A state-aware based robot scheduling method and system

Abstract

This application relates to the field of robot management technology, and in particular to a robot scheduling method and system based on state awareness. The method includes: acquiring a set of control tasks; analyzing the set of control tasks to determine task types; acquiring environmental perception data and robot operation data in real time; analyzing the operation data to determine the real-time operating state; determining the real-time priority of each control task based on the real-time operating state, the environmental perception data, and the task type; acquiring resource availability status in real time; determining the set of preemptible resources per unit time based on the resource availability status; allocating the set of preemptible resources according to the real-time priority; and scheduling the robot to operate according to the allocation result. This achieves optimized matching of resources and tasks, executes necessary resource preemption operations, improves response speed, throughput, and overall operational efficiency, and ensures that tasks are completed in an orderly manner according to dynamic priorities.

Description

Technical Field

[0001] This application relates to the field of robot management technology, and in particular to a state-aware robot scheduling method and system. Background Technology

[0002] With the widespread application of robotics technology, robot systems are becoming increasingly prevalent in scenarios such as industrial automation, service robots, and smart warehousing. These systems typically operate in dynamic, resource-constrained heterogeneous environments, involving various resource constraints. In this environment, robots need to handle a set of concurrent control tasks, including high-priority and low-priority tasks.

[0003] However, traditional robot scheduling methods mainly rely on static priority mechanisms, such as first-in-first-out (FIFO) queues for resource allocation. This allocation method, when an urgent task arises, still requires waiting for the preceding task to complete, causing scheduling delays and reducing overall system throughput and resource utilization. Summary of the Invention

[0004] This application provides a state-aware robot scheduling method and system to solve the above problems.

[0005] Firstly, this application provides a state-aware robot scheduling method, the method comprising: Obtain the control task set; analyze the control task set to determine the task type; Real-time acquisition of environmental perception data and robot operation data; analysis of the operation data to determine the real-time operating status; The real-time priority of each control task is determined based on the real-time operating status, the environmental perception data, and the task type. Real-time acquisition of resource carrying status; and determination of the set of preemptible resources per unit time based on the resource carrying status. Based on the real-time priority, the set of preemptible resources is allocated, and the robot is scheduled to operate according to the allocation result.

[0006] This solution acquires a control task set, effectively identifying the complete range of concurrent tasks and ensuring all concurrent tasks are included in the scheduling scope. Analyzing the control task set determines task types, ensuring each task has a clear type definition, thus avoiding scheduling decisions based on vague or unclassified task information and supporting targeted processing of priority calculation and resource allocation. Real-time acquisition of environmental perception data and robot operation data enables immediate perception of dynamic threats in the external environment, providing an environmental basis for priority adjustment; and accurately capturing changes in the robot's own state. Analyzing operation data determines real-time operating status, allowing timely responses to environmental threats or resource fluctuations, providing immediate data support for dynamic priority adjustment, and improving the timeliness and accuracy of decision-making. Based on real-time operating status, environmental perception data, and task type, the real-time priority of each control task is determined, ensuring that priorities reflect multi-dimensional factors, solving the problem of static priority rigidity, and achieving dynamic optimization of task urgency, thereby providing a precise ranking basis for resource allocation and avoiding scheduling decision bias. Real-time acquisition of resource availability status is used to determine the set of preemptible resources per unit time, providing an available resource pool for efficient resource preemption. This ensures rapid resource release when resources are limited, improving resource utilization and scheduling flexibility. Based on real-time priorities, the set of preemptible resources is allocated, and robots are scheduled to operate according to the allocation results. This achieves optimized matching of resources and tasks, executes necessary resource preemption operations, improves response speed, throughput, and overall operational efficiency, and ensures tasks are completed in an orderly manner according to dynamic priorities.

[0007] Optionally, determining the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type includes: Acquire and analyze user input to determine the urgency of the interaction; Analyze the environmental perception data to determine the distance and trajectory of obstacles; The environmental conditions are determined based on the real-time operating status, the distance to the obstacle, and the movement trajectory. The real-time priority of each control task is determined based on the urgency of the interaction, the real-time running status, the impact of the environmental status, and the task type.

