Method for cooperative scheduling of a cluster of inspection robots in a high-altitude environment and related device

Abstract

This application discloses a collaborative scheduling method and related equipment for inspection robot swarms in high-altitude environments, belonging to the field of UAV swarm collaborative scheduling technology. This application divides the target inspection area into multiple task units and collects the real-time status of each inspection robot within the target inspection area and the environmental status of its environment. Based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status, an executable candidate task set is generated. A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and based on the score calculation result, the task allocation result for collaborative scheduling of the multiple inspection robots is determined. Based on the task allocation result, a unique and valid task execution result is determined and the task status is advanced. Risk classification and early warning information are then performed on the task execution result, thereby improving the collaborative scheduling effect of multiple inspection robots.

Description

Technical Field

[0001] This application relates to the field of collaborative scheduling technology for unmanned aerial vehicle (UAV) swarms, and in particular to a collaborative scheduling method and related equipment for inspection robot swarms in high-altitude environments. Background Technology

[0002] Infrastructure inspection tasks in high-altitude areas (such as power transmission lines, road slopes, bridges and culverts, tunnel entrances, and utility tunnels) typically present challenges such as complex terrain, frequent weather changes, low visibility at night, high risks associated with manual inspections, and slow emergency response. Traditional manual inspections pose significant safety hazards under conditions such as wind, snow, low temperatures, and icy roads, and their efficiency is insufficient to meet emergency needs.

[0003] While existing unmanned inspection robot systems can replace manual labor to some extent, they still generally suffer from the following shortcomings in harsh high-altitude environments: First, multi-robot swarm collaboration often relies on stable communication networks or central scheduling, which can easily lead to problems such as task duplication, task state bifurcation, and task suspension when communication links are broken or severely delayed. Second, existing task scheduling often relies primarily on the shortest distance or shortest time, making it difficult to incorporate multiple constraints such as energy constraints, refueling uncertainty, risk thresholds, and communication reachability into the feasibility assessment. Third, under conditions such as wind and snow cover, low-temperature battery degradation, and terrain obstruction, the task execution process requires real-time replanning, but existing systems lack a "rolling scheduling closed loop driven by environmental state changes," resulting in a high probability of execution failure after task allocation. Fourth, task completion determination generally relies on uploaded data or arrival confirmation, lacking a verifiable completion evidence mechanism, and cannot avoid conflicting results submitted by multiple robots after link recovery.

[0004] Therefore, there is an urgent need for a collaborative scheduling method for unmanned inspection robot clusters that can operate stably in harsh high-altitude environments, so that it can maintain the consistency of task status under conditions of unstable communication and limited energy, avoid repeated execution, and be able to replan the task scheduling strategy in real time according to environmental changes. Summary of the Invention

[0005] The main purpose of this application is to provide a collaborative scheduling method and related equipment for inspection robot clusters in high-altitude environments, aiming to solve the technical problem that the existing collaborative scheduling effect of unmanned inspection robot clusters operating stably in harsh high-altitude environments is poor.

[0006] To achieve the above objectives, this application provides a method for collaborative scheduling of inspection robot swarms in high-altitude environments, the method comprising the following steps: The target inspection area is divided into multiple task units, and the real-time status of each inspection robot in the target inspection area and the environmental status of its environment are collected. Based on the preset risk map and preset communication map, as well as the real-time status and the environmental status, a set of executable candidate tasks is generated; A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and the task allocation result of the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result. Based on the task allocation results, a unique and valid task execution result is determined and the task status is advanced. The task execution result is then risk-classified and early warning information is output.

[0007] In one embodiment, the step of generating a set of executable candidate tasks based on a preset risk map and a preset communication map, as well as the real-time state and the environmental state, includes: Based on the real-time status and the environmental status, the available energy and power consumption requirements of each inspection robot in the future time window are predicted, and based on the prediction results, an energy constraint relationship between the available energy and the power consumption requirements is constructed. Based on the preset risk map and preset communication map, the risk cost and communication accessibility of each inspection robot in the future time window are calculated, and a joint constraint relationship between risk and communication is constructed based on the calculation results. Based on the energy constraint relationship and the risk and communication joint constraint relationship, a set of executable candidate tasks is generated.

[0008] In one embodiment, the step of generating a set of executable candidate tasks based on the energy constraint relationship and the joint risk and communication constraint relationship includes: Obtain the pre-configured task attributes and constraint parameters for each task unit, and predict the estimated time consumption of the tasks to be executed by each inspection robot in each task unit based on the task attributes and constraint parameters. Based on the energy constraint relationship, the risk and communication joint constraint relationship, and the estimated time consumption, a multi-constraint joint feasibility determination is made for the task to be executed. Based on the judgment results, candidate tasks that meet the joint feasibility requirements of multiple constraints are identified, and a set of executable candidate tasks is generated.

[0009] In one embodiment, the step of calculating a comprehensive benefit score for candidate combinations of multiple inspection robots in the set of executable candidate tasks includes: Construct a redundant penalty factor to punish multiple inspection robots for repeatedly performing the same task; Based on the energy constraint relationship and the risk and communication joint constraint relationship, as well as the task priority attribute and task coverage ratio in the task attributes, a multi-constraint benefit scoring function is constructed. Based on the multi-constraint benefit scoring function, a comprehensive benefit score is calculated for the candidate combinations of multiple inspection robots in the candidate task set.

[0010] In one embodiment, the step of determining the task allocation result for the collaborative scheduling of the multiple inspection robots based on the scoring calculation result includes: Based on the scoring calculation results, the inspection robot with the highest performance score for each task unit is determined. Based on the inspection robot with the highest score, a unique advance right lock record is assigned to each task unit. Based on the unique advance right lock record, the task allocation result of the collaborative scheduling of the multiple inspection robots is determined, so that only one inspection robot is allowed to execute the same task unit in the same time window.

[0011] In one embodiment, after the step of determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the unique advance right lock record, the method includes: Based on the task allocation results, the communication quality, remaining energy, and health status of the inspection robots performing the corresponding tasks are monitored in real time. When any real-time monitored indicator is lower than the corresponding preset event trigger threshold, the event trigger condition is determined to be met. When the event triggering condition is met, the advance right lock record held by the inspection robot is revoked, the task status of the corresponding task unit is rolled back to pending assignment, and the task version number is incremented and updated. Based on the current environmental conditions, the set of executable candidate tasks is redefined, and the comprehensive benefit score and advancement rights are recalculated. When the inspection robot that has obtained advancement rights is interrupted, it writes the execution trajectory and inspection data into the local cache structure. After the communication is restored, the completion evidence in the local cache is uploaded to the central scheduling system.

[0012] In one embodiment, the step of risk-classifying the task execution result and outputting early warning information includes: A comprehensive score is calculated for the completion evidence of the task execution result, wherein the comprehensive score is obtained by weighting the data quality score, trajectory coverage score and time window legality score; The legality of the completed evidence shall be verified, including at least verifying whether the inspection robot that submitted the evidence is the robot holding the right to advance and whether the task version number in the evidence is consistent with the current task version number. When multiple inspection robots submit evidence of completion for the same task unit, the evidence with the highest comprehensive score is selected as the only valid evidence, and the task status of the corresponding task unit is set to completed, while the remaining evidence is archived. When the evidence is invalid or the overall score is lower than the preset task completion threshold, the task status of the corresponding task unit will be rolled back to pending assignment and a rolling replanning will be triggered. Based on a preset risk threshold, the task execution results are classified into risk levels, abnormal events are categorized and spatially aggregated, and warning information is output that includes at least the risk level, event type, affected area, and timestamp.

