Inspection decision optimization method and device for multiple inspection tasks of power system, terminal equipment and storage medium

Abstract

The invention discloses a routing inspection decision optimization method and device for multiple routing inspection tasks of an electric power system, terminal equipment and a storage medium, and belongs to the technical field of routing inspection decision optimization. The method comprises the following steps: obtaining task information of all to-be-scheduled electric power routing inspection tasks, generating an initial routing inspection path and determining a corresponding execution robot; performing routing inspection according to the initial routing inspection path, obtaining real-time state data and real-time map topological data of each execution robot, obtaining a conflict probability among the execution robots, calculating an adjustment cost according to a residual preset task importance level, residual electric quantity and task execution timeliness information of a conflict robot, and performing routing inspection according to the adjustment cost. And finally, according to the adjustment cost, a plurality of preset adjustment strategy combinations and a Nash equilibrium solution method, determining an optimal adjustment strategy combination, and continuing to perform inspection after adjustment. According to the invention, the problem of low inspection efficiency due to adoption of a fixed scheduling strategy in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of inspection decision optimization technology, and in particular to an inspection decision optimization method, device, terminal equipment and storage medium for multiple inspection tasks in power systems. Background Technology

[0002] In the inspection and maintenance of modern power systems, in order to meet the requirements of high frequency, high coverage and high reliability of inspection tasks, multiple robots are often introduced to work together to complete the allocation and path planning of large-scale inspection tasks.

[0003] Currently, traditional methods often employ fixed scheduling strategies and lack effective conflict detection and resolution mechanisms during the path planning stage. This leads to congestion in densely populated robot work areas and results in low inspection efficiency. Summary of the Invention

[0004] This invention provides a method, device, terminal equipment, and storage medium for optimizing inspection decisions for multiple inspection tasks in a power system. It can solve the problem that existing technologies use fixed scheduling strategies and lack effective conflict detection and resolution mechanisms in the path planning stage, which leads to congestion in areas with dense robot operations and low inspection efficiency.

[0005] An embodiment of the present invention provides a method for optimizing inspection decisions for multiple inspection tasks in a power system, comprising: Obtain task information for all pending power inspection tasks; Based on the above task information, generate initial inspection paths and determine the execution robot corresponding to each initial inspection path. The power inspection tasks are inspected according to the initial inspection path described above, and the real-time status data and real-time map topology data of each robot are obtained. Based on the initial inspection path, real-time status data, real-time map topology data, and preset spatiotemporal conflict detection model, the conflict probability between each execution robot is obtained. Acquire conflict robots whose conflict probability exceeds a preset probability threshold, the remaining preset task importance level of the current remaining power inspection tasks, and the remaining power supply. Based on the remaining preset task importance level, remaining battery power, and task execution time information, the adjustment cost for each conflicting robot is calculated. Based on the aforementioned adjustment costs, several preset adjustment strategy combinations, and the Nash equilibrium solution method, the optimal adjustment strategy combination is determined. After adjusting the corresponding conflict robot according to the above optimal adjustment strategy combination, the inspection continues.

[0006] Furthermore, the aforementioned task information includes: inspection targets, spatial locations of the inspection targets, execution dependencies between inspection tasks, task execution time information, and task types; Based on the aforementioned task information, the process of generating initial inspection paths and determining the corresponding execution robot for each initial inspection path includes: Clustering is performed based on the spatial location of all power inspection tasks to obtain several sub-regions, and a group of execution robots is assigned to each sub-region. For each sub-region, based on the task execution time information of each power inspection task in the above sub-region, several task groups with different task urgency levels are obtained. Based on the above task types and the corresponding robot groups for each sub-area, assign a corresponding robot to each power inspection task within each task group. For each execution robot, an initial inspection path is generated based on the inspection target, spatial location, and execution dependencies of the power inspection task.

[0007] Furthermore, based on the task execution timeliness information of each power inspection task within the aforementioned sub-region, several task groups with different levels of urgency are obtained, including: The deadline for each power inspection task is extracted from the above task execution time information. The urgency of each power inspection task is calculated based on the above task deadlines. Based on the above task urgency and the above preset task importance level, the task timeliness index is calculated. Based on the above-mentioned timeliness indicators and preset timeliness thresholds, the urgency of each power inspection task is divided into several task groups with different urgency levels.

[0008] Furthermore, based on the initial inspection path, real-time status data, real-time map topology data, and the preset spatiotemporal conflict detection model, the conflict probability between each executing robot is obtained, including: Obtain historical conflict data for each execution robot; The aforementioned historical conflict data, initial inspection path, real-time status data, task execution time information, and real-time map topology data are input into the preset spatiotemporal conflict detection model to obtain the aforementioned conflict probabilities.

[0009] Furthermore, after adjusting the corresponding conflict robot according to the above optimal adjustment strategy combination, and continuing the inspection, the process also includes: Obtain the distribution density characteristics of the above-mentioned sub-regions, the time distribution characteristics of the above-mentioned task groups, the time delay rate of all the above-mentioned power inspection tasks to be scheduled, and the frequency of conflict occurrence. The aforementioned distribution density characteristics, time distribution characteristics, time series time extension rate, and conflict occurrence frequency are stored as descriptive features of the power inspection tasks to be scheduled. This allows for the calculation of the similarity between each stored descriptive feature and the descriptive features of the new power inspection tasks to be scheduled, and the selection of the corresponding inspection decision for the new power inspection tasks to be scheduled based on the aforementioned similarity.