[0008] This solution acquires and analyzes user input to determine the urgency of interactions, addressing the issue of a single dimension in priority calculation. It ensures that task scheduling responds to user interactions in real time and avoids operational risks caused by delays in user requests. Priority decisions are no longer rigidly based on static task types but incorporate external interaction factors, improving reliability in user-driven scenarios. Analyzing environmental perception data determines obstacle distances and movement trajectories, addressing the problem of neglecting environmental state factors. This enhances the real-time performance and accuracy of priority calculations, preventing task failures in dense obstacle environments due to a lack of dynamic data, thereby reducing safety hazards. Based on real-time operating status, obstacle distances, and movement trajectories, the solution determines the impact of environmental states, enabling collaborative assessment of the environment and robot states. This ensures that scheduling decisions adapt to dynamic changes and avoids congestion or inefficiency caused by unreasonable resource allocation. Based on interaction urgency, real-time operating status, environmental state impact, and task type, the solution determines the real-time priority of each control task, improving throughput and robustness in resource-constrained environments while avoiding resource waste caused by unreasonable interruptions of low-priority tasks.

[0009] Optionally, determining the environmental state influence based on the real-time operating status, the obstacle distance, and the movement trajectory includes: Based on the real-time operating status, a preset safe distance threshold is obtained; By comparing the safe distance threshold with the distance to the obstacle, it is determined whether there is any operational impact at present; If there are no operational impacts at present, analyze the real-time operating status to determine the robot's direction of movement; Analyze the motion trajectory to determine the real-time motion angle between the obstacle's motion direction and the robot's motion direction; The influence of environmental conditions is determined based on the real-time motion angle.

[0010] This solution obtains preset safe distance thresholds based on real-time operational status, providing a quantitative standard to ensure that environmental impact assessments are based on a unified safety boundary, avoiding assessment biases caused by inconsistent distance thresholds. By comparing the safe distance thresholds with obstacle distances, it determines whether operational impacts exist, quickly assesses whether obstacles pose an immediate threat, and filters out low-risk scenarios that do not require in-depth analysis. If no operational impact exists, the real-time operational status is analyzed to determine the robot's movement direction, ensuring that the environmental impact assessment considers the robot's dynamic behavior. The motion trajectory is analyzed to determine the real-time motion angle between the obstacle's movement direction and the robot's movement direction, quantifying the relative motion trend between the obstacle and the robot, providing numerical basis for mapping environmental impact levels, and enhancing the objectivity of threat assessment. Based on the real-time motion angle, the environmental impact is determined, achieving a simple and effective classification of potential collision risks.

[0011] Optionally, determining the influence of environmental conditions based on the real-time motion angle includes: Analyze the real-time motion angle; when the real-time motion angle is less than 90 degrees, determine the relative velocity of the obstacle based on the motion trajectory. Obtain obstacle size data; calculate the effective collision cross section based on the obstacle size data and the obstacle distance; Calculate the dynamic threat coefficient based on the relative velocity of the obstacle and the effective collision cross section; The dynamic threat coefficient is mapped to the environmental state impact value.

[0012] This solution analyzes real-time motion angles. When the real-time motion angle is less than 90 degrees, it determines the relative velocity of obstacles based on the stated motion trajectory, quantifying the rate at which obstacles approach the robot, providing a crucial velocity dimension input for threat calculation. It acquires obstacle size data, quantifies the size characteristics of obstacles, and ensures that threat assessment comprehensively considers physical size factors. Based on obstacle size data and obstacle distance, it calculates the effective collision cross-section, quantifying the spatial probability of collision, providing a crucial dimension input for the dynamic threat coefficient. Based on the relative velocity of obstacles and the effective collision cross-section, it calculates the dynamic threat coefficient, combining velocity and spatial factors to quantify the real-time threat level of obstacles to the robot. Finally, it maps the dynamic threat coefficient to environmental state impact values, realizing the transformation from numerical calculation to decision output.

[0013] Optionally, determining the set of preemptible resources per unit time based on the resource carrying status includes: Analyze resource utilization status to identify the fragmentable and idle resources occupied by low-priority tasks; Get the task execution progress of low-priority tasks; Calculate the interruption loss based on the task execution progress; Based on the interruption loss, a resource preemptibility score is determined; Based on the resource preemptibility score, the set of preemptible resources per unit time is determined according to the shardable resources and the idle resources.