[0013] Furthermore, to achieve the above objectives, this application also provides a collaborative scheduling device for inspection robot clusters in high-altitude environments, the device comprising: The data acquisition module is used to divide the target inspection area into multiple task units and collect the real-time status of each inspection robot in the target inspection area and the environmental status of its environment. The evaluation module is used to generate a set of executable candidate tasks based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status. The calculation module is used to calculate the comprehensive benefit score of multiple inspection robots in the executable candidate task set, and determine the task allocation result of the collaborative scheduling of the multiple inspection robots based on the score calculation result. The early warning module is used to determine the unique and valid task execution result based on the task allocation result, advance the task status, and perform risk classification and output early warning information for the task execution result.

[0014] Furthermore, to achieve the above objectives, this application also provides a collaborative scheduling device for inspection robot clusters in high-altitude environments. The collaborative scheduling device for inspection robot clusters in high-altitude environments includes: a memory, a processor, and a collaborative scheduling method program for inspection robot clusters in high-altitude environments stored in the memory and executable on the processor. The collaborative scheduling method program for inspection robot clusters in high-altitude environments is configured to implement the steps of the collaborative scheduling method for inspection robot clusters in high-altitude environments as described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer-readable storage medium storing a program for a collaborative scheduling method of a cluster of inspection robots in a high-altitude environment. When the program is executed by a processor, it implements the steps of the collaborative scheduling method of a cluster of inspection robots in a high-altitude environment as described above.

[0016] One or more technical solutions proposed in this application have at least the following technical effects: By dividing the target inspection area into multiple task units, and collecting the real-time status of each inspection robot within the target inspection area and the environmental status of its environment; generating a set of executable candidate tasks based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status; calculating a comprehensive benefit score for candidate combinations of multiple inspection robots in the set of executable candidate tasks, and determining the task allocation result for the collaborative scheduling of the multiple inspection robots based on the score calculation result; determining a unique and effective task execution result based on the task allocation result and advancing the task status, and performing risk classification and outputting early warning information for the task execution result. In other words, by dividing the target inspection area into multiple task units, fine-grained decomposition of complex terrain is achieved, enabling large-scale inspection tasks to be quantitatively managed. Simultaneously, the real-time status of each robot and the environmental status are collected. The system provides dynamic and accurate input data to ensure precise scheduling decisions. It also combines pre-set risk and communication maps to generate a set of executable candidate tasks, fully considering the unique terrain risks, weather changes, and communication limitations of high-altitude areas. This ensures the safety and feasibility of task planning from the outset. By calculating the comprehensive benefit score of candidate combinations and determining task allocation results, it enables optimized resource allocation among multiple robots, avoiding task conflicts and duplication of work. It also determines a unique and effective task execution result and advances the task status, ensuring the traceability and consistency of the task loop. Finally, it performs risk classification on the task execution results and outputs early warning information, providing intuitive decision support for maintenance personnel. This allows for the timely detection and handling of safety hazards in high-altitude environments, thereby achieving intelligent, adaptive, and collaborative scheduling of inspection robot clusters in high-altitude environments, effectively improving inspection coverage and data reliability. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating the first embodiment of the collaborative scheduling method for inspection robot clusters in high-altitude environments according to this application; Figure 2 This is a schematic diagram of the filtering mechanism for the multi-constraint executable candidate set in an embodiment of this application; Figure 3 This is a schematic diagram of the module structure of the collaborative scheduling device for a cluster of inspection robots in a high-altitude environment, as described in an embodiment of this application. Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the collaborative scheduling method of inspection robot clusters in high-altitude environments in the embodiments of this application.

[0020] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0022] Reference Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the collaborative scheduling method for inspection robot clusters in high-altitude environments according to this application.

[0023] In this embodiment, a collaborative scheduling method for inspection robot clusters in high-altitude environments is proposed, the method comprising the following steps: S10, the target inspection area is divided into multiple task units, and the real-time status of each inspection robot in the target inspection area and the environmental status of its environment are collected. S20, Based on the preset risk map and preset communication map, as well as the real-time status and the environmental status, generate a set of executable candidate tasks; S30, perform a comprehensive benefit score calculation on the candidate combinations of multiple inspection robots in the executable candidate task set, and determine the task allocation result of the collaborative scheduling of the multiple inspection robots based on the score calculation result. S40, based on the task allocation result, determine the unique and valid task execution result and advance the task status, and perform risk classification and output early warning information for the task execution result.

[0024] In this embodiment, the target inspection area refers to a high-altitude geographical area that needs to be inspected regularly, such as a section of the Qinghai-Tibet Railway about 50 kilometers long, or a high-altitude mine slope area. Dividing the target inspection area into multiple task units means dividing a continuous large area into several sub-regions according to a certain strategy (such as terrain features, inspection point density, robot endurance, etc.). Each sub-region is an independent task unit, for example, 1 kilometer as a unit, or divided according to the ridgeline. Each unit has a unique identifier (such as unit ID: U001, U002).

[0025] Simultaneously, it is also necessary to collect the real-time status of each inspection robot within the target inspection area. Inspection robots refer to unmanned ground vehicles or quadruped robots equipped with sensors (such as high-definition cameras, infrared thermal imagers, and gas detectors) and communication modules. Their real-time status includes: current position (latitude and longitude coordinates), current battery percentage, current task status (idle, executing, returning), and health status (whether each sensor is online, motor temperature, etc.).

[0026] This also requires collecting environmental data about the robot's surroundings, such as temperature, wind speed, visibility, and snowfall intensity, obtained through the robot's own sensors or nearby weather stations.

[0027] Understandably, after obtaining the relevant data, it is necessary to generate a set of executable candidate tasks based on the preset risk map and preset communication map, as well as the real-time status and the environmental status. The preset risk map is a rasterized layer, with each grid storing the risk level of the location. The risk level is pre-assessed based on historical geological disaster data, slope analysis, avalanche paths, etc. For example, a red grid indicates high risk (possible landslide). The preset communication map is also a rasterized layer, with each grid storing the communication signal strength of the location, such as the signal quality (0-100%) based on base station coverage prediction or actual measurement.

[0028] This involves combining the robot's real-time location, remaining energy, and current environmental conditions (such as severe weather increasing risk) to select which tasks the robot is capable of traversing, have acceptable path risks, and are generally accessible via communication from all pending tasks, thus forming a set of executable candidate tasks. For example, if robot A currently has 80% battery and is located near U001, and the risk map shows that the path to U002 has snow (high risk), but the path to U003 is safe, then U003 will be added to robot A's candidate set.

[0029] Furthermore, after determining the set of executable candidate tasks, a comprehensive benefit score needs to be calculated for the candidate combinations of multiple inspection robots within the set. Since there are multiple robots and multiple tasks, the overall benefit of each possible task allocation combination needs to be evaluated. This comprehensive benefit score is a quantitative value that comprehensively considers factors such as task priority, coverage ratio, robot energy consumption, path risk, and communication assurance. For example, for the combined scheduling of robot A executing task U003 and robot B executing task U004, the total benefit score is calculated. By comparing the scores of all feasible combinations, the task allocation result of the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result, i.e., deciding which robot will specifically execute which task. This result is output in the form of a task allocation table, such as allocating A->U003, B->U004, and C->U005.

[0030] It should be noted that during task execution, multiple robots may attempt to report the completion of the same task simultaneously. The system needs to ensure, through mechanisms (such as advance right locking), that only one completion result is valid, to avoid duplicate counting. The task status changes from pending assignment to in execution, and finally to completed.