[0010] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments; This invention provides a power system multi-inspection task inspection decision optimization device, comprising: The system includes a task information acquisition module, an inspection path generation module, an inspection data acquisition module, a conflict probability detection module, a conflict robot data acquisition module, a module adjustment cost calculation module, an optimal strategy determination module, and an inspection decision optimization module. The aforementioned task information acquisition module is used to acquire task information for all power inspection tasks to be scheduled. The aforementioned inspection path generation module is used to generate an initial inspection path and determine the execution robot corresponding to each initial inspection path based on the aforementioned task information. The aforementioned inspection data acquisition module is used to inspect each power inspection task according to the initial inspection path and acquire the real-time status data and real-time map topology data of each executing robot. The aforementioned conflict probability detection module is used to obtain the conflict probability between each execution robot based on the initial inspection path, real-time status data, real-time map topology data, and the preset spatiotemporal conflict detection model. The aforementioned conflict robot data acquisition module is used to acquire conflict robots whose conflict probability exceeds a preset probability threshold, the remaining preset task importance level of the current remaining power inspection task, and the remaining power. The aforementioned adjustment cost calculation module is used to calculate the adjustment cost of each conflicting robot based on the remaining preset task importance level, remaining power, and task execution time information. The aforementioned optimal strategy determination module is used to determine the optimal combination of adjustment strategies based on the aforementioned adjustment cost, several preset combinations of adjustment strategies, and the Nash equilibrium solution method. The inspection decision optimization module is used to adjust the corresponding conflict robot according to the above optimal adjustment strategy combination and then continue the inspection.

[0011] Furthermore, the aforementioned task information includes: inspection targets, spatial locations of the inspection targets, execution dependencies between inspection tasks, task execution time information, and task types; The aforementioned inspection path generation module includes: Spatial clustering unit, task group division unit, robot allocation unit, and initial inspection path generation unit; The aforementioned spatial clustering unit is used to cluster all power inspection tasks according to their spatial locations, resulting in several sub-regions, and assigning a group of execution robots to each sub-region. The aforementioned task group division unit is used to divide each sub-region into several task groups with different levels of urgency based on the task execution time information of each power inspection task within the sub-region. The robot allocation unit described above is used to allocate a corresponding execution robot to each power inspection task in each task group according to the task type and the execution robot group corresponding to each sub-area. The aforementioned initial inspection path generation unit is used to generate an initial inspection path for each execution robot based on the inspection target, spatial location, and execution dependencies of the power inspection task of the execution robot.

[0012] Furthermore, the aforementioned task group division units include: The system includes a task deadline extraction subunit, a task urgency calculation subunit, a task timeliness index calculation subunit, and a task urgency classification subunit. The aforementioned task deadline extraction subunit is used to extract the task deadline of each power inspection task from the aforementioned task execution time information. The aforementioned task urgency calculation subunit is used to calculate the task urgency of each power inspection task based on the aforementioned task deadline. The aforementioned task timeliness index calculation subunit is used to calculate the task timeliness index based on the aforementioned task urgency and the aforementioned preset task importance level. The aforementioned task urgency classification subunit is used to classify the urgency of each power inspection task according to the aforementioned task timeliness indicators and preset timeliness indicator thresholds, resulting in several task groups with different urgency levels.

[0013] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment; The present invention provides a terminal device, 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 inspection decision optimization method for multiple inspection tasks in a power system as described in any embodiment of the present invention.

[0014] Based on the above method embodiments, the present invention provides a corresponding storage medium embodiment; The present invention provides a storage medium 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 inspection decision optimization method for multiple inspection tasks in a power system as described in any embodiment of the present invention.

[0015] The embodiments of the present invention have the following beneficial effects: This invention provides a method, apparatus, terminal device, and storage medium for optimizing inspection decisions for multiple inspection tasks in a power system. The method includes: acquiring task information for all power inspection tasks to be scheduled; generating initial inspection paths and determining the execution robots corresponding to each initial inspection path based on the task information; inspecting each power inspection task according to the initial inspection paths and acquiring real-time status data and real-time map topology data for each execution robot; obtaining the conflict probability between each execution robot based on the initial inspection paths, real-time status data, real-time map topology data, and a preset spatiotemporal conflict detection model; acquiring conflict robots whose conflict probabilities exceed a preset probability threshold, the remaining preset task importance level of the current remaining power inspection tasks, and the remaining power capacity; calculating the adjustment cost of each conflict robot based on the remaining preset task importance level, remaining power capacity, and task execution time information; determining the optimal adjustment strategy combination based on the adjustment cost, several preset adjustment strategy combinations, and the Nash equilibrium solution method; and finally adjusting the corresponding conflict robots according to the optimal adjustment strategy combination and continuing the inspection. Therefore, this invention calculates the conflict probability based on the real-time status data of each robot during the inspection process, and optimizes the current inspection strategy of each robot in real time based on the conflict probability, thereby improving the inspection efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating an inspection decision optimization method for multiple inspection tasks in a power system, provided by an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the structure of a power system multi-inspection task inspection decision optimization device provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0021] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0024] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0025] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0026] See Figure 1 To address the problem that existing technologies employ fixed scheduling strategies and lack effective conflict detection and resolution mechanisms during the path planning phase, leading to congestion in densely populated robot operation areas and low inspection efficiency, this invention provides an embodiment of an inspection decision optimization method for multiple inspection tasks in a power system, comprising: Step S101: Obtain task information for all pending power inspection tasks; Specifically, the aforementioned power inspection tasks can be annual power inspection plans. Task information includes: inspection targets, the spatial location of these targets, the execution dependencies between various inspection tasks, task execution timeliness information, and task type. The task execution timeliness information includes the task deadline.