[0014] This solution analyzes resource utilization status, identifies the fragmentable and idle resources occupied by low-priority tasks, and avoids blind operations in resource-congestion scenarios. It obtains the execution progress of low-priority tasks and introduces real-time feedback on task progress to ensure that subsequent interruption loss calculations are based on actual task status rather than static presets, thus supporting dynamic decision-making. Based on task execution progress, interruption losses are calculated to ensure that resource preemption decisions consider the actual cost of task interruptions, avoiding resource waste caused by unreasonable interruptions. Based on interruption losses, a resource preemptibility score is determined to dynamically assess the feasibility of resource preemption, thereby optimizing resource management and reducing scheduling risks. Based on the resource preemptibility score, and according to fragmentable and idle resources, the set of preemptible resources per unit time is determined, enabling high-priority tasks to acquire the resources they need immediately, thereby improving responsiveness and alleviating resource congestion problems.

[0015] Optionally, determining the real-time priority of each control task based on the interaction urgency, the real-time running status, the environmental state impact, and the task type includes: Analyze the task type to identify whether the task is associated with a preset high-value goal; If high-value targets are associated, obtain the target value weight coefficient; Analyze the real-time operating status to determine the robot's remaining battery power; Analyze the control task set to determine the robot's control operation set; Based on the set of control operations, the estimated energy consumption of the task is determined; Calculate the probability that the task can be completed based on the estimated energy consumption of the task and the remaining power. Each control task is determined based on the interaction urgency, the environmental state influence value, the target value weight coefficient, and the task completion probability.

[0016] This solution analyzes task types and identifies whether tasks are associated with preset high-value objectives, addressing the issue of a single dimension in priority calculation and avoiding reliance solely on static task types for priority allocation. If associated with high-value objectives, it obtains the objective value weight coefficient, strengthening multi-dimensional priority evaluation and dynamically adjusting priority decisions based on task value. It analyzes real-time operating status to determine the robot's remaining battery power, enhancing dynamic response capabilities. Addressing insufficient real-time response capabilities, it adjusts priorities by integrating the robot's own status. It analyzes the control task set to determine the robot's control operation set, supporting efficient resource management and indirectly alleviating resource congestion and inefficient management. Based on the control operation set, it determines the expected energy consumption of tasks, optimizing resource allocation and avoiding waste. Based on the expected energy consumption and remaining battery power, it calculates the probability of task completion, addressing the lack of a dynamic feedback loop and improving decision-making accuracy and efficiency. Based on interaction urgency, environmental state influence, objective value weight coefficient, and task completion probability, it determines the real-time priority of each control task, achieving efficient and non-blocking resource management.

[0017] Optionally, calculating the interruption loss based on the task execution progress includes: Analyze the task execution progress to determine the percentage of work completed; Obtain the value decay curve corresponding to the task type; Calculate the schedule loss factor based on the percentage of completed work and the value decay curve; Obtain historical task data, and determine the historical resource usage level based on the historical task data; Determine the sunk cost of resources based on the historical resource utilization levels. The interruption loss is determined by weighted summation of the schedule loss factor and the sunk cost of resources.

[0018] This solution analyzes task execution progress, determines the percentage of completed work, and ensures that interruption loss assessment is based on actual execution status, avoiding subjective estimation. It obtains the value decay curve corresponding to each task type, ensuring that interruption loss assessment considers the value characteristics of each task type. Based on the percentage of completed work and the value decay curve, it calculates the schedule loss factor, quantifying the schedule value decay caused by interruption, providing a schedule-dimensional loss indicator for interruption loss calculation. It acquires historical task data, determines historical resource occupancy levels, assesses the amount of resources potentially wasted during interruptions, and compensates for the neglect of sunk resource costs. Based on historical resource occupancy levels, it determines sunk resource costs, quantifies the potential waste of consumed resources, incorporates interruption loss into the resource cost dimension, and addresses the problem of interruption decisions not considering sunk resource costs. Finally, it weights and sums the schedule loss factor and sunk resource costs to determine interruption losses, optimizes resource preemption mechanisms, and improves efficiency.