[0031] Simultaneously, the task execution results are risk-classified and early warning information is output. That is, based on the abnormal situations (such as cracks or gas leaks) analyzed from the data returned by the robot, they are divided into red, orange, yellow and blue levels according to preset risk thresholds, and early warning information containing location, time, risk level and event type is generated and sent to the monitoring center or relevant management personnel.

[0032] Specifically, by modeling the inspection target area in a grid or segmented manner, the inspection range is discretized into multiple task units, and each task unit is configured with inspection type, priority, risk threshold, communication threshold and time window constraints, thereby forming a task queue that can be used for cluster scheduling.

[0033] Specifically, firstly, an inspection task set is established based on the inspection target area (such as high-altitude power transmission lines, road slopes, bridge and culvert sections, pipe corridors, tunnel entrances, etc.), and the area is discretized using a gridded or segmented approach, so that the inspection area can be transformed into a set of computable task units.

[0034] Let the spatial range of the inspection area be... Discretize it into a set of task units: ; in: : A collection of task units; : The i-th task unit; N: The total number of task units; Each task unit Each task is assigned a unique task ID (taskid) and its geometric range description (geomref) is recorded to support the mapping and matching of the task area after the robot is localized.

[0035] For each task unit Further configure its inspection attributes and scheduling constraint parameters to make the task schedulable. The task record should include at least the inspection type, inspection cycle, priority, risk threshold, communication threshold, minimum energy margin, and latest completion time.

[0036] The task attribute configuration vector is defined as follows:

[0037] in: Task basic attribute vector; : Task constraint parameter vector; Inspection type; Inspection cycle; Task priority; Risk threshold; Communication threshold; Minimum energy margin; Latest completion timestamp; Through the above configuration, tasks can not only be identified by the scheduling system, but also participate in subsequent multi-constraint judgments such as energy, risk and communication. In order to ensure the consistency of task status in the event of a link failure or delay, the task status is initialized for each task unit, initially set to PENDING, and the task version number is initialized to identify the scheduling round.

[0038] Task status is represented as: ; in: :Task The state at time t; PENDING: pending assignment; LOCKED: the task has been locked by the advance right, but has not yet entered the actual execution stage (e.g., waiting for the robot to arrive at the task area or waiting for the time window to start); RUNNING: in execution; DONE: completed; FAILED: failed and rolled back.

[0039] The task version number is used to distinguish the progress rights record and completion evidence of the same task in different scheduling rounds, thereby avoiding the old evidence from interfering with the new task status after the chain is restored.

[0040] When performing state acquisition and environmental state vector modeling of the inspection robot, the robot's real-time state and environmental state are collected, and a unified environmental state vector EnvState(t) is constructed. This enables the scheduling module to perceive changes in wind and snow, visibility, road surface adhesion, and communication obstruction in high-altitude environments, providing a unified input for subsequent energy prediction, risk assessment, and rolling replanning.

[0041] Specifically, the operating status of each robot is periodically collected to form a robot state vector, which improves the scalability of the state description.

[0042] Wherein, the robot position vector is defined as ; The robot power vector is defined as follows: ; Based on this, the robot state vector is defined as: ; in: : The state vector of the k-th robot at time t; Robot position coordinates; Remaining energy; : Power vector for energy replenishment; Communication quality score; Robot status records include at least robotid, pos, energywh, harvestsolar, harvestwind, commq, health, temp, etc., which enables the scheduling system to determine whether the robot has the ability to perform tasks.

[0043] Specifically, environmental information such as wind speed, snow intensity, visibility, and road surface adhesion is collected. At the same time, risk map version and communication map version are loaded, and a unified environmental state vector is constructed. To avoid the environmental vector being too long, it is split into physical environmental state vector and map index vector.

[0044] The environmental physical state vector is defined as follows: ; The map index vector is defined as follows: ; The final environment state vector is defined as follows: ; in: : Environmental physical state vector; W(t): wind speed; S(t): snow intensity; V(t): visibility; G(t): road surface adhesion; m(t): map index vector; riskmapid: risk map version identifier; commmapid: communication map version identifier; By modeling the above environmental vectors, the scheduling module can correct the risk assessment and communication assessment results in real time when there are sudden changes in wind, snow, visibility or road surface adhesion.

[0045] In this embodiment, the step of generating a set of executable candidate tasks based on a preset risk map, a preset communication map, the real-time status, and the environmental status includes: Based on the real-time status and the environmental status, the available energy and power consumption requirements of each inspection robot in the future time window are predicted, and based on the prediction results, an energy constraint relationship between the available energy and the power consumption requirements is constructed. Based on the preset risk map and preset communication map, the risk cost and communication accessibility of each inspection robot in the future time window are calculated, and a joint constraint relationship between risk and communication is constructed based on the calculation results. Based on the energy constraint relationship and the risk and communication joint constraint relationship, a set of executable candidate tasks is generated.

[0046] In this embodiment, based on the real-time state and the environmental state, the available energy and power consumption requirements of each inspection robot within a future time window are predicted. The future time window refers to a planned period of time starting from the current moment, such as the next 2 hours. The available energy refers to the robot's current remaining power multiplied by a safety factor (such as reserving 20% ​​of the power for return). For example, if the remaining power is 80% and the total capacity is 100Ah, then the available energy is 80%×100Ah×(1-0.2)=64Ah.

[0047] The power consumption requirement refers to the estimated power consumption required for the robot to perform a specific task, including power consumption for travel (estimated based on distance and terrain energy consumption models) and power consumption for operation (power consumption when sensors are turned on). For example, the robot needs to travel 5km to U003, which is expected to consume 30Ah of power, and it will consume 5Ah of power for 20 minutes of operation, for a total requirement of 35Ah.

[0048] In summary, establishing an energy constraint relationship between available energy and power consumption demand means ensuring that the demand is less than the available energy, i.e., 35Ah ≤ 64Ah, to meet the energy constraint. Then, based on the preset risk map and preset communication map, the risk cost and communication accessibility of each inspection robot in the future time window can be calculated.

[0049] The risk cost refers to the risk value accumulated by the robot as it travels along a certain path to the task unit. For example, the total risk cost is obtained by multiplying the risk level (0-1) of each grid on the path by the length of the grid it passes through.

[0050] Among them, communication reachability refers to whether the robot can maintain contact with the scheduling center during task execution. It can be estimated through the communication map whether the average signal strength on the task point and path is higher than the threshold.

[0051] Specifically, establishing a joint constraint relationship between risk and communication means combining these two factors to form a constraint. For example, the risk cost must be lower than a certain upper limit (e.g., total risk < 10), and the communication reachability must meet a minimum requirement (e.g., signal strength > 30%). For instance, if the risk cost of the path to U003 is 8 (acceptable), but the communication signal strength at the mission point is only 20% (below the threshold), then the joint constraint is not met.

[0052] In summary, based on the energy constraint and the joint risk and communication constraint, a set of executable candidate tasks is generated. The system iterates through all tasks to be assigned, checking each robot-task combination (e.g., A-U003, A-U004, B-U003, etc.) to see if it simultaneously satisfies both the energy constraint and the joint risk and communication constraint. Tasks (robots) that satisfy all constraints are then included in the set of executable candidate tasks. For example, A-U003 satisfies the energy constraint but not the communication constraint, so it is excluded; A-U004 satisfies both constraints, so it is added to the candidate set; B-U003 satisfies both constraints, so it is added to the candidate set. Each element in the set is a tuple (robot ID, task unit ID). Specifically, in this embodiment, a rolling decision window is set. (e.g., 2–10 minutes) and scheduling cycle (For example, 30–60 seconds), and perform the state acquisition → prediction → filtering → allocation process once in each scheduling cycle. If a sudden event occurs during the scheduling cycle, rolling replanning is triggered in advance, thereby ensuring that the scheduling strategy can be updated in real time with changes in the environment.