[0027] Step S102: Based on the above task information, generate the initial inspection path and determine the execution robot corresponding to each initial inspection path; Specifically, for each execution robot, an initial inspection path is generated based on a weighted graph search algorithm (such as Dijkstra or A*), thereby obtaining the initial global path for all execution robots.

[0028] In a preferred embodiment, the task information includes: inspection target, spatial location of the inspection target, execution dependency relationship between each inspection task, task execution time information, and task type; Based on the aforementioned task information, the process of generating initial inspection paths and determining the corresponding execution robot for each initial inspection path includes: Clustering is performed based on the spatial location of all power inspection tasks to obtain several sub-regions, and a group of execution robots is assigned to each sub-region. Specifically, based on the spatial location of each power inspection task, clustering algorithms with density awareness, such as K-means++ or DBSCAN, are applied to divide all power inspection tasks into multiple spatial adaptive blocks. Each block is configured with an independent group of execution robots. These adaptive blocks are the aforementioned sub-regions.

[0029] Preferably, if the task density in a certain sub-region is too high, the region boundary adjustment mechanism can be triggered to automatically refine the sub-region or redistribute some tasks to neighboring sub-regions in order to maintain overall scheduling balance.

[0030] For each sub-region, based on the task execution time information of each power inspection task in the above sub-region, several task groups with different task urgency levels are obtained. Specifically, since the task execution time information includes the task deadline, the power inspection tasks in each sub-region are divided into several task groups based on the urgency of the task deadline, and the urgency of different task groups is different.

[0031] Based on the above task types and the corresponding robot groups for each sub-area, assign a corresponding robot to each power inspection task within each task group. Specifically, the task type is used as a "functional label" (such as image recognition, local perception, and high-voltage operation), and matched with the "functional label" of each execution robot in the corresponding execution robot group, thereby determining the execution robot for each power inspection task.

[0032] Preferably, after obtaining the above task information, all power inspection tasks to be scheduled can be transformed into a multi-objective, multi-constraint task graph model based on this information. Node information includes inspection targets (such as transformers, transmission lines, substations, etc.), task execution timeliness information, spatial location via GPS or GIS coordinates, execution dependencies between tasks (such as pre-checks, concurrency limits), and resource requirements (such as whether aerial work robots, infrared camera equipment, etc. are needed). This graph is stored using a graph structure for efficient subsequent parsing and operation. The graph includes a spatial layer, a temporal layer, and a resource layer. The spatial layer represents the geographical proximity between tasks, with nodes corresponding to specific inspection targets and edges representing physical distance or accessibility indicators. The temporal layer focuses on task execution timeliness information, with connections representing the time sequence logic or window overlap of tasks. The resource layer models the matching relationship between tasks and the resources required for execution, including descriptions of dimensions such as equipment type, capability tags, and schedulability.

[0033] Specifically, the internal structure of each layer consists of node sets and edge sets, stored using adjacency matrices or sparse graph structures. To overcome the limitations of independent modeling between space, time, and resources, the concept of cross-dimensional edges is introduced, connecting different layers through heterogeneous edges. For example, a task node at a certain spatial location can be connected to its time window node in the time layer via a cross-dimensional edge, while simultaneously establishing a resource requirement edge with a node in the resource layer. In this way, complex dependencies between tasks can be expressed, such as a task being spatially connected to neighboring tasks, temporally constrained by the completion of preceding tasks, and requiring the sharing of the same robot group. This complex dependency is mapped to a connected subgraph between multiple layers.

[0034] Specifically, by constructing a multi-layered heterogeneous graph model, the semantic expressive power of task modeling in the three dimensions of space, time, and resources is enhanced, effectively supporting subsequent complex scheduling and decision-making logic. Then, a progressive decomposition is performed on this basis to gradually obtain sub-regions, task groups, and execution robots assigned to individual power inspection tasks.

[0035] For each execution robot, an initial inspection path is generated based on the inspection target, spatial location, and execution dependencies of the power inspection task.

