Dynamic cooperative inspection and position adjustment scheduling management method and system based on legged robot

Abstract

The application relates to the technical field of intelligent inspection of power equipment, and particularly discloses a dynamic cooperative inspection position supplement scheduling management method and system based on a foot-type robot, which acquires environment data of a target inspection area, constructs a global inspection scene model based on the environment data, establishes a dynamic wireless communication topology network, performs self-adaptive task division on the global inspection area, and generates a plurality of standardized inspection task units; operation state data of each foot-type robot is acquired, a comprehensive operation capability evaluation model is established, and dynamic inspection task allocation is performed in combination with inspection task attributes of the inspection task units; based on abnormal state rule matching, a nearby robot automatic position supplement mechanism is executed to balance-distribute the inspection task area; a double-layer scheduling mechanism is established based on a cloud scheduling node and an edge scheduling node, and an inspection task knowledge base is established. The application improves the intelligent level, task execution efficiency and complex environment adaptability of the foot-type robot inspection system.

Description

Technical Field

[0001] This invention relates to the field of intelligent inspection technology for power equipment, and more specifically, to a dynamic collaborative inspection and supplementary scheduling management method and system based on a legged robot. Background Technology

[0002] Legged robots are widely used in equipment status inspection and hidden danger investigation in power scenarios such as substations and power transmission channels. At present, the industry generally adopts a scheduling mode of fixed partition and static task allocation for multiple legged robot cluster inspections. That is, the inspection area is divided into fixed blocks in advance, and each legged robot is bound to a dedicated inspection area and inspection point, and completes routine inspections according to the preset route.

[0003] Existing traditional cluster inspection scheduling technologies have many inherent defects, including: fixed partition allocation modes cannot dynamically adjust tasks based on the real-time status of legged robots and the inspection priority of points, easily leading to overlapping inspection points in adjacent areas and duplicate inspections. A large number of invalid inspection jobs significantly reduce the overall inspection efficiency of the cluster, causing unnecessary consumption of equipment computing power, electricity, and communication resources; individual legged robots are prone to hardware failures, posture failures, power depletion, and program crashes during inspections. Furthermore, outdoor power transmission channels and complex substation equipment areas are susceptible to wireless signal obstruction and interference, causing communication loss for legged robots. The scheduling scheme relies heavily on continuous and stable cloud network communication. Once the on-site network is interrupted or the communication link fluctuates, the cloud cannot issue scheduling instructions. The local legged robots have no autonomous scheduling or replacement capabilities, and the cluster inspection work will stop directly, making it impossible to guarantee the continuity of inspection work in harsh working conditions and complex environments. Each legged robot completes the inspection work independently, and the data on hidden dangers, equipment images, and status detection data cannot be shared in real time. Multiple devices are prone to repeatedly identifying and collecting the same hidden dangers and the same equipment status, further aggravating ineffective work. At the same time, it is impossible to form a closed loop of inspection data across the entire domain, which is not conducive to the unified analysis and tracing of power hazards.

[0004] Therefore, it is necessary to provide a dynamic collaborative inspection and replacement scheduling management method and system based on legged robots to solve the above-mentioned technical problems. In order to solve the above problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned shortcomings of existing technologies, this invention provides a dynamic collaborative inspection and replacement scheduling management method and system based on legged robots. This system addresses the industry pain points of existing multi-legged robot inspection scheduling technologies, such as task redundancy and duplication, missed fault detection, network outage and loss of control, and lack of data interoperability, making it difficult to adapt to the large-scale, highly reliable, and fully covered cluster inspection deployment requirements of the power grid.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A dynamic collaborative inspection and replacement scheduling management method based on legged robots includes the following steps: Acquire environmental data of the target inspection area, construct a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network; Based on the full-domain inspection scenario model and dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks to generate several standardized inspection task units. By acquiring the operational status data of each legged robot, a comprehensive operational capability evaluation model for legged robots is established, and dynamic inspection task allocation is carried out in combination with the inspection task attributes of the inspection task unit. By monitoring the inspection status of each legged robot, abnormal status rules are matched, and for legged robots with abnormal status, the inspection task area is evenly distributed based on the automatic replacement mechanism of the nearest neighbor robot. A two-layer scheduling mechanism is established based on cloud scheduling nodes and edge scheduling nodes, and an inspection task knowledge base is established to store the inspection data collected by the legged robot.

[0007] As a further aspect of the present invention, based on the full-domain inspection scenario model and the dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks to generate several standardized inspection task units. The specific steps are as follows: Based on the whole-domain inspection scenario model, spatial feature analysis of the target area is performed to obtain the inspection requirement parameters of the whole-domain inspection area. Adaptive grid clustering is performed based on inspection requirements parameters to perform preliminary inspection grid division for the entire inspection area and determine the size of the inspection grid. By combining dynamic wireless communication topology network optimization and adjustment of inspection grid boundaries, standardized inspection task units are generated.

[0008] As a further aspect of the present invention, by acquiring the operational status data of each legged robot, a comprehensive operational capability evaluation model for the legged robot is established, and dynamic inspection task allocation is performed in conjunction with the inspection task attributes of the inspection task unit. The specific steps are as follows: The system obtains the real-time operating status information of each legged robot in the cluster through a cloud-based scheduling system, and then normalizes the data to form a set of legged robot status data. A comprehensive capability evaluation model is constructed based on the state data set of legged robots, and the comprehensive operational capability evaluation value of each legged robot is output. Extract the inspection task attributes of the inspection task unit, including the spatial location of the inspection task, the priority of the inspection task, the workload of the inspection task, the frequency of the inspection, and the content of the inspection. An inspection task matching model is established based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit. The matching score of the legged robot to perform the inspection task is obtained, and dynamic inspection task allocation is performed according to the matching score.

