at least two terminal nodes; and performs inter-class collaborative management of distributed terminals of the power grid according to the N inter-class collaborative groups. The present invention solves the technical problems of insufficient control accuracy and low management efficiency in the prior art, and achieves the technical effect of improving the operation efficiency and stability of the power grid through real-time monitoring, data analysis, collaborative control and other means.

[0081] Embodiment 2

[0082] Based on the same inventive concept as the method for coordinated management of distributed power grid terminals in the aforementioned embodiment, Figure 2 As shown, the present application provides a power grid distributed terminal collaborative management system, and the system and method embodiments in the present application embodiments are based on the same inventive concept. The system includes:

[0083] A terminal node acquisition module 11, wherein the terminal node acquisition module 11 acquires a plurality of terminal nodes of a distributed terminal of a power grid;

[0084] A data collection module 12, wherein the data collection module 12 collects data from the plurality of terminal nodes to obtain a plurality of terminal monitoring data sets, wherein the plurality of terminal nodes correspond to the plurality of terminal monitoring data sets;

[0085] The anti-risk assessment module 13 obtains an isolated-terminal anti-risk assessment model, inputs the plurality of terminal monitoring data sets into the isolated-terminal anti-risk assessment model to perform anti-risk assessment, and outputs a plurality of anti-risk performance indicators;

[0086] A balanced partitioning module 14, wherein the balanced partitioning module 14 trains an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, and the multiple anti-risk performance indicators are input into the collaborative-terminal anti-risk assessment model to perform balanced partitioning on the multiple terminal nodes, and a balanced partitioning result satisfying a balanced constraint condition is obtained, wherein the balanced partitioning result is N inter-class collaborative groups, and each inter-class collaborative group includes at least two terminal nodes;

[0087] The inter-class collaborative management module 15 performs inter-class collaborative management on the power grid distributed terminals according to the N inter-class collaborative groups.

[0088] Furthermore, the system is also used to implement the following functions:

[0089] Performing anti-risk factor analysis on the plurality of terminal monitoring data sets to obtain a plurality of anti-risk factors, wherein the plurality of anti-risk factors include fault recovery time, fault occurrence frequency, communication stability, and maintenance response speed;

[0090] Generating an anti-risk factor sample set according to the multiple anti-risk factors;

[0091] According to the anti-risk factor sample set and the identification information identifying the anti-risk performance, the isolated-terminal anti-risk assessment model is trained, and the output end of the isolated-terminal anti-risk assessment model is connected to the input end of the inter-class balanced classifier.

[0092] Furthermore, the system is also used to implement the following functions:

[0093] Using the inter-class balanced classifier to initialize and classify the multiple terminal nodes to obtain multiple initialized cooperative groups;

[0094] Inputting the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model, the collaborative-terminal anti-risk assessment model performs collaborative anti-risk performance assessment on the multiple initialized collaborative groups, and outputs multiple collaborative anti-risk performance indicators corresponding to the multiple initialized collaborative groups;

[0095] The inter-class balanced classifier determines whether the multiple collaborative anti-risk performance indicators meet the balanced constraint conditions. If the balanced constraint conditions are not met, the multiple initialized collaborative groups are optimized until the balanced constraint conditions are met, and a balanced division result is output.

[0096] Furthermore, the system is also used to implement the following functions:

[0097] The inter-class balance classifier determines whether the multiple collaborative anti-risk performance indicators meet a balance constraint condition, where the balance constraint condition includes a first constraint condition and a second constraint condition;

[0098] The first constraint condition is that the collaborative anti-risk performance index corresponding to each inter-class collaborative group in the N inter-class collaborative groups is greater than or equal to a preset collaborative anti-risk performance index;

[0099] The second constraint condition is that the variance of the inter-group risk resistance performance indicator of each inter-class cooperative group in the N inter-class cooperative groups is less than the expected variance.

[0100] Furthermore, the system is also used to implement the following functions:

[0101] The inter-class equalization classifier further includes a third constraint condition, wherein the third constraint condition includes a preset discreteness;

[0102] Obtain the distribution status of the risk resistance performance indicators of each inter-class collaboration group;

[0103] Performing a dispersion analysis on the distribution state of the anti-risk performance indicator to obtain the dispersion corresponding to each inter-class collaboration group;

[0104] The inter-class balance classifier performs discrete constraints on the distribution state of the indicators within each inter-class cooperation group according to the preset discreteness.

[0105] Furthermore, the system is also used to implement the following functions:

[0106] Training a plurality of timing network layers, wherein the plurality of timing network layers include a continuous flow peak timing period, an intermittent flow peak timing period, and a flat flow timing period;

[0107] Collect continuous flow collaborative monitoring samples, intermittent flow collaborative monitoring samples, and gentle flow collaborative monitoring samples;

[0108] The multiple time-series network layers are trained according to the continuous traffic collaborative monitoring samples, the intermittent traffic collaborative monitoring samples and the smooth traffic collaborative monitoring samples to construct a collaborative-terminal anti-risk assessment model.

[0109] Furthermore, the system is also used to implement the following functions:

[0110] Determine a first inter-class cooperation group among the N inter-class cooperation groups, wherein the first inter-class cooperation group includes K terminal nodes;

[0111] Identify communication configuration parameters of the K terminal nodes, generate communication protocols corresponding to the K terminal nodes according to the communication configuration parameters, and perform inter-class collaborative management on the K terminal nodes according to the communication protocol.