[0053] In particular, by predicting the available energy within the future rolling time window and modeling the energy consumption of task execution from multiple factors, the scheduling process is equipped with the ability to judge the executability under energy constraints, thereby avoiding task allocation failure under the conditions of low-temperature battery decay and uncertain energy replenishment.

[0054] Specifically, considering the random fluctuations in solar and wind power input, the available energy within the future time window is predicted based on the robot's current remaining energy and replenishment power. To avoid overly broad formulas, the robot's total replenishment power is first defined as follows: ; in: Total power replenishment for the robot; Solar power input; Wind energy input power.

[0055] Based on this, the available energy within the future time window is calculated using a discrete-time prediction method: ; in: Robot K has available energy within a future time window; The robot's current remaining energy; Energy conversion efficiency coefficient; : Sampling interval; M: Number of sampling points, satisfying ; : The j-th sampling time.

[0056] Understandably, the above predictions enable real-time estimation of future energy supply capacity when changes in wind and snow cause fluctuations in replenishment power.

[0057] To accurately describe the impact of slope climb, snow resistance, and low-temperature attenuation on energy consumption in high-altitude environments, a task execution energy consumption demand model is established. To ensure formula simplicity, task energy consumption is decomposed into three parts: movement energy consumption, climb energy consumption, and environmentally-related energy consumption. ; in: The total energy consumption required for robot k to perform task i; Energy consumption for moving on flat ground; Energy consumption for climbing slopes; Environmental energy consumption; Mobile energy consumption is defined as follows: ; The climbing energy consumption is defined as follows: ; in: : The path distance from the robot to the task area; Cumulative elevation gain; Energy consumption coefficient per unit distance; Energy consumption coefficient per unit height difference; To describe the impact of snow accumulation, adhesion, and temperature on energy consumption, environmental factors are defined as follows: ; Among them, environmentally added energy consumption is defined as: ; in: Environmental energy consumption influencing factors; G(t): road surface adhesion; S(t): snow intensity; T(t): temperature; Environmentally-added energy consumption weighting coefficient; : Environmental factor mapping function.

[0058] Understandably, the above decomposition model can be used to correct the task energy consumption estimate in real time based on the on-site environmental conditions, thereby improving the accuracy of scheduling feasibility assessment.

[0059] To prevent the robot from losing contact or being unable to return to the recharge point due to insufficient power after completing a task, a minimum energy margin for the task is introduced. Based on the predicted future available energy and the mission's energy consumption requirements, energy accessibility constraints are constructed: ; in: Minimum safety energy margin for the mission.

[0060] To further simplify the constraint expression, the energy-pressure ratio is defined as follows: ; Energy is considered feasible when the following conditions are met: ; in: : The energy pressure ratio of robot k performing task i.

[0061] Understandably, by introducing the energy pressure ratio, energy feasibility assessment can be transformed into a unified dimensionless constraint, which facilitates its participation in subsequent scoring calculations along with indicators such as risk and communication.

[0062] Furthermore, by introducing risk maps and communication maps, and combining them with real-time environmental state vectors, the execution risk and communication reachability of each task unit within the current time window are evaluated, forming risk constraints and communication constraints, which provide input for the subsequent generation of feasible candidate sets.

[0063] Specifically, the risk cost of the robot performing a task is calculated based on a risk map model. For task i and robot k, the risk cost function is defined as follows: ; in: The risk and cost of robot k performing task i; : Task unit space region; Robot position vector; : Environment state vector; Risk assessment mapping function, used to comprehensively consider risk factors such as terrain slope, snow intensity, wind speed, probability of collapse, and probability of landslide.

[0064] To ensure that the risks of the tasks are controllable, a risk threshold is set for each task. And construct risk constraints: ; in: Task risk upper limit threshold.

[0065] Specifically, to address the issue of unstable communication links caused by high-altitude terrain obstruction, a communication reachability score is calculated during the robot's task execution. For task i and robot k, communication reachability is defined as: ; in: : Communication reachability score, with a value range of [0,1]; : Communication reachability mapping function, used to integrate factors such as signal obstruction, base station distance, and terrain shadow areas.

[0066] In order to ensure that the task execution process can be transmitted back, a communication threshold is set. And construct communication constraints: ; in: Minimum threshold for task communication.

[0067] In this embodiment, the step of generating a set of executable candidate tasks based on the energy constraint relationship and the joint risk and communication constraint relationship includes: Obtain the pre-configured task attributes and constraint parameters for each task unit, and predict the estimated time consumption of the tasks to be executed by each inspection robot in each task unit based on the task attributes and constraint parameters. Based on the energy constraint relationship, the risk and communication joint constraint relationship, and the estimated time consumption, a multi-constraint joint feasibility determination is made for the task to be executed. Based on the judgment results, candidate tasks that meet the joint feasibility requirements of multiple constraints are identified, and a set of executable candidate tasks is generated.

[0068] In this embodiment, the pre-configured task attributes and constraint parameters for each task unit are first obtained. The task attributes include: task type (such as visible light inspection, infrared thermometry, gas detection), task priority (levels 1-5, with level 5 being the highest), and coverage ratio requirement (such as needing to collect images of at least 80% of the area). The constraint parameters include: latest completion time, maximum allowed operation time, etc. For example, the task attribute of U003 is "infrared thermometry", priority level 4, requiring 90% coverage of the area, and a maximum allowed completion time of 2 hours.

[0069] Then, based on the task attributes and constraint parameters, the estimated time for each inspection robot in each task unit to perform its task can be predicted. The estimated time needs to take into account differences between robot models (such as speed and sensor efficiency). For example, robot A is fast but its sensors require preheating, while robot B is slow but its sensors are ready to use immediately. For U003, it is predicted that A will take 1.5 hours and B will take 1.8 hours. The estimated time is used for subsequent feasibility assessment.

[0070] Furthermore, a multi-constraint joint feasibility determination needs to be performed on the task to be executed based on the energy constraint relationship, the risk and communication joint constraint relationship, and the estimated time consumption. This means checking the three constraints together: 1) whether the energy is sufficient (demand ≤ available); 2) whether the risk communication is acceptable; 3) whether the estimated time consumption is within the upper limit of the allowed operation time (e.g., ≤ 2 hours). For example, A-U003: energy is satisfied, risk communication is satisfied, and the estimated time consumption is 1.5 hours ≤ 2 hours, then it is determined to be feasible; B-U003: energy is satisfied, risk communication is satisfied, and the estimated time consumption is 1.8 hours ≤ 2 hours, also feasible.

[0071] Finally, based on the assessment results, candidate tasks that meet the multi-constraint joint feasibility requirements are identified, and a set of executable candidate tasks is generated. (Robot, task) pairs that satisfy all conditions are added to this set. This set serves as the basis for subsequent comprehensive benefit scoring calculations.

[0072] Specifically, by jointly determining energy feasibility, risk constraints, communication constraints, and time window constraints, an executable candidate task set FeasibleSet is generated, preventing unexecutable task combinations from entering the subsequent scheduling scoring stage, thereby reducing the computational complexity of real-time scheduling and improving scheduling reliability.

[0073] In this process, the estimated time for the robot to perform the task is calculated to obtain the estimated completion time: ; in: The estimated time for robot k to execute task i; : The path distance from the robot to the task area; The robot's current speed or estimated average speed; Task operation time (such as taking pictures, infrared scanning, sampling, etc.); The task time window constraint is defined as follows: ; in: The latest completion timestamp for the task.

