Multi-robot dog dynamic inspection scheduling method and system in electric energy meter calibration scene

Abstract

This invention discloses a dynamic inspection scheduling method and system for multiple robotic dogs in the context of electricity meter verification, belonging to the field of inspection scheduling technology. This invention adapts to the characteristics of this scenario, such as the coexistence of routine inspection tasks and meter retrieval tasks, the need for continuous and interconnected execution of some tasks, and varying requirements for the robotic dog's load capacity and external equipment for different tasks. It constructs a dynamic inspection scheduling method for multiple robotic dogs, enabling unified coordination of the task execution process of multiple inspection robotic dogs, improving the overall execution efficiency of inspection and anomaly handling tasks, and enhancing the robotic dog system's ability to continuously guarantee inspections and rapidly handle anomalies in the electricity meter verification pipeline. A polling-based dynamic scheduling mechanism is adopted to quickly update the scheduling results when system states change, such as new tasks, task cancellations, robotic dogs returning to idle after completing tasks, and robotic dogs completing charging. This enables timely replanning of the multi-robotic dog task sequence, ensuring the stable operation of the electricity meter verification pipeline.

Description

Technical Field

[0001] This invention belongs to the field of inspection and scheduling technology, and in particular relates to a method and system for dynamic inspection and scheduling of multiple robot dogs in the scenario of electricity meter calibration. Background Technology

[0002] As a critical metering device in the power grid system, the accuracy of electricity meters directly affects the reliability of electricity metering and settlement. Newly manufactured electricity meters and recycled meters must undergo comprehensive functional verification on an electricity meter verification line to ensure the metering accuracy of newly manufactured meters. Simultaneously, the condition of recycled meters is assessed to determine their reusability, improving their utilization rate. With the increasing automation of the electricity meter verification process, the operational efficiency of the verification line has significantly improved. However, this high efficiency also places higher demands on the timeliness of daily inspections and handling of anomalies. Currently, the electricity meter verification line requires employees to regularly inspect the equipment at various points along the line to ensure its normal operation, and to handle unexpected anomalies as needed, such as retrieving unqualified meters or fixing line blockages. Such anomalies usually occur dynamically during the operation of the production line and, in some cases, have high requirements for processing timeliness. If they are not handled in time, they may cause local equipment to stop, abnormal connection between upstream and downstream processes, or even a decrease in the operating efficiency of the entire production line.

[0003] With their excellent mobility, adaptability to complex environments, and ability to carry various external devices, inspection robot dogs can perform tasks such as equipment inspection, anomaly identification, and return of unqualified meters in electricity meter calibration workshops. They have become an important technological tool for replacing manual inspections, improving the operational monitoring capabilities and anomaly response efficiency of calibration lines. However, applying multiple inspection robot dogs to dynamic inspection task processing scenarios in electricity meter calibration lines still faces a series of problems that urgently need to be solved: First, in the meter calibration pipeline, inspection tasks are significantly dynamic. Besides the pre-known routine inspection tasks, additional inspection tasks are generated intermittently during system operation, such as picking up non-compliant meters, triggering new inspection tasks based on the latest collected inspection data, or confirming anomalies. Therefore, the set of tasks to be executed is not statically given but changes in real time with the pipeline's operational status. Existing scheduling methods oriented towards static task sets struggle to respond promptly to such dynamic tasks.

[0004] Secondly, the different task types in meter verification scenarios exhibit significant differences, leading to more complex scheduling constraints. Some routine inspection tasks only require the inspection robot to go to a designated location for status checks or data collection, while other meter retrieval tasks involve both picking up and retrieving meters. Especially for the task of retrieving faulty meters, after the robot completes the retrieval, it also triggers a subsequent task of returning the retrieved meters to the meter recycling station, and this subsequent task must be completed consecutively by the same robot. Furthermore, due to the limited carrying space of the inspection robot and the fact that different task types may require the robot to carry different external devices, the scheduling process not only needs to determine whether the robot's current external devices meet the task requirements and include its reloading behavior at the transfer station in the scheduling considerations, but also needs to determine whether its current remaining load space is sufficient to complete the subsequent task. When the robot is already carrying a certain number of meters and the remaining load space is insufficient, it may need to go to the meter recycling station to return the already carried meters before executing the task. Therefore, the task allocation process is no longer a simple matching problem between robot dogs and tasks, but requires further coordination and scheduling by combining clothing change decisions, return decisions, and task execution paths, which significantly increases the complexity of scheduling solutions.