[0019] Optionally, determining the resource preemptibility score based on the interruption loss includes: Determine the basic preemption cost based on the aforementioned interruption loss; Analyze the task types to determine critical protection weights; Based on the resource carrying capacity status, determine the resource scarcity coefficient; Based on the basic preemption cost, the critical protection weight, and the resource scarcity coefficient, a resource preemptibility score is generated.

[0020] This solution determines the basic preemption cost based on the interruption loss, avoiding additional conversion or computational overhead and ensuring consistency between preemption cost assessment and interruption loss, thereby simplifying the data flow for scheduling decisions. It analyzes task types to determine criticality protection weights, ensuring that highly critical tasks receive higher weights in the scoring, thus reducing their likelihood of being preempted. Based on resource utilization status, it determines the resource stress coefficient, quantifying the degree of resource stress and reflecting overall resource availability. Based on the basic preemption cost, criticality protection weights, and resource stress coefficient, it generates a resource preemptibility score, thereby quantifying the preemptibility of tasks.

[0021] Optionally, comparing the safe distance threshold and the obstacle distance to determine whether there is an operational impact includes: Based on the real-time operating status, determine the robot's current acceleration; Calculate the distance difference between the obstacle distance and the safe distance threshold; Calculate the minimum safe braking distance based on the current acceleration and the distance difference; When the distance to an obstacle is less than the minimum safe braking distance, it is determined that there is an operational impact.

[0022] This solution determines the robot's current acceleration based on its real-time operating status, providing real-time motion data and ensuring safety assessments are based on the latest information. It calculates the distance difference between the obstacle and the safe distance threshold, providing crucial input for calculating the minimum safe braking distance and effectively assessing whether the obstacle's position poses a potential risk. Based on the current acceleration and distance difference, the minimum safe braking distance is calculated, ensuring that it considers both real-time motion and the relative position of the obstacle. When the obstacle distance is less than the minimum safe braking distance, it is determined that there is an operational impact, ensuring timely response in high-risk scenarios and preventing safety accidents caused by delays.

[0023] Secondly, this application provides a state-aware robot scheduling system, the system comprising: The task analysis module is used to acquire a set of control tasks; analyze the set of control tasks, and determine the task type. The status analysis module is used to acquire environmental perception data and robot operation data in real time; analyze the operation data to determine the real-time operating status; The priority determination module is used to determine the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type. The resource determination module is used to acquire the resource carrying status in real time and determine the set of preemptible resources per unit time based on the resource carrying status. The scheduling module is used to allocate the set of preemptible resources according to the real-time priority, and schedule the robot to operate according to the allocation result. Attached Figure Description

[0024] 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, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of an application scenario provided in an embodiment of this application.

[0026] Figure 2 This is a flowchart illustrating a state-aware robot scheduling method provided in one embodiment of this application.

[0027] Figure 3 This is a schematic diagram of a state-aware robot scheduling system provided in one embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0029] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0030] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0031] Traditional robot scheduling methods mainly rely on static priority mechanisms, such as first-in-first-out (FIFO) queues for resource allocation. This allocation method, when an urgent task arises, still requires waiting for preceding tasks to complete, causing scheduling delays and reducing overall system throughput and resource utilization.

[0032] Based on this, this application provides a state-aware robot scheduling method and system. It acquires a control task set, effectively identifies the complete range of concurrent tasks, and ensures that all concurrent tasks are included in the scheduling scope. The control task set is analyzed to determine task types, ensuring that each task has a clear type definition. This avoids scheduling decisions based on fuzzy or unclassified task information and supports targeted processing for priority calculation and resource allocation. Real-time acquisition of environmental perception data and robot operation data enables immediate perception of dynamic threats in the external environment, providing an environmental basis for priority adjustment; and accurately capturing changes in the robot's own state. Analysis of operation data determines the real-time operating status, allowing timely responses to environmental threats or resource fluctuations, providing immediate data support for dynamic priority adjustment, and improving the timeliness and accuracy of decision-making. Based on the real-time operating status, environmental perception data, and task type, the real-time priority of each control task is determined, ensuring that the priority reflects multi-dimensional factors, solving the problem of static priority rigidity, and achieving dynamic optimization of task urgency. This provides a precise ranking basis for resource allocation and avoids scheduling decision bias. Real-time acquisition of resource availability status is used to determine the set of preemptible resources per unit time, providing an available resource pool for efficient resource preemption. This ensures rapid resource release when resources are limited, improving resource utilization and scheduling flexibility. Based on real-time priorities, the set of preemptible resources is allocated, and robots are scheduled to operate according to the allocation results. This achieves optimized matching of resources and tasks, executes necessary resource preemption operations, improves response speed, throughput, and overall operational efficiency, and ensures tasks are completed in an orderly manner according to dynamic priorities.