[0009] As a further aspect of the present invention, based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit, the matching score of the legged robot performing the inspection task is analyzed. The specific steps are as follows: The real-time position data of the legged robot and the spatial position of the inspection task are obtained respectively. The distance between the inspection task position and the current position of the legged robot is calculated as the distance to be moved by the legged robot. Based on the comprehensive operational capability evaluation value, inspection task priority, inspection task workload, and distance to be moved, an inspection task matching model is constructed, and a matching score for the legged robot to perform inspection tasks is output.

[0010] As a further aspect of the present invention, dynamic inspection task allocation is performed based on the matching score, and the specific steps are as follows: For each inspection task, the matching scores of all legged robots are sorted in descending order, and after conditional filtering, a candidate legged robot set is formed based on the sorting results. The inspection tasks are sorted according to their priority, and the legged robot with the highest matching score is selected from the candidate legged robot set based on the task sorting.

[0011] As a further aspect of the present invention, a unit occupancy lock mechanism and a timestamp verification mechanism are employed in the dynamic task allocation process, specifically: Set a unique identifier for each standardized inspection task unit; When a standardized inspection task unit is occupied by a one-legged robot, a unit occupancy lock mechanism is implemented to lock the inspection point and prevent other legged robots from repeating the task. The time stamp verification mechanism is used to determine whether the standardized inspection task unit is within the valid inspection cycle, and the completed standardized inspection task unit is removed from the set of tasks to be inspected.

[0012] As a further aspect of the present invention, by monitoring the inspection status of each legged robot and performing abnormal status rule matching, the inspection task area is evenly allocated based on the automatic replacement mechanism of the nearest neighboring robot for legged robots with abnormal status. The specific steps are as follows: By monitoring the inspection status of each legged robot and matching it with preset abnormal status rules, the legged robots with abnormal status and their corresponding standardized inspection task units that have not completed the inspection are marked. For standardized inspection task units that have not completed their inspections, an automatic replacement mechanism is implemented using the nearest neighboring robot to determine the replacement legged robot and the replacement inspection path.

[0013] As a further aspect of the present invention, the two-layer scheduling mechanism specifically includes a first-layer scheduling mechanism based on cloud scheduling nodes and a second-layer scheduling mechanism based on edge scheduling nodes.

[0014] The dynamic collaborative inspection and replacement scheduling management system based on legged robots includes a scene model building module, a task unit division module, a dynamic allocation module for inspection tasks, an anomaly replacement module, and a two-layer scheduling module. The scenario model building module is used to acquire environmental data of the target inspection area, build a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network. The task unit division module is used to adaptively divide the whole-domain inspection area into tasks based on the whole-domain inspection scenario model and dynamic wireless communication topology network, generating several standardized inspection task units. The inspection task dynamic allocation module is used to obtain the operating status data of each legged robot, establish a comprehensive operation capability evaluation model of the legged robot, and dynamically allocate inspection tasks in combination with the inspection task attributes of the inspection task unit. The anomaly compensation module is used to monitor the inspection status of each legged robot, perform anomaly status rule matching, and for legged robots with anomalies, it uses the nearest neighbor robot automatic compensation mechanism to evenly distribute the inspection task area. The dual-layer scheduling module is used to establish a dual-layer scheduling mechanism based on cloud scheduling nodes and edge scheduling nodes, and to establish an inspection task knowledge base to store the inspection data collected by the legged robot.

[0015] A readable storage medium storing a computer program, which, when executed by a processor, is used to implement the steps of the above-described dynamic collaborative inspection and replacement scheduling management method based on a legged robot.

[0016] The technical effects and advantages of this invention, based on a dynamic collaborative inspection and supplementary scheduling management method and system using legged robots, are as follows: By acquiring environmental data of the target inspection area and establishing a global inspection scenario model, this invention enables precise perception of the spatial structure, equipment distribution, and environmental characteristics of the inspection area, providing a reliable data foundation for subsequent task division and path planning. Combined with a dynamic wireless communication topology network for dynamic grid division of the inspection area, the generated standardized inspection task units can adaptively adjust according to actual environmental requirements, improving the rationality of inspection area division and avoiding task overlap or area omissions. By establishing a comprehensive operational capability evaluation model for legged robots and dynamically allocating tasks based on inspection task attributes, the optimal execution target can be selected according to the battery status, hardware status, mobility, and current load of different legged robots, improving the utilization rate of inspection resources and task execution efficiency.

[0017] This invention utilizes rule matching for abnormal states and an automatic replacement mechanism with the nearest neighboring machine to quickly migrate tasks when a legged robot malfunctions, experiences communication failures, or runs out of power. This avoids inspection interruptions caused by single-machine malfunctions and improves the system's continuous operation capability. A two-layer scheduling mechanism, constructed through cloud and edge scheduling nodes, allows for flexible switching of scheduling modes based on communication status. This enables global optimization management under normal network conditions and autonomous on-site scheduling under weak communication or network outage conditions, enhancing the stability and anti-interference capabilities of the inspection system. Furthermore, by combining historical inspection data, equipment status data, and task execution data with an inspection task knowledge base for storage and analysis, the invention improves the intelligence level, task execution efficiency, and adaptability to complex environments of the legged robot inspection system. Attached Figure Description

