A dispatching operation and maintenance method, device and equipment of a power communication network and a storage medium

Abstract

The application discloses a kind of dispatching operation and maintenance methods, device and equipment of electric power communication network and storage medium, applied to dispatching operation and maintenance technical field, method includes obtaining the equipment state and service priority of several equipment in electric power communication network;Based on equipment state and service priority, obtain equipment operation data;Equipment operation data is handled, and load prediction result is obtained;And topological analysis is carried out to electric power communication network, and topological analysis result is obtained;Load prediction result and topological analysis result are analyzed, and power grid dispatching scheme is obtained;The equipment extracted is handled, and portrait construction is carried out, and equipment portrait is obtained;Equipment portrait is handled, generates operation and maintenance task list, and power grid dispatching scheme and operation and maintenance task list are handled, and dispatching operation and maintenance scheme is obtained.The method provided in the application solves the technical problems that traditional electric power communication network relies on manual experience, passive alarm response, and cannot predict equipment performance degradation and link hidden danger in advance.

Description

Technical Field

[0001] This invention relates to the field of dispatching and maintenance technology, and in particular to a dispatching and maintenance method, apparatus, equipment and storage medium for power communication networks. Background Technology

[0002] The power communication network carries production control services such as relay protection, stability control, and dispatch automation. Efficient dispatch and maintenance can ensure the real-time availability and reliability of communication links, making it the first line of defense for power grid security.

[0003] In existing technologies, the power communication network dispatch and maintenance system has long relied on a passive response mechanism. Traditional maintenance starts with alarm triggering, which is a passive response mode. It lacks the ability to predict the trend of equipment performance degradation and link hidden risks in advance, and cannot generate pre-control strategies before the fault occurs, thus directly threatening the safety of power grid operation. Summary of the Invention

[0004] This invention provides a scheduling and operation maintenance method, device, equipment, and storage medium for power communication networks. It solves the technical problem that traditional power communication networks rely on manual experience and passive alarm response, making it impossible to predict equipment performance degradation and link risks in advance. It realizes proactive predictive intelligent operation and maintenance of power communication networks, and can identify potential equipment failures in advance.

[0005] To address the aforementioned technical problems, this invention provides a method for scheduling and maintaining a power communication network, the method comprising: Obtain the device status and service priority of several devices in the target power communication network; Based on the device status and service priority of the aforementioned devices, and based on a determined data collection frequency, the device operation data of the aforementioned devices is obtained; The equipment operation data is processed for load forecasting to obtain load forecasting results; the target power communication network is subjected to topology analysis to obtain topology analysis results; the load forecasting results and the topology analysis results are fused and analyzed to obtain a power grid dispatching scheme. For each extracted device, the CPU utilization, memory usage, and link error rate are used to construct a vector, and the resulting multi-dimensional feature vector sequence is used to construct a profile to obtain a device profile; anomaly detection is performed on the device profile to identify potential faults; A maintenance task list is generated based on the potential faults, and the power grid dispatching scheme and the maintenance task list are time-series aligned to obtain the dispatching maintenance scheme. Based on the aforementioned scheduling and maintenance plan, standardized operation and maintenance operations are performed on several devices in the target power communication network, and execution feedback results are generated.

[0006] As one preferred embodiment, the process of performing load forecasting processing on the equipment operating data to obtain load forecasting results includes: Acquire historical load time series data from the equipment operation data; perform missing value imputation and outlier smoothing on the historical load time series data to obtain a preprocessed load sequence; Based on the preprocessed load sequence, a fixed-length continuous subsequence is extracted as sample features using the sliding window method to construct a training sample set; The pre-constructed gated recurrent unit neural network is trained using the training sample set to obtain a load prediction model; The preprocessed load sequence prior to the current time is input into the load prediction model with the same sliding window length to obtain the load prediction result.

[0007] As one preferred embodiment, the topology analysis of the target power communication network to obtain the topology analysis results includes: All devices in the target power communication network are abstracted as nodes in graph theory, and the physical links between all devices are abstracted as edges in graph theory to construct an undirected network topology graph. Determine the network connectivity of the undirected network topology graph and the degree centrality of each node in the undirected network topology graph; The network connectivity is compared and analyzed with a preset connectivity threshold to generate topology connectivity alarm information; Nodes whose degree centrality exceeds a preset centrality threshold are marked as critical path nodes, and topological critical path information is generated. The topology connectivity alarm information and the topology critical path information are used together as the topology analysis result.

[0008] As one preferred embodiment, the step of fusing the load forecast results and the topology analysis results to obtain a power grid dispatching scheme includes: Determine the first information entropy of the load prediction result and the second information entropy of the topology analysis result, respectively; Based on the first information entropy of the load forecast result and the second information entropy of the topology analysis result, the fusion analysis result is obtained; Obtain the power grid dispatching procedures, parse the condition judgment rules and action execution rules in the power grid dispatching procedures, and build a knowledge base; Using the fusion analysis results as input, the condition judgment rules in the knowledge base are traversed to match the current power grid state node; the corresponding action execution rules are extracted from the current power grid state node to generate the power grid dispatching scheme.

