AI-Prediction-Based Power Communication Network Scheduling Optimization Method and Device

Abstract

The present application provides an AI prediction-based power communication network scheduling optimization method and device, which relates to the technical field of smart grids and includes: performing node monitoring of the power communication network, using the real-time monitoring results as the execution data set, predicting the node status based on the node prediction model, and establishing a prediction data set; establishing a comprehensive evaluation network based on the node revenue function and the revenue ratio, using the execution data set and the prediction data set as input data, performing collaborative effect optimization with the comprehensive evaluation network, and establishing a node action and collaborative weight allocation scheme based on the optimization results to optimize the network scheduling. Through the present application, the technical problem in the existing power communication network scheduling method that the power communication network cannot be accurately regulated in a timely manner according to the operating conditions, resulting in poor working performance and efficiency of the communication network, can be solved, and the technical effects of improving the reliability and efficiency of the power communication network and ensuring the stability and high quality of the power communication service can be achieved.

Description

Technical Field

[0001] This application relates to the technical field of smart grids, and in particular, to a method and device for optimizing the scheduling of a power communication network based on AI prediction. Background Art

[0002] As the power system's dependence on the communication network continues to increase, the requirements for network performance and efficiency are also getting higher and higher. However, in the face of complex and changing operating conditions, due to the dynamicity and complexity of the network, traditional scheduling methods are difficult to respond to changes in network traffic in real time, resulting in problems such as uneven resource allocation, which limit the performance and efficiency of the network.

[0003] Currently, due to the complex and changing operating conditions of power communication, there is a technical problem in the existing power communication network scheduling methods that the power communication network cannot be accurately regulated in a timely manner according to the operating conditions, resulting in poor working performance and efficiency of the entire communication network. Summary of the Invention

[0004] The purpose of this application is to provide a method and device for optimizing the scheduling of a power communication network based on AI prediction, so as to solve the technical problem that in the existing power communication network scheduling methods, due to the complex and changing operating conditions of power communication, the power communication network cannot be accurately regulated in a timely manner according to the operating conditions, resulting in poor working performance and efficiency of the entire communication network.

[0005] In view of the above problems, this application provides a method and device for optimizing the scheduling of a power communication network based on AI prediction.

[0006] In a first aspect, the present application provides an AI prediction-based power communication network scheduling optimization method, which is implemented through an AI prediction-based power communication network scheduling optimization device, including: deploying a monitoring sensor group at the nodes of the power communication network, establishing a node dataset with the monitoring sensor group, and each node data in the node dataset includes network traffic data, node load data, latency data, fault data, and energy consumption data; synchronously acquiring external environment data, where the external environment data includes weather environment data and task data, establishing a node prediction model based on the external environment data and the node dataset, and the node prediction model corresponds to each node of the power communication network one by one; performing node monitoring of the power communication network, using the real-time monitoring results as the execution dataset and taking the real-time monitoring results as the reference data, predicting the node status based on the node prediction model, and establishing a prediction dataset; reading the connectivity data of the nodes, invoking the node collaboration relationship with the connectivity data, establishing a node collaboration weight, and configuring a node revenue function with the node collaboration weight; configuring the revenue ratio of the execution state and the prediction state, establishing a comprehensive evaluation network based on the node revenue function and the revenue ratio, using the execution dataset and the prediction dataset as input data, optimizing the collaboration effect of the power communication network with the comprehensive evaluation network, and establishing a node action and collaboration weight allocation scheme based on the optimization result; optimizing the power communication network scheduling through the node action and collaboration weight allocation scheme.

[0007] Second aspect, the present application also provides a power communication network scheduling optimization device based on AI prediction, which is used to execute a power communication network scheduling optimization method as described in the first aspect, including: a node dataset establishment module, which is used to deploy a monitoring sensor group at the nodes of the power communication network, and establish a node dataset with the monitoring sensor group. Each node data in the node dataset includes network traffic data, node load data, latency data, fault data, and energy consumption data; a node prediction model establishment module, which is used to synchronously obtain external environment data. The external environment data includes weather environment data and task data, and establish a node prediction model according to the external environment data and the node dataset. The node prediction models correspond one-to-one with the nodes of the power communication network; a node status prediction module, which is used to perform node monitoring of the power communication network, use the real-time monitoring results as the execution dataset, and use the real-time monitoring results as the reference data, and perform node status prediction based on the node prediction model to establish a prediction dataset; a node revenue function configuration module, which is used to read the connectivity data of the nodes, call the node collaboration relationship with the connectivity data, establish a node collaboration weight, and configure a node revenue function with the node collaboration weight; a collaboration effect optimization module, which is used to configure the revenue ratio of the execution status and the prediction status, establish a comprehensive evaluation network based on the node revenue function and the revenue ratio, use the execution dataset and the prediction dataset as input data, perform collaboration effect optimization of the power communication network with the comprehensive evaluation network, and establish a node action and collaboration weight allocation scheme based on the optimization result; a network scheduling optimization module, which is used to perform power communication network scheduling optimization through the node action and collaboration weight allocation scheme.