[0033] Figure 1 This is a schematic diagram illustrating an application scenario provided by this application, showing the application of the method provided in this application during robot scheduling.

[0034] Specifically, the method provided in this application is applied to any server. The server interacts with the robot control system, sensor system, monitoring system, and resource monitoring module to obtain a set of control tasks from the robot control system. The control task set is analyzed to determine the task type. Environmental perception data is acquired in real time through the robot's onboard sensor system, and the robot's operational data is acquired in real time through the monitoring system. The operational data is analyzed to determine the real-time operational status. Based on the real-time operational status, environmental perception data, and task type, the real-time priority of each control task is determined. The resource monitoring module acquires the resource carrying status in real time. Based on the resource carrying status, the set of preemptible resources per unit time is determined, providing an available resource pool for efficient resource preemption. This ensures that resources can be quickly released when resources are limited, improving resource utilization and scheduling flexibility. Based on the real-time priority, the set of preemptible resources is allocated. The robot is scheduled to operate based on the allocation results, achieving optimized matching of resources and tasks, executing necessary resource preemption operations, improving response speed, throughput, and overall operational efficiency, and ensuring that tasks are completed in an orderly manner according to dynamic priorities.

[0035] For specific implementation details, please refer to the following examples.

[0036] Figure 2 This is a flowchart illustrating a state-aware robot scheduling method according to an embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes: S201. Obtain the control task set; analyze the control task set to determine the task type; The control task set can be a set of concurrent tasks that need to be processed by the robot.

[0037] Task type can be a category label that controls the task.

[0038] Specifically, the control task set is obtained from the robot control system. The control task set is analyzed, its metadata (such as task description fields) is parsed, and pre-defined classification rules (which store the mapping relationship between metadata and task types and are used to analyze the control task set to determine the task type) are applied based on domain-specific knowledge (such as industry standard documents and technical manuals) and practical experience. For example, if the task description includes an obstacle avoidance field, the task type is a safety task; if the task description includes a navigation field, the task type is a navigation task.

[0039] S202. Acquire environmental perception data and robot operation data in real time; analyze the operation data to determine the real-time operating status; Environmental perception data can be dynamic information about the external environment acquired in real time through sensor systems.

[0040] Operational data can be the robot's own state parameters obtained through a monitoring system.

[0041] Real-time operating status can be a discrete state identifier that reflects the current operating status of the robot, derived from the analysis of operating data.

[0042] Specifically, the robot acquires environmental perception data (such as obstacle distance and movement trajectory) in real time through its onboard sensor system (such as lidar and infrared sensors); at the same time, it acquires the robot's operating data (such as remaining battery power and CPU utilization) through a monitoring system.

[0043] Analyze the operational data and apply the rule engine (used to analyze the operational data to determine the real-time operational status) to determine the real-time operational status. For example, when the remaining power is lower than the preset power threshold set according to the robot battery safety specifications (used to compare the remaining power in the rule engine), it is marked as a low power state; when the CPU utilization is higher than the preset load threshold set according to the stress test data (used to compare the CPU utilization in the rule engine), it is marked as a high load state.

[0044] S203. Determine the real-time priority of each control task based on the real-time operating status, environmental perception data, and task type. A control task can be a single task within a set of control tasks.

[0045] Real-time priority can be determined by combining task type, real-time running status, and environmental perception data to determine the urgency of the task.

[0046] Specifically, the task type (such as high or low priority classification) is read as the basic weight; then, combined with real-time operating status (such as low battery status will reduce the weight of non-urgent tasks) and environmental perception data (such as obstacles being too close will increase the weight of obstacle avoidance tasks), a weighted summation logic is used to determine the real-time priority of each control task. For example, for obstacle avoidance tasks, when environmental perception data shows that an obstacle is approaching and the operating status is that the battery is sufficient, the real-time priority is raised to a high level.