[0009] As a preferred embodiment, before performing load forecasting processing on the equipment operating data, the dispatching and maintenance method of the power communication network further includes: The device operation data is acquired, and the field missing rate of the device operation data is calculated as an integrity indicator, the difference between the acquisition timestamp and the current time is calculated as a timeliness indicator, and the data difference of multiple source channels of the same device is calculated as a consistency indicator. When the integrity index, the timeliness index, or the consistency index is lower than their respective preset thresholds, supplementary data collection is performed to obtain supplementary data. The supplementary data and the equipment operation data are timestamped and filled with fields to obtain optimized equipment operation data; The optimized equipment operation data is used as input data for the load forecasting process.

[0010] As one preferred embodiment, the anomaly detection of the device profile to identify potential faults includes: The multidimensional feature vector sequence is input into a pre-trained bidirectional gated cyclic unit time-series prediction model to obtain the device state prediction value. Based on the predicted equipment status and the equipment operation data, determine the absolute value of the prediction residual; Obtain historical residual sequences of several devices, determine the mean and standard deviation of the historical residual sequences, and determine an anomaly detection threshold based on a linear combination of the mean and the standard deviation; The absolute value of the predicted residual is compared with the anomaly determination threshold. When the absolute value of the predicted residual is greater than the anomaly determination threshold, an anomaly is determined to exist, and a potential fault is generated, which includes the device identifier, anomaly type, severity level, prediction window, scope of impact, and handling complexity.

[0011] As one preferred embodiment, after generating the execution feedback result, the scheduling and operation method of the power communication network further includes: Obtain the execution feedback result and the device operation data; calculate the state difference between the device state parameters after execution in the execution feedback result and the device state parameters before execution in the device operation data; The state difference is compared with a preset expected deviation range; when the state difference is within the expected deviation range, a positive feedback signal is generated and sent to the knowledge base to increase the node weight of the current matching path in the knowledge base; When the state difference exceeds the expected deviation range, a negative feedback signal is generated and sent to the knowledge base to reduce the node weight of the current matching path in the knowledge base and trigger a secondary decision to regenerate the power grid dispatching scheme and operation and maintenance task list.

[0012] The present invention also provides a dispatching and maintenance device for a power communication network, comprising: The acquisition module is used to acquire the device status and service priority of several devices in the target power communication network. The determination module is used to obtain the device operation data of the several devices based on the device status and service priority of the several devices and based on a determined data collection frequency; The scheduling module is used to perform load forecasting processing on the equipment operation data to obtain load forecasting results; and to perform topology analysis on the target power communication network to obtain topology analysis results; and to perform fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid scheduling scheme. The detection module is used to construct vectors from the extracted CPU utilization, memory usage, and link error rate of each device, and to construct a device profile from the obtained multi-dimensional feature vector sequence; and to perform anomaly detection on the device profile to identify potential faults. The operation and maintenance module is used to generate an operation and maintenance task list based on the potential faults, and to align the power grid dispatching scheme and the operation and maintenance task list in time to obtain a dispatching operation and maintenance scheme. The execution module is used to perform standardized operation and maintenance operations on several devices in the target power communication network according to the scheduling and operation and maintenance plan and generate execution feedback results.

[0013] The present invention also provides a dispatching and maintenance device for a power communication network, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the dispatching and maintenance method for the power communication network as described above.

[0014] The present invention further provides a computer-readable storage medium storing a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the power communication network scheduling and maintenance method as described above.

[0015] Compared with the prior art, the beneficial effects of the present invention are at least one of the following: This invention acquires the device status and service priority of several devices in a target power communication network; based on the device status and service priority of the devices, and a determined data acquisition frequency, obtains the device operation data of the devices; performs load forecasting processing on the device operation data to obtain load forecasting results; performs topology analysis on the target power communication network to obtain topology analysis results; performs fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid dispatching scheme; constructs vectors from the extracted CPU utilization, memory occupancy, and link error rate of each device, and constructs a profile from the obtained multi-dimensional feature vector sequence to obtain a device profile; performs anomaly detection on the device profile to obtain potential faults; generates an operation and maintenance task list based on the potential faults, and aligns the power grid dispatching scheme and the operation and maintenance task list in time to obtain a dispatching operation and maintenance scheme; performs standardized operation and maintenance operations on several devices in the target power communication network according to the dispatching operation and maintenance scheme and generates execution feedback results.