[0008] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0009] By using a monitoring sensor group to perform node monitoring of a power communication network, taking the real-time monitoring results as an execution data set, and using the real-time monitoring results as reference data, predicting the node status based on a node prediction model, and establishing a prediction data set; then reading the connectivity data of the nodes, calling the node cooperation relationship with the connectivity data, establishing a node cooperation weight, and configuring a node revenue function with the node cooperation weight; further configuring the revenue ratio of the execution status and the prediction status, establishing a comprehensive evaluation network based on the node revenue function and the revenue ratio, using the execution data set and the prediction data set as input data, optimizing the cooperation effect of the power communication network with the comprehensive evaluation network, and establishing a node action and cooperation weight allocation scheme based on the optimization result; finally, optimizing the power communication network scheduling through the node action and cooperation weight allocation scheme; it can improve the timeliness and accuracy of resource regulation in the power communication network, enable the power communication network to better adapt to the changing environment and communication requirements, achieve the technical effects of improving the reliability and efficiency of the entire power communication network, and ensuring the stability and high quality of power communication services.

[0010] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of this application, nor is it used to limit the scope of this application. Other features of this application will become easily understandable through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in this application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary, and for those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0012] Figure 1 It is a schematic flowchart of a power communication network scheduling optimization method based on AI prediction of this application;

[0013] Figure 2 It is a schematic flowchart of establishing a node action and cooperation weight allocation scheme based on the optimization result in a power communication network scheduling optimization method based on AI prediction of this application;

[0014] Figure 3 It is a schematic structural diagram of a power communication network scheduling optimization device based on AI prediction of this application.

[0015] Description of the reference numerals:

[0016] Node dataset establishment module 11, node prediction model establishment module 12, node status prediction module 13, node revenue function configuration module 14, collaborative effect optimization module 15, network scheduling optimization module 16. Specific implementation manners

[0017] By providing a power communication network scheduling optimization method and device based on AI prediction, this application solves the technical problem that in the existing power communication network scheduling methods, due to the complex and changeable operating conditions of power communication, it is impossible to accurately regulate the resources of the power communication network in a timely manner according to the operating conditions, resulting in poor working performance and efficiency of the entire communication network. It can improve the timeliness and accuracy of resource regulation in the power communication network, enable the power communication network to better adapt to the changing environment and communication requirements, achieve the technical effects of improving the reliability and efficiency of the entire power communication network, and ensuring the stability and high quality of power communication services.

[0018] Next, the technical solutions in this application will be described clearly and completely with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited by the example embodiments described here. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application. Additionally, it should be noted that for the sake of description, only the parts related to this application are shown in the drawings rather than all of them.

[0019] Embodiment 1. Please refer to the attached Figure 1 , this application provides a power communication network scheduling optimization method based on AI prediction, which is applied to a power communication network scheduling optimization device based on AI prediction, and specifically includes the following steps:

[0020] Step 1: Deploy a monitoring sensor group at the nodes of the power communication network, and establish a node dataset with the monitoring sensor group. Each node data in the node dataset includes network traffic data, node load data, delay data, fault data, and energy consumption data.

[0021] Specifically, multiple communication nodes of the power communication network are obtained, and a monitoring sensor group is deployed at the multiple communication nodes. These sensors can collect key operation data, and the sensors can be configured according to the expected monitored data type. Then, based on the multiple sensor groups, operation data collection is performed on the multiple communication nodes to obtain a node data set. Each node data in the node data set includes network traffic data, node load data, latency data, fault data, and energy consumption data. Among them, network traffic data refers to the network traffic situation of each node, including the number and rate of incoming and outgoing data packets. This data is crucial for analyzing the network usage and predicting future network demands; node load data involves the usage of the processing capacity of each node, including CPU utilization rate, memory usage rate, etc. By monitoring the node load, potential overload problems can be detected in a timely manner, and network resource allocation can be adjusted accordingly; latency data reflects the time taken for data to be transmitted in the network. Monitoring latency helps identify bottlenecks in the network, thereby optimizing the data transmission path and reducing communication latency; fault data records the fault situations that occur at the nodes, including fault types, occurrence times, durations, etc. This data is of great significance for diagnosing the network health status and preventing future faults; energy consumption data monitors the energy consumption situation of each node, which helps optimize energy usage and reduce operating costs. By monitoring and obtaining the node data set, it provides data support for subsequent optimization analysis of power communication network scheduling.

[0022] Step 2: Synchronously obtain external environment data. The external environment data includes weather environment data and task data. A node prediction model is established based on the external environment data and the node data set. The node prediction model corresponds one-to-one with the nodes of the power communication network.

[0023] Specifically, at the same time node, an environmental monitoring sensor group is used to synchronously obtain external environment data. The external environment data includes weather environment data and task data. Among them, the weather environment data includes temperature, humidity, wind speed, rainfall, solar radiation, etc. These factors may affect the operation of the power communication network, especially for outdoor equipment. For example, high temperature or extreme weather may cause a decline in equipment performance or an increase in the failure rate; task data includes power transactions, remote monitoring, emergency responses, etc. Changes in task data will affect network traffic and resource requirements.

