Method and system for automatically generating and distributing electricity utilization inspection task

Abstract

The invention discloses a power utilization inspection task automatic generation and distribution method and system, and relates to the technical field of power supply and distribution, and the method comprises the steps: carrying out the data collection of regional nodes, building a regional node data set, carrying out the power utilization risk recognition, creating M power utilization inspection tasks, and building a first task priority; activating a backtracking extraction channel, carrying out electricity consumption information backtracking extraction, establishing M electricity consumption information sequences, and establishing a second task priority; and performing multi-agent adaptation analysis according to the first task priority and the second task priority, and establishing a power utilization inspection task distribution result. The method solves the technical problems of low task execution efficiency and slow response speed caused by dependence on manual arrangement and lack of intelligent analysis and real-time response capabilities of the power utilization inspection task in the prior art, and achieves the purposes of automatically generating the power utilization inspection task by fusing multi-source data, dynamically adjusting the task priority in combination with real-time data, and improving the efficiency of power utilization inspection. And the task execution efficiency and the response speed are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply and distribution, and in particular to a method and system for automatically generating and distributing power inspection tasks. Background Art

[0002] As power systems continue to grow in complexity, particularly in the context of large-scale and smart grids, traditional electricity inspection tasks rely heavily on manual scheduling, resulting in low efficiency and limited flexibility in responding to dynamic changes in the power system. Existing technologies typically allocate tasks based on static rules and experience, lacking intelligent analysis of real-time data. This results in inaccurate task prioritization, delayed inspection execution, and an inability to respond promptly to emergencies such as equipment failures and power fluctuations. Summary of the Invention

[0003] The present application provides a method and system for automatically generating and distributing electricity inspection tasks, which is used to solve the technical problems in the prior art that electricity inspection tasks rely on manual arrangements, lack intelligent analysis and real-time response capabilities, resulting in low task execution efficiency and slow response speed.

[0004] The first aspect of the present application provides a method for automatically generating and distributing electricity inspection tasks, the method comprising: collecting data from regional nodes to establish a regional node dataset, the regional node dataset comprising an electricity dataset and an external dataset; using the regional node dataset to identify electricity risks and create M electricity inspection tasks; performing a joint task analysis based on the external environment and electricity data on the M electricity inspection tasks to establish a first task priority; activating a backtracking extraction channel and using the backtracking extraction channel to backtrack extract electricity intelligence for the M electricity inspection tasks to establish M electricity intelligence sequences; performing a joint task priority analysis of the M electricity inspection tasks on the M electricity intelligence sequences to establish a second task priority; performing a multi-agent adaptation analysis of the electricity inspection tasks based on the first task priority and the second task priority to establish an electricity inspection task distribution result.

[0005] The second aspect of the present application provides an automatic generation and distribution system for electricity inspection tasks, the system comprising: a regional node data acquisition module, the regional node data acquisition module is used to collect data from regional nodes and establish a regional node data set, the regional node data set includes an electricity data set and an external data set; a regional electricity risk identification module, the regional electricity risk identification module is used to use the regional node data set to identify electricity risks and create M electricity inspection tasks; a first joint task analysis module, the first joint task analysis module is used to perform a joint task analysis on the M electricity inspection tasks based on the external environment and electricity data, and establish a first joint task analysis module. A task priority; an electricity intelligence backtracking module, the electricity intelligence backtracking module is used to activate the backtracking extraction channel, use the backtracking extraction channel to perform backtracking extraction of electricity intelligence for M electricity inspection tasks, and establish M electricity intelligence sequences; a second joint task analysis module, the second joint task analysis module is used to perform joint task priority analysis of M electricity inspection tasks on the M electricity intelligence sequences, and establish a second task priority; a multi-agent adaptation analysis module, the multi-agent adaptation analysis module is used to perform multi-agent adaptation analysis of electricity inspection tasks based on the first task priority and the second task priority, and establish an electricity inspection task distribution result.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The method and system for automatically generating and distributing electricity inspection tasks provided in the present application relate to the field of power supply and distribution technology. A regional node data set is established through data collection, electricity risk identification is performed and inspection tasks are created. Task priorities are determined through joint task analysis, a backtracking extraction channel is activated to generate an electricity intelligence sequence, and a second priority analysis is performed. Combined with the first and second priorities, task distribution is optimized using multi-agent adaptive analysis to improve the execution efficiency and intelligence level of inspection tasks. This solves the technical problem in the prior art that electricity inspection tasks rely on manual arrangements, lack intelligent analysis and real-time response capabilities, resulting in low task execution efficiency and slow response speed. This achieves the technical effect of automatically generating electricity inspection tasks by fusing multi-source data, and dynamically adjusting task priorities in combination with real-time data to improve task execution efficiency and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1A flowchart of a method for automatically generating and distributing electricity inspection tasks provided in an embodiment of the present application; Figure 2 Schematic diagram of the structure of the automatic generation and distribution system of electricity inspection tasks provided in an embodiment of the present application.

[0009] Explanation of the accompanying drawings: regional node data collection module 11, regional electricity risk identification module 12, first joint task analysis module 13, electricity intelligence backtracking module 14, second joint task analysis module 15, multi-agent adaptation analysis module 16. DETAILED DESCRIPTION

[0010] The present application provides a method and system for automatically generating and distributing electricity inspection tasks, which is used to solve the technical problems in the prior art that electricity inspection tasks rely on manual arrangements, lack intelligent analysis and real-time response capabilities, resulting in low task execution efficiency and slow response speed.