[0005] Furthermore, the highly efficient and continuous operation of the meter calibration pipeline results in varying timeliness requirements for different tasks. For example, non-compliant meters need to be picked up before a given time limit, otherwise it will affect the calibration efficiency of subsequent meters; data collection for certain inspection tasks needs to be completed within a specified time window, otherwise it will affect subsequent status judgment or anomaly handling. Therefore, task scheduling in this scenario must consider not only whether the task can be completed, but also whether it can be completed within its corresponding time limit. Scheduling methods based solely on shortest distance or simple priority are usually insufficient to accommodate the time window constraints of different tasks.

[0006] Finally, the power management of the robot dog in the electricity meter verification scenario is even more complex. Due to the limited battery capacity of the robot dog, even when fully charged, it is usually insufficient to support the continuous completion of all tasks during long-term operation of the verification pipeline. Therefore, the scheduling process of the robot dog must simultaneously consider task execution and charging planning. Furthermore, since different tasks require the robot dog to carry different external devices, it may also be carrying different numbers of electricity meters during execution. These factors all affect the actual power consumption assessment of the robot dog. If the power consumption of the robot dog in different states cannot be accurately estimated, and charging decisions cannot be considered in conjunction with the task execution path, it is easy for the robot dog to run out of power during task execution, affecting task continuity and the feasibility of the scheduling plan.

[0007] From the perspective of existing related technologies, one type of method mainly relies on static rules or local priorities for task scheduling. Although the computation speed is fast, it is usually only applicable to scenarios with relatively stable task sets and simple task constraints. It is difficult to handle problems such as dynamic task arrival, significant differences in task types, different task time window requirements, and real-time changes in robot dog status in the electricity meter verification pipeline. Another type of method focuses on path planning itself, usually only considering the robot's travel path from the starting point to the destination, without further considering the scheduling problems that may be involved before and after task execution, such as electricity meter return, external equipment replacement, and mid-journey charging. Therefore, it is difficult to adapt to the robot dog capacity constraints and external equipment matching constraints caused by differences in task types in the electricity meter verification scenario. A third type of method, although considering the robot dog's power management and charging strategy, often treats charging as an additional process independent of task scheduling, failing to integrate charging pile occupancy, task node access order, task execution timeliness, and the robot dog's current load and external equipment status into the same scheduling framework.

[0008] Therefore, it is necessary to propose a new dynamic scheduling method for multiple inspection robot dogs in the electricity meter verification pipeline scenario. This method should be able to adapt to the characteristics of dynamic task arrival, heterogeneous task types, differences in task timeliness, and continuous changes in robot dog status in this scenario. It should comprehensively solve the problems of task allocation, path planning, and power management for multiple inspection robot dogs, thereby improving the efficiency and stability of robot dog inspection scheduling in the electricity meter verification scenario. Summary of the Invention

[0009] The technical problem this invention aims to solve is: addressing the challenges of dynamically generated inspection and anomaly handling tasks in a power meter calibration pipeline scenario, including significant differences in task types, inconsistent task timeliness requirements, and continuous changes in the power, load, and external device status of the inspection robot. This invention provides a dynamic multi-robot inspection scheduling method for power meter calibration scenarios. This method adapts to the characteristics of this scenario, such as the coexistence of routine inspection tasks and meter retrieval tasks, the need for continuous and interconnected execution of some tasks, and varying requirements for robot load space and external devices across different tasks. It comprehensively solves the collaborative scheduling problems of multiple robot dogs in task allocation, task execution path planning, and power management, thereby improving the continuous inspection capability and rapid anomaly handling capability of the multi-robot system in the power meter calibration pipeline. It reduces the efficiency impact on the calibration process caused by untimely scheduling responses, task interruptions, or mismatched robot status, ensuring the efficient and stable operation of the power meter calibration pipeline.

[0010] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A dynamic inspection and scheduling system for multiple robot dogs in the context of electricity meter verification includes a path planning module, a status monitoring module, a task scheduling module, and a device communication module. The path planning module is used to build a static shortest path library, the status monitoring module is used to collect and identify system status changes in real time, the task scheduling module is used to perform task matching and robot dog scheduling calculations, and the device communication module is used to issue scheduling instructions and send and receive data.