[0036] Specifically, based on a weighted graph search algorithm (such as Dijkstra's or A*), and combining the inspection target, spatial location, and execution dependencies of each robot, a corresponding initial inspection path is generated. It should be noted that determining the inspection path based on a weighted graph search algorithm is existing technology and will not be elaborated upon here.

[0037] Preferably, before generating the initial inspection path, a task collaboration tree structure can be constructed based on the aforementioned progressive decomposition results. Starting from the task groups in the sub-regions, the structure is refined downwards according to the aforementioned task groups to a single inspection target, and a binding relationship is established with the corresponding robot. This tree structure defines the pre- and post-execution dependencies between tasks through topological sorting, while also incorporating resource scheduling constraints, such as exclusive scheduling strategies for similar robots and charging station sharing strategies. Then, based on this task collaboration tree structure, a weighted graph search algorithm is used to generate the initial inspection path.

[0038] In this preferred embodiment, based on task information, corresponding execution machines are assigned to power inspection tasks of the same type and within the same task group, and initial inspection paths are generated for these execution robots.

[0039] In another preferred embodiment, based on the task execution timeliness information of each power inspection task within the aforementioned sub-region, several task groups with different levels of urgency are obtained, including: The deadline for each power inspection task is extracted from the above task execution time information. The urgency of each power inspection task is calculated based on the above task deadlines. Specifically, the urgency of a task is calculated using the following formula; the higher the value, the more urgent the task. In the formula, This indicates the urgency of power line inspection task i. This indicates the deadline for power inspection task i. Indicates the current time.

[0040] Preferably, the above current time This can refer to the time spent dividing tasks into groups based on their urgency.

[0041] Based on the above task urgency and the above preset task importance level, the task timeliness index is calculated. Specifically, each power inspection task is pre-set with a priority level to indicate its importance. Then, corresponding weighting coefficients are assigned to the priority level and the urgency level, and a weighted calculation is performed to obtain the task timeliness index. The task timeliness index is calculated using the following formula: In the formula, This represents the timeliness indicator for power inspection task i. This represents the weighting coefficient corresponding to the urgency of the task. This represents the weighting coefficient corresponding to the preset task importance level. This indicates the preset task importance level of power inspection task i.

[0042] Based on the above-mentioned timeliness indicators and preset timeliness thresholds, the urgency of each power inspection task is divided into several task groups with different urgency levels.

[0043] Preferably, two preset timeliness indicator thresholds of different sizes can be set as follows: and ,and If there is a timeliness indicator for a certain power inspection task that is not less than If a power inspection task has a timeliness index of less than [a certain threshold], it will be classified as an emergency task group. and greater than If a power inspection task has a timeliness index of no more than [a certain value], it will be classified as a high-priority task group. If so, it will be classified as a regular task group.

[0044] In this preferred embodiment, based on the timeliness information of the power inspection tasks in each sub-region, several task groups with different levels of urgency are obtained.

[0045] Step S103: Perform inspections on each power inspection task according to the initial inspection path described above, and obtain real-time status data and real-time map topology data of each executing robot. Specifically, after generating the initial inspection path, each robot begins its inspection according to its own initial path, while simultaneously acquiring real-time status data and real-time map topology data. This map topology data and real-time status data can be acquired based on the robot's onboard perception system (such as LiDAR, vision modules, and IMU).

[0046] Preferably, the map topology data mentioned above includes: each inspection target and the obstacles at its location.

[0047] Step S104: Based on the initial inspection path, real-time status data, real-time map topology data and preset spatiotemporal conflict detection model, obtain the conflict probability between each execution robot; Specifically, the aforementioned pre-defined spatiotemporal conflict detection model is a neural network model that can predict the probability of conflict between each executing robot based on the initial inspection path, real-time status data, and real-time map topology data.

[0048] In a preferred embodiment, the process of obtaining the conflict probability between each execution robot based on the initial inspection path, real-time status data, real-time map topology data, and a preset spatiotemporal conflict detection model includes: Obtain historical conflict data for each execution robot; Specifically, the aforementioned historical conflict data refers to a structured dataset of path conflict events that occurred during the current inspection of the corresponding robot while performing its inspection tasks. This dataset includes conflict scenario characteristics (such as the spatial location and map topology of the conflict) and the robot's historical state at the time of the conflict.

[0049] The aforementioned historical conflict data, initial inspection path, real-time status data, task execution time information, and real-time map topology data are input into the preset spatiotemporal conflict detection model to obtain the aforementioned conflict probabilities.

[0050] Specifically, the aforementioned real-time status data includes real-time position and real-time speed. The historical conflict data, initial inspection path, real-time status data, task execution time information, and real-time map topology data of all executing robots are used as input data and fed into a pre-trained preset spatiotemporal conflict detection model to obtain the conflict probability.

[0051] Specifically, the pre-defined spatiotemporal conflict detection model predicts collision risks by detecting spatiotemporal overlap of paths (such as two robots simultaneously passing through the same area), combined with dynamic factors such as speed fluctuations and real-time data updates. For example, the probability of a collision between robots A and B at intersection point C is 80%.