[0011] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0012] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0013] Example 1, as Figure 1 As shown, the present application provides a method for automatically generating and distributing electricity inspection tasks, the method comprising: P10: Collect data from regional nodes and establish a regional node dataset, which includes an electricity consumption dataset and an external dataset.

[0014] Specifically, comprehensive data collection is first required for regional nodes. Regional nodes are key regional units in the power system responsible for data collection, monitoring, and control, including substations, distribution networks, and user-side electricity meters. The goal of data collection is to obtain multi-dimensional information related to electricity usage to support subsequent risk identification and task generation. Specifically, data collection is divided into two main parts: electricity usage datasets and external datasets.

[0015] Electricity usage datasets primarily include data related to power consumption and equipment operation, such as power consumption, usage time, equipment status (e.g., on / off status, load factor), and voltage and current parameters. This data is typically collected in real time by smart meters, sensors, or power monitoring systems (SCADA systems) and stored in a database. The core value of electricity usage datasets lies in their ability to reflect actual power usage at regional nodes, providing fundamental data support for subsequent risk identification.

[0016] External datasets cover external environmental factors related to electricity usage, such as weather data (such as temperature, humidity, and wind speed), holiday information, and economic activity data (such as industrial production indices and commercial activity data). This data is typically obtained through third-party data sources (such as meteorological bureaus and national statistics bureaus) or public data interfaces. External datasets provide contextual information that influences electricity usage. For example, high temperatures may increase electricity load, while holidays may lead to changes in electricity usage patterns.

[0017] During data collection, various technical measures can be employed to ensure data integrity and accuracy. For example, the Internet of Things (IoT) can be used to enable real-time data transmission from smart meters and sensors, data cleaning techniques can be used to address missing or outliers, and data integration techniques can be used to uniformly store data from different sources into regional node datasets. Furthermore, to improve data collection efficiency, edge computing can be employed to perform preliminary data processing locally at regional nodes, reducing data transmission latency and bandwidth pressure.

[0018] Ultimately, through this data collection and integration, a complete regional node dataset was established. This dataset not only includes electricity usage data and external data, but also links metadata such as timestamps and geographic locations to form a structured data set, providing a reliable data foundation for subsequent electricity risk identification and task generation.

[0019] P20: Use the regional node dataset to identify electricity consumption risks and create M electricity consumption inspection tasks.

[0020] Optionally, based on regional node datasets, electricity usage risks can be identified through data analysis and model algorithms, and corresponding electricity inspection tasks can be generated. For example, first, data such as electricity usage, equipment status, voltage and current from the electricity usage dataset is combined with information such as weather, holidays, and economic activities from external datasets to construct a multi-dimensional electricity usage behavior analysis model. For example, time series analysis techniques can be used to identify abnormal fluctuations in electricity usage; machine learning algorithms (such as isolation forests and support vector machines) can be used to detect abnormalities in equipment operating status; and association rule mining can be used to discover potential relationships between electricity usage behavior and the external environment. During the risk identification process, a risk scoring model can be used to quantitatively assess the electricity usage risk of each regional node. Factors such as abnormal electricity usage, equipment failure probability, and external environmental risks are used as inputs. A weighted calculation is used to determine a risk score for each node. Based on the score, the risk is categorized into different levels (e.g., high, medium, and low), and specific electricity inspection tasks are generated for high-risk nodes.

[0021] Each electricity inspection task includes a task objective (specifying the area, node, or device to be inspected), a risk type (specifying the specific risk category, such as overload risk or equipment failure risk), a priority (determining the urgency of the task based on the risk score), and inspection content (listing specific items to be inspected, such as device status and line load). Ultimately, through the above analysis and task generation process, M electricity inspection tasks are created. These tasks serve as input for subsequent steps. Through prioritization and intelligent distribution, efficient execution of electricity inspections is achieved.

[0022] P30: Perform joint task analysis on M electricity inspection tasks based on external environment and electricity consumption data, and establish the first task priority.

[0023] Furthermore, step P30 in the embodiment of the present application further includes: P31: Call the electricity consumption dataset and external dataset in the regional node dataset, divide the electricity consumption dataset and external dataset into data hierarchies respectively, and establish first-level data and second-level data. The first-level data includes power grid anomaly data, equipment operation status data, and user electricity consumption behavior data. The second-level data includes meteorological data, geographical environment data, social environment data, and power public opinion data; P32: After performing associative reasoning on the first-level data and the second-level data, perform joint task analysis and establish the first task priority.

[0024] It should be understood that a joint task analysis based on external environment and electricity usage data is performed on the M electricity inspection tasks to establish a first task priority. This process can scientifically sort electricity inspection tasks by comprehensively considering the internal operating status of the power system and external environmental factors, ensuring that high-risk electricity issues are prioritized, thereby improving the safety and stability of the power system.

[0025] First, the power consumption dataset and external dataset from the regional node dataset are accessed and hierarchically divided into two categories, creating first-level and second-level data. The first-level data includes grid anomaly data, equipment operating status data, and user electricity usage behavior data, which directly reflect the operating status of the power system and user electricity usage behavior. The second-level data includes meteorological data, geographic environment data, social environment data, and power public opinion data, which provide information about the external environment that influences electricity usage behavior. This hierarchical division provides a clear data structure foundation for subsequent associative reasoning and joint task analysis.