[0016] Compared with existing technologies, this invention determines differentiated data collection frequencies by combining equipment status and service priorities to obtain accurate equipment operation data. It then performs load forecasting on the operation data, completes topology analysis of the entire communication network, and merges the two results to generate an adaptive power grid dispatching scheme. Simultaneously, it extracts equipment CPU utilization, memory usage, and link error rate to construct a multi-dimensional feature vector sequence to generate a dynamic equipment profile. Through anomaly detection in the profile, it identifies hidden faults during performance degradation. Finally, it aligns the maintenance task list generated from these faults with the power grid dispatching scheme, executing standardized maintenance operations within the window of opportunity that does not affect core production operations such as relay protection and stability control, forming an execution feedback loop. This breaks the traditional power communication network's maintenance model of relying on manual experience and passively waiting for alarms, solving the technical pain point of being unable to predict equipment performance degradation and link risks in advance. It achieves an intelligent maintenance upgrade of the power communication network from passive response to proactive prediction, effectively ensuring the real-time availability of communication links and the safety of power grid operation. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a power communication network scheduling and maintenance method according to one embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a power communication network dispatching and maintenance device in one embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a power communication network dispatching and maintenance equipment in one embodiment of the present invention; Figure label: Among them, 11. Acquisition module; 12. Determination module; 13. Scheduling module; 14. Detection module; 15. Operation and maintenance module; 16. Execution module; 21. Processor; 22. Memory. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] One embodiment of the present invention provides a method for scheduling and operation and maintenance of a power communication network. For details, please refer to [link to relevant documentation]. Figure 1 , Figure 1 The diagram shown is a flowchart illustrating a power communication network scheduling and maintenance method according to one embodiment of the present invention. The method includes: S1: Obtain the device status and service priority of several devices in the target power communication network; S2: Based on the device status and service priority of the several devices, and based on a determined data collection frequency, obtain the device operation data of the several devices; S3: Perform load forecasting processing on the equipment operation data to obtain load forecasting results; perform topology analysis on the target power communication network to obtain topology analysis results; perform fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid dispatching scheme; S4: Construct vectors from the CPU utilization, memory usage, and link error rate of each extracted device, and construct a profile from the resulting multi-dimensional feature vector sequence to obtain a device profile; perform anomaly detection on the device profile to identify potential faults; S5: Generate an operation and maintenance task list based on the potential faults, and align the power grid dispatching scheme and the operation and maintenance task list in time sequence to obtain the dispatching operation and maintenance scheme; S6: Perform standardized operation and maintenance operations on several devices in the target power communication network according to the scheduling and operation and maintenance plan, and generate execution feedback results.

[0021] Specifically, the system proactively discovers access devices and establishes a device awareness engine through a multi-protocol adapter to perceive device status and identify device type and service priority.

[0022] The device awareness engine detects power communication devices in the network by sending discovery messages. After receiving device response messages, it parses the device identifier, device model, protocol type, and physical location information to establish a device resource ledger. The multi-protocol adapter has a built-in power-specific protocol stack, including the IEC 61850 protocol stack, Modbus protocol stack, and SNMP protocol stack. It automatically matches the corresponding protocol stack to establish a communication session based on the protocol type field in the device response message.

[0023] The data collection frequency is determined based on equipment type and service priority. High-frequency data collection is used for core service equipment, while low-frequency data collection is used for edge monitoring equipment. Status data is collected according to the collection frequency, and the collection timestamp is recorded. Equipment types include relay protection devices, stability control devices, dispatch data network equipment, and management information equipment.

[0024] The service priority is determined based on the real-time and reliability requirements of the services carried by the equipment. Among them, relay protection services and stability control services have the highest priority, dispatch data network services have the second highest priority, and management information services have the lowest priority. The high-frequency acquisition frequency is set to the second level, and the low-frequency acquisition frequency is set to the minute level.

[0025] Equipment operation data includes equipment operation status data, link status data, environmental status data, and alarm status data.

[0026] Before performing load forecasting processing on the equipment operation data, the power communication network scheduling and maintenance method further includes: acquiring the equipment operation data, calculating the field missing rate of the equipment operation data as an integrity index, calculating the difference between the acquisition timestamp and the current time as a timeliness index, and calculating the data difference degree of the same equipment's multi-source channels as a consistency index; when the integrity index, the timeliness index, or the consistency index is lower than their respective preset thresholds, supplementary acquisition processing is performed to obtain supplementary acquisition data; the supplementary acquisition data is timestamped and fields are filled in with the equipment operation data to obtain optimized equipment operation data; and the optimized equipment operation data is used as input data for the load forecasting processing.

[0027] When any of the above three indicators deviates from the preset threshold, the device will automatically increase its collection frequency and re-collect data for the missing time period, filling the missing time period with the supplementary data.