[0011] A method for dynamic inspection scheduling of multiple robotic dogs in an energy meter verification scenario, applied to a dynamic inspection scheduling system for multiple robotic dogs in an energy meter verification scenario, includes the following steps: Step S1: After the system starts, it first initializes the polling time, obtains the static scene information of the electricity meter calibration pipeline, and uses the global shortest path planning algorithm to build a static shortest path library; Step S2: Update the next polling time based on the preset polling mechanism; Step S3: If the current time reaches the next polling time of the algorithm and the system state changes, proceed to step S4; otherwise, return to step S2. Step S4: If there is an idle machine dog in the current system, proceed to step S5; otherwise, return to step S2. Step S5: Read the real-time system status and construct a set of information on tasks to be assigned, a set of status for idle robot dogs, and a set of status for charging piles; Step S6: Perform a comprehensive cost calculation on the robot dog-task combinations in the idle robot dog set and the task set to be assigned, and obtain the comprehensive time cost corresponding to each robot dog-task combination; Step S7: Select the robot dog-task combination with the lowest overall time cost as the optimal combination, and assign the planned path corresponding to this combination to the corresponding robot dog, and update the task information set, robot dog status set and charging pile status set. Step S8: If all tasks have been assigned, proceed to step S9; otherwise, return to step S4. Step S9: If there are still idle robot dogs with no assigned tasks, proceed to step S10; otherwise, proceed to step S11. Step S10: For all idle robot dogs, dispatch them to charging piles or meter recycling stations according to their current power status and load conditions; Step S11: If the meter calibration process is closed, the program ends; otherwise, return to step S3.

[0012] Preferably, in step S1, the system initialization step is as follows: Step S101: Initialize the global polling time; Step S102: Obtain static scene information of the electricity meter calibration line, including the location and size of the equipment, channel restrictions, obstacle size, charging pile location, replacement station location and electricity meter recycling station location, and draw the layout plan; Step S103: Create a two-dimensional planar model of the workshop interior using the grid map method; Step S104: Based on the two-dimensional plane model in step S103, the static shortest path between the task node, charging pile node, swapping station node, and meter recycling station node is calculated using a global shortest path planning algorithm to construct a static shortest path library. The global shortest path planning algorithm uses the A* algorithm.

[0013] Preferably, in step S3, the change in system state includes at least one of the following situations: the robot dog becomes idle again after completing its task, a new task is added, a task is canceled, the robot dog completes charging, or other events that cause a change in the schedulable state of the system.

[0014] Preferably, in step S5, the specific steps for constructing the current set of tasks to be assigned, the set of idle robot dog states, and the set of charging pile states are as follows: Step S501: Obtain the current set of tasks to be assigned; for each task, obtain information such as task node, load space required by the task, increase in weight carried by the robot dog caused by completing the task, type of external device required by the task, specific robot dog number required by the task, power required by the task, task time, and task time window, and construct a task information set. Step S502: Obtain the current set of idle robot dogs; for each idle robot dog, obtain its current node, total load space, used load space, current weight of the battery meter it is carrying, current type of external device, and remaining battery power, and construct a robot dog status set; Step S503: Obtain the charging pile set; for each charging pile, obtain its planned charging occupancy time window information and construct the charging pile status set.

[0015] Preferably, in step S6, the comprehensive cost calculation for the robot dog-task combination includes the following steps: Step S601: For a certain robot dog-task combination, take the static shortest path from the current node of the robot dog to the task node in the static shortest path library as the initial planned path, and construct a set of planned paths based on the initial planned path; Step S602: If the task requires execution by a specific robot dog, and the current robot dog does not match the specific robot dog, then proceed to step S603; otherwise, proceed to step S604. Step S603: Set the planned paths for the robot dog-task combination. Set it to an empty set and proceed to step S6012; Step S604: If the current load capacity of the robot dog can meet the load requirements of the task, then proceed to step S605; otherwise, proceed to step S606. Step S605: If the current external device of the robot dog cannot complete the task, and the robot dog is currently carrying a meter that has failed inspection, then proceed to step S606; otherwise, proceed to step S607. Step S606: For the initial planned path of the robot dog-task combination, insert a meter recycling station node into the path before the task node. This meter recycling station is the one that minimizes the increase in path length after insertion. Step S607: If the current type of external device of the robot dog can complete the task, proceed to step S609; otherwise, proceed to step S608. Step S608: For the current planned path of the robot dog-task combination, insert a changing station node in the planned path before the task node. This changing station is the one that minimizes the path length increment after insertion. Step S609: If the robot dog has enough remaining battery power to complete the task and go to the nearest charging station according to the current planned path, then proceed to step S6011; otherwise, proceed to step S6010. Step S6010: For the current planned path, enumerate the locations where charging piles can be inserted and the optional charging piles, and construct an augmented path; based on the time when the robot dog arrives at the charging pile, the remaining power at the time of arrival, and the planned charging time window of the charging pile, determine the earliest feasible charging time window for the robot dog at the corresponding charging pile; retain feasible augmented paths that meet the power constraints, and delete redundant augmented paths. Step S6011: Calculate the comprehensive time cost for each path in the planned path set, wherein the comprehensive time cost includes at least the time it takes for the robot dog to complete the relevant actions at each node, the time it takes for the robot dog to walk along the path, the waiting time when the robot dog arrives earlier than the task time window, and the late penalty cost when it arrives later than the task time window. Step S6012: If the planned path set is empty, then set the overall time cost of the robot dog-task combination to infinity; otherwise, select the minimum overall time cost in the planned path set as the final overall time cost of the robot dog-task combination.