[0026] Next, associative reasoning is performed on the first- and second-level data, followed by joint task analysis and the establishment of a first task priority. Associative reasoning uses algorithms to perform logical and statistical analysis on various types of data, identifying correlations between them and helping the system determine which factors are most critical to power system operation under specific conditions. For example, equipment operating status data can be correlated with meteorological data to analyze the impact of extreme weather on equipment failures; user electricity usage data can be correlated with social environmental data to study changes in electricity usage patterns during holidays or special events. Based on these associative reasoning results, M electricity inspection tasks are prioritized, taking into account both electricity usage data and external environmental factors. Finally, the first task priority is established based on the results of the joint task analysis. This priority is determined based on the analysis of comprehensive data, ensuring that the most urgent or important tasks are prioritized. This allows the system to intelligently adjust task priorities based on the combined external environment and electricity usage data, achieving efficient allocation and execution of power inspection tasks.

[0027] Furthermore, association reasoning is performed on the first-level data and the second-level data. Step P30-2 of the embodiment of the present application further includes: P32-1a: After feature normalization of the first-level data and the second-level data, perform time alignment; P32-2a: Configure the task-device-environment knowledge graph and use the graph neural network of the task-device-environment knowledge graph to perform associative reasoning. The graph neural network is as follows: ;in, Representation Task The updated feature vector of Characterization The trainable weight matrix of the layer, Representation Task The set of neighbor nodes of Representation Task With neighbors The correlation degree, Characterizing Neighbors In the The eigenvector of the round iteration, Characterizing Neighbors The task priority influencing factor, Represents the urgency of the task, is the task urgency adjustment coefficient, Characterize nonlinear activation functions, Characterize the task risk factor, Characterizes the task risk impact coefficient.

[0028] Optionally, further associative reasoning is performed on the first-level data and the second-level data to better determine the priority of the task. First, after the features of the first-level data and the second-level data are standardized, time alignment is performed. Feature standardization is to eliminate the dimensional differences between different data dimensions so that the data can be compared and analyzed at the same scale. For example, power grid anomaly data, equipment operating status data, user electricity consumption behavior data, meteorological data, geographical environment data, etc. are uniformly converted into standardized numerical values. Time alignment ensures that the time series of different data sets remain consistent so that subsequent associative reasoning can be performed based on the same time point. For example, the timestamps of electricity consumption data and meteorological data are aligned to ensure that the data analyzed is within the same time period. This step provides data consistency and comparability for subsequent associative reasoning.

[0029] Next, a task-device-environment knowledge graph is constructed and associative reasoning is performed using a graph neural network. The task-device-environment knowledge graph is a structured knowledge representation method that models the relationships between power inspection tasks, device status, and external environmental factors in a graph format. A graph neural network (GNN) captures the complex relationships between tasks, devices, and environments by iteratively updating the nodes and edges in the knowledge graph.

[0030] Through these components, graph neural networks can integrate multi-dimensional information to accurately infer task priorities. Each task's feature vector represents its current state and is continuously updated as the task progresses. By calculating the relationship between tasks and their neighboring nodes, the system can dynamically adjust priorities based on the mutual influence between tasks.

[0031] Furthermore, graph neural networks utilize a trainable weight matrix to adjust the weighting of node features, allowing the system to optimize task allocation based on the relationships between tasks and devices. Furthermore, a task's neighbor nodes represent other tasks or devices closely related to the current task. The higher the degree of correlation between tasks, the greater the impact of neighboring tasks on the current task. Therefore, by calculating the correlation between tasks and their neighboring nodes, the system can appropriately adjust task priorities.

[0032] During the associative reasoning process, task priorities can be further adjusted based on the feature vectors of neighboring nodes and their task priority influencing factors. The task's urgency adjustment coefficient ensures that urgent tasks are prioritized, while the nonlinear activation function helps capture the complex relationship between task priority and risk. By comprehensively considering task risk factors and risk impact coefficients, the system can assess the potential risks of each task and further adjust its priority based on the task's risk level, ensuring that high-risk tasks are handled promptly.

[0033] Through these steps, the graph neural network integrates multi-dimensional information between tasks, equipment, and the environment to accurately derive task priorities. Ultimately, it rationally adjusts task priorities based on their importance, urgency, and risk factors, ensuring that tasks are executed when they are most needed in the power system, improving system stability and efficiency.

[0034] Furthermore, association reasoning is performed on the first-level data and the second-level data. Step P30-2 of the embodiment of the present application further includes: P32-1b: Establish a double-layer compensation mechanism, use the double-layer compensation mechanism to perform double-layer compensation analysis on the first-level data and the second-level data, and complete association reasoning based on the double-layer compensation analysis results.

[0035] Among them, the double-layer compensation analysis of the first-level data and the second-level data is performed using the double-layer compensation mechanism, including: P32-11b: using the first-level data as the central data, using the first-level data to identify the probability of abnormal power consumption, and establishing a first abnormal power consumption probability identification result; P32-12b: using the second-level data as the local compensation data, compensating and correcting the first abnormal power consumption probability identification result based on the local compensation data, and calculating the comprehensive abnormal probability; P32-13b: using the second-level data as the central data, using the second-level data to identify the probability of abnormal power consumption, and establishing a second abnormal power consumption probability identification result; P32-14b: using the first-level data as the lag compensation data, compensating and correcting the second abnormal power consumption probability identification result based on the lag compensation data, and calculating the comprehensive abnormal probability; P32-15b: performing weighted fusion based on the comprehensive abnormal probabilities calculated twice to complete the double-layer compensation analysis.

[0036] In a possible embodiment of the present application, a two-tier compensation mechanism can be introduced to optimize the associative reasoning of the first-tier data and the second-tier data, thereby enhancing the accuracy and comprehensiveness of risk assessment. First, the two-tier compensation mechanism compensates for the possible deviations of a single data source by combining the first-tier data and the second-tier data, thereby ensuring the multi-dimensionality and globality of the analysis results. When the first-tier data is used as the center, the second-tier data can compensate for the potential impact of environmental factors on the status of the power grid and equipment, and avoid misjudgments due to deviations in environmental factors; conversely, when the second-tier data is used as the center, the first-tier data makes the risk assessment more comprehensive by compensating for the feedback effect of equipment and user behavior on environmental risks.