[0018] Figure 1 A flowchart illustrating the dynamic collaborative inspection and replacement scheduling management method based on a legged robot provided in an embodiment of the present invention; Figure 2 The diagram below shows a system block diagram of a dynamic collaborative inspection and replacement scheduling management system based on a legged robot, as provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, and not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0020] like Figure 1 The diagram shown is a flowchart of a dynamic collaborative inspection and replacement scheduling management method based on a legged robot provided in an embodiment of the present invention. Figure 1The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. Steps S1 to S5 are detailed as follows: Step S1: Obtain environmental data of the target inspection area, construct a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network; The environmental data includes three-dimensional terrain information of the inspection area, equipment distribution information, inspection area boundary information, communication obstruction area information, and restricted area information. Based on the full-domain inspection scenario model, the locations of equipment to be inspected, high-risk inspection areas, and areas with poor communication are determined. The system collects the communication signal strength, communication link delay, and network connection status of each legged robot in real time, generates a dynamic wireless communication topology network based on the communication relationship between each legged robot, and establishes backup communication links based on areas with communication anomalies. Step S2: Based on the full-domain inspection scenario model and dynamic wireless communication topology network, adaptive task division is performed on the full-domain inspection area to generate several standardized inspection task units. Among them, the size of the inspection grid is adjusted according to the density of equipment, inspection priority and terrain complexity. Each inspection task unit corresponds to at least one inspection point, inspection frequency and inspection content, so that the inspection task unit can be split, merged or migrated according to scheduling needs. Step S3: By acquiring the operating status data of each legged robot, establish a comprehensive operational capability evaluation model for the legged robot, and dynamically allocate inspection tasks in combination with the inspection task attributes of the inspection task unit. The operational status data includes remaining power data, hardware status data, real-time location data, movement speed data, data of completed inspection points, and remaining operational capacity data. Step S4: By monitoring the inspection status of each legged robot, abnormal status rule matching is performed. For legged robots with abnormal status, the inspection task area is evenly allocated based on the automatic replacement mechanism of the nearest neighbor robot. The path planning algorithm generates the corresponding inspection path for the supplementary legged robot, enabling the supplementary legged robot to automatically take over the unfinished inspection tasks corresponding to the abnormal legged robot, and achieve continuous coverage of the inspection area. Step S5: Establish a two-layer scheduling mechanism based on cloud scheduling nodes and edge scheduling nodes, and establish an inspection task knowledge base to store the inspection data collected by the legged robot. Under normal communication conditions, the cloud scheduling node performs tasks such as inspection task division, inspection task allocation, supplementary scheduling, and data aggregation. When cloud communication anomalies or network interruptions are detected, the system switches to the edge scheduling node to perform legged robot status monitoring, inspection task reallocation, and fault compensation scheduling. The inspection data includes equipment image data, hazard characteristic data, status detection data, inspection trajectory data, and point inspection records. Based on historical inspection information in the inspection task knowledge base, the system performs repeated inspections and judgments on the points to be inspected, and enables collaborative sharing of inspection data among multiple legged robots.

[0021] Preferably, based on the full-domain inspection scenario model and the dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks to generate several standardized inspection task units. The specific steps are as follows: Based on the full-domain inspection scenario model, spatial feature analysis is performed on the target area to calculate the inspection requirement parameters of the full-domain inspection area; the inspection requirement parameters include equipment density, inspection priority, and terrain complexity. Adaptive grid clustering is performed based on inspection requirements parameters to perform preliminary inspection grid division for the entire inspection area and determine the size of the inspection grid. By combining dynamic wireless communication topology network optimization and adjustment of inspection grid boundaries, standardized inspection task units are generated.

[0022] In this embodiment of the invention, based on a full-domain inspection scenario model and a dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks, generating several standardized inspection task units. The core objective is to dynamically adjust the inspection area division method according to the inspection environment, equipment distribution, and communication conditions, avoiding problems such as uneven inspection tasks, duplicate paths, and unstable communication caused by traditional fixed area division. The specific implementation process is as follows: Based on the established full-domain inspection scenario model, spatial feature analysis is performed on the target inspection area to obtain multi-dimensional parameters reflecting the area's inspection needs, and the inspection requirement parameters for the entire inspection area are calculated. The full-domain inspection scenario model includes data such as the target area's three-dimensional spatial information, equipment spatial distribution information, inspection point information, obstacle distribution information, road traffic information, and historical inspection records. By performing spatial mapping and feature extraction on the above data, the full-domain inspection area is divided into multiple spatial units to be analyzed, and the corresponding inspection requirement parameters are calculated for each spatial unit.

[0023] Inspection requirements parameters include equipment density, inspection priority, and terrain complexity. Equipment density represents the distribution of equipment to be inspected within a unit of space, and is determined by statistically analyzing the number of devices, device types, and distances between devices within each spatial unit. When an area contains a large number of critical devices or the inspection targets are concentrated, its equipment density is high, requiring the division into smaller inspection grids to improve the inspection accuracy of the legged robot. For areas with fewer devices, the inspection range can be appropriately expanded to improve inspection efficiency.

[0024] Inspection priority is used to indicate the importance of different areas, and is calculated comprehensively based on equipment operating status, historical equipment failure records, equipment importance level, and current operational risks. For example, for important power transmission equipment, core substation equipment, and equipment areas with historical anomalies, the inspection priority of the corresponding areas is increased, so that subsequently generated inspection task units can be preferentially assigned to legged robots with stronger operational capabilities. For ordinary equipment areas, the inspection priority is decreased to reduce the consumption of inspection resources.

[0025] Terrain complexity describes the difficulty of a legged robot's movement when performing inspection tasks in different areas. It is calculated by analyzing factors such as 3D terrain information, slope changes, number of obstacles, length of passable paths, and spatial confinement in the global inspection scene model. When the terrain complexity is high, the difficulty of path planning for the legged robot increases. Therefore, it is necessary to reduce the coverage area of ​​a single inspection grid to avoid a single legged robot bearing excessive movement distance and inspection load.