[0052] Specifically, the model contains an input layer, an embedding layer, an attention layer, a hidden layer, and an output layer. The input layer is used to fuse the input data and transform it into a feature vector, which is then fed into the embedding layer to capture the correlation between the initial inspection paths. The attention layer is used to focus on spatiotemporal overlapping features (such as distance and time). The hidden layer is then used to fit the relationship between the features and the conflict probability. Finally, the output layer outputs the final conflict probability.

[0053] Specifically, when training the preset spatiotemporal conflict detection model, a virtual inspection environment is first constructed using a simulation platform. Then, different inspection paths are set, and corresponding robots are arranged to carry out inspections. This is to obtain real-time simulation status data, simulation task execution time information, and real-time simulation map topology data of each robot during the simulation inspection process. Then, combined with the historical conflict data of the corresponding robots, training samples are constructed. Finally, the training samples are labeled with the actual conflict probability.

[0054] Specifically, when labeling the probability of actual conflict, the label can be determined by judging the spatial distance and time difference between the robots, based on actual needs. For example, if two robots will both pass through the same road segment within a preset time window, and the time difference does not exceed a preset time difference threshold (e.g., 30 seconds), then they are considered to have a high conflict, and the probability of actual conflict is set between 0.8 and 1.0. If they will both pass through the same road segment within a preset time window, but the time difference exceeds the preset time difference threshold; or if the time difference does not exceed the preset time difference threshold, but the distance between them is greater than a safe distance threshold (e.g., 5 meters), then they are considered to have a medium conflict, and the probability of actual conflict is set between 0.3 and 0.7. If the distance between them is greater than the safe distance threshold, and the time difference is greater than the preset time difference threshold, then they are labeled as having no conflict, and the probability of actual conflict is set between 0 and 0.2.

[0055] Specifically, at the start of training, training samples are input into the model, which predicts a simulated conflict probability. Then, a loss function is calculated between the predicted simulated conflict probability and the labeled true conflict probability. The model is then judged to have converged. If converged, the model training is complete; if not, the internal parameters are adjusted and training continues. Preferably, the loss function can be set as mean squared error loss.

[0056] In this preferred embodiment, the probability of conflict between each execution robot is predicted based on the initial inspection path, real-time status data, real-time map topology data, and a preset spatiotemporal conflict detection model.

[0057] Step S105: Obtain the conflict robots whose conflict probability exceeds the preset probability threshold, the remaining preset task importance level of the current remaining power inspection tasks, and the remaining power. Specifically, robots whose conflict probability exceeds a preset probability threshold are identified as conflict robots. The preset task importance level of the remaining power inspection tasks to be performed by these conflict robots is obtained as the remaining preset task importance level. At the same time, the current remaining power of these conflict robots is also obtained.

[0058] Step S106: Based on the remaining preset task importance level, remaining battery power, and task execution time information, calculate the adjustment cost for each conflicting robot. Specifically, the remaining time for the remaining power inspection task is first determined based on the task execution time information. Then, the remaining power, remaining task time, and remaining preset task importance level are normalized. The adjustment cost is calculated by weighting the normalized values. The normalized remaining power ranges from 0 to 1, with higher values ​​indicating more remaining power. The normalized remaining preset task importance level also ranges from 0 to 1, with higher values ​​indicating higher importance. The normalized remaining task time (i.e., the flexibility of the remaining task time) also ranges from 0 to 1, with higher values ​​indicating more time. The adjustment cost for the conflict robot is calculated using the following formula: In the formula, Conflict robots The cost of adjustment Conflict robots The weight coefficients corresponding to the importance levels of the remaining preset tasks. Conflict robots The normalized values ​​of the remaining preset task importance levels. This represents the weighting coefficient corresponding to the remaining battery power. Conflict robots The normalized value of the remaining battery power. This represents the weighting coefficient corresponding to the remaining time elasticity value of the task. Conflict robots The normalized value of the remaining task time.

[0059] Step S107: Based on the above adjustment costs, several preset adjustment strategy combinations, and the Nash equilibrium solution method, determine the optimal adjustment strategy combination; Specifically, conflicting robots are grouped together, and the optimal strategy combination for this group of robots is determined using a game theory equilibrium algorithm, namely Nash equilibrium. The aforementioned preset adjustment strategy combinations are obtained based on a preset adjustment strategy set, which includes several adjustment strategies such as "maintain path," "adjust path," "minor avoidance," and "cooperative waiting." Subsequently, corresponding adjustment strategies are assigned to each conflicting robot, resulting in several preset adjustment strategy combinations. Each preset adjustment strategy combination corresponds to an adjustment scheme for one robot.

[0060] Specifically, for each conflict robot, a game type threshold is set. Conflict robots with a game type exceeding this threshold are classified as zero-sum games, while the remaining conflict robots are classified as cooperative games. Then, the payoff value of each conflict robot in each group is calculated using the Nash equilibrium method. For example, for conflict robot A in a zero-sum game, choosing the "stay on the path" payoff is the decrease in its own adjustment cost, while the payoff of conflict robot B is the increase in its own adjustment cost. The payoff values ​​of the remaining cooperative game conflict robots are calculated based on the overall payoff.