[0074] Specifically, a joint feasibility determination is performed on the candidate task combination (i,k), and a set of feasibility constraints is constructed: ; in: : The set of feasibility constraints for robot k to perform task i; Energy-pressure ratio; Risk and cost; Communication score; when all the above constraints are met, the combination (i,k) is considered to belong to the executable candidate set.

[0075] Among them, combinations that satisfy the constraint set are added to the executable candidate set: ; in: : The set of executable candidate combinations at time t; through this step, only executable combinations are scored and sorted to avoid a large number of invalid combinations in the scheduling calculation.

[0076] Specifically, an example of the output data structure (FeasibleSet record) can be found in the output logic diagram. Figure 2 .

[0077] { "timestamp":1710001200, "feasiblepairs":[ {"taskid":"T012","robotid":"R03"}, {"taskid":"T012","robotid":"R05"}, {"taskid":"T018","robotid":"R01"} ] } In this embodiment, the step of calculating the comprehensive benefit score for candidate combinations of multiple inspection robots in the executable candidate task set includes: Construct a redundant penalty factor to punish multiple inspection robots for repeatedly performing the same task; Based on the energy constraint relationship and the risk and communication joint constraint relationship, as well as the task priority attribute and task coverage ratio in the task attributes, a multi-constraint benefit scoring function is constructed. Based on the multi-constraint benefit scoring function, a comprehensive benefit score is calculated for the candidate combinations of multiple inspection robots in the candidate task set.

[0078] In this embodiment, a redundancy penalty factor is first constructed to penalize multiple inspection robots for repeatedly performing the same task. In collaborative scheduling, resource waste caused by multiple robots performing the same task should be avoided. The redundancy penalty factor is a coefficient, for example, 0.8. When the same task is assigned to multiple robots, its benefit is multiplied by this coefficient and discounted, thereby reducing the total combined score. For example, if task U003 is assigned to two robots, its total benefit contribution will be reduced by 20%, guiding the scheduling to select non-repeating assignments.

[0079] Secondly, based on the energy constraint relationship, the risk and communication joint constraint relationship, and the task priority attribute and task coverage ratio in the task attributes, a multi-constraint benefit scoring function is constructed. The revenue scoring function is a mathematical expression that sums multiple factors with weights. For example, the revenue Rij for a robot i performing task j can be designed as: Rij = w1*(task priority) + w2*(expected coverage ratio) - w3*(energy cost normalization) - w4*(risk cost normalization) + w5*(communication quality normalization).

[0080] Here, w1~w5 are weighting coefficients, set through experiments or expert experience. Energy cost is calculated based on the ratio of actual energy consumption to available energy, risk cost is normalized based on the total path risk, and communication quality is taken as the signal strength at the task point. Task priority is the original value (1-5), and coverage ratio is the percentage (0-100%) that the robot is expected to cover.

[0081] Then, based on the multi-constraint benefit scoring function, a comprehensive benefit score can be calculated for candidate combinations of multiple inspection robots in the candidate task set. However, since there are multiple robots and multiple tasks, it is necessary to calculate the total benefit for each possible allocation combination. For example, the candidate set contains robots A and B and tasks U003, U004, and U005. One possible combination is A->U003, B->U004 (U005 is currently unavailable), then the total benefit = RAU003 + RBU004. Another combination is A->U003, B->U005, the total benefit = RAU003 + RBU005. The system iterates through all feasible combinations (using the Hungarian algorithm or a heuristic algorithm) and calculates the total score for each combination.

[0082] Specifically, by constructing a multi-constraint benefit scoring function, candidate task combinations are sorted, and a unique advancement right lock record is assigned to each task, so that only one robot is allowed to advance the same task within the same time window, thereby avoiding repeated execution and ensuring the consistency of task states. Corresponding data expressions can also be constructed based on the following content.

[0083] Specifically, to address the problem of multiple robots potentially performing the same task repeatedly, a redundancy penalty factor is constructed: ; in: Task redundancy penalty factor; Task status; : The last time the task was completed; : The current task is locked; Redundancy penalty mapping function; when the task is in RUNNING or is locked, Take a larger value to suppress duplicate allocation.

[0084] For the candidate task combination (i,k), a comprehensive reward scoring function is defined: ; in: Overall return score; Basic revenue items; : Penalty item.

[0085] The basic revenue item is defined as follows: ; The penalty term is defined as follows: ; in: Task priority; Coverage gain (indicates the degree of blind spot compensation or inspection coverage improvement); Energy-pressure ratio; Risk and cost; Communication score; Redundancy penalty factor; Weighting coefficients; through the above scoring function, a scheduling tendency is achieved that prioritizes high priority, high coverage benefits, low energy consumption pressure, low risk, and high communication quality.

[0086] In this embodiment, the step of determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the scoring calculation result includes: Based on the scoring calculation results, the inspection robot with the highest performance score for each task unit is determined. Based on the inspection robot with the highest score, a unique advance right lock record is assigned to each task unit. Based on the unique advance right lock record, the task allocation result of the collaborative scheduling of the multiple inspection robots is determined, so that only one inspection robot is allowed to execute the same task unit in the same time window.

[0087] In this embodiment, firstly, based on the scoring calculation results, the inspection robot with the highest score for each task unit is determined. Then, the combined scores calculated in the previous step are analyzed to extract which robot contributes the highest benefit when performing each task unit. For example, for task U003, among all feasible combinations, the benefit is 85 points when assigned to A, 80 points when assigned to B, and 60 points when assigned to C. Therefore, A is determined to be the robot with the highest score for U003. Similarly, B has the highest score for U004, and C has the highest score for U005.

[0088] Furthermore, based on the inspection robot with the highest score, a unique advance right lock record can be assigned to each task unit. This unique advance right lock record is an atomic operation, inserting a record into the system database for each task unit, recording the robot ID holding the lock and the locking timestamp. For example, the system writes in the task allocation table: the advance right of U003 is locked to robot A, with a status of "locked". Based on the unique advance right lock record, the task allocation result of the collaborative scheduling of multiple inspection robots can be determined, i.e., the final task allocation table. This result is sent to each robot through the communication network, informing A to go to U003, B to go to U004, and C to go to U005. At the same time, this locking mechanism ensures that only one inspection robot is allowed to execute the same task unit within the same time window. Other robots, even if they reach U003, will be informed by the system that they do not have the right to execute, thereby avoiding resource waste and task conflicts.

[0089] Specifically, for each task i, the robot with the highest score is selected from the set of executable candidate robots: ; in: The robot with the right to advance task i has the robot number. To ensure the uniqueness of task progress, a progress right lock record is generated: ; in: : Record of task progress control lock; : Lock the start and end times of the time window; Task version number; Lock the summary or signature for reconciliation arbitration; after the lock record is generated, the task status is updated to RUNNING, and only the locked bot is allowed to submit evidence of task progress.

[0090] In this embodiment, after the step of determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the unique advance right locking record, the method includes: Based on the task allocation results, the communication quality, remaining energy, and health status of the inspection robots performing the corresponding tasks are monitored in real time. When any real-time monitored indicator is lower than the corresponding preset event trigger threshold, the event trigger condition is determined to be met. When the event triggering condition is met, the advance right lock record held by the inspection robot is revoked, the task status of the corresponding task unit is rolled back to pending assignment, and the task version number is incremented and updated. Based on the current environmental conditions, the set of executable candidate tasks is redefined, and the comprehensive benefit score and advancement rights are recalculated. When the inspection robot that has obtained advancement rights is interrupted, it writes the execution trajectory and inspection data into the local cache structure. After the communication is restored, the completion evidence in the local cache is uploaded to the central scheduling system.