[0016] Preferably, step S7 includes the following steps: Step S701: From all robot dog-task combinations, select the feasible combination with the lowest final overall time cost as the current optimal combination; Step S702: Assign the planned path corresponding to the optimal combination to the corresponding robot dog, delete the corresponding task from the set of tasks to be assigned, and delete the corresponding robot dog from the set of idle robot dogs; Step S703: Update the task information set, robot dog status set, and charging pile status set.

[0017] Preferably, step S10 includes the following steps: Step S1001: For an idle robot dog that is not carrying a power meter, if its current remaining power is less than a set threshold, it is dispatched to the charging station that it can reach and can start charging earliest; if its current remaining power is greater than or equal to the set threshold, its current state remains unchanged. Step S1002: For an idle robot dog carrying an electricity meter, if its remaining power is sufficient to reach the nearest electricity meter recycling station, it is dispatched to the nearest electricity meter recycling station to perform the electricity meter return task; if its remaining power is insufficient to reach the nearest electricity meter recycling station, it is first dispatched to the charging station that it can reach and can start charging earliest, and after charging is completed, it goes to the electricity meter recycling station closest to its charging station; after the robot dog completes the electricity meter return task, if its remaining power is still lower than the set threshold, it is dispatched to the charging station to charge.

[0018] Through the above design scheme, the present invention can bring the following beneficial effects: (1) In view of the characteristics of dynamic generation of inspection tasks and anomaly handling tasks in the power meter calibration pipeline scenario, significant differences in task types and inconsistent timeliness requirements, this invention constructs a dynamic inspection scheduling method for multiple robot dogs, which can coordinate the task execution process of multiple inspection robot dogs in a unified manner, improve the overall execution efficiency of inspection tasks and anomaly handling tasks, and enhance the robot dog system's ability to continuously inspect and ensure the power meter calibration pipeline and its ability to quickly handle anomalies.

[0019] (2) In the scheduling process, the present invention considers factors such as the current remaining load space of the robot dog, the type of external equipment, the status of the back meter and the remaining power. It can not only complete the allocation of routine inspection tasks, but also take into account meter picking tasks, meter return tasks and the robot dog's replacement needs for different types of tasks. This makes the task allocation no longer limited to simple matching of robot dog and task, but can dynamically construct feasible execution paths according to the conditions required for task execution, thereby improving the consistency and feasibility of scheduling results with actual operation scenarios.

[0020] (3) The present invention considers the charging planning and task scheduling process in a unified manner. During the task execution path construction stage, feasible paths are screened by combining the remaining power of the robot dog and the time window occupied by the charging pile, and charging nodes are dynamically inserted when necessary. This avoids the problem of charging scheduling delay caused by treating charging as an independent additional process in the prior art, thereby reducing the risk of task interruption caused by insufficient power during the task execution of the robot dog.

[0021] (4) The present invention adopts a polling dynamic scheduling mechanism, which can quickly update the scheduling results when the system state changes such as adding new tasks, canceling tasks, the robot dog completing a task and becoming idle again, and the robot dog completing charging. This enables timely replanning of multiple robot dog task sequences, thereby improving the scheduling method's responsiveness to dynamic events, reducing the impact on the normal operation of the electricity meter verification pipeline caused by untimely task response, task interruption, or mismatched robot dog states, and ensuring the efficient and stable operation of the electricity meter verification process. Attached Figure Description

[0022] The present invention will be further described below with reference to the accompanying drawings and specific embodiments: Figure 1 This is a flowchart illustrating the overall process of the dynamic inspection and scheduling method for multiple robot dogs in a fixed electricity meter scenario according to an embodiment of the present invention. Figure 2 This is a flowchart illustrating the comprehensive cost calculation and path construction for the robot dog-task combination in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios of this invention will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not constitute a limitation thereof.

[0024] This invention provides a method for dynamic inspection scheduling of multiple robot dogs in the context of electricity meter calibration.

[0025] A method for dynamic inspection scheduling of multiple robot dogs in the context of electricity meter verification is applied to a dynamic inspection scheduling system for multiple robot dogs in the context of electricity meter verification. The system has a built-in path planning module, status monitoring module, task scheduling module and device communication module, and is respectively connected to the electricity meter verification production line, several inspection robot dogs, production line charging piles and electricity meter recycling station. Among them, the path planning module is used to build a static shortest path library, the status monitoring module is used to collect and identify system status changes in real time, the task scheduling module is used to perform task matching and robot dog scheduling calculations, and the device communication module is used to issue scheduling instructions and send and receive peripheral data.