[0047] S204. Obtain resource carrying status in real time, and determine the set of preemptible resources per unit time based on resource carrying status; Resource carrying status can be the carrying status during system operation.

[0048] A preemptible resource set can be a set of resources that can be safely interrupted or reallocated within a unit of time.

[0049] Specifically, the resource monitoring module acquires the resource utilization status in real time. It applies resource filtering rules (used to quickly filter out resource items that can be preempted by high-priority tasks) to examine each resource item (an independently schedulable logical resource unit, such as a single CPU core) in the resource utilization status. If the resource utilization rate (the proportion of a resource item currently occupied by a task) is lower than a preset threshold configured based on test data (indicating that the resource is not being fully utilized and can be preempted), or if the occupying task (the task instance currently using the resource item) is of low priority (the task has a low real-time priority label), it is marked as a set of preemptible resources per unit time.

[0050] S205. Allocate a set of preemptible resources according to real-time priority, and schedule the robot to operate according to the allocation result.

[0051] The allocation result can be generated after allocating the set of preemptible resources according to real-time priority.

[0052] Specifically, the control task set is traversed, and the set of preemptible resources (such as idle CPU) is allocated to the highest priority task (the task with the highest real-time priority label, such as obstacle avoidance task and emergency stop task). If resources are insufficient, resources of low priority tasks (the task with the lowest real-time priority label, such as log upload) are preempted. Then, according to the allocation results, scheduling instructions are generated and sent to the robot control system to schedule the robot to operate. For example, after allocating CPU resources to the obstacle avoidance task, the robot immediately performs obstacle avoidance actions.

[0053] This solution acquires a control task set, effectively identifying the complete range of concurrent tasks and ensuring all concurrent tasks are included in the scheduling scope. Analyzing the control task set determines task types, ensuring each task has a clear type definition, thus avoiding scheduling decisions based on vague or unclassified task information and supporting targeted processing of priority calculation and resource allocation. Real-time acquisition of environmental perception data and robot operation data enables immediate perception of dynamic threats in the external environment, providing an environmental basis for priority adjustment; and accurately capturing changes in the robot's own state. Analyzing operation data determines real-time operating status, allowing timely responses to environmental threats or resource fluctuations, providing immediate data support for dynamic priority adjustment, and improving the timeliness and accuracy of decision-making. Based on real-time operating status, environmental perception data, and task type, the real-time priority of each control task is determined, ensuring that priorities reflect multi-dimensional factors, solving the problem of static priority rigidity, and achieving dynamic optimization of task urgency, thereby providing a precise ranking basis for resource allocation and avoiding scheduling decision bias. Real-time acquisition of resource availability status is used to determine the set of preemptible resources per unit time, providing an available resource pool for efficient resource preemption. This ensures rapid resource release when resources are limited, improving resource utilization and scheduling flexibility. Based on real-time priorities, the set of preemptible resources is allocated, and robots are scheduled to operate according to the allocation results. This achieves optimized matching of resources and tasks, executes necessary resource preemption operations, improves response speed, throughput, and overall operational efficiency, and ensures tasks are completed in an orderly manner according to dynamic priorities.

[0054] In some embodiments, user input is acquired and analyzed to determine the urgency of the interaction; environmental perception data is analyzed to determine the distance to obstacles and their movement trajectories; the impact of environmental conditions is determined based on the real-time operating status, the distance to obstacles, and their movement trajectories; and the real-time priority of each control task is determined based on the urgency of the interaction, the real-time operating status, the impact of environmental conditions, and the task type.

[0055] Interaction urgency can be indicated by a label representing the level of urgency entered by the user.

[0056] Obstacle distance can be the straight-line distance from the robot to the nearest obstacle.

[0057] The trajectory of motion can be a description of the direction and speed of the obstacle's movement.

[0058] Environmental condition impact can be a label indicating the urgency of the impact of environmental factors on the task.

[0059] Specifically, the robot's user interaction module receives user input in real time. Based on the user input, it queries keyword matching rules (used to convert the semantics of the user input into quantifiable interaction urgency) to determine the interaction urgency. For example, when the user input contains urgent or immediate keywords (such as emergency help), the interaction urgency is determined to be high; when the user input is a regular request (such as starting route guidance), the interaction urgency is determined to be medium; and when the user input is a non-time-sensitive request (such as uploading logs), the interaction urgency is determined to be low.