[0028] Timestamp alignment uses nearest neighbor interpolation to match the timestamps of the supplementary data with the timestamps of the existing status data, and takes the existing data point with the smallest time difference as the alignment benchmark. For missing fields, forward filling is used to fill the missing positions with the most recent valid collection value of the field. It also includes format unification, which converts the supplementary data and status data into a unified message format, including device identifier, timestamp, status field and quality label.

[0029] The load prediction process for the equipment operation data includes: acquiring historical load time series data from the equipment operation data; performing missing value imputation and outlier smoothing on the historical load time series data to obtain a preprocessed load sequence; based on the preprocessed load sequence, extracting fixed-length continuous subsequences as sample features using the sliding window method to construct a training sample set; using the training sample set to supervise the training of a pre-constructed gated recurrent unit neural network to obtain a load prediction model; and inputting the preprocessed load sequence before the current time into the load prediction model with the same sliding window length to obtain the load prediction result.

[0030] Specifically, a hybrid forecasting model based on historical trend extrapolation and real-time load overlay can also be used.

[0031] The load forecasting model takes a preprocessed load sequence as input and outputs the predicted load values ​​for future periods. The load forecasting model uses exponential smoothing, and the calculation formula is as follows: In the formula For the first Forecast load values ​​for each time period, in Mbps; This represents the actual load value for the t-th time period, in Mbps. Let be the predicted load value for the t-th time period, in Mbps; It is a smoothing coefficient, dimensionless, with a value ranging from 0.1 to 0.5. It is determined based on the fluctuation characteristics of the load sequence, taking a larger value when the fluctuation is severe and a smaller value when the fluctuation is stable.

[0032] A topology analysis is performed on the target power communication network to obtain the topology analysis results, including: abstracting all devices in the target power communication network as nodes in graph theory, and abstracting the physical links between all devices as edges in graph theory to construct an undirected network topology graph; determining the network connectivity and degree centrality of each node in the undirected network topology graph; comparing and analyzing the network connectivity with a preset connectivity threshold to generate topology connectivity alarm information; marking nodes whose degree centrality exceeds the preset centrality threshold as critical path nodes to generate topology critical path information; and using the topology connectivity alarm information and the topology critical path information together as the topology analysis results.

[0033] Specifically, topology analysis abstracts the power communication network as an undirected graph, where nodes represent communication devices and edges represent physical links.

[0034] The formula for calculating connectivity is: In the formula The network connectivity is dimensionless. This refers to the actual number of links, expressed in units of links. This represents the total number of network nodes, expressed in units.

[0035] When the connectivity is lower than a preset threshold, the network is determined to be at risk of being split, and a topology connectivity alarm is generated.

[0036] The preset connectivity threshold can be determined based on statistical analysis of historical operating data or on network redundancy design requirements. During the design phase, power communication networks determine redundancy requirements according to service reliability levels. For example, for backbone networks carrying relay protection services, N-1 redundancy is typically required, meaning the network remains connected even after any link is interrupted. Based on this requirement, the lower limit of connectivity can be derived.

[0037] The process of determining the degree centrality of each node in an undirected network topology graph is as follows: count the number of physical links directly connected to the node to obtain the original degree of the node; count the total number of all physical links in the entire network topology graph; divide the original degree of the node by the total number of links in the network, and the resulting ratio is the degree centrality of the node.

[0038] This ratio reflects the density of connections a node has in the network. The higher the ratio, the more links the node is connected to, the more central its position is in the network topology, and the greater the impact of its failure on the overall network connectivity.

[0039] The preset centrality threshold can be determined based on the design requirements of network hierarchical classification or based on the service carrying level.

[0040] Power communication networks are typically designed in layers: backbone, aggregation, and access. The importance of equipment at different layers varies, and their centrality thresholds should also differ. For example, backbone nodes, which handle core forwarding functions for the entire network, usually have high degree centrality. The degree centrality of all nodes in the network can be sorted in descending order, and the top 10% to 20% quantile values ​​can be used as the threshold for determining key backbone nodes. Access layer nodes generally have lower degree centrality, and a lower threshold can be set separately for them.

[0041] Different devices in a power communication network carry different service levels. Devices carrying production control services such as relay protection and stability control are far more important than those carrying management information services. A corresponding centrality threshold can be set for each device based on its highest service level.

[0042] The load forecasting results and the topology analysis results are fused to obtain a power grid dispatching scheme, including: determining the first information entropy of the load forecasting results and the second information entropy of the topology analysis results respectively; obtaining a fusion analysis result based on the first information entropy of the load forecasting results and the second information entropy of the topology analysis results; acquiring a power grid dispatching procedure, parsing the condition judgment rules and action execution rules in the power grid dispatching procedure, and constructing a knowledge base; using the fusion analysis result as input, traversing the condition judgment rules in the knowledge base, and matching to obtain the current power grid state node; extracting the corresponding action execution rules from the current power grid state node, and generating the power grid dispatching scheme.