[0091] In this embodiment, after task allocation is completed, the system enters the real-time monitoring phase. Based on the task allocation results, the communication quality, remaining energy, and health status of the inspection robot performing the corresponding task are monitored in real time. Communication quality is measured by indicators such as the signal strength and packet loss rate transmitted by the robot; remaining energy is the real-time battery percentage; and health status includes motor temperature and sensor online status. Simultaneously, event trigger thresholds are preset, such as communication quality below 20%, remaining energy below 15%, and a fault code appearing in the health status. When any monitored indicator falls below the corresponding threshold, the event trigger condition is determined to be met.

[0092] When the event trigger condition is met, the system executes an exception handling procedure. First, it revokes the advance right lock record held by the inspection robot, that is, it removes the robot's lock on the task unit from the database, so that it no longer has the sole execution right for the task. At the same time, it rolls back the task status of the corresponding task unit to pending assignment, indicating that the task needs to be rescheduled.

[0093] In addition, the task version number is incremented. The task version number is an integer that increments each time the task status changes (such as reassignment) and is used for subsequent evidence verification. For example, U003 originally had a version number of 1, but after an exception occurred, it was rolled back and updated to version number 2.

[0094] Understandably, this necessitates re-determining the set of executable candidate tasks based on the current environmental state, and recalculating the comprehensive benefit score and allocating the advancement rights. This is equivalent to triggering a rolling replanning, which reschedules U003 along with other tasks to be assigned based on the latest robot state and environmental information.

[0095] It's worth noting that a local caching mechanism is additionally designed to handle communication interruptions. When communication is interrupted during task execution, the robot holding propulsion rights cannot upload data or receive new instructions in real time. In this situation, the robot writes its execution trajectory and inspection data into a local cache structure, such as storing it on an onboard solid-state drive, as a file to be synchronized. Once communication is restored, the robot uploads the locally cached completion evidence to the central scheduling system, which then verifies the data based on the evidence. This ensures that task data is not lost even in harsh communication environments.

[0096] Specifically, by making real-time judgments on events such as sudden changes in environmental state, robot malfunctions, communication interruptions, and insufficient energy, the system automatically revokes the task advancement rights and rolls back the task state. Then, under new environmental constraints, it regenerates a feasible candidate set and reallocates the advancement rights, thereby achieving real-time adaptive scheduling in harsh environments.

[0097] This includes continuous monitoring of the robot's and environment's states, and defining event triggering conditions: ; in: : Event trigger flag for robot k at time t; Communication quality score; Remaining energy; Health indicators; Event trigger threshold; : Logical OR operation; when the event trigger condition is met, the system considers that the robot's current task is at risk of failure.

[0098] Specifically, when the robot holding the advancement rights for task i triggers an event, the lock record is revoked and the task state is rolled back: ; At the same time, the task version number is incremented and updated. ; in: : Task version number; the version number increments to distinguish different scheduling rounds and avoid old evidence from incorrectly advancing to new states.

[0099] In particular, after the advance right is revoked, the energy prediction, risk assessment, communication assessment and feasible candidate set filtering process are re-executed, and the benefit scoring and advance right locking allocation are re-performed, so that the system can automatically adjust the scheduling strategy when there are sudden changes in wind and snow, increased communication blockage or sudden changes in energy supply.

[0100] Specifically, by designing task execution and offline caching upload in case of network interruption, the robot is allowed to continue executing tasks under network interruption conditions, and the execution trajectory and inspection data are cached locally. This allows the system to continue to complete inspection tasks even without network access. Once the network is restored, the system will submit completion evidence and enter the arbitration and reconciliation process.

[0101] Among them, the robot that obtains the right to advance and lock the record performs inspections according to the task type, including image acquisition, infrared scanning, structural anomaly detection, snow cover detection and other operations, and generates trajectory records and data summaries.

[0102] When the robot's communication quality falls below a threshold, it enters offline execution mode and writes the execution results to a local cache structure. ; in: Offline caching completion evidence; trajhash: trajectory summary hash; datahash: inspection data summary hash; quality: data quality score; : Task version number; Through the caching structure, the robot can ensure the traceability of evidence even in the event of a chain break.

[0103] In this embodiment, the step of risk-classifying the task execution result and outputting early warning information includes: A comprehensive score is calculated for the completion evidence of the task execution result, wherein the comprehensive score is obtained by weighting the data quality score, trajectory coverage score and time window legality score; The legality of the completed evidence shall be verified, including at least verifying whether the inspection robot that submitted the evidence is the robot holding the right to advance and whether the task version number in the evidence is consistent with the current task version number. When multiple inspection robots submit evidence of completion for the same task unit, the evidence with the highest comprehensive score is selected as the only valid evidence, and the task status of the corresponding task unit is set to completed, while the remaining evidence is archived. When the evidence is invalid or the overall score is lower than the preset task completion threshold, the task status of the corresponding task unit will be rolled back to pending assignment and a rolling replanning will be triggered. Based on a preset risk threshold, the task execution results are classified into risk levels, abnormal events are categorized and spatially aggregated, and warning information is output that includes at least the risk level, event type, affected area, and timestamp.

[0104] In this embodiment, after the robot completes the task, it uploads the completion evidence (including inspection images, sensor data, trajectory files, etc.) to the system, and performs a comprehensive score calculation on the completion evidence of the task execution result.

[0105] The overall score is calculated by weighting three sub-items: 1) Data quality score: assesses image clarity, data integrity, etc., such as by calculating the average gradient of the image and checking the integrity of the data packets; 2) Trajectory coverage score: assesses whether the robot's actual driving path covers the key areas required by the task unit, for example, a coverage rate of 95% will receive a high score; 3) Time window validity score: checks whether the task completion time is within the allowed time window, and points are deducted if the time is exceeded. Overall score = w6 * Data quality + w7 * Trajectory coverage + w8 * Time window validity.

[0106] Secondly, the legality of the completed evidence is verified. This verification includes: 1) whether the robot submitting the evidence is the current holder of the task unit's advancement rights (this needs to be compared with the locked records); 2) whether the task version number carried in the evidence is consistent with the current system version number (to prevent interference from old version task data). If the verification passes, the evidence is legal.

[0107] It should be noted that since multiple robots may submit completion evidence for the same task unit (e.g., multiple robots submitting simultaneously after communication is restored), the system must select the evidence with the highest comprehensive score as the only valid evidence. This evidence will be used to advance the status of the corresponding task unit as completed. Evidence from other robots will only be saved as archived data and will not affect the task status.

[0108] If the evidence is invalid (e.g., submitted by someone other than the holder) or the overall score is lower than the preset task completion threshold (e.g., 60 points), the task is deemed not to have been completed effectively. The system will roll back the task status of the corresponding task unit to pending assignment and trigger rolling replanning, that is, reschedule the robot to execute the task.

[0109] Finally, the system performs real-time analysis on the data collected during task execution and classifies the task execution results into risk levels based on preset risk thresholds. For example, if a crack width exceeds 5mm, it is classified as high risk (red); if the gas concentration exceeds the limit but does not reach the emergency value, it is classified as medium risk (orange).

[0110] Simultaneously, abnormal events are classified and spatially aggregated, i.e., the anomaly type is identified (cracks, gas leaks, landslides, etc.), and adjacent anomalies of the same type are merged into event areas. The final output includes warning information that at least includes the risk level, event type, affected area, and timestamp. For example: [Red Warning] 2025-03-10 14:30, a crack was found on the eastern slope of unit U003, affecting the grid area (120,345)-(122,348). Please take immediate action. This information is pushed to relevant personnel via SMS, platform pop-ups, etc.