[0060] The system uses a distance calculation algorithm to process environmental perception data, extract obstacle coordinates, and calculate the straight-line distance from the robot to the nearest obstacle to determine the obstacle distance. At the same time, a motion tracking algorithm built based on the principle of tracking moving objects using the reference optical flow method is used to identify the direction and speed of the obstacle's movement and determine the motion trajectory.

[0061] The real-time operating status, obstacle distance, and motion trajectory are input into an impact assessment model (used to map the impact of environmental status) built based on expert experience to determine the impact of environmental status. For example, when the real-time operating status shows a stationary speed, a small obstacle distance, and a motion trajectory pointing towards the robot, the impact of environmental status is considered high threat.

[0062] By integrating interaction urgency, real-time operating status, environmental status impact, and task type, the decision engine performs priority decisions to determine the real-time priority of each control task. For example, if the task type is a security task and the interaction urgency or environmental status impact is high, the real-time priority is set to the highest; if the task type is a non-critical task and the remaining battery power is low in the real-time operating status, the real-time priority is set to low.

[0063] This solution acquires and analyzes user input to determine the urgency of interactions, addressing the issue of a single dimension in priority calculation. It ensures that task scheduling responds to user interactions in real time and avoids operational risks caused by delays in user requests. Priority decisions are no longer rigidly based on static task types but incorporate external interaction factors, improving reliability in user-driven scenarios. Analyzing environmental perception data determines obstacle distances and movement trajectories, addressing the problem of neglecting environmental state factors. This enhances the real-time performance and accuracy of priority calculations, preventing task failures in dense obstacle environments due to a lack of dynamic data, thereby reducing safety hazards. Based on real-time operating status, obstacle distances, and movement trajectories, the solution determines the impact of environmental states, enabling collaborative assessment of the environment and robot states. This ensures that scheduling decisions adapt to dynamic changes and avoids congestion or inefficiency caused by unreasonable resource allocation. Based on interaction urgency, real-time operating status, environmental state impact, and task type, the solution determines the real-time priority of each control task, improving throughput and robustness in resource-constrained environments while avoiding resource waste caused by unreasonable interruptions of low-priority tasks.

[0064] In some embodiments, a preset safe distance threshold is obtained based on the real-time operating status; the safe distance threshold and the obstacle distance are compared to determine whether there is an operational impact; if there is no operational impact, the real-time operating status is analyzed to determine the robot's movement direction; the movement trajectory is analyzed to determine the real-time movement angle between the obstacle's movement direction and the robot's movement direction; and the environmental state impact is determined based on the real-time movement angle.

[0065] The safe distance threshold can be a preset benchmark value used to determine whether an obstacle poses a threat. It is stored in the server in advance and called when needed.

[0066] Operational impact can be a Boolean value indicating whether an obstacle poses an immediate threat to the robot's operation.

[0067] The robot's direction of motion can be the current direction of movement extracted from the real-time operating status.

[0068] Obstacles can be objects in the environment that pose a threat to the robot's movement.

[0069] The direction of movement can be the direction in which the obstacle moves.

[0070] The real-time motion angle can be the absolute difference between the direction of obstacle movement and the direction of robot movement.

[0071] Specifically, a preset safe distance threshold (used as a benchmark reference value to determine whether an obstacle poses a threat) is directly read from the real-time operating status. The obstacle distance is then compared with the preset safe distance threshold. Based on the comparison result, it is determined whether there is an operational impact. For example, if the obstacle distance is less than or equal to the preset safe distance threshold, it is determined that there is an operational impact; if the obstacle distance is greater than the preset safe distance threshold, it is determined that there is no operational impact.

[0072] If there are no operational influences at present, the real-time operating status is analyzed to obtain the orientation angle (an angle value representing the robot's movement direction, used to quantify the robot's movement direction) and determine the robot's movement direction. The movement direction angle of the obstacle is extracted from the movement trajectory as the obstacle's movement direction; then the absolute difference between the robot's movement direction angle and the obstacle's movement direction angle is calculated; and then the absolute difference is normalized to a real-time movement angle within the range of 0° to 180°.