[0043] Specifically, the fusion analysis employs a weighted fusion algorithm, which weights and fuses the load forecast results with the topology analysis results. The calculation formula is as follows: In the formula Dimensionless for fusion analysis results; For the first The weighting coefficients of the intermediate results of the analysis are dimensionless and satisfy the following conditions: ; For the first Intermediate results of the analysis, dimensionless; The number of intermediate results is expressed in units of "number". In this formula, both the load forecast results and the topology analysis results are intermediate results.

[0044] Weighting coefficient The weighting coefficient is determined based on the information entropy of each intermediate result in the analysis. The higher the information entropy, the higher the uncertainty of the intermediate result in the analysis, and the smaller the weighting coefficient is assigned.

[0045] Weighting coefficient The information entropy is determined based on the information entropy of each intermediate analysis result. The formula for calculating information entropy is: In the formula For the first The information entropy of the intermediate results of the analysis is dimensionless. For the first The probability value of each state interval is dimensionless; K is the number of state intervals, in units of 1.

[0046] The state interval division uses the equal-frequency binning method to divide the historical data distribution of the intermediate analysis results into... There are 3 intervals, each containing the same number of samples.

[0047] Obtain the power grid dispatching procedures, parse the condition judgment rules and action execution rules in the power grid dispatching procedures, construct the condition judgment rules and action execution rules into a structured decision tree, and use the structured decision tree as a knowledge base.

[0048] Specifically, the power grid dispatching procedure text is read, and condition judgment rules and action execution rules are extracted through natural language processing. Condition judgment rules include equipment load conditions, link status conditions, time window conditions, and service type conditions; action execution rules include load adjustment actions, route switching actions, power factor correction actions, and alarm reporting actions.

[0049] The fusion analysis results are used as input features. Based on the input features, the condition judgment rules in the knowledge base are traversed to match the current power grid state node. The corresponding action execution rules in the knowledge base are extracted at the current power grid state node as a set of scheduling actions. Based on the set of scheduling actions, a scheduling scheme including load adjustment, route switching or power factor correction is generated.

[0050] At the same time, the feasibility of each scheduling action needs to be assessed. The feasibility assessment is based on the current status of the target equipment, resource availability, and operational constraints.

[0051] Vectors are constructed for the CPU utilization, memory usage, and link error rate of each extracted device, and the resulting multidimensional feature vector sequence is used to construct a profile to obtain a device profile. Specifically, the acquired device operation data is aggregated according to the device dimension, and the CPU utilization, memory usage, and link error rate of each device at a specific time point are extracted. These three values ​​reflect the device's computing load, storage pressure, and communication quality, respectively.

[0052] For each device, the health profile of the device is formed by the three-dimensional feature vector consisting of CPU utilization, memory usage and link error rate at each time point, in ascending order of the collected timestamps. This profile is essentially a two-dimensional matrix, with rows representing time steps and columns representing the three health feature dimensions, recording the complete evolution of the health status of a device over a period of time.

[0053] Anomaly detection is performed on the device profile to identify potential faults. This includes: inputting the multidimensional feature vector sequence into a pre-trained bidirectional gated recurrent unit time-series prediction model to obtain device state prediction values; determining the absolute value of the prediction residual based on the device state prediction values ​​and the device operating data; acquiring several historical residual sequences of the device, determining the mean and standard deviation of the historical residual sequences, and determining an anomaly judgment threshold based on a linear combination of the mean and the standard deviation; comparing the absolute value of the prediction residual with the anomaly judgment threshold, and determining that an anomaly exists when the absolute value of the prediction residual is greater than the anomaly judgment threshold, and generating potential faults that include device identifier, anomaly type, severity level, prediction window, impact range, and handling complexity.

[0054] A time-series prediction model is established based on the historical evolution trajectory of equipment health profiles, and the predicted values ​​of equipment status for future periods are output.

[0055] The bidirectional gated recurrent unit (GRU) time series prediction model employs a deep time series network based on GRUs. The network structure includes an input layer, a bidirectional GRU layer, an attention mechanism layer, and a fully connected output layer.

[0056] The input layer receives a time series of health feature vectors, with an input dimension of [missing information]. The time step is T.

[0057] The input vector is represented as: In the formula, X is the input matrix with dimension 1. ; Let be the health feature vector at time step t, with dimension . ; For the health feature dimension, the value is 3, corresponding to the three features of CPU utilization, memory usage and link error rate; This represents the time step, measured in steps, with a value of 60, corresponding to historical data from the past 60 time steps; superscript. This indicates the matrix transpose.

[0058] The bidirectional gated loop unit layer contains a forward-gated loop unit sublayer and a backward-gated loop unit sublayer, with each sublayer having the same number of hidden units. The gating loop unit contains three core steps: updating the gate, resetting the gate, and setting the candidate hidden state.