[0026] Adaptive grid clustering is performed based on equipment density, inspection priority, and terrain complexity to initially divide the entire inspection area into inspection grids. Specifically, an initial grid set is first established based on the spatial boundaries of the target area, and each initial grid is associated with its corresponding inspection requirement parameters to form grid data units containing spatial location, equipment information, and inspection requirement characteristics. Subsequently, cluster analysis is performed based on the similarity of inspection requirement parameters among the grids, grouping areas with similar inspection requirement characteristics into the same inspection grid.

[0027] During the grid clustering process, the inspection grid size is dynamically adjusted based on inspection requirement parameters. When a region simultaneously meets the criteria of high equipment density, high inspection priority, and high terrain complexity, the inspection grid size for that region is reduced, shrinking the coverage area of ​​a single inspection task unit and thus increasing inspection frequency and detection accuracy. Conversely, when a region has fewer devices, lower risk, and open terrain, the inspection grid size is expanded, enabling a single legged robot to complete a larger inspection task in one go, improving overall inspection efficiency. This approach allows the inspection grid size to automatically adjust with environmental changes, rather than using a fixed area division method.

[0028] After initial grid division, the grid boundaries were further optimized and adjusted using a dynamic wireless communication topology network. Since multiple legged robots rely on the wireless communication network for status uploading, task reception, and collaborative scheduling during inspections, communication stability directly impacts the inspection task's performance. Communication adaptability analysis was performed on the initially generated inspection grid by reading node connection relationships, communication signal strength, link latency, and network coverage information from the dynamic wireless communication topology network.

[0029] When a patrol grid is detected to be in an area with poor communication quality, the boundary of that patrol grid is adjusted. For example, some weak communication areas can be divided into adjacent task units with stable communication coverage, or the task range corresponding to the weak communication area can be reduced, so that the legged robot performing the task can maintain effective communication. At the same time, for areas with stable communication coverage and spatial continuity, adjacent patrol grids can be appropriately merged to reduce the number of tasks and improve the scheduling efficiency of the legged robot.

[0030] The embodiments of the present invention realize the transformation of the inspection area from a fixed division to an adaptive division based on environmental conditions, equipment requirements and communication capabilities, so that the generated inspection task units are more in line with actual inspection needs, improve the task balancing ability and collaborative inspection efficiency among multiple legged robots, and provide a reliable data foundation for subsequent dynamic task allocation and automatic fault compensation.

[0031] Preferably, by acquiring the operational status data of each legged robot, a comprehensive operational capability evaluation model for the legged robot is established, and dynamic inspection tasks are allocated in combination with the inspection task attributes of the inspection task units. The specific steps are as follows: The system obtains real-time operational status information of each legged robot within the cluster through a cloud-based scheduling system, and then performs normalization processing to form a set of legged robot status data. ,in, This is the remaining battery power data. This is hardware status data. For real-time location data, For movement speed data, For the completed inspection point data, Data on remaining operational capacity; A comprehensive capability evaluation model is constructed based on the state data set of legged robots, and the comprehensive operational capability evaluation value of each legged robot is output. The specific formula of the comprehensive capability evaluation model is as follows: ; In the formula: The comprehensive operational capability evaluation value of legged robot i. The weighting coefficients for the remaining battery power data. These are the weighting coefficients for the hardware status data. The weighting coefficients for real-time location data. The weighting coefficients for the movement speed data, The weighting coefficients for the completed inspection point data. These are the weighting coefficients for the remaining operational capacity data; Extract the inspection task attributes of the inspection task unit, including the spatial location of the inspection task, the priority of the inspection task, the workload of the inspection task, the frequency of the inspection, and the content of the inspection. An inspection task matching model is established based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit. The matching score of the legged robot to perform the inspection task is obtained, and dynamic inspection task allocation is performed according to the matching score.

[0032] Preferably, based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit, the matching score of the legged robot performing the inspection task is analyzed. The specific steps are as follows: The real-time position data of the legged robot and the spatial position of the inspection task are obtained respectively. The distance between the inspection task position and the current position of the legged robot is calculated as the distance to be moved by the legged robot. Based on the comprehensive operational capability evaluation value, inspection task priority, inspection task workload, and distance to be moved, an inspection task matching model is constructed, outputting a matching score for the legged robot's execution of inspection tasks; the calculation formula for the inspection task matching model is: ; In the formula: Matching score for legged robot i performing inspection task j. The weighting coefficients for the overall performance evaluation value. The comprehensive operational capability evaluation value of legged robot i. This is a weighting coefficient for the priority of inspection tasks. The priority of the j-th inspection task is... This is a weighting coefficient for the workload of inspection tasks. For the workload of the j-th inspection task, For the weighting coefficient of the distance to be moved by the legged robot, The distance to be moved when legged robot i performs inspection task j.

[0033] Dynamic inspection task allocation is performed based on the matching score. The specific steps are as follows: Inspection tasks are assigned based on matching scores. Specifically, for each inspection task, the matching scores of all legged robots are sorted in descending order, and after conditional filtering, a set of candidate legged robots is formed based on the sorting results. The conditional filtering criteria are: the comprehensive operation capability evaluation value reaches a preset threshold; the remaining battery power of the legged robot meets the task execution requirements; the distance between the legged robot and the task location meets the path constraints; and the current task load does not exceed the maximum carrying capacity. The inspection tasks are sorted according to their priority, and the legged robot with the highest matching score is selected from the candidate legged robot set based on the task sorting.

[0034] In addition, a unit occupancy lock mechanism and a timestamp verification mechanism are used in the dynamic task allocation process; among which: Set a unique identifier for each standardized inspection task unit; When a standardized inspection task unit is occupied by a one-legged robot, a unit occupancy lock mechanism is implemented to lock the inspection point and prevent other legged robots from repeating the task. The time stamp verification mechanism is used to determine whether the standardized inspection task unit is within the valid inspection cycle, and the completed standardized inspection task unit is removed from the set of tasks to be inspected.