[0061] Specifically, based on the adjustment cost, the payoff value of the conflict robots under each preset adjustment strategy combination is calculated. Then, the Nash equilibrium is verified combination by combination, i.e., an optimal adjustment strategy combination is found such that, under this optimal combination, in the same group of conflict robots, any individual robot cannot further increase its payoff by changing its preset adjustment strategy. The payoffs for zero-sum games and cooperative games are calculated using the following formulas: Specifically, for a zero-sum game, a baseline adjustment cost is defined for each conflict robot, representing the cost when it chooses the "stay on the path" strategy and other conflict robots also adopt the corresponding default strategies. Therefore, for each conflict robot, its payoff under the preset adjustment strategy combination is equal to the difference between its baseline cost and the current adjustment cost: In the formula, This represents the payoff of conflict robot i' under the preset adjustment strategy combination S. This represents the baseline adjustment cost corresponding to the conflict robot i'.

[0062] Specifically, for cooperative games, under a preset combination of adjustment strategies, first calculate the sum of the current adjustment costs of all conflicting robots and the sum of the baseline adjustment costs, then calculate the payoff of the cooperative game by the difference between the baseline total cost and the current total cost. In the formula, This represents the current total cost of all conflicting robots under the preset adjustment strategy combination. This represents the baseline total cost of all conflicting robots under the preset adjustment strategy combination.

[0063] Step S108: After adjusting the corresponding conflict robot according to the above optimal adjustment strategy combination, continue the inspection.

[0064] Specifically, after each conflict robot has made adjustments according to the optimal combination of adjustment strategies, it will continue to conduct inspections.

[0065] In a preferred embodiment, after adjusting the corresponding conflict robot according to the above-described optimal adjustment strategy combination and continuing the inspection, the process further includes: Obtain the distribution density characteristics of the above-mentioned sub-regions, the time distribution characteristics of the above-mentioned task groups, the time delay rate of all the above-mentioned power inspection tasks to be scheduled, and the frequency of conflict occurrence. Specifically, the above distribution density characteristics are obtained by statistically analyzing the distribution density of power inspection tasks in all sub-regions; the above time distribution characteristics are obtained by constructing a histogram of time demand distribution for all task groups; the above time extension rate is obtained by comparing the deviation between the actual completion time and the corresponding planned completion time of each power inspection task; and the above conflict frequency is obtained by statistically analyzing the number of conflicts occurring per unit time.

[0066] The aforementioned distribution density characteristics, time distribution characteristics, time series time extension rate, and conflict occurrence frequency are stored as descriptive features of the power inspection tasks to be scheduled. This allows for the calculation of the similarity between each stored descriptive feature and the descriptive features of the new power inspection tasks to be scheduled, and the selection of the corresponding inspection decision for the new power inspection tasks to be scheduled based on the aforementioned similarity.

[0067] Specifically, the distribution density characteristics, time distribution characteristics, time series time extension rate, and conflict occurrence frequency are used as descriptive features of the current power inspection task and stored in a structured manner. When a new power inspection task needs to be optimized, the most similar historical optimization decision can be found by matching the similarity between the two. Then, the inspection plan of these new power inspection tasks can be executed directly according to this historical optimization decision or after fine-tuning this historical optimization decision.

[0068] Preferably, the above process is essentially a meta-task migration mechanism. This mechanism improves the execution efficiency and intelligence of new power inspection plans through experience transfer and rapid adaptation, and has good practical value and wide engineering applicability.

[0069] In this preferred embodiment, the descriptive features of the currently decided power inspection task are stored in a structured manner to facilitate faster decision optimization for subsequent new power inspection tasks.

[0070] Preferably, this invention transforms power line inspection tasks from static planning into a multi-level decision-making process, possessing strong adaptability and scalability. By enhancing the expressive power of the task structure through graph modeling, it achieves precise decomposition of spatial, temporal, and functional dimensions, effectively improving the rationality of task allocation and execution efficiency. The multi-level collaborative tree structure clearly defines the dependencies between tasks and the resource scheduling logic, improving system coordination capabilities. The multi-stage decision adjustment mechanism, combined with conflict prediction and local perception optimization, significantly enhances the safety and dynamic adaptability of path execution.

[0071] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0072] like Figure 2 As shown, an embodiment of the present invention provides an inspection decision optimization device for multiple inspection tasks in a power system, comprising: The system includes a task information acquisition module, an inspection path generation module, an inspection data acquisition module, a conflict probability detection module, a conflict robot data acquisition module, a module adjustment cost calculation module, an optimal strategy determination module, and an inspection decision optimization module. The aforementioned task information acquisition module is used to acquire task information for all power inspection tasks to be scheduled. The aforementioned inspection path generation module is used to generate an initial inspection path and determine the execution robot corresponding to each initial inspection path based on the aforementioned task information. The aforementioned inspection data acquisition module is used to inspect each power inspection task according to the initial inspection path and acquire the real-time status data and real-time map topology data of each executing robot. The aforementioned conflict probability detection module is used to obtain the conflict probability between each execution robot based on the initial inspection path, real-time status data, real-time map topology data, and the preset spatiotemporal conflict detection model. The aforementioned conflict robot data acquisition module is used to acquire conflict robots whose conflict probability exceeds a preset probability threshold, the remaining preset task importance level of the current remaining power inspection task, and the remaining power. The aforementioned adjustment cost calculation module is used to calculate the adjustment cost of each conflicting robot based on the remaining preset task importance level, remaining power, and task execution time information. The aforementioned optimal strategy determination module is used to determine the optimal combination of adjustment strategies based on the aforementioned adjustment cost, several preset combinations of adjustment strategies, and the Nash equilibrium solution method. The inspection decision optimization module is used to adjust the corresponding conflict robot according to the above optimal adjustment strategy combination and then continue the inspection.