[0059] When the predicted residual is greater than the anomaly detection threshold, an anomaly is detected, and potential faults are generated.

[0060] The severity level of potential faults is determined by the multiple by which the predicted residual exceeds the threshold, the prediction window is determined by the output step size of the time series prediction model, the scope of impact is determined by the equipment topology correlation, and the complexity of handling is determined by statistics of historical operation and maintenance cases.

[0061] Device topology correlation is calculated based on the degree centrality of the device in the network topology. Degree centrality is the ratio of the number of links directly connected to the device to the total number of links in the network.

[0062] A maintenance task list is generated based on the identified fault hazards, and the power grid dispatching scheme and the maintenance task list are time-series aligned to obtain a dispatching maintenance scheme. Specifically, the process of generating the maintenance task list based on the fault hazards includes: extracting six fields of information from the fault hazards obtained from anomaly detection: equipment identifier, anomaly type, severity level, prediction window, impact range, and handling complexity; and weighting and summing the three factors of severity level, impact range, and handling complexity to calculate a comprehensive threat score.

[0063] The severity level has the highest weight, followed by the scope of impact, and the complexity of handling has the lowest weight. The higher the overall threat score, the greater the threat that the potential fault poses to the power production business.

[0064] The overall threat score is compared with preset high and low thresholds: threats above the high threshold are marked as urgent priority, those between the high and low thresholds are marked as important priority, and those below the low threshold are marked as general priority. Priority labels are then associated with corresponding handling procedures, required tools, and security risk descriptions to form a structured operational task entry.

[0065] Repeat the above steps for all potential faults and obtain a list of maintenance tasks.

[0066] Extract the list of exclusive resources occupied by the power grid dispatch scheme and the expected execution timestamp; extract the resource requirements list for each task from the operation and maintenance task list, including physical ports, logical channels, bandwidth resources and computing resources.

[0067] Calculate the intersection between the exclusive resource list of the power grid dispatch scheme and the resource requirement list of the operation and maintenance task list. If the intersection is empty, it means that there is no resource conflict and the process can proceed directly to the next step. If the intersection is not empty, mark the intersection part as the set of conflicting resources and determine that there is a resource conflict.

[0068] Conflicting resources are arbitrated and prioritized according to the service priority rules of the power communication network: relay protection services take precedence over dispatch data network services, and dispatch data network services take precedence over management information services. Higher-priority tasks gain priority access to resources and receive an earlier execution timestamp, while lower-priority tasks are executed later. Simultaneously, dependencies between tasks are checked; if a maintenance task depends on the equipment status after the scheduling plan is executed, the maintenance task can only start after the scheduling plan is completed. After arbitration, a revised expected execution timestamp is obtained.

[0069] The power grid dispatching scheme and operation and maintenance task list that do not have resource conflicts are merged with the scheme and list that have been revised with execution timestamps after time-series arbitration to form a unified execution schedule. This schedule clarifies the execution order, execution time and resource allocation of each dispatching operation and operation and maintenance task, which is the final dispatching and operation and maintenance scheme.

[0070] Based on the aforementioned scheduling and maintenance plan, standardized operation and maintenance operations are performed on several devices in the target power communication network, and execution feedback results are generated.

[0071] After generating the execution feedback results, the scheduling and operation method of the power communication network further includes: Obtain the execution feedback result and the device operation data; calculate the state difference between the device state parameters after execution in the execution feedback result and the device state parameters before execution in the device operation data; The state difference is compared with a preset expected deviation range; when the state difference is within the expected deviation range, a positive feedback signal is generated and sent to the knowledge base to increase the node weight of the current matching path in the knowledge base; When the state difference exceeds the expected deviation range, a negative feedback signal is generated and sent to the knowledge base to reduce the node weight of the current matching path in the knowledge base and trigger a secondary decision to regenerate the power grid dispatching scheme and operation and maintenance task list.

[0072] Another embodiment of the present invention provides a dispatching and maintenance device for a power communication network. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2 The diagram shown is a structural schematic of a power communication network dispatching and maintenance device according to one embodiment of the present invention. The device includes: Module 11 is used to acquire the device status and service priority of several devices in the target power communication network; The determination module 12 is used to obtain the device operation data of the several devices based on the device status and service priority of the several devices and based on a determined data collection frequency; The scheduling module 13 is used to perform load forecasting processing on the equipment operation data to obtain load forecasting results; and to perform topology analysis on the target power communication network to obtain topology analysis results; and to perform fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid scheduling scheme. The detection module 14 is used to construct vectors from the extracted CPU utilization, memory usage, and link error rate of each device, and to construct a profile from the obtained multi-dimensional feature vector sequence to obtain a device profile; and to perform anomaly detection on the device profile to obtain potential faults. The operation and maintenance module 15 is used to generate an operation and maintenance task list based on the potential faults, and to align the power grid dispatching scheme and the operation and maintenance task list in time to obtain a dispatching operation and maintenance scheme. The execution module 16 is used to perform standardized operation and maintenance operations on several devices in the target power communication network according to the scheduling and operation and maintenance plan and generate execution feedback results.