[0035] In this embodiment of the invention, the operational status information of each legged robot in the cluster is acquired in real time through a cloud-based scheduling system. The acquired data is then preprocessed and normalized to form a set of legged robot status data. Since different types of operational status data have different data dimensions, it is necessary to standardize and transform the collected multidimensional status data so that all status parameters are mapped to a unified evaluation range.

[0036] After constructing the state data set, a comprehensive capability evaluation model for legged robots is built based on multi-dimensional state parameters to quantitatively calculate the current comprehensive operational capability of each legged robot. The comprehensive capability evaluation model integrates energy status, equipment reliability, spatial position, motion capability, task completion status, and remaining operational capability by setting weight coefficients corresponding to different state parameters, and outputs a comprehensive operational capability evaluation value for each legged robot.

[0037] The real-time operating status of different legged robots is converted into a unified evaluation index by a comprehensive operational capability evaluation value. For example, a legged robot with high remaining battery power, good hardware condition, fast movement speed, and low current task load has a high comprehensive operational capability evaluation value and is suitable for undertaking inspection tasks that are far away, have a large workload, or have a high priority; while a legged robot with low battery power, hardware malfunction, or a large current task load will have its task allocation priority reduced.

[0038] Extract the inspection task attributes corresponding to each inspection task unit. These attributes include the spatial location of the inspection task, its priority, workload, frequency, and content. Specifically, the spatial location describes the area where the task is located, providing a basis for calculating the legged robot's movement distance; the priority indicates the task's importance, ensuring that important equipment and high-risk areas are inspected first; the workload describes the difficulty of task execution, including the number of inspection points, the amount of data collected, and the estimated execution time; the frequency indicates the task's execution cycle; and the content describes the types of data that the task needs to collect.

[0039] After obtaining the comprehensive operational capability evaluation value and inspection task attributes of the legged robot, an inspection task matching model is established to calculate the suitability of each legged robot for different inspection tasks. Real-time position data of the legged robot and spatial position data of the inspection task are acquired. The distance between the current position of the legged robot and the target inspection task position is determined using a spatial distance calculation method, and this distance is used as the distance the legged robot needs to move when performing the task. Based on the comprehensive operational capability evaluation value, inspection task priority, inspection task workload, and the distance to be moved, a task matching model is constructed to calculate the matching score between the legged robot and the inspection task. This matching model ensures that legged robots with higher comprehensive capabilities, closer proximity to the task area, and the ability to handle high workloads receive higher matching scores, while avoiding time waste and energy consumption due to excessive movement distances.

[0040] After obtaining the matching scores between all legged robots and inspection tasks, dynamic inspection task allocation is performed based on the matching scores. Specifically, for each inspection task to be executed, the matching scores corresponding to all legged robots are obtained and sorted in descending order according to the score size to form a candidate list of legged robots.

[0041] Subsequently, the sorted legged robots were screened based on certain criteria to form a candidate legged robot set. The screening criteria included: the comprehensive operational capability evaluation value reaching a preset threshold, meaning the legged robot possessed the basic ability to complete the inspection task; remaining battery power meeting the task execution requirements, meaning the legged robot could complete the task and had the ability to return or continue inspection; the distance between the current position of the legged robot and the task position meeting path constraints to avoid affecting inspection efficiency due to excessive movement distance; and the current task load not exceeding the maximum carrying capacity to avoid task congestion caused by a single legged robot undertaking too many tasks.

[0042] After the candidate set is filtered, the tasks to be assigned are sorted according to the priority of the inspection tasks. For high-priority inspection tasks, the legged robot with the highest matching score is selected from the candidate legged robot set for execution; for low-priority tasks, the legged robot with higher remaining capacity is selected for execution while ensuring task completion, thereby achieving a reasonable allocation of inspection resources.

[0043] In addition, to prevent multiple legged robots from repeatedly executing the same inspection task, a unit occupancy lock mechanism and a timestamp verification mechanism are introduced during the dynamic task allocation process. A unique identifier is assigned to each standardized inspection task unit. When a legged robot acquires and begins executing an inspection task unit, that task unit is locked, preventing other legged robots from requesting the task again.

[0044] The task allocation time, execution time, and completion time are recorded through a timestamp verification mechanism to determine whether the current inspection task unit is within a valid inspection cycle. When a task unit completes its inspection, the corresponding task status is updated to "completed" and it is removed from the set of tasks to be inspected. When a task unit exceeds the valid inspection cycle or an abnormality occurs during the execution of the legged robot, the task lock is released, allowing the task to re-enter the dynamic allocation process.

[0045] This invention implements dynamic inspection task allocation based on real-time capability perception and task attribute analysis of legged robots, enabling legged robots with different capability levels to undertake the inspection tasks with the highest matching degree. At the same time, it combines task locking and time verification mechanisms to avoid duplicate inspections, thereby improving the task balance, autonomous collaboration and overall inspection efficiency in the multi-legged robot cluster inspection process.

[0046] Preferably, by monitoring the inspection status of each legged robot, abnormal status rule matching is performed. For legged robots with abnormal status, the inspection task area is evenly allocated based on the automatic replacement mechanism of the nearest neighbor robot. The specific steps are as follows: By monitoring the inspection status of each legged robot and matching it with preset abnormal status rules, the legged robots with abnormal status and their corresponding standardized inspection task units that have not completed the inspection are marked. The inspection status includes posture status, hardware status, power status and communication status. The preset abnormal status rules include posture failure, hardware failure, shutdown due to low power, and communication failure.

[0047] For standardized inspection task units that have not completed their inspections, an automatic replacement mechanism using the nearest neighboring robot is implemented to determine the replacement legged robot and the replacement inspection path. The replacement objectives of the automatic replacement mechanism using the nearest neighboring robot include minimizing the additional mileage of the legged robot and minimizing power consumption.