[0024] Next, multiple node prediction models are constructed based on multiple power communication network nodes. The node prediction models correspond one-to-one with the power communication network nodes. The node prediction models are used to predict the node operation data in the future time period. An appropriate prediction model can be selected according to the actual situation, such as a BP neural network model, a time series analysis model, a random forest, etc. Among them, the input of the node prediction model is the external environment data and node data under the current node, and the output is the node prediction data in the future period of time. The construction method of the node prediction model is as follows. First, a first network node is randomly selected from multiple power communication network nodes, and then a sample training data set is obtained based on the first network node through retrieval. Then, the node prediction model constructed based on the BP neural network is supervised and trained using the sample training data set. The backpropagation algorithm and the loss function can be used for supervised training to obtain the first node prediction model that meets the convergence conditions, and multiple node prediction models are constructed using the same method. By using a neural network to construct a node prediction model, the accuracy and efficiency of node data prediction can be improved.

[0025] Step 3: Perform node monitoring of the power communication network, use the real-time monitoring results as the execution data set, and use the real-time monitoring results as the reference data to predict the node status based on the node prediction model, and establish a prediction data set.

[0026] Specifically, use the monitoring sensor group to perform node monitoring of the power communication network to obtain real-time monitoring results, including network traffic, node load, delay, faults, and energy consumption, etc., and use the real-time monitoring results as the execution data set; on the other hand, monitor and obtain real-time external environment data, including weather environment data and task data. Then, input the real-time monitoring results and real-time external environment data into the node prediction model to predict the node status and obtain a prediction data set, where the prediction data includes the predicted status of the node in the future period of time. By obtaining the prediction data set, it provides data support for subsequent node collaborative scheduling optimization analysis of the power communication network.

[0027] Step 4: Read the connectivity data of the nodes, call the node collaboration relationship with the connectivity data, establish the node collaboration weight, and configure the node revenue function with the node collaboration weight.

[0028] Specifically, read the connectivity data of the nodes, such as reading the connectivity data of the nodes through a network management system (NMS) or directly from network devices. These data usually include the connection status between nodes, link capacity, communication delay, link quality, etc. Then call the node collaboration relationship with the connectivity data, that is, based on the connectivity data, analyze the collaboration relationship between nodes, determine which nodes are critical in data transmission, and the interdependence relationship; further establish node collaboration weights according to the node collaboration relationship, that is, assign a weight value to the collaboration relationship between each pair of nodes according to the node collaboration relationship. The weight value can be determined based on factors such as the bandwidth, delay, and reliability of the link. Expert systems, statistical analysis, or machine learning methods can be used to calculate the weights. The weight value reflects the degree of interdependence between nodes, that is, the tightness of the collaboration relationship. Then configure the node revenue function according to the node collaboration weights, where the revenue function can quantify the network value and contribution of each node and reflect its importance in the network. In this way, network administrators can better understand the performance and importance of each node in the network and optimize the allocation and scheduling of network resources accordingly. For example, in the case of limited resources, resources can be preferentially allocated according to the node revenue, or when a fault occurs, the nodes that have the greatest impact on the overall network performance can be restored first. This data- and model-driven decision-making method helps to improve the intelligence level and adaptive ability of the power communication network.

[0029] Step Five: Configure the revenue ratio of the execution state and the prediction state, establish a comprehensive evaluation network based on the node revenue function and the revenue ratio, use the execution data set and the prediction data set as input data, optimize the collaboration effect of the power communication network with the comprehensive evaluation network, and establish a node action and collaboration weight allocation scheme based on the optimization result.

[0030] Specifically, first, it is necessary to determine the relative importance of the execution state (i.e., the current actual state) and the prediction state (i.e., the future predicted state) in the comprehensive evaluation, and configure the benefit ratios of the execution state and the prediction state according to the relative importance, where the benefit ratio is proportional to the relative importance. Then, based on the node benefit function and the benefit ratio, a comprehensive evaluation network is established. The comprehensive evaluation network can be a multi-objective optimization model that combines the benefits of the execution state and the prediction state and is used to evaluate the overall performance of the power communication network. Then, the execution data set and the prediction data set are used as input data to input into the comprehensive evaluation network, and the collaborative effect optimization of the power communication network is carried out through the comprehensive evaluation network. The comprehensive evaluation network contains an objective optimization function to obtain the optimization result. Then, based on the optimization result, a node action and collaborative weight allocation scheme is established. Among them, the node action defines the specific operations that each node should take, such as increasing / decreasing bandwidth, adjusting routing strategies, performing maintenance tasks, etc.; the collaborative weight allocation adjusts the collaborative relationship between nodes according to the output of the comprehensive evaluation network to optimize the overall performance of the network.

[0031] Step Six: Optimize the power communication network scheduling through the node action and collaborative weight allocation scheme.