[0111] Specifically, by performing quality scoring, version consistency verification, and legality verification of the completion evidence submitted by the robot, and by implementing an account reconciliation arbitration mechanism in the event of multiple evidence conflicts, the system ultimately determines the only valid task completion result and advances the task status, thereby ensuring that the task execution result can still converge to a consistent state even under the condition of a broken chain.

[0112] The comprehensive score is calculated for the completion evidence submitted by the robot: ; in: : Comprehensive score of evidence of task i's completion; Data quality score (clarity, infrared validity, anomaly detection reliability, etc.); Track coverage score (whether the task area is covered); Time window validity score (whether it was completed within the locked time window); Weighting coefficient.

[0113] Among these measures, the legality of evidence is determined, and only evidence that meets the requirements of advancing the right to proceed and version consistency can advance the task status.

[0114] Defining the determination of the legality of evidence: ; in: Indicators of evidence legality; : Task advancement rights held by the robot; The task version number recorded in the evidence; : Current task version number; : Logical AND operation.

[0115] Among them, setting a task completion threshold The task is considered complete when the evidence is valid and the score meets the requirements. ; When multiple robots submit evidence for the same task, the arbitration rule shall be used to select the only valid evidence. The winning robot was ultimately chosen. Evidence advancing the task status is used to advance the task status; other evidence is archived but does not change the task status.

[0116] When the task completion determination is met, the task status is pushed: .

[0117] Specifically, if the task completion determination is invalid or the evidence conflict cannot be resolved through arbitration, a rollback will be executed and a rolling replanning will be triggered: .

[0118] In addition, it is necessary to spatially aggregate and hierarchically map the results of task completion, risk costs and abnormal events to form clear risk levels, affected areas and time markers, and support multiple early warning information release methods to achieve a closed loop of emergency response in high-altitude and harsh environments.

[0119] This involves classifying and mapping the fusion risk values ​​of the task area or grid area to form an early warning level. Let the risk score for the task area be... The risk level is defined as follows: ; in: The warning level of grid or region g at time t; : Integrated risk score (which can be calculated by overlaying the risk map output with abnormal events); Risk classification threshold.

[0120] This involves classifying and judging abnormal events, such as landslides, snow-covered roads, strong wind hazards, communication blackout areas, and equipment failures, and spatially aggregating similar events to form a set of impact areas. ; in: : Set of areas affected by the event; g: Spatial grid or task area unit.

[0121] Understandably, the warning results are output in the form of structured data, including risk score, risk level, event type, affected area and timestamp, and can be released through wireless broadcast, satellite link, command center system interface or mobile terminal push.

[0122] This embodiment divides the target inspection area into multiple task units and collects the real-time status of each inspection robot within the target inspection area and the environmental status of its surrounding environment. Based on a preset risk map and a preset communication map, as well as the real-time status and environmental status, a set of executable candidate tasks is generated. A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and the task allocation result for the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result. Based on the task allocation result, a unique and valid task execution result is determined and the task status is advanced. Risk classification and early warning information are then output for the task execution result. In other words, by dividing the target inspection area into multiple task units, fine-grained decomposition of complex terrain is achieved, enabling quantifiable management of large-scale inspection tasks. Simultaneously, the real-time status and environmental status of each robot are collected, thereby providing dynamic... Accurate input data ensures precise scheduling decisions. Combined with pre-set risk and communication maps, a set of executable candidate tasks is generated, fully considering the unique terrain risks, weather variations, and communication limitations of high-altitude areas. This guarantees the safety and feasibility of task planning from the outset. By calculating the comprehensive benefit score of candidate combinations and determining task allocation results, resource optimization among multiple robots can be achieved, avoiding task conflicts and duplication of work. A unique and effective task execution result is determined, and task status is advanced, ensuring the traceability and consistency of the task loop. Finally, risk classification of task execution results and output of early warning information provide intuitive decision support for maintenance personnel, enabling timely detection and handling of safety hazards in high-altitude environments. This achieves intelligent, adaptive, and collaborative scheduling of inspection robot clusters in high-altitude environments, effectively improving inspection coverage and data reliability.

[0123] Furthermore, this application also proposes a collaborative scheduling device for inspection robot clusters in high-altitude environments, referring to... Figure 3 The high-altitude environment inspection robot cluster collaborative scheduling device includes: The data acquisition module 10 is used to divide the target inspection area into multiple task units and collect the real-time status of each inspection robot in the target inspection area and the environmental status of its environment. Evaluation module 20 is used to generate a set of executable candidate tasks based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status. The calculation module 30 is used to calculate the comprehensive benefit score of the candidate combinations of multiple inspection robots in the executable candidate task set, and determine the task allocation result of the collaborative scheduling of the multiple inspection robots based on the score calculation result. The early warning module 40 is used to determine the unique and valid task execution result and advance the task status based on the task allocation result, and to classify the risk of the task execution result and output early warning information.

[0124] This embodiment divides the target inspection area into multiple task units and collects the real-time status of each inspection robot within the target inspection area and the environmental status of its surrounding environment. Based on a preset risk map and a preset communication map, as well as the real-time status and environmental status, a set of executable candidate tasks is generated. A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and the task allocation result for the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result. Based on the task allocation result, a unique and valid task execution result is determined and the task status is advanced. Risk classification and early warning information are then output for the task execution result. In other words, by dividing the target inspection area into multiple task units, fine-grained decomposition of complex terrain is achieved, enabling quantifiable management of large-scale inspection tasks. Simultaneously, the real-time status and environmental status of each robot are collected, thereby providing dynamic... Accurate input data ensures precise scheduling decisions. Combined with pre-set risk and communication maps, a set of executable candidate tasks is generated, fully considering the unique terrain risks, weather variations, and communication limitations of high-altitude areas. This guarantees the safety and feasibility of task planning from the outset. By calculating the comprehensive benefit score of candidate combinations and determining task allocation results, resource optimization among multiple robots can be achieved, avoiding task conflicts and duplication of work. A unique and effective task execution result is determined, and task status is advanced, ensuring the traceability and consistency of the task loop. Finally, risk classification of task execution results and output of early warning information provide intuitive decision support for maintenance personnel, enabling timely detection and handling of safety hazards in high-altitude environments. This achieves intelligent, adaptive, and collaborative scheduling of inspection robot clusters in high-altitude environments, effectively improving inspection coverage and data reliability.

[0125] It should be noted that each module in the above-mentioned device can be used to implement each step in the above-mentioned method and achieve the corresponding technical effect. This embodiment will not elaborate further here.

[0126] Reference Figure 4 , Figure 4 This is a schematic diagram of the hardware operating environment of the device involved in the embodiments of this application.

[0127] like Figure 4As shown, the device may include: a processor 1001, such as a CPU; a communication bus 1002; a user interface 1003; a network interface 1004; and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (e.g., a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0128] Those skilled in the art will understand that Figure 4 The structure shown does not constitute a limitation on the device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0129] like Figure 4 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a program for collaborative scheduling of inspection robot clusters in high-altitude environments.

[0130] exist Figure 4 In the device shown, the network interface 1004 is mainly used for data communication with an external network; the user interface 1003 is mainly used for receiving user input commands; the device calls the high-altitude environment inspection robot cluster collaborative scheduling method program stored in the memory 1005 through the processor 1001, and performs the following operations: The target inspection area is divided into multiple task units, and the real-time status of each inspection robot in the target inspection area and the environmental status of its environment are collected. Based on the preset risk map and preset communication map, as well as the real-time status and the environmental status, a set of executable candidate tasks is generated; A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and the task allocation result of the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result. Based on the task allocation results, a unique and valid task execution result is determined and the task status is advanced. The task execution result is then risk-classified and early warning information is output.

[0131] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0132] The above description is merely a specific embodiment 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. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0133] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the high-altitude environment inspection robot cluster collaborative scheduling method in the above embodiments.