[0048] This invention monitors the inspection status of each legged robot, identifies anomalies using preset abnormal status rules, and automatically fills in the gaps for incomplete inspection tasks corresponding to the abnormal legged robot by using the nearest neighboring robot, thus achieving dynamic and balanced allocation of inspection task areas. The inspection status information of each legged robot in the cluster is obtained in real time through cloud scheduling nodes or edge scheduling nodes. The inspection status includes posture status, hardware status, battery status, and communication status. Specifically, posture status is used to determine whether the current movement posture of the legged robot is stable, such as whether there is tipping, joint imbalance, or inability to move normally; hardware status is used to detect whether there are any abnormalities in the actuators, inspection sensors, and control modules of the legged robot; battery status is used to determine whether the remaining energy of the legged robot is sufficient to perform the current inspection task; and communication status is used to detect whether the data transmission between the legged robot and the scheduling node is normal.

[0049] The acquired inspection status data is matched with preset abnormal status rules to identify legged robots with abnormal operating conditions. Abnormal status rules include posture failure rules, hardware failure rules, low battery shutdown rules, and communication failure rules. When a legged robot is detected to meet any of these abnormal rule conditions, it is marked as an abnormal legged robot, and the information of its corresponding standardized inspection task unit is further retrieved. By analyzing the task execution record of the legged robot, completed and incomplete inspection task areas are distinguished. Replacement task requests are generated only for standardized inspection task units that have not completed inspections, and the location coordinates, inspection priority, remaining workload, and inspection content of the corresponding task area are recorded, providing a basis for subsequent automatic replacement.

[0050] For incomplete inspection tasks generated by abnormal legged robots, an automatic replacement mechanism using the nearest neighboring robot is activated. Candidate legged robots are selected from the set of currently online and normal legged robots, excluding those with abnormal states, insufficient power, or excessive workload. Based on the spatial distance between the current position of each candidate legged robot and the inspection task area to be replaced, the additional travel distance required for each candidate legged robot to perform the replacement task is calculated. A comprehensive evaluation of the candidate legged robots is then conducted, taking into account their remaining power, expected energy consumption, and task execution capabilities.

[0051] The shortest possible travel distance and lowest power consumption are used as the criteria for replacement robots. Prioritized are legged robots that are closest to the abnormal task area and have sufficient remaining operational capacity. After determining the replacement robot, the replacement inspection path is replanned based on its current location, the distribution of unfinished inspection points, and environmental constraints, enabling it to quickly reach the target area and complete the remaining inspection tasks. Once the replacement robot is determined, the unfinished standardized inspection task units corresponding to the abnormal robot are migrated, their execution status is updated, and a task association is established between the abnormal robot and the replacement robot. Simultaneously, the task load of other legged robots in the cluster is reassessed. If the replacement robot's workload exceeds a preset load threshold, further task splitting or secondary scheduling is performed to achieve a balanced allocation of inspection areas and task resources.

[0052] The embodiments of the present invention can realize automatic detection of abnormal states of legged robots, rapid location of unfinished task areas, and autonomous replacement scheduling of nearby legged robots, so that the inspection task will not be interrupted due to the failure of a single legged robot, thereby improving the continuous operation capability and fault recovery capability of the multi-legged robot collaborative inspection system.

[0053] Preferably, a two-layer scheduling mechanism is established based on cloud scheduling nodes and edge scheduling nodes; the two-layer scheduling mechanism includes a first-layer scheduling mechanism based on cloud scheduling nodes and a second-layer scheduling mechanism based on edge scheduling nodes; The first-level scheduling mechanism is as follows: under normal communication conditions, the cloud scheduling node performs the division of inspection tasks, allocation of inspection tasks, supplementary scheduling, and data aggregation. The two-layer scheduling mechanism is as follows: when cloud communication abnormalities or network interruptions are detected, the system switches to the edge scheduling node to perform legged robot status monitoring, inspection task reallocation, and fault compensation scheduling.

[0054] This invention establishes a two-layer scheduling mechanism based on cloud-based scheduling nodes and edge-based scheduling nodes. By combining global cloud-based scheduling with autonomous edge-based scheduling, it achieves stable collaborative operation of multiple legged robots in complex inspection environments. The two-layer scheduling mechanism includes a first-layer scheduling mechanism based on cloud-based scheduling nodes and a second-layer scheduling mechanism based on edge-based scheduling nodes. The corresponding scheduling level is dynamically selected according to the communication status, ensuring that inspection tasks can be continuously executed in both normal and abnormal communication environments.

[0055] Under normal communication conditions, a single-layer scheduling mechanism based on cloud-based scheduling nodes is employed. The cloud-based scheduling nodes acquire real-time operational status data, inspection task execution status, and environmental data of the inspection area uploaded by each legged robot. Combined with a comprehensive inspection scenario model, the target inspection area is analyzed, and the inspection area is dynamically divided based on device distribution, inspection priority, and task workload, generating corresponding standardized inspection task units. The cloud-based scheduling nodes calculate the matching relationship between different legged robots and inspection tasks based on the comprehensive operational capability evaluation value of each legged robot and the inspection task attributes, selecting the appropriate legged robot to perform the corresponding task, thus achieving dynamic allocation of inspection tasks. During inspection execution, if a legged robot is detected to have abnormal posture, hardware failure, insufficient power, or communication anomalies, the cloud-based scheduling nodes, based on the unfinished inspection task area corresponding to the abnormal legged robot and considering the location, remaining operational capacity, and task load of other online legged robots, perform supplementary scheduling to ensure continuous task execution even in the event of a failure. The cloud-based scheduling node aggregates and processes all inspection data collected by the legged robots, including inspection image data, equipment status data, task completion status, and anomaly records, providing data support for subsequent inspection optimization.