[0073] See Figure 3 This is a schematic diagram of the structure of a power communication network dispatching and maintenance device provided in an embodiment of the present invention. The power communication network dispatching and maintenance device provided in this embodiment includes a processor 21, a memory 22, and a computer program stored in the memory 22 and configured to be executed by the processor 21. When the processor 21 executes the computer program, it implements the steps described in the above embodiment of the power communication network dispatching and maintenance method, for example... Figure 1 The steps S1 to S6 described above; or, when the processor 21 executes the computer program, it implements the functions of each module in the above-described device embodiments, such as the acquisition module 11.

[0074] For example, the computer program can be divided into one or more modules, which are stored in the memory 22 and executed by the processor 21 to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the dispatching and maintenance equipment of the power communication network. For example, the computer program can be divided into an acquisition module 11, a determination module 12, a dispatching module 13, etc., with the specific functions of each module as follows: Module 11 is used to acquire the device status and service priority of several devices in the target power communication network; The determination module 12 is used to obtain the device operation data of the several devices based on the device status and service priority of the several devices and based on a determined data collection frequency; The scheduling module 13 is used to perform load forecasting processing on the equipment operation data to obtain load forecasting results; and to perform topology analysis on the target power communication network to obtain topology analysis results; and to perform fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid scheduling scheme. The detection module 14 is used to construct vectors from the extracted CPU utilization, memory usage, and link error rate of each device, and to construct a profile from the obtained multi-dimensional feature vector sequence to obtain a device profile; and to perform anomaly detection on the device profile to obtain potential faults. The operation and maintenance module 15 is used to generate an operation and maintenance task list based on the potential faults, and to align the power grid dispatching scheme and the operation and maintenance task list in time to obtain a dispatching operation and maintenance scheme. The execution module 16 is used to perform standardized operation and maintenance operations on several devices in the target power communication network according to the scheduling and operation and maintenance plan and generate execution feedback results.

[0075] The dispatching and maintenance equipment for the power communication network may include, but is not limited to, processor 21 and memory 22. Those skilled in the art will understand that the schematic diagram is merely an example of the dispatching and maintenance equipment for the power communication network and does not constitute a limitation on the equipment. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the dispatching and maintenance equipment for the power communication network may also include input / output devices, network access devices, buses, etc.

[0076] The processor 21 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor 21 is the control center of the power communication network's dispatching and maintenance equipment, connecting various parts of the power communication network's dispatching and maintenance equipment through various interfaces and lines.

[0077] The memory 22 can be used to store the computer programs and / or modules. The processor 21 implements various functions of the power communication network dispatching and maintenance equipment by running or executing the computer programs and / or modules stored in the memory 22 and calling the data stored in the memory 22. The memory 22 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0078] If the modules integrated into the dispatching and maintenance equipment of the power communication network are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0079] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0080] Accordingly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform steps in the power communication network scheduling and maintenance method of the above embodiments, for example... Figure 1 Steps S1 to S6 as described above.

[0081] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for scheduling and maintaining a power communication network, characterized in that, include: Obtain the device status and service priority of several devices in the target power communication network; Based on the device status and service priority of the aforementioned devices, and based on a determined data collection frequency, the device operation data of the aforementioned devices is obtained; The equipment operation data is processed for load forecasting to obtain load forecasting results; A topology analysis is performed on the target power communication network to obtain the topology analysis results; the load forecast results and the topology analysis results are then fused and analyzed to obtain a power grid dispatching scheme. The CPU utilization, memory usage, and link error rate of each extracted device are used to construct a vector, and the resulting multi-dimensional feature vector sequence is used to construct a profile to obtain a device profile. Anomaly detection was performed on the device profile to identify potential faults; A maintenance task list is generated based on the potential faults, and the power grid dispatching scheme and the maintenance task list are time-series aligned to obtain the dispatching maintenance scheme. Based on the aforementioned scheduling and maintenance plan, standardized operation and maintenance operations are performed on several devices in the target power communication network, and execution feedback results are generated.

2. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, The process of performing load forecasting on the equipment operating data to obtain load forecasting results includes: Acquire historical load time series data from the equipment operation data; perform missing value imputation and outlier smoothing on the historical load time series data to obtain a preprocessed load sequence; Based on the preprocessed load sequence, a fixed-length continuous subsequence is extracted as sample features using the sliding window method to construct a training sample set; The pre-constructed gated recurrent unit neural network is trained using the training sample set to obtain a load prediction model; The preprocessed load sequence prior to the current time is input into the load prediction model with the same sliding window length to obtain the load prediction result.

3. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, The topology analysis of the target power communication network, to obtain the topology analysis results, includes: All devices in the target power communication network are abstracted as nodes in graph theory, and the physical links between all devices are abstracted as edges in graph theory to construct an undirected network topology graph. Determine the network connectivity of the undirected network topology graph and the degree centrality of each node in the undirected network topology graph; The network connectivity is compared and analyzed with a preset connectivity threshold to generate topology connectivity alarm information; Nodes whose degree centrality exceeds a preset centrality threshold are marked as critical path nodes, and topological critical path information is generated. The topology connectivity alarm information and the topology critical path information are used together as the topology analysis result.

4. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, The process of fusing the load forecast results and the topology analysis results to obtain a power grid dispatching scheme includes: Determine the first information entropy of the load prediction result and the second information entropy of the topology analysis result, respectively; Based on the first information entropy of the load forecast result and the second information entropy of the topology analysis result, the fusion analysis result is obtained; Obtain the power grid dispatching procedures, parse the condition judgment rules and action execution rules in the power grid dispatching procedures, and build a knowledge base; Using the fusion analysis results as input, the condition judgment rules in the knowledge base are traversed to match the current power grid state node; the corresponding action execution rules are extracted from the current power grid state node to generate the power grid dispatching scheme.

5. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, Before performing load forecasting processing on the equipment operating data, the dispatching and maintenance method of the power communication network further includes: The device operation data is acquired, and the field missing rate of the device operation data is calculated as an integrity indicator, the difference between the acquisition timestamp and the current time is calculated as a timeliness indicator, and the data difference of multiple source channels of the same device is calculated as a consistency indicator. When the integrity index, the timeliness index, or the consistency index is lower than their respective preset thresholds, supplementary data collection is performed to obtain supplementary data. The supplementary data and the equipment operation data are timestamped and filled with fields to obtain optimized equipment operation data; The optimized equipment operation data is used as input data for the load forecasting process.

6. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, The anomaly detection performed on the device profile to identify potential faults includes: The multidimensional feature vector sequence is input into a pre-trained bidirectional gated cyclic unit time-series prediction model to obtain the device state prediction value. Based on the predicted equipment status and the equipment operation data, determine the absolute value of the prediction residual; Obtain historical residual sequences of several devices, determine the mean and standard deviation of the historical residual sequences, and determine an anomaly detection threshold based on a linear combination of the mean and the standard deviation; The absolute value of the predicted residual is compared with the anomaly determination threshold. When the absolute value of the predicted residual is greater than the anomaly determination threshold, an anomaly is determined to exist, and a potential fault is generated, which includes the device identifier, anomaly type, severity level, prediction window, scope of impact, and handling complexity.

7. The dispatching and maintenance method for a power communication network as described in claim 1, characterized in that, After generating the execution feedback results, the scheduling and operation method of the power communication network further includes: Obtain the execution feedback result and the device operation data; calculate the state difference between the device state parameters after execution in the execution feedback result and the device state parameters before execution in the device operation data; The state difference is compared with a preset expected deviation range; when the state difference is within the expected deviation range, a positive feedback signal is generated and sent to the knowledge base to increase the node weight of the current matching path in the knowledge base; When the state difference exceeds the expected deviation range, a negative feedback signal is generated and sent to the knowledge base to reduce the node weight of the current matching path in the knowledge base and trigger a secondary decision to regenerate the power grid dispatching scheme and operation and maintenance task list.

8. A dispatching and maintenance device for a power communication network, characterized in that, include: The acquisition module is used to acquire the device status and service priority of several devices in the target power communication network. The determination module is used to obtain the device operation data of the several devices based on the device status and service priority of the several devices and based on a determined data collection frequency; The scheduling module is used to perform load forecasting processing on the equipment operation data to obtain load forecasting results; and to perform topology analysis on the target power communication network to obtain topology analysis results; and to perform fusion analysis on the load forecasting results and the topology analysis results to obtain a power grid scheduling scheme. The detection module is used to construct vectors from the extracted CPU utilization, memory usage, and link error rate of each device, and to construct a profile from the obtained multi-dimensional feature vector sequence to obtain a device profile. Anomaly detection was performed on the device profile to identify potential faults; The operation and maintenance module is used to generate an operation and maintenance task list based on the potential faults, and to align the power grid dispatching scheme and the operation and maintenance task list in time to obtain a dispatching operation and maintenance scheme. The execution module is used to perform standardized operation and maintenance operations on several devices in the target power communication network according to the scheduling and operation and maintenance plan and generate execution feedback results.

9. A dispatching and maintenance device for a power communication network, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the scheduling and operation method of the power communication network as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the scheduling and operation and maintenance method of the power communication network as described in any one of claims 1 to 7.

Images