[0037] In practice, the first layer of data is used as the central data to identify the probability of abnormal power usage and establish the first abnormal power usage probability identification result. At this point, the first layer of data primarily focuses on the state of the power grid, equipment operation, and user electricity usage behavior. It provides important information directly related to the power system. By analyzing this data, potential anomalies in the power usage process can be identified, such as equipment failures, power grid anomalies, or unusual power usage patterns.

[0038] Next, the second-level data is used as local compensation data to compensate for the first power consumption anomaly probability identification result and calculate the overall anomaly probability. The second-level data includes meteorological data, geographic environmental data, social environmental data, and power public opinion data, which can reflect the impact of the external environment on power consumption behavior. Because the first-level data may be localized, compensation with the second-level data can effectively correct the potential impact of environmental factors on the power grid and equipment status, avoiding misjudgments caused by deviations from a single data source.

[0039] Furthermore, the second-level data is used as the core data to identify the probability of abnormal electricity usage and establish a second abnormal electricity usage probability identification result. The second-level data can identify risk factors that may affect electricity usage behavior from the perspective of the external environment, such as the impact of extreme weather or social events on electricity usage patterns.

[0040] The first-level data is then used as hysteresis compensation data to compensate for the second-level power consumption anomaly probability identification results and calculate the overall anomaly probability. Because second-level data may have hysteresis, such as delayed updates of meteorological data or social event data, compensating with first-level data can account for the feedback effects of equipment and user behavior on environmental risks, making risk assessment more comprehensive.

[0041] Furthermore, a weighted fusion is performed based on the combined anomaly probabilities from the two calculations to complete a two-layer compensation analysis. This weighted fusion comprehensively considers the compensation results of the first-layer and second-layer data to obtain the final result for identifying abnormal power usage. This process not only eliminates the bias and lag inherent in a single data source but also leverages the complementary strengths of the two layers of data, providing a more accurate and comprehensive basis for subsequent associative reasoning.

[0042] Through the double-layer compensation mechanism, the scientificity and reliability of the probability identification of abnormal power consumption can be effectively improved, providing strong support for the safe operation and risk prevention and control of the power system.

[0043] P40: Activate the retrospective extraction channel, use the retrospective extraction channel to retrospectively extract the electricity consumption intelligence of M electricity consumption inspection tasks, and establish M electricity consumption intelligence sequences.

[0044] Furthermore, step P40 in the embodiment of the present application further includes: P41: Use M electricity inspection tasks to locate regional nodes, and obtain the previous round of backtracking time and the previous round of electricity inspection tasks of the located regional nodes; P42: Perform task similarity calculations on the M electricity inspection tasks and the previous round of electricity inspection tasks to generate similarity trust values; P43: Calculate the burstiness index and periodicity index of the M electricity inspection tasks; P44: Use the similarity trust value and the previous round of backtracking time to create a basic backtracking time with the M electricity inspection tasks, and perform basic backtracking time compensation based on the burstiness index and the periodicity index to create a backtracking time mapped to the M electricity inspection tasks to complete the backtracking extraction of electricity intelligence.

[0045] Optionally, activate the retrospective extraction channel and use it to retrospectively extract electricity usage intelligence from M electricity inspection tasks, establishing M electricity usage intelligence sequences. This process provides a reference for current electricity inspection tasks by retrospectively analyzing historical electricity inspection tasks and related data, improving the efficiency and accuracy of task execution.

[0046] First, we use M electricity inspection tasks to locate regional nodes and obtain the previous round of backtracking duration and electricity inspection tasks for the located regional nodes. Regional nodes are power system nodes associated with electricity inspection tasks. By locating a regional node, we can obtain the node's backtracking duration and task execution status in the previous electricity inspection round. This historical data provides the basis for subsequent task similarity calculations and backtracking duration adjustments.

[0047] Next, a task similarity calculation is performed between the M electricity inspection tasks and the tasks from the previous round to generate a similarity trust value. For example, the similarity between the current task and the previous round of tasks (such as task type, regional node, electricity usage data, etc.) can be compared to assess the similarity between the two and generate a similarity trust value. The similarity trust value reflects the degree of correlation between the current task and the previous round of tasks; the higher the similarity, the greater the trust value. This step provides a basis for adjusting the lookback period. For example, if the tasks of a certain regional node are very similar to those of the previous round, the execution status of the previous round of tasks can be used to infer the potential risks and execution difficulty of the current task, thereby improving task planning and allocation.

[0048] Furthermore, the burstiness and periodicity indicators for the M electricity inspection tasks are calculated. The burstiness indicator measures the burstiness of the electricity inspection task, such as the frequency of sudden failures or abnormal events. The periodicity indicator measures the periodic characteristics of the task, such as seasonal peak electricity consumption or periodic equipment maintenance. These two indicators reflect the temporal characteristics of the task and provide a reference for compensating the backtracking time.

[0049] Next, based on the similar trust value and the previous round of backtracking duration, a basic backtracking duration is created that is mapped to the M electricity inspection tasks. This basic backtracking duration is a backtracking time range preliminarily determined based on historical data and task similarity, providing a reference framework that represents the time span between the previous round of inspection and the current task. Then, the basic backtracking duration is compensated based on the burstiness index and periodicity index to create the final backtracking duration. For example, for tasks with higher burstiness, the backtracking duration is appropriately extended to capture more bursts; for tasks with stronger periodicity, the backtracking duration is adjusted according to the periodic characteristics to cover key time points. Through this step, the backtracking extraction of electricity consumption intelligence is completed, providing a scientific time range and reference data for the M electricity inspection tasks.