[0032] Specifically, finally, optimize the power communication network scheduling through the node action and collaborative weight allocation scheme, such as deploying the scheme, configuring network devices, and adjusting the configuration of network devices according to the scheme, such as routers, switches, and other communication devices, to ensure the stability of the network and the continuity of services; execute specific node actions according to the optimization scheme, such as adjusting bandwidth, optimizing routing paths, performing maintenance tasks, etc.; adjust the collaborative relationship between nodes according to the collaborative weight allocation scheme, implement dynamic weight adjustment, and dynamically adjust the collaborative weight according to the real-time changes of network traffic to optimize the overall performance of the network. Through the above steps, the power communication network can achieve efficient management of resources and ensure that the network can maintain the best performance when facing different operating conditions. This data- and model-driven decision-making method helps to improve the intelligence level and adaptive ability of the network, thereby improving the performance and efficiency of the entire power communication network.

[0033] The described power communication network scheduling optimization method based on AI prediction is applied to a power communication network scheduling optimization device based on AI prediction, which can solve the technical problem that due to the complex and changeable operation conditions of power communication, the existing power communication network scheduling methods cannot accurately regulate the resources of the power communication network in a timely manner according to the operation conditions, resulting in poor working performance and efficiency of the entire communication network. By using the monitoring sensor group to perform node monitoring of the power communication network, taking the real-time monitoring results as the execution data set, and using the real-time monitoring results as the reference data, predicting the node status based on the node prediction model, and establishing a prediction data set; then reading the connectivity data of the nodes, calling the node collaboration relationship with the connectivity data, establishing the node collaboration weight, and configuring the node revenue function with the node collaboration weight; further configuring the revenue ratio of the execution state and the prediction state, establishing a comprehensive evaluation network based on the node revenue function and the revenue ratio, using the execution data set and the prediction data set as input data, optimizing the collaboration effect of the power communication network with the comprehensive evaluation network, and establishing a node action and collaboration weight allocation scheme based on the optimization results; finally, optimizing the power communication network scheduling through the node action and collaboration weight allocation scheme; it can improve the timeliness and accuracy of resource regulation in the power communication network, enable the power communication network to better adapt to the changing environment and communication requirements, achieve the technical effects of improving the reliability and efficiency of the entire power communication network, and ensuring the stability and high quality of power communication services.

[0034] Further, reading the connectivity data of the nodes, calling the node collaboration relationship with the connectivity data, establishing the node collaboration weight, and configuring the node revenue function with the node collaboration weight, step four of this application includes:

[0035] Establish the node collaboration weight, and the formula is as follows: ; where represents the collaboration weight of any two nodes and , represents the communication capacity of any two nodes and , represents the communication delay of any two nodes and , represents the communication quality parameter of any two nodes and , is a comprehensive function; configure the node revenue function as follows: ; where represents the node revenue function of the th node, represents the action of node , represents the processing node The action sets of other nodes except Characterize the node and The benefit function generated by the collaboration between them.

[0036] Specifically, first, obtain the calculation formula for the node collaboration weight. In the node collaboration weight expression, Characterize any two nodes and The collaboration weight of Characterize any two nodes and The communication capacity of Characterize any two nodes and The communication delay of Characterize any two nodes and The communication quality parameter of Is a comprehensive function used to calculate the collaboration weight; the purpose of this function is to combine factors such as communication capacity, communication delay, and communication quality to form a weight value that can reflect the collaboration relationship between nodes. Through this collaboration weight expression that comprehensively considers multiple factors, the power communication network can more accurately evaluate the collaboration relationship between nodes, thereby achieving more effective scheduling optimization and resource management.

[0037] Obtain the calculation formula for the node benefit function. In the node benefit function, Characterize the th node benefit function Characterize the action of the node Characterize the action sets of other nodes except the processing node Characterize the node and The benefit function generated by the collaboration between them. The purpose of this function is to combine the actions of the node, the actions of other nodes, and the collaboration relationship between nodes to form a benefit value that can reflect the value of the node. Through this node benefit function that comprehensively considers multiple factors, the power communication network can more accurately evaluate the value and contribution of the node, thereby achieving more effective scheduling optimization and resource management. This method helps to improve the performance and efficiency of the network and ensure the stability and high quality of power communication services.

[0038] Furthermore, configure the benefit ratio of the execution state and the prediction state. Step five of this application includes:

[0039] ​​Perform reliability analysis on the execution status based on the execution dataset to establish a reliability evaluation result; perform uncertainty analysis on the prediction status based on the prediction dataset to establish an uncertainty evaluation result; obtain the stability requirements of the power communication network, and configure the benefit ratio according to the stability requirements, reliability evaluation result, and uncertainty evaluation result.