[0112] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0113] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0114] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for collaborative management of distributed power grid terminals, characterized in that: The method comprises: Acquire multiple terminal nodes of a distributed terminal of a power grid; Performing data collection on the multiple terminal nodes to obtain multiple terminal monitoring data sets, wherein the multiple terminal nodes correspond to the multiple terminal monitoring data sets; Acquire an isolated-terminal anti-risk assessment model, input the plurality of terminal monitoring data sets into the isolated-terminal anti-risk assessment model to perform anti-risk assessment, and output a plurality of anti-risk performance indicators; Training an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, inputting the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division on the multiple terminal nodes, and obtaining a balanced division result that satisfies a balanced constraint condition, wherein the balanced division result is N inter-class collaborative groups, each of which includes at least two terminal nodes; Performing inter-class collaborative management on the power grid distributed terminals according to the N inter-class collaborative groups; Inputting the plurality of terminal monitoring data sets into the isolated-terminal anti-risk assessment model to perform anti-risk assessment, the method comprising: Performing anti-risk factor analysis on the plurality of terminal monitoring data sets to obtain a plurality of anti-risk factors, wherein the plurality of anti-risk factors include fault recovery time, fault occurrence frequency, communication stability, and maintenance response speed; Generating an anti-risk factor sample set according to the multiple anti-risk factors; According to the anti-risk factor sample set and the identification information identifying the anti-risk performance, training the isolated-terminal anti-risk assessment model, and connecting the output end of the isolated-terminal anti-risk assessment model to the input end of the inter-class balanced classifier; Inputting the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model to perform balanced division on the multiple terminal nodes, and obtaining a balanced division result that satisfies a balanced constraint condition, the method comprising: Using the inter-class balanced classifier to initialize and classify the multiple terminal nodes to obtain multiple initialized cooperative groups; Inputting the multiple anti-risk performance indicators into the collaborative-terminal anti-risk assessment model, the collaborative-terminal anti-risk assessment model performs collaborative anti-risk performance assessment on the multiple initialized collaborative groups, and outputs multiple collaborative anti-risk performance indicators corresponding to the multiple initialized collaborative groups; The inter-class balanced classifier determines whether the multiple collaborative anti-risk performance indicators meet the balanced constraint conditions. If the balanced constraint conditions are not met, the multiple initialized collaborative groups are optimized until the balanced constraint conditions are met, and a balanced division result is output.

2. The method according to claim 1, characterized in that The inter-class balance classifier determines whether the multiple collaborative anti-risk performance indicators meet a balance constraint condition, where the balance constraint condition includes a first constraint condition and a second constraint condition; The first constraint condition is that the collaborative anti-risk performance index corresponding to each inter-class collaborative group in the N inter-class collaborative groups is greater than or equal to a preset collaborative anti-risk performance index; The second constraint condition is that the variance of the inter-group risk resistance performance indicator of each inter-class cooperative group in the N inter-class cooperative groups is less than the expected variance.

3. The method according to claim 2, characterized in that The inter-class equalization classifier further includes a third constraint condition, wherein the third constraint condition includes a preset discreteness; Obtain the distribution status of the risk resistance performance indicators of each inter-class collaboration group; Performing a dispersion analysis on the distribution state of the anti-risk performance indicator to obtain the dispersion corresponding to each inter-class collaboration group; The inter-class balance classifier performs discrete constraints on the distribution state of the indicators within each inter-class cooperation group according to the preset discreteness.

4. The method according to claim 1, characterized in that Training a collaborative-terminal anti-risk assessment model, the method includes: Training a plurality of timing network layers, wherein the plurality of timing network layers include a continuous flow peak timing period, an intermittent flow peak timing period, and a flat flow timing period; Collect continuous flow collaborative monitoring samples, intermittent flow collaborative monitoring samples, and gentle flow collaborative monitoring samples; The multiple time-series network layers are trained according to the continuous traffic collaborative monitoring samples, the intermittent traffic collaborative monitoring samples and the smooth traffic collaborative monitoring samples to construct a collaborative-terminal anti-risk assessment model.

5. The method according to claim 1, characterized in that The method further includes: performing inter-class collaborative management on the distributed terminals of the power grid according to the N inter-class collaborative groups; Determine a first inter-class cooperation group among the N inter-class cooperation groups, wherein the first inter-class cooperation group includes K terminal nodes; Identify communication configuration parameters of the K terminal nodes, generate communication protocols corresponding to the K terminal nodes according to the communication configuration parameters, and perform inter-class collaborative management on the K terminal nodes according to the communication protocol.

6. A power grid distributed terminal collaborative management system, characterized in that: The system is used to execute the method according to any one of claims 1 to 5, and the system comprises: A terminal node acquisition module, wherein the terminal node acquisition module acquires a plurality of terminal nodes of a distributed terminal of a power grid; A data collection module, wherein the data collection module collects data from the plurality of terminal nodes to obtain a plurality of terminal monitoring data sets, wherein the plurality of terminal nodes correspond to the plurality of terminal monitoring data sets; an anti-risk assessment module, wherein the anti-risk assessment module obtains an isolated-terminal anti-risk assessment model, inputs the plurality of terminal monitoring data sets into the isolated-terminal anti-risk assessment model to perform anti-risk assessment, and outputs a plurality of anti-risk performance indicators; A balanced partitioning module, wherein the balanced partitioning module trains an inter-class balanced classifier, wherein the inter-class balanced classifier is embedded with a collaborative-terminal anti-risk assessment model, and the multiple anti-risk performance indicators are input into the collaborative-terminal anti-risk assessment model to perform balanced partitioning on the multiple terminal nodes, and a balanced partitioning result satisfying a balanced constraint condition is obtained, wherein the balanced partitioning result is N inter-class collaborative groups, and each inter-class collaborative group includes at least two terminal nodes; An inter-class collaborative management module is used to perform inter-class collaborative management on the power grid distributed terminals according to the N inter-class collaborative groups.

Images