[0050] Through the above steps, the retrospective extraction channel can comprehensively consider factors such as historical data, task similarity, suddenness and periodicity, and provide accurate retrospective analysis support for electricity inspection tasks, thereby improving the efficiency and accuracy of task execution.

[0051] P50: Perform joint task priority analysis of M electricity inspection tasks for M electricity intelligence sequences and establish the second task priority.

[0052] Specifically, a joint task priority analysis of M power consumption inspection tasks is performed on M power consumption intelligence sequences to establish a second task priority. This means that through a comprehensive analysis of historical power consumption intelligence, task priority allocation is further optimized to ensure that tasks are prioritized according to importance and urgency.

[0053] First, feature extraction is performed on M power consumption intelligence sequences to obtain the key features of each task, including power consumption anomaly probability, task urgency, task risk factor, suddenness index, periodicity index, and similarity trust value. These features can fully reflect the nature and potential impact of the task. Then, a task priority scoring model is constructed based on the extracted features. The model can be expressed as ,in, Indicates a task Priority score, Indicates a task The probability of abnormal power consumption, Indicates a task The urgency of Indicates a task risk factors, Indicates a task The suddenness index, Indicates a task The cyclical indicator Indicates a task Similar trust values ​​of are the weight coefficients of each feature, which are determined by expert experience or machine learning methods.

[0054] Then, score the tasks based on their priority , sort the M electricity inspection tasks and establish a second task priority. The higher the score, the higher the task priority, indicating that the task should be executed first. Finally, during the actual execution process, the task priority is dynamically adjusted based on real-time data and task execution status. For example, when the sudden indicator or risk factor of a task changes, its priority score is recalculated and the task priority ranking is updated. Through the above steps, the joint task priority analysis can comprehensively consider the multi-dimensional characteristics of electricity inspection tasks, provide scientific and reasonable priority ranking for task execution, and thus improve the efficiency and effectiveness of electricity inspections.

[0055] P60: Perform multi-agent adaptation analysis on the electricity inspection task based on the first task priority and the second task priority, and establish the electricity inspection task distribution result.

[0056] Furthermore, step P60 of the embodiment of the present application further includes: P61: Configure the inspection task agent and the inspection personnel agent; P62: Use the inspection task agent to perform task detection resource competition analysis for the electricity inspection task, and generate a first adaptation analysis result; P63: Based on the inspection personnel agent, perform personnel task adaptation analysis for the electricity inspection task, and generate a second adaptation analysis result; P64: Use the first adaptation analysis result, the second adaptation analysis result, the first task priority, and the second task priority to perform task distribution sorting and optimization for the electricity inspection task, and establish the electricity inspection task distribution result.

[0057] It should be understood that the multi-agent adaptation analysis of the electricity inspection task is performed based on the first task priority and the second task priority. Through the synergy of the multi-agent system, the allocation and execution of tasks are optimized, and the electricity inspection task distribution results are established to ensure that tasks can be efficiently processed according to priority, resource availability and personnel capabilities.

[0058] First, configure the inspection task agent and the inspector agent. The inspection task agent represents the virtual agent responsible for managing tasks in the system, capable of analyzing and optimizing the task execution process. The inspector agent, on the other hand, represents the actual person or equipment performing the task and is responsible for interacting with the task allocation system to determine the actual execution plan for the assigned task.

[0059] Next, the inspection task agent performs a task detection resource contention analysis for the electricity inspection task, evaluating the resources required for each task (such as equipment, tools, and time) and identifying potential resource contention between tasks. By analyzing the task resource requirements, the agent determines which tasks can be successfully executed within the same time period and which tasks may need to be postponed due to resource conflicts. Based on this analysis, the first adaptation analysis result is generated, reflecting the compatibility between tasks and resources.

[0060] Next, the inspector agent performs a human-task fit analysis for the electricity inspection task, evaluating each inspector's skills, available time, workload, and other factors to determine which individuals are capable of performing which tasks. By matching individual capabilities with task requirements, the agent assigns the most suitable individual to each task. The system then generates a second fit analysis result, assigning the appropriate individual to each task, taking into account individual workload and priority.

[0061] Finally, the first and second adaptation analysis results, along with the first and second task priorities, are used to prioritize and optimize the distribution of electricity inspection tasks, ultimately establishing a distribution result. For example, by comprehensively considering task priority, resource compatibility, and personnel compatibility, an optimization algorithm (such as a heuristic algorithm or machine learning model) is employed to prioritize and allocate tasks, ensuring that high-priority tasks are prioritized while maximizing resource utilization and personnel efficiency. Through these steps, multi-agent adaptation analysis enables the scientific distribution of electricity inspection tasks, improving the efficiency and effectiveness of task execution.

[0062] Furthermore, the embodiment of the present application also includes: Continuously monitor the regional nodes to establish a monitoring data set; determine whether the monitoring data set contains monitoring data that exceeds a preset threshold; when the monitoring data set contains monitoring data that exceeds a preset threshold, generate a dynamic update instruction; and dynamically update the power inspection task distribution result using the monitoring data according to the dynamic update instruction.

[0063] In a possible embodiment of the present application, the real-time and adaptability of the electricity inspection task distribution results can be ensured through continuous monitoring and dynamic update mechanism of regional nodes.