[0026] Including the following steps: S1: Initialize the polling time, obtain static scene information of the electricity meter calibration pipeline, and build a static shortest path library; The system initialization includes the following steps: S101: Initialize global polling time .

[0027] S102: Obtain information on the internal layout of the calibration workshop, including the location and size of equipment, passage restrictions, obstacle dimensions, and charging pile locations, and draw a layout plan. The steps involved in constructing the static shortest path library are as follows: S103: Two-dimensional planar modeling of the workshop interior using the grid mapping method: discretizing the workshop interior into... grid nodes Construct a square grid map with the following map matrix:

[0028] Among them, when node When a node is an obstacle, its label variable at its corresponding position in the map matrix. Similarly, when node When it is a channel node, When node When it is a task node, When node When it is a charging pile node, When node When it is a transshipment station node, When node When it is a node for electricity meter recycling station, .

[0029] S104: Select obstacle nodes, passage nodes, device nodes, and charging pile nodes from all nodes in the grid map to construct an obstacle node set. Channel node set Task Node Set Charging pile node set , Set of changing station nodes The set of nodes for the electricity meter recycling station, Z, is shown in formulas (2)-(7):

[0030] Relevant algorithms (such as the A* algorithm, which is a heuristic optimal path search algorithm and the most mainstream pathfinding scheme in games, robots, autonomous driving, and map navigation) are used to calculate the paths of each task node, charging pile node, swapping station node, and meter recycling station node (i.e., any two nodes). The static shortest path between ( ).

[0031] S2: Update the next polling time. ,in This indicates the sampling step size. The system checks the system status every time this step size is applied to see if there are any status changes; the preferred interval is 1 second.

[0032] S3: If the current time reaches the next polling time of the algorithm and the system state changes (such as the robot dog completing a task, adding a new task, or canceling a task, causing the robot dog to be idle), then proceed to step S4; otherwise, return to step S2.

[0033] S4: If there is an idle machine dog in the current system, proceed to step S5; otherwise, return to step S2.

[0034] Among them, the idle robot dog is a robot dog that is not performing any inspection tasks and is in a standby state.

[0035] S5: Read the real-time status of the system.

[0036] The steps for reading the real-time status of the system include: S501: Get the current set of tasks to be assigned For each task Obtain task information and construct a task information set. .

[0037]

[0038] in For task nodes; The required payload space for the robot dog to complete the task; The increase in weight carried by the robot dog to complete the task; The type of robot dog attachment required for the task; when the task mandates that it be performed by a specific robot dog. Number the corresponding robot dog; otherwise... It is a null value; The amount of power required for the robot dog to complete the relevant actions at each task node; The time taken for the task; The task time window is defined as the time window during which the robot dog must reach the task node and begin executing the task.

[0039] S502: Get the current set of idle robot dogs For each idle robot dog Obtain the robot dog's state and construct a robot dog state set. .

[0040]

[0041] in, This refers to the current node where the robot dog is located. This represents the robot dog's current total load capacity. This refers to the current load space used by the robot dog. The weight of the electricity meter currently being carried by the robot dog; This refers to the current type of external device used by the robot dog. This is the robot dog's current remaining battery power.

[0042] S503: Obtain a collection of charging stations For each charging station Obtain the status of charging piles and construct a charging pile status set. .

[0043]

[0044] Among them, set Recorded charging stations The time window for all ongoing or upcoming robot dog charging events.

[0045] S6: Calculate the total cost of all robot dog-task combinations.

[0046] The overall cost of the computer dog-task combination includes the following steps: S601: Initialize the robot dog - task combination plan path This is the static shortest path from the robot dog's current node to the task node, found in the shortest path library. A set of planned paths is constructed. and initialized to .

[0047] S602: For the idle robot dog in the combination and tasks If the robot dog does not match the specific robot dog required for the task, i.e. Not null and If the condition is met, proceed to step S603; otherwise, proceed to step S604.

[0048] S603: The set of planned paths for this robot dog-task combination. Set it to an empty set and proceed to step S6012.

[0049] S604: If the robot dog's current load capacity can meet the task's load requirements, i.e. If the condition is met, proceed to step S605; otherwise, proceed to step S606.

[0050] S605: If the robot dog's current external device type cannot complete the task ( ), and the robot dog is currently carrying a faulty electricity meter ( If the condition is met, proceed to step S606; otherwise, proceed to step S607.