[0040] Specifically, use the execution dataset to perform reliability analysis on the current operating status of the power communication network, including indicators such as the network failure rate, fault recovery time, and the impact of faults on network performance. Based on the analysis results, establish a reliability evaluation result to reflect the current stability and reliability level of the network; on the other hand, use the prediction dataset to perform uncertainty analysis on the future operating status of the power communication network, including fluctuations in network traffic, the probability of faults, prediction errors of network performance, etc. According to the analysis results, establish an uncertainty evaluation result to reflect the uncertainty and potential risks of the network's future state.

[0041] Then obtain the stability requirements of the power communication network, that is, determine the stability requirements of the network according to the operation requirements and safety standards of the power communication network, including the allowable incidence rate of faults, the minimum standard of network performance, etc.; finally, configure the benefit ratio of the execution status and prediction status according to the stability requirements, reliability evaluation result, and uncertainty evaluation result. The configured benefit ratio reflects the degree of emphasis of the network on current performance and future predictions. For example, if the network has higher requirements for the stability of current performance, the benefit ratio of the execution status may be higher; if the network has greater concerns about future uncertainties, the benefit ratio of the prediction status may be higher. Through this process, the power communication network can achieve a comprehensive evaluation of the execution status and prediction status, thereby more reasonably configuring the benefit ratio and guiding the scheduling optimization decision of the network. This data-driven method helps to improve the intelligent level of network management and ensure that the network can maintain stable and efficient operation in the face of complex and changing environments.

[0042] Furthermore, establish a comprehensive evaluation network based on the node benefit function and benefit ratio. Step five of this application includes:

[0043] Configure the calculation layer of the comprehensive evaluation network according to the node benefit function as follows: ; where represents the optimization objective function, represents the weight matrix, is the action set, is the total number of nodes in the power communication network, is the adjustment coefficient, which is used to balance the network performance and the cost of weight adjustment in the objective function, is the edge set between nodes, used to represent node pairs that may have a collaborative relationship in the power communication network. is the cost function for weight adjustment, representing the cost required to adjust the collaborative weight during the optimization process; a proportional collaboration layer is established based on the benefit ratio, and an integrated evaluation network is established based on the proportional collaboration layer and the calculation layer.

[0044] Specifically, first, configure the calculation layer of the integrated evaluation network according to the node benefit function, where the calculation layer is embedded with an optimization function. In the optimization function expression, represents the optimization objective function, which is the core of the power communication network scheduling optimization and reflects the optimization objective of network performance and efficiency; represents the weight matrix, is the action set, is the total number of nodes in the power communication network, is the adjustment coefficient, used to balance the network performance and the cost of weight adjustment in the objective function, is the edge set between nodes, used to represent node pairs that may have a collaborative relationship in the power communication network. is the cost function for weight adjustment, representing the cost required to adjust the collaborative weight during the optimization process. By constructing the optimization objective function, the power communication network scheduling optimization can more accurately evaluate and optimize the network performance and efficiency, providing support for optimizing the collaborative effect of the power communication network.

[0045] A proportional collaboration layer is established based on the benefit ratio. The proportional collaboration layer reflects the degree of emphasis of the network on the current performance and future prediction. The elements of the proportional collaboration layer represent the collaborative weights of two nodes and can be adjusted according to the benefit ratio; finally, an integrated evaluation network is established based on the proportional collaboration layer and the calculation layer.

[0046] Furthermore, based on the optimization result, establish a node action and collaborative weight allocation scheme. As shown in Figure 2 Step five of this application includes:

[0047] Obtain the node historical data set of the power communication network, and establish a state evaluation model for the nodes based on the node historical data set; perform state analysis on each node based on the prediction data set and the state evaluation model to generate the first state analysis result; perform execution state prediction on each node based on the optimization result and the state evaluation model to establish the second state analysis result; generate an optimization feedback according to the first state analysis result and the second state analysis result, and perform optimization result compensation with the optimization feedback.

[0048] Specifically, collect the operation historical data of each node in the power communication network to obtain a node historical data set. The historical data includes network traffic, node load, latency, faults, energy consumption, etc.; then establish a state evaluation model for the nodes based on the node historical data set. The state evaluation model can quantify the value and contribution of the nodes and reflect their importance in the network. First, select a suitable model according to the goal of the evaluation model, such as linear regression, support vector machine, random forest, etc., considering the prediction ability and computational complexity of the model, as well as the interpretability of the model. Then use the historical data to perform supervised training on the selected model to learn the patterns and rules in the data and obtain the state evaluation model.

[0049] Next, input the prediction data set into the state evaluation model to perform state analysis on each node, generating a first state analysis result for reflecting the predicted state of the nodes; on the other hand, input the optimization result into the state evaluation model to perform state analysis on each node, outputting a second state analysis result, that is, obtaining the predicted state of the optimization result; then perform deviation analysis based on the first state analysis result and the second state analysis result, and generate an optimization feedback according to the deviation analysis result. The optimization feedback should include adjustment suggestions for the existing scheduling strategy to reduce the deviation; finally, compensate the optimization result with the optimization feedback. For example, according to the optimization feedback, adjust the node actions, resource allocation, scheduling strategy, etc. in the optimization result. By executing the optimization feedback to compensate the optimization result, the accuracy of the optimization result can be improved, the efficient management of resources can be achieved, and it can be ensured that the network can maintain the best performance when facing different operating conditions.