[0064] First, regional nodes are continuously monitored, and a monitoring dataset is created based on the collected real-time data. This dataset includes real-time power data (such as current, voltage, power, and load) from each regional node, as well as external environmental data (such as weather and geographic conditions). Through continuous monitoring, the operating status of the power system and any potential changes are monitored in real time, providing data support for subsequent task adjustments.

[0065] Next, the monitoring data set is analyzed to determine whether any data exceeds preset thresholds. These thresholds are typically determined by historical data and system standards, representing normal data fluctuations. When monitoring data exceeds these thresholds, the system deems it a potential anomaly or risk, necessitating adjustments to task allocation. This mechanism enables timely identification of potential issues and the implementation of measures, avoiding delays in responding to power system failures or anomalies.

[0066] Next, when data in the monitoring dataset exceeds a preset threshold, the system generates a dynamic update instruction. This instruction triggers dynamic adjustments to task distribution, ensuring that task execution responds to real-time data changes. For example, if the node load in a certain area increases abnormally, task allocation is updated based on this dynamic change, and additional resources and personnel are dispatched to that area for inspection.

[0067] Finally, according to dynamic update instructions, the monitoring data is used to dynamically update the electricity inspection task distribution results. By re-evaluating task priorities, resource adaptability, and personnel adaptability, and incorporating the latest monitoring data, the task distribution strategy is optimized to ensure that high-priority tasks are handled promptly and resources and personnel are allocated appropriately. Through these steps, the dynamic adjustment of electricity inspection task distribution results is achieved, thereby improving the flexibility and responsiveness of task execution, ensuring that tasks can be quickly adjusted and efficiently executed in the event of emergencies or environmental changes.

[0068] Furthermore, the steps of the embodiment of the present application also include: Establish a mapping warning signal with the electricity inspection task distribution result; and use the mapping warning signal to issue a warning.

[0069] Specifically, by establishing and utilizing mapping warning signals, the system's warning capability and response speed can be further enhanced, so that potential electricity risks can be discovered and handled in advance, ensuring that tasks can be adjusted in a timely manner to avoid system failures or abnormalities.

[0070] First, a mapping warning signal corresponding to the distribution results of power inspection tasks is established. The mapping warning signal is established based on the relationship between the task distribution results and the operating status of the power system. When the task distribution results change, especially when the allocation of high-risk tasks is adjusted, the system will generate a corresponding warning signal. These warning signals can be set based on multi-dimensional factors such as task priority, equipment status, and changes in the external environment. For example, when the inspection task priority of a node in a certain area increases, or there is a possibility of equipment failure, the corresponding warning signal will be triggered to alert the system and operators that there may be potential risks in the area.

[0071] Furthermore, the system uses the generated mapped warning signals to issue early warning alerts. When a warning signal is triggered, a pre-set warning mechanism issues an alert, notifying relevant personnel or systems to respond. This alert can be delivered through real-time notifications within the system, text messages, or emails, ensuring that responsible personnel are immediately aware of potential power risks and can make appropriate decisions and adjustments. The purpose of early warning alerts is to provide early notification of potential problems, enabling inspectors or the system to take timely and effective measures to prevent further escalation of risks or power system failures.

[0072] Through this series of steps, the system can predict and warn of potential risks during the distribution of power inspection tasks. The combination of mapped warning signals and task distribution results not only enhances the system's emergency response capabilities but also ensures more intelligent and flexible task allocation, enabling the power system to promptly respond to potential risks and anomalies, ensuring the stability and security of power supply.

[0073] In summary, the embodiments of the present application have at least the following technical effects: This application collects data from regional nodes to establish a regional node dataset containing electricity usage data and external data, identifies electricity usage risks, and creates multiple electricity inspection tasks. Through joint task analysis, a first priority for tasks is established. A backtracking extraction channel is activated to extract electricity usage intelligence for tasks and generate intelligence sequences for second priority analysis. Combining the first and second priority levels of tasks, multi-agent adaptation analysis is performed to optimize task distribution.

[0074] The technical effect of automatically generating electricity inspection tasks by fusing multi-source data and dynamically adjusting task priorities based on real-time data has been achieved, thereby improving task execution efficiency and response speed.

[0075] Example 2, based on the same inventive concept as the method for automatically generating and distributing electricity inspection tasks in the previous embodiment, such as Figure 2 As shown, the present application provides a system for automatically generating and distributing electricity inspection tasks. The system and method embodiments in the present application are based on the same inventive concept. The system includes: The regional node data collection module 11 is used to collect data from regional nodes and establish a regional node data set. The regional node data set includes an electricity consumption data set and an external data set.

[0076] The regional electricity risk identification module 12 is used to use the regional node data set to identify electricity risks and create M electricity inspection tasks.

[0077] The first joint task analysis module 13 is used to perform a joint task analysis on the M electricity inspection tasks based on the external environment and electricity consumption data, and establish a first task priority.

[0078] The power consumption information backtracking module 14 is used to activate the backtracking extraction channel, use the backtracking extraction channel to backtrack and extract the power consumption information of M power consumption inspection tasks, and establish M power consumption information sequences.

[0079] The second joint task analysis module 15 is used to perform a joint task priority analysis of the M electricity inspection tasks on the M electricity intelligence sequences to establish a second task priority.

[0080] The multi-agent adaptation analysis module 16 is used to perform multi-agent adaptation analysis of the electricity inspection task based on the first task priority and the second task priority, and establish the electricity inspection task distribution result.