[0051] S606: Initial planned path for the robot dog-task combination , at the task node First, insert a meter recycling node into this path. ,Right now .in The node that minimizes the path length increment after insertion is the meter recycling station node. ,in This indicates the slave node read from the shortest path library. To the node The shortest path length.

[0052] S607: If the robot dog's current external device type can complete this task ( If the condition is met, proceed to step S609; otherwise, proceed to step S608.

[0053] S608: Current planned path for the robot dog-task combination , at the task node Forward, towards Insert a changing station node , Change to .in The transshipment station node that minimizes the path length increment after insertion, i.e. ,in The second-to-last node in the planned path before inserting the replacement station node, which may be the current node of the robot dog or the electricity meter recycling station node.

[0054] S609: If the robot dog has enough remaining battery power to complete the task according to the current planned path and proceed to the nearest charging station ( If the condition is met, then proceed to step S6011; otherwise, proceed to step S6010. This represents the total amount of electricity required for the robot dog to complete its task and reach the nearest charging station according to the current planned path, and is calculated using the following formula:

[0055] Indicates the planned path The first in Each node.

[0056] The first term of the formula represents the power consumption of the robot dog's actions at each node in the planned path (current node). except) if For electricity meter recycling station nodes ,in The electricity consumption per unit weight carried by the robot dog at the electricity meter recycling station; if For the changing station node ,in The power consumption of the robot dog during one clothing change operation at the clothing station; if For task nodes .

[0057] The second term of the formula represents the total power consumption required for the robot dog to walk along the planned path, and the power consumption for walking is related to the weight carried by the robot dog. This indicates the power consumption per unit distance the robot dog travels without any load or external devices. This represents the additional power consumption per unit weight carried by the robot dog for each unit of distance it travels. This indicates that the robot dog is at the slave node. Walk to the node The total weight carried during the process, including the weight of the robot dog's external attachments. and the weight of the electricity meter carried on his back .

[0058] The third term in the formula represents the robot dog's movement from the task node. Walk to the nearest charging station The power consumption during walking is also affected by the weight carried by the robot dog.

[0059] The fourth term in the formula represents the extra power consumption of the robot dog to maintain its balance, which is independent of the robot dog's state. This represents the power consumption of the robot dog to maintain balance per unit of time. This indicates the total time taken for the robot dog to complete its task and reach the nearest charging station from the current moment.

[0060] S6010: For the planned path There are a total of A location where a charging station can be inserted, Indicates different insertion positions. For a given insertion position... and a certain charging station Constructing augmenting paths ,in Indicates charging station The node it is located at. The computer dog is on the path. Arrival at charging stations The time, and based on that time and the charging station The planned charging time window (based on the charging pile status set) (Obtained from the charging station), computer robot dog at the charging station The earliest feasible charging time window is determined. This time window satisfies the following conditions: 1. The start time is no earlier than the time the robot dog arrives at the charging station; 2. The length of the time window is the time it takes for the robot dog to fully charge at that charging station; 3. The time window does not overlap with the charging time windows of other robot dogs already scheduled to charge at that charging station. If the robot dog currently has enough remaining battery power to reach the charging station... If, after a full charge, the robot dog has enough power to complete the subsequent path and task execution from the charging station, then the augmented path is considered feasible. If feasible, then Add to the planned path collection In the middle. After adding all feasible augmentation paths, delete. The original planned path Furthermore, for any and If the two augmenting paths are inserted at the same position and simultaneously satisfy condition 1. The increase in path length due to inserting a charging station is less than or equal to 2. The robot dog in If the charging station inserted into the device can complete charging earlier, then it is considered... To create a redundant augmenting path, from Remove from the middle.

[0061] S6011: For the set of planned paths Calculate the overall time cost for all paths in the given set. For a given path... Comprehensive time cost Calculate as follows: The computer dog starts from the current node according to the path Reaching the task node Time consumption :

[0062] in Indicates the robot dog is at the node The time taken to complete the relevant actions, if For electricity meter recycling station nodes ,in The time taken for the robot dog to unload a unit of load space at the meter recycling station; if For the changing station node ,in The time taken for the robot dog to perform a single costume change at the costume station. The walking speed of the robot dog; if For charging piles This represents the total time it takes for the robot dog to arrive at the charging station and complete the charging process.

[0063] The overall time cost of calculating the robot dog :

[0064] in This indicates that the robot dog is earlier than the task time window. The waiting time generated when reaching the task node (later than) (0 upon arrival) This indicates the delay time when the robot dog arrives after the end time of the task time window. The penalty cost per unit of lateness.