[0050] Furthermore, according to the first state analysis result and the second state analysis result, generate an optimization feedback. The present application further includes the following steps:

[0051] Perform difference analysis on the same nodes of the first state analysis result and the second state analysis result to establish a difference penalty; perform state threshold discrimination on the second state analysis result, and generate a state threshold trigger penalty based on the state threshold discrimination result; generate the optimization feedback according to the difference penalty and the state threshold trigger penalty.

[0052] Specifically, first, compare the data of the same nodes in the first state analysis result and the second state analysis result, identify and quantify the differences between the two analysis results to obtain a difference analysis result; then, based on the result of the difference analysis, establish a difference penalty mechanism, including adjusting weights, increasing adjustment costs, restricting certain operations, etc. The purpose of the difference penalty is to encourage network managers to reduce the difference between prediction and reality during the adjustment process. On the other hand, perform a state threshold discrimination on the second state analysis result to determine the key thresholds of network performance, such as the maximum allowable delay, the minimum available bandwidth, etc., and then generate a state threshold trigger penalty based on the state threshold discrimination result; finally, generate the optimization feedback according to the difference penalty and the state threshold trigger penalty, where the optimization feedback includes adjustment suggestions for the existing scheduling strategy to reduce differences and avoid state threshold triggers.

[0053] Furthermore, this application also includes the following steps:

[0054] Establish a temporary backtracking window, perform collaborative backtracking data collection based on the temporary backtracking window to generate node collaborative taboos; use the node collaborative taboos to constrain the node collaborative relationship and reconstruct the node collaborative weights.

[0055] Specifically, establish a temporary backtracking window, that is, define a time window for backtracking analysis. The temporary backtracking window can be a fixed time period or can be dynamically adjusted as needed; then perform collaborative backtracking data collection based on the temporary backtracking window to obtain a collaborative backtracking data set, that is, within the temporary backtracking window, collect the collaborative relationship data between nodes in the network, including the communication traffic between nodes, the packet exchange situation, the task collaboration record, etc.; then generate node collaborative taboos according to the collaborative backtracking data set, that is, analyze the collaborative backtracking data, identify node pairs or collaborative relationships that may have problems, and generate node collaborative taboos. The node collaborative taboos represent prohibiting or restricting certain collaborative operations between nodes within a certain period of time. Further, introduce node collaborative taboos into the node collaborative relationship as a constraint condition to ensure that violations of node collaborative taboos are avoided during collaborative operations; finally, adjust the original node collaborative weights according to the node collaborative taboos. The reconstructed node collaborative weights reflect the restrictions and optimizations of the collaborative relationship. Through this process, the power communication network can better manage the collaborative relationship between nodes, avoid potential problems, and optimize the use of network resources.

[0056] In summary, the power communication network scheduling optimization method based on AI prediction provided by this application has the following technical effects:

[0057] By using a monitoring sensor group to perform node monitoring of a power communication network, taking the real-time monitoring results as an execution data set, and using the real-time monitoring results as reference data, predicting the node status based on a node prediction model, and establishing a prediction data set; then reading the connectivity data of the nodes, calling the node collaboration relationship with the connectivity data, establishing a node collaboration weight, and configuring a node revenue function with the node collaboration weight; further configuring the revenue ratio of the execution state and the prediction state, establishing a comprehensive evaluation network based on the node revenue function and the revenue ratio, using the execution data set and the prediction data set as input data, optimizing the collaboration effect of the power communication network with the comprehensive evaluation network, and establishing a node action and collaboration weight allocation scheme based on the optimization results; finally, optimizing the power communication network scheduling through the node action and collaboration weight allocation scheme; it can improve the timeliness and accuracy of resource regulation in the power communication network, enable the power communication network to better adapt to the changing environment and communication requirements, achieve the technical effects of improving the reliability and efficiency of the entire power communication network, and ensuring the stability and high quality of power communication services.

[0058] Embodiment 2. Based on the same inventive concept as the method for optimizing the scheduling of a power communication network based on AI prediction in the foregoing embodiment, the present application also provides an apparatus for optimizing the scheduling of a power communication network based on AI prediction. Please refer to the attached Figure 3 , including:

[0059] The node dataset establishment module 11 is used to deploy a monitoring sensor group at the nodes of the power communication network, establish a node dataset with the monitoring sensor group, and each node data in the node dataset includes network traffic data, node load data, delay data, fault data, and energy consumption data; the node prediction model establishment module 12 is used to synchronously obtain external environment data, where the external environment data includes weather environment data and task data, and establish a node prediction model according to the external environment data and the node dataset, and the node prediction model corresponds to each node of the power communication network one by one; the node status prediction module 13 is used to perform node monitoring on the power communication network, use the real-time monitoring result as the execution dataset, and use the real-time monitoring result as the reference data, and perform node status prediction based on the node prediction model to establish a prediction dataset; the node revenue function configuration module 14 is used to read the connectivity data of the nodes, call the node cooperation relationship with the connectivity data, establish a node cooperation weight, and configure the node revenue function with the node cooperation weight; the cooperation effect optimization module 15 is used to configure the revenue ratio of the execution state and the prediction state, establish a comprehensive evaluation network based on the node revenue function and the revenue ratio, use the execution dataset and the prediction dataset as input data, perform cooperation effect optimization on the power communication network with the comprehensive evaluation network, and establish a node action and cooperation weight allocation scheme based on the optimization result; the network scheduling optimization module 16 is used to perform power communication network scheduling optimization through the node action and cooperation weight allocation scheme.