[0073] In a preferred embodiment, the task information includes: inspection target, spatial location of the inspection target, execution dependency relationship between each inspection task, task execution time information, and task type; The aforementioned inspection path generation module includes: Spatial clustering unit, task group division unit, robot allocation unit, and initial inspection path generation unit; The aforementioned spatial clustering unit is used to cluster all power inspection tasks according to their spatial locations, resulting in several sub-regions, and assigning a group of execution robots to each sub-region. The aforementioned task group division unit is used to divide each sub-region into several task groups with different levels of urgency based on the task execution time information of each power inspection task within the sub-region. The robot allocation unit described above is used to allocate a corresponding execution robot to each power inspection task in each task group according to the task type and the execution robot group corresponding to each sub-area. The aforementioned initial inspection path generation unit is used to generate an initial inspection path for each execution robot based on the inspection target, spatial location, and execution dependencies of the power inspection task of the execution robot.

[0074] In another preferred embodiment, the task group division unit includes: The system includes a task deadline extraction subunit, a task urgency calculation subunit, a task timeliness index calculation subunit, and a task urgency classification subunit. The aforementioned task deadline extraction subunit is used to extract the task deadline of each power inspection task from the aforementioned task execution time information. The aforementioned task urgency calculation subunit is used to calculate the task urgency of each power inspection task based on the aforementioned task deadline. The aforementioned task timeliness index calculation subunit is used to calculate the task timeliness index based on the aforementioned task urgency and the aforementioned preset task importance level. The aforementioned task urgency classification subunit is used to classify the urgency of each power inspection task according to the aforementioned task timeliness indicators and preset timeliness indicator thresholds, resulting in several task groups with different urgency levels.

[0075] It should be noted that the device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort. The above schematic diagram is merely an example of a power system multi-inspection task inspection decision optimization device and does not constitute a limitation on a power system multi-inspection task inspection decision optimization device. It may include more or fewer components than illustrated, or combine certain components, or use different components.

[0076] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0077] Another embodiment of the present invention provides a terminal device, 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 inspection decision optimization method for multiple inspection tasks in a power system as described in any embodiment of the present invention.

[0078] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the device. The aforementioned terminal devices may be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. These devices may include, but are not limited to, processors and memory. The processor 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. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the device, connecting various parts of the device via various interfaces and lines. The aforementioned memory can be used to store the aforementioned computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned device by running or executing the computer programs and / or modules stored in the aforementioned memory, and by calling data stored in the memory. The aforementioned memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory 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.

[0079] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0080] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the inspection decision optimization method for multiple inspection tasks in a power system as described in any embodiment of the present invention.

[0081] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0082] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for optimizing inspection decisions in a power system with multiple inspection tasks, characterized in that, include: Obtain task information for all pending power inspection tasks; Based on the task information, an initial inspection path is generated and the corresponding execution robot for each initial inspection path is determined. The power inspection tasks are inspected according to the initial inspection path, and the real-time status data and real-time map topology data of each robot are obtained. Based on the initial inspection path, real-time status data, real-time map topology data, and the preset spatiotemporal conflict detection model, the conflict probability between each execution robot is obtained. Acquire conflict robots whose conflict probability exceeds a preset probability threshold, the remaining preset task importance level of the current remaining power inspection tasks, and the remaining power supply. Based on the remaining preset task importance level, remaining battery power, and task execution time information, the adjustment cost of each conflicting robot is calculated. Based on the adjustment cost, several preset adjustment strategy combinations, and the Nash equilibrium solution method, the optimal adjustment strategy combination is determined. After adjusting the corresponding conflict robot according to the optimal adjustment strategy combination, the inspection continues.

2. The method for optimizing inspection decisions for multiple inspection tasks in a power system according to claim 1, characterized in that, The task information includes: the inspection target, the spatial location of the inspection target, the execution dependencies between each inspection task, the task execution time information, and the task type; The step of generating an initial inspection path and determining the execution robot corresponding to each initial inspection path based on the task information includes: Clustering is performed based on the spatial location of all power inspection tasks to obtain several sub-regions, and a group of execution robots is assigned to each sub-region. For each sub-region, based on the task execution time information of each power inspection task within the sub-region, several task groups with different levels of urgency are obtained. Based on the task type and the corresponding robot group for each sub-region, assign a corresponding robot to each power inspection task within each task group. For each execution robot, an initial inspection path is generated based on the inspection target, spatial location, and execution dependencies of the power inspection task.