[0081] Furthermore, the first joint task analysis module 13 is further configured to perform the following steps: Call the electricity consumption dataset and the external dataset in the regional node dataset, divide the electricity consumption dataset and the external dataset into data hierarchies, and establish first-level data and second-level data. The first-level data includes power grid anomaly data, equipment operation status data, and user electricity consumption behavior data. The second-level data includes meteorological data, geographical environment data, social environment data, and power public opinion data. After associative reasoning is performed on the first-level data and the second-level data, perform joint task analysis and establish the first task priority.

[0082] Furthermore, the first joint task analysis module 13 is further configured to perform the following steps: After feature normalization of the first-level data and the second-level data, time alignment is performed. A task-device-environment knowledge graph is configured, and association reasoning is performed using the graph neural network of the task-device-environment knowledge graph. The graph neural network is as follows: ;in, Representation Task The updated feature vector of Characterization The trainable weight matrix of the layer, Representation Task The set of neighbor nodes of Representation Task With neighbors The correlation degree, Characterizing Neighbors In the The eigenvector of the round iteration, Characterizing Neighbors The task priority influencing factor, Represents the urgency of the task, is the task urgency adjustment coefficient, Characterize nonlinear activation functions, Characterize the task risk factor, Characterizes the task risk impact coefficient.

[0083] Furthermore, the first joint task analysis module 13 is further configured to perform the following steps: A double-layer compensation mechanism is established, and double-layer compensation analysis of the first-level data and the second-level data is performed using the double-layer compensation mechanism, and association reasoning is completed based on the double-layer compensation analysis results.

[0084] Furthermore, the first joint task analysis module 13 is further configured to perform the following steps: Taking the first-level data as the central data, the first-level data is used to identify the probability of abnormal electricity consumption and establish a first abnormal electricity consumption probability identification result; taking the second-level data as the local compensation data, the first abnormal electricity consumption probability identification result is compensated and corrected based on the local compensation data, and the comprehensive abnormal probability is calculated; taking the second-level data as the central data, the second-level data is used to identify the probability of abnormal electricity consumption and establish a second abnormal electricity consumption probability identification result; taking the first-level data as the lag compensation data, the second abnormal electricity consumption probability identification result is compensated and corrected based on the lag compensation data, and the comprehensive abnormal probability is calculated; weighted fusion is performed based on the comprehensive abnormal probabilities calculated twice to complete the double-layer compensation analysis.

[0085] Furthermore, the power consumption information backtracking module 14 is further configured to perform the following steps: M electricity inspection tasks are used to locate regional nodes, and the previous round of backtracking duration and the previous round of electricity inspection tasks of the located regional nodes are obtained; task similarity calculations are performed on the M electricity inspection tasks and the previous round of electricity inspection tasks to generate similarity trust values; the burstiness index and periodicity index of the M electricity inspection tasks are calculated; the basic backtracking duration of the M electricity inspection tasks is created with the similarity trust value and the previous round of backtracking duration, and the basic backtracking duration is compensated based on the burstiness index and the periodicity index to create a backtracking duration mapped to the M electricity inspection tasks to complete the backtracking extraction of electricity intelligence.

[0086] Furthermore, the multi-agent adaptation analysis module 16 is further configured to perform the following steps: Configure an inspection task agent and an inspection personnel agent; use the inspection task agent to perform task detection resource competition analysis for the electricity inspection task to generate a first adaptation analysis result; based on the inspection personnel agent, perform personnel task adaptation analysis for the electricity inspection task to generate a second adaptation analysis result; use the first adaptation analysis result, the second adaptation analysis result, the first task priority, and the second task priority to perform task distribution sorting and optimization for the electricity inspection task to establish an electricity inspection task distribution result.

[0087] Furthermore, the system further includes a task monitoring module, configured to perform the following steps: Continuously monitor the regional nodes to establish a monitoring data set; determine whether the monitoring data set contains monitoring data that exceeds a preset threshold; when the monitoring data set contains monitoring data that exceeds a preset threshold, generate a dynamic update instruction; and dynamically update the power inspection task distribution result using the monitoring data according to the dynamic update instruction.

[0088] Furthermore, the system further includes a pre-alarm output module for performing the following steps: Establish a mapping warning signal with the electricity inspection task distribution result; and use the mapping warning signal to issue a warning.

[0089] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0090] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0091] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for automatically generating and distributing electricity inspection tasks, characterized in that: The method comprises: Collecting data from regional nodes to establish a regional node data set, wherein the regional node data set includes an electricity consumption data set and an external data set; Using the regional node dataset to identify electricity consumption risks, M electricity consumption inspection tasks are created; Conduct joint task analysis on M electricity inspection tasks based on external environment and electricity consumption data, and establish the first task priority; Activate a retrospective extraction channel, use the retrospective extraction channel to retrospectively extract electricity consumption intelligence for M electricity consumption inspection tasks, and establish M electricity consumption intelligence sequences; Perform joint task priority analysis of M power consumption inspection tasks on M power consumption intelligence sequences and establish the second task priority; A multi-agent adaptation analysis of the electricity inspection task is performed based on the first task priority and the second task priority, and a distribution result of the electricity inspection task is established.

2. The method for automatically generating and distributing electricity inspection tasks according to claim 1, wherein: The joint task analysis of the M electricity inspection tasks based on the external environment and electricity consumption data to establish a first task priority includes: Calling the electricity consumption dataset and the external dataset in the regional node dataset, dividing the electricity consumption dataset and the external dataset into data hierarchies, respectively, to establish first-level data and second-level data, wherein the first-level data includes power grid anomaly data, equipment operation status data, and user electricity consumption behavior data, and the second-level data includes meteorological data, geographical environment data, social environment data, and power public opinion data; After associative reasoning is performed on the first-level data and the second-level data, joint task analysis is performed to establish the first task priority.