[0060] Further, the power communication network scheduling optimization device based on AI prediction is also used for:

[0061] Establish a node cooperation weight, and the formula is as follows: ; where represents the cooperation weight of any two nodes and , represents the communication capacity of any two nodes and , represents the communication delay of any two nodes and , represents the communication quality parameter of any two nodes and , is a comprehensive function; configure the node revenue function, as follows: ; where represents the node revenue function of the th node, represents the action of node , represents the processing node The action sets of other nodes characteristic node and the benefit function jointly generated therebetween

[0062] Furthermore, the AI prediction-based power communication network scheduling optimization device is further configured to:

[0063] Configure the computing layer of the comprehensive evaluation network according to the node revenue function as follows: ; where represents the optimization objective function represents the weight matrix is the action set is the total number of nodes in the power communication network is the adjustment coefficient, which is used to balance the network performance and the cost of weight adjustment in the objective function is the edge set between nodes, which is used to represent the node pairs that may have a collaborative relationship in the power communication network is the cost function of weight adjustment, which represents the cost required to adjust the collaborative weight during the optimization process; establish a proportional collaboration layer based on the revenue ratio, and establish a comprehensive evaluation network based on the proportional collaboration layer and the computing layer

[0064] Furthermore, the AI prediction-based power communication network scheduling optimization device is further configured to:

[0065] Obtain the node historical data set of the power communication network, and establish a state evaluation model for the nodes with the node historical data set; perform state analysis on each node based on the prediction data set and the state evaluation model to generate a first state analysis result; perform execution state prediction on each node based on the optimization result and the state evaluation model to establish a second state analysis result; generate an optimization feedback according to the first state analysis result and the second state analysis result, and perform optimization result compensation with the optimization feedback

[0066] Furthermore, the AI prediction-based power communication network scheduling optimization device is further configured to:

[0067] Perform differential analysis on the same nodes of the first state analysis result and the second state analysis result to establish a differential penalty; perform state threshold discrimination on the second state analysis result, and generate a state threshold trigger penalty based on the state threshold discrimination result; generate the optimization feedback according to the differential penalty and the state threshold trigger penalty

[0068] Furthermore, the AI prediction-based power communication network scheduling optimization device is further configured to:

[0069] Perform reliability analysis on the execution status based on the execution dataset to establish a reliability evaluation result; perform uncertainty analysis on the prediction status based on the prediction dataset to establish an uncertainty evaluation result; obtain the stability requirements of the power communication network, and configure the revenue ratio according to the stability requirements, reliability evaluation result, and uncertainty evaluation result.

[0070] Further, the power communication network scheduling optimization device based on AI prediction is further configured to:

[0071] Establish a temporary backtracking window, perform collaborative backtracking data collection based on the temporary backtracking window to generate node collaborative taboos; use the node collaborative taboos to constrain the node collaborative relationship and reconstruct the node collaborative weights.

[0072] In this specification, each embodiment is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The method and specific examples of a power communication network scheduling optimization method based on AI prediction in the foregoing Embodiment 1 are equally applicable to the power communication network scheduling optimization device based on AI prediction in this embodiment. Through the foregoing detailed description of a power communication network scheduling optimization method based on AI prediction, those skilled in the art can clearly know the power communication network scheduling optimization device based on AI prediction in this embodiment. Therefore, for the sake of brevity of the specification, it will not be elaborated here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0073] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and 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 equivalent technologies, the present application is also intended to include these changes and variations.