[0056] Upon detecting cloud communication anomalies or network interruptions, the system automatically switches to a two-layer scheduling mechanism based on edge scheduling nodes. These edge scheduling nodes, deployed at the inspection site, possess local data processing and autonomous task decision-making capabilities, replacing the cloud in performing basic scheduling functions when normal cloud communication is unavailable. Specifically, the edge scheduling nodes first monitor the status of legged robots in the current area in real time, acquiring information such as the robot's location, battery level, hardware status, communication status, and task execution progress. They then re-analyze the remaining inspection task units based on locally cached inspection task information. Subsequently, the edge scheduling nodes reallocate incomplete inspection tasks based on the legged robots' current operating capabilities and remaining task requirements, selecting legged robots that meet the task execution conditions to continue performing the inspection tasks. When a legged robot malfunctions, the edge scheduling nodes do not need to wait for a cloud response; they can directly perform fault replacement scheduling based on the location relationships of neighboring legged robots, remaining battery power, and current task load, determining a replacement legged robot and generating a replacement inspection path to ensure continuous coverage of the inspection area.

[0057] Through the aforementioned two-layer scheduling mechanism, this embodiment of the invention achieves a collaborative mode of centralized cloud-based optimized scheduling under normal communication conditions and autonomous rapid response at the edge under abnormal communication conditions. This mechanism reduces the impact of network fluctuations on the execution of inspection tasks, improves the autonomous operation capability of legged robot swarms in complex environments, weak communication areas, and temporary network outages, and enables inspection tasks to have higher continuity, reliability, and fault resistance.

[0058] Example 2: A dynamic collaborative inspection and replacement scheduling management system based on a legged robot, including a scene model building module, a task unit division module, a dynamic allocation module for inspection tasks, an anomaly replacement module, and a two-layer scheduling module; The scenario model building module is used to acquire environmental data of the target inspection area, build a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network. The task unit division module is used to adaptively divide the whole-domain inspection area into tasks based on the whole-domain inspection scenario model and dynamic wireless communication topology network, generating several standardized inspection task units. The inspection task dynamic allocation module is used to obtain the operating status data of each legged robot, establish a comprehensive operation capability evaluation model of the legged robot, and dynamically allocate inspection tasks in combination with the inspection task attributes of the inspection task unit. The anomaly compensation module is used to monitor the inspection status of each legged robot, perform anomaly status rule matching, and for legged robots with anomalies, it uses the nearest neighbor robot automatic compensation mechanism to evenly distribute the inspection task area. The dual-layer scheduling module is used to establish a dual-layer scheduling mechanism based on cloud scheduling nodes and edge scheduling nodes, and to establish an inspection task knowledge base to store the inspection data collected by the legged robot.

[0059] like Figure 2 The diagram shown is a system block diagram of a dynamic collaborative inspection and replacement scheduling management system based on a legged robot, according to an embodiment of the present invention. This system can be used to execute... Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0060] A readable storage medium storing a computer program, which, when executed by a processor, is used to implement the steps of the dynamic collaborative inspection and replacement scheduling management method based on a legged robot as described in any of the above claims.

[0061] The readable storage medium can be a computer storage medium or a communication medium. A communication medium includes any medium that facilitates the transfer of computer programs from one location to another. A computer storage medium can be any available medium accessible to a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application-Specific Integrated Circuit (ASIC). Alternatively, the ASIC can be located in a user equipment. Of course, the processor and the readable storage medium can also exist as discrete components in a communication device. The readable storage medium can be a read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0062] The present invention also provides a program product including executable instructions stored in a readable storage medium. At least one processor of the device can read the executable instructions from the readable storage medium, and the at least one processor executes the executable instructions to cause the device to implement the methods provided in the various embodiments described above.

[0063] In the embodiments of the above-described device, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly manifested as execution by a hardware processor, or execution by a combination of hardware and software modules within the processor.

[0064] Through the above embodiments, this invention, by acquiring environmental data of the target inspection area and establishing a global inspection scenario model, can achieve accurate perception of the spatial structure, equipment distribution, and environmental characteristics of the inspection area, providing a reliable data foundation for subsequent task division and path planning. By combining a dynamic wireless communication topology network for dynamic grid division of the inspection area, the generated standardized inspection task units can be adaptively adjusted according to actual environmental requirements, improving the rationality of inspection area division and avoiding task overlap or area omissions. By establishing a comprehensive operational capability evaluation model for legged robots and dynamically allocating tasks based on inspection task attributes, the optimal execution target can be selected according to the battery status, hardware status, mobility, and current load of different legged robots, improving the utilization rate of inspection resources and task execution efficiency.

[0065] This invention, through abnormal state rule matching and automatic replacement mechanism with the nearest neighboring machine, can quickly complete task migration when the legged robot malfunctions, experiences communication abnormalities, or lacks energy, avoiding inspection interruptions caused by single-machine failures and improving the system's continuous operation capability. A two-layer scheduling mechanism is constructed through cloud scheduling nodes and edge scheduling nodes, enabling flexible switching of scheduling modes based on communication status. This achieves global optimization management in normal network environments and on-site autonomous scheduling in weak communication or network outage situations, enhancing the stability and anti-interference capability of the inspection system. By combining an inspection task knowledge base to store and analyze historical inspection data, equipment status data, and task execution data, the intelligence level, task execution efficiency, and adaptability to complex environments of the legged robot inspection system are improved. The above descriptions are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0066] In conclusion, the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic collaborative inspection and replacement scheduling management method based on legged robots, characterized in that, Includes the following steps: Acquire environmental data of the target inspection area, construct a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network; Based on the full-domain inspection scenario model and dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks to generate several standardized inspection task units. By acquiring the operational status data of each legged robot, a comprehensive operational capability evaluation model for legged robots is established, and dynamic inspection task allocation is carried out in combination with the inspection task attributes of the inspection task unit. By monitoring the inspection status of each legged robot, abnormal status rules are matched, and for legged robots with abnormal status, the inspection task area is evenly distributed based on the automatic replacement mechanism of the nearest neighbor robot. A two-layer scheduling mechanism is established based on cloud scheduling nodes and edge scheduling nodes, and an inspection task knowledge base is established to store the inspection data collected by the legged robot.

2. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 1, characterized in that, Based on the full-domain inspection scenario model and dynamic wireless communication topology network, the full-domain inspection area is adaptively divided into tasks to generate several standardized inspection task units. The specific steps are as follows: Based on the whole-domain inspection scenario model, spatial feature analysis of the target area is performed to obtain the inspection requirement parameters of the whole-domain inspection area. Adaptive grid clustering is performed based on inspection requirements parameters to perform preliminary inspection grid division for the entire inspection area and determine the size of the inspection grid. By combining dynamic wireless communication topology network optimization and adjustment of inspection grid boundaries, standardized inspection task units are generated.

3. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 1, characterized in that, By acquiring the operational status data of each legged robot, a comprehensive operational capability evaluation model for legged robots is established. Dynamic inspection task allocation is then performed by combining the inspection task attributes of the inspection task units. The specific steps are as follows: The system obtains the real-time operating status information of each legged robot in the cluster through a cloud-based scheduling system, and then normalizes the data to form a set of legged robot status data. A comprehensive capability evaluation model is constructed based on the state data set of legged robots, and the comprehensive operational capability evaluation value of each legged robot is output. Extract the inspection task attributes of the inspection task unit, including the spatial location of the inspection task, the priority of the inspection task, the workload of the inspection task, the frequency of the inspection, and the content of the inspection. An inspection task matching model is established based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit. The matching score of the legged robot to perform the inspection task is obtained, and dynamic inspection task allocation is performed according to the matching score.

4. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 3, characterized in that, Based on the comprehensive operational capability evaluation value and the inspection task attributes of the inspection task unit, the matching score of the legged robot performing the inspection task is analyzed. The specific steps are as follows: The real-time position data of the legged robot and the spatial position of the inspection task are obtained respectively. The distance between the inspection task position and the current position of the legged robot is calculated as the distance to be moved by the legged robot. Based on the comprehensive operational capability evaluation value, inspection task priority, inspection task workload, and distance to be moved, an inspection task matching model is constructed, and a matching score for the legged robot to perform inspection tasks is output.

5. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 4, characterized in that, Dynamic inspection task allocation is performed based on the matching score. The specific steps are as follows: For each inspection task, the matching scores of all legged robots are sorted in descending order, and after conditional filtering, a candidate legged robot set is formed based on the sorting results. The inspection tasks are sorted according to their priority, and the legged robot with the highest matching score is selected from the candidate legged robot set based on the task sorting.

6. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 5, characterized in that, The dynamic task allocation process employs a unit occupancy lock mechanism and a timestamp verification mechanism, specifically: Set a unique identifier for each standardized inspection task unit; When a standardized inspection task unit is occupied by a one-legged robot, a unit occupancy lock mechanism is implemented to lock the inspection point and prevent other legged robots from repeating the task. The time stamp verification mechanism is used to determine whether the standardized inspection task unit is within the valid inspection cycle, and the completed standardized inspection task unit is removed from the set of tasks to be inspected.

7. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 1, characterized in that, By monitoring the inspection status of each legged robot and performing abnormal status rule matching, the inspection task area is evenly allocated based on the nearest neighbor robot automatic replacement mechanism for legged robots with abnormal status. The specific steps are as follows: By monitoring the inspection status of each legged robot and matching it with preset abnormal status rules, the legged robots with abnormal status and their corresponding standardized inspection task units that have not completed the inspection are marked. For standardized inspection task units that have not completed their inspections, an automatic replacement mechanism is implemented using the nearest neighboring robot to determine the replacement legged robot and the replacement inspection path.

8. The dynamic collaborative inspection and replacement scheduling management method based on a legged robot according to claim 1, characterized in that, The two-layer scheduling mechanism consists of a first-layer scheduling mechanism based on cloud scheduling nodes and a second-layer scheduling mechanism based on edge scheduling nodes.

9. A dynamic collaborative inspection and replacement scheduling management system based on a legged robot, applied to the dynamic collaborative inspection and replacement scheduling management method based on a legged robot as described in any one of claims 1-8, characterized in that, It includes a scene model building module, a task unit division module, a dynamic allocation module for inspection tasks, an anomaly compensation module, and a two-layer scheduling module; The scenario model building module is used to acquire environmental data of the target inspection area, build a full-domain inspection scenario model based on the environmental data, and establish a dynamic wireless communication topology network. The task unit division module is used to adaptively divide the whole-domain inspection area into tasks based on the whole-domain inspection scenario model and dynamic wireless communication topology network, generating several standardized inspection task units. The inspection task dynamic allocation module is used to obtain the operating status data of each legged robot, establish a comprehensive operation capability evaluation model of the legged robot, and dynamically allocate inspection tasks in combination with the inspection task attributes of the inspection task unit. The anomaly compensation module is used to monitor the inspection status of each legged robot, perform anomaly status rule matching, and for legged robots with anomalies, it uses the nearest neighbor robot automatic compensation mechanism to evenly distribute the inspection task area. The dual-layer scheduling module is used to establish a dual-layer scheduling mechanism based on cloud scheduling nodes and edge scheduling nodes, and to establish an inspection task knowledge base to store the inspection data collected by the legged robot.

10. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it is used to implement the steps of the dynamic collaborative inspection and replacement scheduling management method based on a legged robot as described in any one of claims 1-8.

Images