3. The method for automatically generating and distributing electricity inspection tasks according to claim 2, wherein: The performing association reasoning on the first-level data and the second-level data includes: After performing feature normalization on the first-level data and the second-level data, performing time alignment; Configure the task-device-environment knowledge graph and use the graph neural network of the task-device-environment knowledge graph to perform association reasoning. The graph neural network is as follows: ； in, Representation Task The updated feature vector of Characterization The trainable weight matrix of the layer, Representation Task The set of neighbor nodes of Representation Task With neighbors The correlation degree, Characterizing Neighbors In the The eigenvector of the round iteration, Characterizing Neighbors The task priority influencing factor, Represents the urgency of the task, is the task urgency adjustment coefficient, Characterize nonlinear activation functions, Characterize the task risk factor, Characterizes the task risk impact coefficient.

4. The method for automatically generating and distributing electricity inspection tasks according to claim 2, wherein: The performing association reasoning on the first-level data and the second-level data includes: A double-layer compensation mechanism is established, and double-layer compensation analysis of the first-level data and the second-level data is performed using the double-layer compensation mechanism, and association reasoning is completed based on the double-layer compensation analysis results.

5. The method for automatically generating and distributing electricity inspection tasks according to claim 4, wherein: The performing of the double-layer compensation analysis of the first-level data and the second-level data by using the double-layer compensation mechanism includes: Taking the first-level data as central data, using the first-level data to perform power consumption anomaly probability identification, and establishing a first power consumption anomaly probability identification result; Using the second-level data as local compensation data, performing compensation correction on the first power consumption abnormality probability identification result based on the local compensation data, and calculating the comprehensive abnormality probability; Taking the second-level data as central data, using the second-level data to perform power consumption anomaly probability identification, and establishing a second power consumption anomaly probability identification result; Using the first-level data as hysteresis compensation data, performing compensation correction on the second power consumption abnormality probability identification result based on the hysteresis compensation data, and calculating the comprehensive abnormality probability; A weighted fusion is performed based on the comprehensive abnormality probabilities calculated twice to complete the double-layer compensation analysis.

6. The method for automatically generating and distributing electricity inspection tasks according to claim 1, wherein: The method of using the retrospective extraction channel to retrospectively extract the electricity consumption intelligence of M electricity consumption inspection tasks to establish M electricity consumption intelligence sequences includes: Use M electricity inspection tasks to locate regional nodes and obtain the last round of backtracking time and the last round of electricity inspection tasks of the located regional nodes; Perform task similarity calculations on M electricity inspection tasks and the previous round of electricity inspection tasks to generate similarity trust values; Calculate the suddenness and periodicity indicators of M electricity inspection tasks; The basic backtracking duration of the M electricity inspection tasks is created using the similar trust value and the previous backtracking duration, and the basic backtracking duration is compensated based on the suddenness index and the periodicity index to create a backtracking duration mapped to the M electricity inspection tasks to complete the backtracking extraction of electricity intelligence.

7. The method for automatically generating and distributing electricity inspection tasks according to claim 1, wherein: The multi-agent adaptive analysis of the electricity inspection task based on the first task priority and the second task priority to establish the electricity inspection task distribution result includes: Configure inspection task agents and inspection personnel agents; Using the inspection task agent to perform task detection resource competition analysis for the electricity inspection task, and generate a first adaptation analysis result; Performing a personnel task adaptation analysis of the electricity inspection task based on the inspection personnel agent to generate a second adaptation analysis result; The first adaptation analysis result, the second adaptation analysis result, the first task priority, and the second task priority are used to perform task distribution sorting and optimization of the electricity inspection task to establish an electricity inspection task distribution result.

8. The method for automatically generating and distributing electricity inspection tasks according to claim 1, wherein: After the electricity inspection task distribution result is established, the following steps are included: Continuously monitor the regional nodes and establish a monitoring data set; Determining whether the monitoring data set contains monitoring data exceeding a preset threshold; When there is monitoring data in the monitoring data set that exceeds a preset threshold, a dynamic update instruction is generated; The power consumption inspection task distribution result is dynamically updated using the monitoring data according to the dynamic update instruction.

9. The method for automatically generating and distributing electricity inspection tasks according to claim 1, wherein: After the electricity inspection task distribution result is established, the method further includes: Establish a mapping warning signal with the distribution results of electricity inspection tasks; The mapped warning signal is used to issue a warning.

10. The automatic generation and distribution system of electricity inspection tasks is characterized by: The system comprises: A regional node data collection module, which is used to collect data from regional nodes and establish a regional node data set, wherein the regional node data set includes an electricity consumption data set and an external data set; A regional electricity risk identification module, configured to identify electricity risks using the regional node dataset and create M electricity inspection tasks; a first joint task analysis module, configured to perform a joint task analysis on the M electricity inspection tasks based on the external environment and electricity consumption data, and establish a first task priority; An electricity consumption information backtracking module, which is used to activate a backtracking extraction channel, use the backtracking extraction channel to backtrack and extract electricity consumption information for M electricity inspection tasks, and establish M electricity consumption information sequences; a second joint task analysis module, the second joint task analysis module being configured to perform a joint task priority analysis of the M electricity inspection tasks on the M electricity intelligence sequences, and establish a second task priority; A multi-agent adaptation analysis module is used to perform multi-agent adaptation analysis of the electricity inspection task based on the first task priority and the second task priority, and establish the electricity inspection task distribution result.

Images