Claims

1. A power communication network dispatch optimization method based on AI prediction, characterized in that: include: Deploy a monitoring sensor group at a node of the electric power communication network, and establish a node data set with the monitoring sensor group, wherein each node data in the node data set includes network traffic data, node load data, delay data, fault data and energy consumption data; Synchronously acquiring external environment data, the external environment data including weather environment data and task data, and establishing a node prediction model according to the external environment data and the node data set, wherein the node prediction model corresponds one-to-one to the power communication network node; Perform node monitoring of the power communication network, use the real-time monitoring results as the execution data set, and use the real-time monitoring results as the benchmark data to perform node state prediction based on the node prediction model to establish a prediction data set; Reading the connectivity data of the node, calling the node coordination relationship with the connectivity data, establishing the node coordination weight, and configuring the node benefit function with the node coordination weight; Configure the profit ratio of the execution state and the prediction state, establish a comprehensive evaluation network based on the node profit function and the profit ratio, use the execution data set and the prediction data set as input data, use the comprehensive evaluation network to optimize the synergy effect of the power communication network, and establish a node action and synergy weight allocation plan based on the optimization result; Optimizing the dispatch of the electric power communication network through the node actions and the collaborative weight allocation scheme; The reading of the connectivity data of the node, calling the node coordination relationship with the connectivity data, establishing the node coordination weight, and configuring the node benefit function with the node coordination weight also includes: Establish node collaboration weights, the formula is as follows: ； in, Representing any two nodes and The collaborative weight of Representing any two nodes and communication capacity, Representing any two nodes and The communication delay, Representing any two nodes and Communication quality parameters, is a comprehensive function; Configure the node profit function as follows: ； in, Characterization The node profit function of nodes is Characterization Node Actions, Characterization Processing Node The action set of other nodes, Characterization Node and The benefit function generated by the synergy between them; The establishing of a comprehensive evaluation network based on the node benefit function and benefit ratio also includes: The calculation layer of the comprehensive evaluation network is configured according to the node benefit function as follows: ； in, Characterize the optimization objective function, Representation weight matrix, is a collection of actions, is the total number of nodes in the power communication network, is the adjustment coefficient used to balance the network performance and weight adjustment cost in the objective function. is the edge set between nodes, which is used to characterize the node pairs that may have a collaborative relationship in the power communication network. is the cost function for weight adjustment, which represents the cost required to adjust the collaborative weights during the optimization process; A proportional coordination layer is established based on the income ratio, and a comprehensive evaluation network is established based on the proportional coordination layer and the calculation layer.

2. The power communication network scheduling optimization method based on AI prediction according to claim 1 is characterized in that: The node action and collaborative weight distribution scheme is established based on the optimization result, and further includes: Acquire a node historical data set of the power communication network, and establish a node status evaluation model based on the node historical data set; Performing a state analysis of each node based on the prediction data set and the state evaluation model to generate a first state analysis result; Predicting the execution state of each node based on the optimization result and the state evaluation model, and establishing a second state analysis result; Optimization feedback is generated according to the first state analysis result and the second state analysis result, and optimization result compensation is performed using the optimization feedback.

3. The power communication network scheduling optimization method based on AI prediction according to claim 2 is characterized in that: The generating optimization feedback according to the first state analysis result and the second state analysis result also includes: Performing a same-node difference analysis on the first state analysis result and the second state analysis result, and establishing a difference penalty; Performing state threshold determination on the second state analysis result, and generating a state threshold trigger penalty based on the state threshold determination result; The optimization feedback is generated according to the difference penalty and the state threshold trigger penalty.

4. The power communication network scheduling optimization method based on AI prediction according to claim 1 is characterized in that: The profit ratio of the configuration execution state and the prediction state also includes: Performing reliability analysis on the execution status based on the execution data set to establish a reliability evaluation result; Performing uncertainty analysis on the predicted state based on the predicted data set to establish an uncertainty evaluation result; The stability requirement of the electric power communication network is obtained, and the profit ratio is configured according to the stability requirement, the reliability evaluation result, and the uncertainty evaluation result.

5. The power communication network scheduling optimization method based on AI prediction according to claim 1, characterized in that: Also includes: Establishing a temporary backtracking window, performing collaborative backtracking data collection based on the temporary backtracking window, and generating node collaborative taboos; The node collaboration taboos are used to constrain the node collaboration relationship and reconstruct the node collaboration weights.

6. Power communication network dispatch optimization device based on AI prediction, characterized in that: The steps for implementing the power communication network dispatch optimization method based on AI prediction as described in any one of claims 1 to 5 include: A node data set establishment module, used to deploy a monitoring sensor group at a node of the electric power communication network, and establish a node data set with the monitoring sensor group, wherein each node data in the node data set includes network flow data, node load data, delay data, fault data and energy consumption data; A node prediction model establishment module is used to synchronously acquire external environment data, the external environment data includes weather environment data and task data, and establish a node prediction model according to the external environment data and the node data set, wherein the node prediction model corresponds to the power communication network node one by one; A node state prediction module is used to perform node monitoring of the power communication network, use the real-time monitoring results as the execution data set, and use the real-time monitoring results as the benchmark data to perform node state prediction based on the node prediction model and establish a prediction data set; A node benefit function configuration module is used to read the connectivity data of the node, call the node coordination relationship with the connectivity data, establish the node coordination weight, and configure the node benefit function with the node coordination weight; A synergy effect optimization module is used to configure the profit ratio of the execution state and the prediction state, establish a comprehensive evaluation network based on the node profit function and the profit ratio, use the execution data set and the prediction data set as input data, use the comprehensive evaluation network to optimize the synergy effect of the power communication network, and establish a node action and synergy weight allocation plan based on the optimization result; The network scheduling optimization module is used to optimize the scheduling of the power communication network through the node action and collaborative weight allocation scheme.

Images