3. The method for optimizing inspection decisions for multiple inspection tasks in a power system according to claim 2, characterized in that, Based on the task execution timeliness information of each power inspection task within the sub-region, several task groups with different levels of urgency are obtained, including: The task deadline for each power inspection task is extracted from the task execution time information. The urgency of each power inspection task is calculated based on the task deadline. Based on the task urgency and the preset task importance level, the task timeliness index is calculated. Based on the task timeliness index and the preset timeliness index threshold, the urgency of each power inspection task is divided into several task groups with different urgency levels.

4. The method for optimizing inspection decisions for multiple inspection tasks in a power system according to claim 3, characterized in that, The step of obtaining the conflict probability between each execution robot based on the initial inspection path, real-time status data, real-time map topology data, and a preset spatiotemporal conflict detection model includes: Obtain historical conflict data for each execution robot; The historical conflict data, initial inspection path, real-time status data, task execution time information, and real-time map topology data are input into a preset spatiotemporal conflict detection model to obtain the conflict probability.

5. The method for optimizing inspection decisions for multiple inspection tasks in a power system according to claim 4, characterized in that, After adjusting the corresponding conflict robot according to the optimal adjustment strategy combination, and continuing the inspection, the process further includes: Obtain the distribution density characteristics of the several sub-regions, the temporal distribution characteristics of the task group, the time delay rate of all power inspection tasks to be scheduled, and the frequency of conflict occurrence; The distribution density feature, the time distribution feature, the time extension rate, and the frequency of conflict occurrence are stored as descriptive features of the power inspection task to be scheduled. Then, by calculating the similarity between each stored descriptive feature and the descriptive features of the new power inspection task to be scheduled, the corresponding inspection decision is selected for the new power inspection task to be scheduled based on the similarity.

6. A power system multi-inspection task inspection decision optimization device, characterized in that, include: The system includes a task information acquisition module, an inspection path generation module, an inspection data acquisition module, a conflict probability detection module, a conflict robot data acquisition module, a module adjustment cost calculation module, an optimal strategy determination module, and an inspection decision optimization module. The task information acquisition module is used to acquire task information for all power inspection tasks to be scheduled. The inspection path generation module is used to generate an initial inspection path and determine the execution robot corresponding to each initial inspection path based on the task information. The inspection data acquisition module is used to inspect each power inspection task according to the initial inspection path, and acquire the real-time status data and real-time map topology data of each execution robot. The conflict probability detection module is used to obtain the conflict probability between each execution robot based on the initial inspection path, real-time status data, real-time map topology data and preset spatiotemporal conflict detection model. The conflict robot data acquisition module is used to acquire conflict robots whose conflict probability exceeds a preset probability threshold, the remaining preset task importance level of the current remaining power inspection tasks, and the remaining power. The adjustment cost calculation module is used to calculate the adjustment cost of each conflicting robot based on the remaining preset task importance level, remaining power, and task execution time information. The optimal strategy determination module is used to determine the optimal combination of adjustment strategies based on the adjustment cost, several preset combinations of adjustment strategies, and the Nash equilibrium solution method. The inspection decision optimization module is used to adjust the corresponding conflict robot according to the optimal adjustment strategy combination and then continue the inspection.

7. The inspection decision optimization device for multiple inspection tasks in a power system according to claim 6, characterized in that, The task information includes: the inspection target, the spatial location of the inspection target, the execution dependencies between each inspection task, the task execution time information, and the task type; The inspection path generation module includes: Spatial clustering unit, task group division unit, robot allocation unit, and initial inspection path generation unit; The spatial clustering unit is used to cluster all power inspection tasks according to their spatial locations to obtain several sub-regions, and to assign an execution robot group to each sub-region. The task group division unit is used to divide each sub-region into several task groups with different levels of urgency based on the task execution time information of each power inspection task within the sub-region. The robot allocation unit is used to allocate a corresponding execution robot to each power inspection task in each task group according to the task type and the execution robot group corresponding to each sub-area. The initial inspection path generation unit is used to generate an initial inspection path for each execution robot based on the inspection target, spatial location, and execution dependencies of the power inspection task of the execution robot.

8. The inspection decision optimization device for multiple inspection tasks in a power system according to claim 7, characterized in that, The task group division unit includes: The system includes a task deadline extraction subunit, a task urgency calculation subunit, a task timeliness index calculation subunit, and a task urgency classification subunit. The task deadline extraction subunit is used to extract the task deadline of each power inspection task from the task execution time information. The task urgency calculation subunit is used to calculate the task urgency of each power inspection task based on the task deadline. The task timeliness index calculation subunit is used to calculate the task timeliness index based on the task urgency and the preset task importance level. The task urgency classification subunit is used to classify the urgency of each power inspection task according to the task timeliness index and the preset timeliness index threshold, so as to obtain several task groups with different urgency levels.

9. A terminal device, characterized in that, The system includes 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 an inspection decision optimization method for multiple inspection tasks in a power system as described in any one of claims 1 to 5.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the inspection decision optimization method for multiple inspection tasks in a power system as described in any one of claims 1 to 5.

Images