[0065] S6012: If If the set is empty, the combined time cost of the robot dog-task combination is set to infinity. Otherwise, a set of planned paths is selected. The minimum of the overall time cost is taken as the final overall time cost of the robot dog-task combination.

[0066] S7: Select the robot dog-task combination with the lowest overall time cost, i.e. and and the corresponding planned path Assigned to robot dog The task Remove from the set of tasks to be assigned. Remove the idle robot dog from the set and update the task information set, robot dog status set, and charging pile status set.

[0067] S8: If all tasks have been assigned, the task set If the result is empty, proceed to step S9; otherwise, return to step S4.

[0068] S9: If there are still idle robot dogs with unassigned tasks, proceed to step S10; otherwise, proceed to step S11.

[0069] S10: For all idle robot dogs, identify their current power status and load, and dispatch them to charging piles or meter recycling stations.

[0070] The step of scheduling the robot dog to a charging station or meter recycling station based on the idle robot dog quantity status and load condition includes the following steps: S1001: For an idle robot dog that is not carrying a power meter, if its current remaining power is less than a set threshold, it will be dispatched to the charging station that it can reach and can start charging earliest; if its current remaining power is greater than or equal to the set threshold, its current state will remain unchanged. S1002: For an idle robot dog carrying an electricity meter, if its remaining battery power is sufficient to reach the nearest electricity meter recycling station, it will be dispatched to the nearest electricity meter recycling station to perform the electricity meter return task; if its remaining battery power is insufficient to reach the nearest electricity meter recycling station, it will first be dispatched to the charging station that it can reach and can start charging earliest, and after charging is completed, it will go to the electricity meter recycling station closest to its charging station. After the robot dog completes the electricity meter return task, if its remaining battery power is still lower than the set threshold, it will continue to be dispatched to the charging station for charging.

[0071] S11: If the meter calibration process is closed, the program ends; otherwise, return to step S3.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

[0073] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-robot dynamic inspection and scheduling system for electricity meter calibration scenarios, characterized in that: It includes a path planning module, a status monitoring module, a task scheduling module, and a device communication module; The path planning module is used to build a static shortest path library, the status monitoring module is used to collect and identify system status changes in real time, the task scheduling module is used to perform task matching and robot dog scheduling calculations, and the device communication module is used to issue scheduling instructions and send and receive data.

2. A method for dynamic inspection and scheduling of multiple robot dogs in an energy meter calibration scenario, characterized in that, The multi-robot dynamic inspection and scheduling system applied to the electricity meter verification scenario described in claim 1 includes the following steps: Step S1: After the system starts, the polling time is initialized first to obtain the static scene information of the electricity meter calibration pipeline, and the static shortest path library is constructed using the global shortest path planning algorithm; Step S2: Update the next polling time based on the preset polling mechanism; Step S3: If the current time reaches the next polling time of the algorithm and the system state changes, proceed to step S4; otherwise, return to step S2. Step S4: If there is an idle machine dog in the current system, proceed to step S5; otherwise, return to step S2. Step S5: Read the real-time system status and construct a set of information on tasks to be assigned, a set of status for idle robot dogs, and a set of status for charging piles; Step S6: Perform a comprehensive cost calculation on the robot dog-task combinations in the idle robot dog set and the task set to be assigned, and obtain the comprehensive time cost corresponding to each robot dog-task combination; Step S7: Select the robot dog-task combination with the lowest overall time cost as the optimal combination, and assign the planned path corresponding to this combination to the corresponding robot dog, and update the task information set, robot dog status set and charging pile status set. Step S8: If all tasks have been assigned, proceed to step S9; otherwise, return to step S4. Step S9: If there are still idle robot dogs with no assigned tasks, proceed to step S10; otherwise, proceed to step S11. Step S10: For all idle robot dogs, dispatch them to charging piles or meter recycling stations according to their current power status and load conditions; Step S11: If the meter calibration process is closed, the program ends; otherwise, return to step S3.

3. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that, In step S1, the system initialization steps are as follows: Step S101: Initialize the global polling time; Step S102: Obtain static scene information of the electricity meter calibration line, including the location and size of the equipment, channel restrictions, obstacle size, charging pile location, replacement station location and electricity meter recycling station location, and draw the layout plan; Step S103: Create a two-dimensional planar model of the workshop interior using the grid map method; Step S104: Based on the two-dimensional plane model in step S103, the static shortest path between the task node, charging pile node, swapping station node, and meter recycling station node is calculated using a global shortest path planning algorithm to construct a static shortest path library. The global shortest path planning algorithm uses the A* algorithm.

4. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that, In step S3: the change in system state includes at least one of the following situations: the robot dog becomes idle again after completing its task, a new task is added, a task is canceled, the robot dog completes charging, or other events that cause the system's schedulable state to change.

5. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that: In step S5, the specific steps for constructing the current set of tasks to be assigned, the set of idle robot dog states, and the set of charging pile states are as follows: Step S501: Obtain the current set of tasks to be assigned; for each task, obtain information such as task node, load space required by the task, increase in weight carried by the robot dog caused by completing the task, type of external device required by the task, specific robot dog number required by the task, power required by the task, task time, and task time window, and construct a task information set. Step S502: Obtain the current set of idle robot dogs; for each idle robot dog, obtain its current node, total load space, used load space, current weight of the battery meter it is carrying, current type of external device, and remaining battery power, and construct a robot dog status set; Step S503: Obtain the charging pile set; for each charging pile, obtain its planned charging occupancy time window information and construct the charging pile status set.

6. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that: In step S6, a comprehensive cost calculation is performed on the robot dog-task combination, including the following steps: Step S601: For a certain robot dog-task combination, take the static shortest path from the current node of the robot dog to the task node in the static shortest path library as the initial planned path, and construct a set of planned paths based on the initial planned path; Step S602: If the task requires execution by a specific robot dog, and the current robot dog does not match the specific robot dog, then proceed to step S603; otherwise, proceed to step S604. Step S603: Set the planned paths for the robot dog-task combination. Set it to an empty set and proceed to step S6012; Step S604: If the current load capacity of the robot dog can meet the load requirements of the task, then proceed to step S605; otherwise, proceed to step S606. Step S605: If the current external device of the robot dog cannot complete the task, and the robot dog is currently carrying a meter that has failed inspection, then proceed to step S606; otherwise, proceed to step S607. Step S606: For the initial planned path of the robot dog-task combination, insert a meter recycling station node into the path before the task node. This meter recycling station is the one that minimizes the increase in path length after insertion. Step S607: If the current type of external device of the robot dog can complete the task, proceed to step S609; otherwise, proceed to step S608. Step S608: For the current planned path of the robot dog-task combination, insert a changing station node in the planned path before the task node. This changing station is the one that minimizes the path length increment after insertion. Step S609: If the robot dog has enough remaining battery power to complete the task and go to the nearest charging station according to the current planned path, then proceed to step S6011; otherwise, proceed to step S6010. Step S6010: For the current planned path, enumerate the locations where charging piles can be inserted and the optional charging piles, and construct an augmented path; based on the time when the robot dog arrives at the charging pile, the remaining power at the time of arrival, and the planned charging time window of the charging pile, determine the earliest feasible charging time window for the robot dog at the corresponding charging pile; retain feasible augmented paths that meet the power constraints, and delete redundant augmented paths. Step S6011: Calculate the comprehensive time cost for each path in the planned path set, wherein the comprehensive time cost includes at least the time it takes for the robot dog to complete the relevant actions at each node, the time it takes for the robot dog to walk along the path, the waiting time when the robot dog arrives earlier than the task time window, and the late penalty cost when it arrives later than the task time window. Step S6012: If the planned path set is empty, then set the overall time cost of the robot dog-task combination to infinity; otherwise, select the minimum overall time cost in the planned path set as the final overall time cost of the robot dog-task combination.

7. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that: Step S7 includes the following steps: Step S701: From all robot dog-task combinations, select the feasible combination with the lowest final overall time cost as the current optimal combination; Step S702: Assign the planned path corresponding to the optimal combination to the corresponding robot dog, delete the corresponding task from the set of tasks to be assigned, and delete the corresponding robot dog from the set of idle robot dogs; Step S703: Update the task information set, robot dog status set, and charging pile status set.

8. The multi-robot dynamic inspection and scheduling method for electricity meter verification scenarios according to claim 2, characterized in that: Step S10 includes the following steps: Step S1001: For an idle robot dog that is not carrying a power meter, if its current remaining power is less than a set threshold, it is dispatched to the charging station that it can reach and can start charging earliest; if its current remaining power is greater than or equal to the set threshold, its current state remains unchanged. Step S1002: For an idle robot dog carrying an electricity meter, if its remaining power is sufficient to reach the nearest electricity meter recycling station, it is dispatched to the nearest electricity meter recycling station to perform the electricity meter return task; if its remaining power is insufficient to reach the nearest electricity meter recycling station, it is first dispatched to the charging station that it can reach and can start charging earliest, and after charging is completed, it goes to the electricity meter recycling station closest to its charging station; after the robot dog completes the electricity meter return task, if its remaining power is still lower than the set threshold, it is dispatched to the charging station to charge.

Images