[0134] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0135] The aforementioned computer-readable storage medium may be included in the collaborative scheduling equipment for inspection robot clusters in high-altitude environments; or it may exist independently and not be installed in the collaborative scheduling equipment for inspection robot clusters in high-altitude environments.

[0136] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the high-altitude environment inspection robot cluster collaborative scheduling device, enable the high-altitude environment inspection robot cluster collaborative scheduling device to: The target inspection area is divided into multiple task units, and the real-time status of each inspection robot in the target inspection area and the environmental status of its environment are collected. Based on the preset risk map and preset communication map, as well as the real-time status and the environmental status, a set of executable candidate tasks is generated; A comprehensive benefit score is calculated for candidate combinations of multiple inspection robots in the executable candidate task set, and the task allocation result of the collaborative scheduling of the multiple inspection robots is determined based on the score calculation result. Based on the task allocation results, a unique and valid task execution result is determined and the task status is advanced. The task execution result is then risk-classified and early warning information is output.

[0137] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0138] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings; for example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0139] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0140] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described collaborative scheduling method for inspection robot clusters in high-altitude environments, thereby solving the technical problems of collaborative scheduling methods for inspection robot clusters in high-altitude environments. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the collaborative scheduling method for inspection robot clusters in high-altitude environments provided in the above embodiments, and will not be repeated here.

[0141] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0142] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0143] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (e.g., ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0144] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for collaborative scheduling of inspection robot swarms in high-altitude environments, characterized in that, The method includes the following steps: dividing the target inspection area into multiple task units, and collecting the real-time status of each inspection robot within the target inspection area and the environmental status of its environment; generating an executable candidate task set based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status; calculating a comprehensive benefit score for candidate combinations of multiple inspection robots in the executable candidate task set, and determining the task allocation result for the collaborative scheduling of the multiple inspection robots based on the score calculation result; determining a unique and valid task execution result based on the task allocation result and advancing the task status, and performing risk classification and outputting early warning information for the task execution result.

2. The method as described in claim 1, characterized in that, The step of generating a set of executable candidate tasks based on a preset risk map, a preset communication map, the real-time state, and the environmental state includes: predicting the available energy and power consumption requirements of each inspection robot within a future time window based on the real-time state and the environmental state, and constructing an energy constraint relationship between the available energy and the power consumption requirements based on the prediction results; calculating the risk cost and communication accessibility of each inspection robot within a future time window based on the preset risk map and the preset communication map, and constructing a joint risk and communication constraint relationship based on the calculation results; and generating a set of executable candidate tasks based on the energy constraint relationship and the joint risk and communication constraint relationship.

3. The method as described in claim 2, characterized in that, The step of generating a set of executable candidate tasks based on the energy constraint relationship and the risk and communication joint constraint relationship includes: obtaining the pre-configured task attributes and constraint parameters of each task unit, and predicting the estimated time consumption of the tasks to be executed by each inspection robot in each task unit based on the task attributes and constraint parameters; performing a multi-constraint joint feasibility determination on the tasks to be executed based on the energy constraint relationship, the risk and communication joint constraint relationship and the estimated time consumption; and determining candidate tasks that meet the multi-constraint joint feasibility requirements based on the determination result, and generating a set of executable candidate tasks.

4. The method as described in claim 3, characterized in that, The step of calculating the comprehensive benefit score for candidate combinations of multiple inspection robots in the executable candidate task set includes: constructing a redundancy penalty factor to penalize multiple inspection robots for repeatedly executing the same task; constructing a multi-constraint benefit scoring function based on the energy constraint relationship and the risk and communication joint constraint relationship, as well as the task priority attribute and task coverage ratio in the task attributes; and calculating the comprehensive benefit score for candidate combinations of multiple inspection robots in the candidate task set based on the multi-constraint benefit scoring function.

5. The method as described in claim 4, characterized in that, The step of determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the scoring calculation result includes: determining the inspection robot with the highest task score corresponding to each task unit based on the scoring calculation result; assigning a unique advance right lock record to each task unit based on the inspection robot with the highest score; and determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the unique advance right lock record, so that only one inspection robot is allowed to execute the same task unit in the same time window.

6. The method as described in claim 5, characterized in that, After the step of determining the task allocation result of the collaborative scheduling of the multiple inspection robots based on the unique advance right lock record, the method includes: real-time monitoring of the communication quality, remaining energy, and health status of the inspection robots performing the corresponding tasks based on the task allocation result; when any real-time monitored indicator is lower than the corresponding preset event trigger threshold, the event trigger condition is determined to be met; when the event trigger condition is met, the advance right lock record held by the inspection robot is revoked, the task status of the corresponding task unit is rolled back to pending allocation, and the task version number is incremented and updated; based on the current environmental state, the set of executable candidate tasks is re-determined, and the comprehensive benefit score and advance right allocation are recalculated, wherein when the inspection robot that has obtained the advance right experiences a communication interruption, it writes the execution trajectory and inspection data into a local cache structure, and uploads the completion evidence in the local cache to the central scheduling system after communication is restored.

7. The method as described in claim 1, characterized in that, The steps of risk classification and early warning information output for the task execution results include: calculating a comprehensive score for the completion evidence of the task execution results, wherein the comprehensive score is obtained by weighting data quality score, trajectory coverage score, and time window legality score; verifying the legality of the completion evidence, wherein the legality verification includes at least verifying whether the inspection robot submitting the evidence is the robot holding the advancement right and whether the task version number in the evidence is consistent with the current task version number; when multiple inspection robots submit completion evidence for the same task unit, the evidence with the highest comprehensive score is selected as the only valid evidence, the task status of the corresponding task unit is set to completed, and the remaining evidence is archived; when the evidence is illegal or the comprehensive score is lower than the preset task completion threshold, the task status of the corresponding task unit is rolled back to pending allocation and a rolling replanning is triggered; according to the preset risk threshold, risk classification is performed on the task execution results, abnormal events are classified and spatially aggregated, and early warning information including at least risk level, event type, affected area, and timestamp is output.

8. A collaborative scheduling device for a cluster of inspection robots in high-altitude environments, characterized in that, The high-altitude environment inspection robot cluster collaborative scheduling device includes: a data acquisition module, used to divide the target inspection area into multiple task units and acquire the real-time status of each inspection robot in the target inspection area and the environmental status of its environment; an evaluation module, used to generate a set of executable candidate tasks based on a preset risk map and a preset communication map, as well as the real-time status and the environmental status; a calculation module, used to perform a comprehensive benefit score calculation on the candidate combinations of multiple inspection robots in the set of executable candidate tasks, and determine the task allocation result of the collaborative scheduling of the multiple inspection robots based on the score calculation result; and an early warning module, used to determine the unique and valid task execution result based on the task allocation result and advance the task status, and perform risk classification on the task execution result and output early warning information.

9. A collaborative scheduling device for a cluster of inspection robots in high-altitude environments, characterized in that, The high-altitude environment inspection robot cluster collaborative scheduling device includes: a memory, a processor, and a high-altitude environment inspection robot cluster collaborative scheduling method program stored in the memory and executable on the processor. The high-altitude environment inspection robot cluster collaborative scheduling method program is configured to implement the steps of the high-altitude environment inspection robot cluster collaborative scheduling method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a program for implementing a collaborative scheduling method for a cluster of inspection robots in a high-altitude environment. The program for implementing the collaborative scheduling method for a cluster of inspection robots in a high-altitude environment is executed by a processor to implement the steps of the collaborative scheduling method for a cluster of inspection robots in a high-altitude environment as described in any one of claims 1 to 7.

Images