[0073] The real-time motion angle is directly mapped to the environmental state impact level. For example, when the real-time motion angle is small, the environmental state impact is determined to be high threat (meaning that the obstacle's movement direction is highly coincident with the robot's movement direction, and the threat is high); when the real-time motion angle is large, the environmental state impact is determined to be low threat.

[0074] This solution obtains preset safe distance thresholds based on real-time operational status, providing a quantitative standard to ensure that environmental impact assessments are based on a unified safety boundary, avoiding assessment biases caused by inconsistent distance thresholds. By comparing the safe distance thresholds with obstacle distances, it determines whether operational impacts exist, quickly assesses whether obstacles pose an immediate threat, and filters out low-risk scenarios that do not require in-depth analysis. If no operational impact exists, the real-time operational status is analyzed to determine the robot's movement direction, ensuring that the environmental impact assessment considers the robot's dynamic behavior. The motion trajectory is analyzed to determine the real-time motion angle between the obstacle's movement direction and the robot's movement direction, quantifying the relative motion trend between the obstacle and the robot, providing numerical basis for mapping environmental impact levels, and enhancing the objectivity of threat assessment. Based on the real-time motion angle, the environmental impact is determined, achieving a simple and effective classification of potential collision risks.

[0075] In some embodiments, the real-time motion angle is analyzed. When the real-time motion angle is less than 90 degrees, the relative velocity of the obstacle is determined based on the motion trajectory. Obtain obstacle size data. Calculate the effective collision cross section based on the obstacle size data and obstacle distance. Calculate the dynamic threat coefficient based on the obstacle relative velocity and effective collision cross section. Map the dynamic threat coefficient to an environmental state impact value.

[0076] The relative velocity of an obstacle can be the velocity value of the obstacle relative to the robot.

[0077] Obstacle size data can be the physical size information of the obstacle.

[0078] The effective collision cross section can be a numerical value representing the size of the potential collision area.

[0079] The dynamic threat coefficient can be a numerical coefficient that represents the real-time strength of the current threat.

[0080] The environmental status impact value can be a level value used to represent the overall impact of the environmental status.

[0081] Specifically, the real-time motion angle is analyzed, and it is compared whether the real-time motion angle is less than 90 degrees. When the real-time motion angle is less than 90 degrees, the speed value of the obstacle is extracted from the motion trajectory as the relative speed of the obstacle relative to the robot, that is, the relative speed of the obstacle (representing the rate at which the obstacle approaches or moves away from the robot).

[0082] Obstacle size data (such as width, height, or equivalent diameter) is obtained from the environmental perception system. Based on the obstacle size data, the projected area of ​​the obstacle (the two-dimensional projection size of the obstacle's physical size in different directions) is determined through geometric simplification (reducing the three-dimensional obstacle to a two-dimensional planar shape). Then, a geometric model established by collision detection theory is applied to scale the projected area of ​​the obstacle in combination with the obstacle distance to calculate the effective collision cross section. For example, the closer the distance, the larger the effective collision cross section.

[0083] The dynamic threat coefficient is determined by combining the relative velocity of obstacles and the effective collision cross-section through a multiplication operation (e.g., high-speed obstacles at close range result in a high dynamic threat coefficient). A preset mapping rule (used to convert the dynamic threat coefficient into an environmental state impact value) is then used to map the dynamic threat coefficient to an environmental state impact value. For example, a high dynamic threat coefficient corresponds to a high environmental state impact value, and a low dynamic threat coefficient corresponds to a low environmental state impact value.

[0084] This solution analyzes real-time motion angles. When the real-time motion angle is less than 90 degrees, it determines the relative velocity of obstacles based on the stated motion trajectory, quantifying the rate at which obstacles approach the robot, providing a crucial velocity dimension input for threat calculation. It acquires obstacle size data, quantifies the size characteristics of obstacles, and ensures that threat assessment comprehensively considers physical size factors. Based on obstacle size data and obstacle distance, it calculates the effective collision cross-section, quantifying the spatial probability of collision, providing a crucial dimension input for the dynamic threat coefficient. Based on the relative velocity of obstacles and the effective collision cross-section, it calculates the dynamic threat coefficient, combining velocity and spatial factors to quantify the real-time threat level of obstacles to the robot. Finally, it maps the dynamic threat coefficient to environmental state impact values, realizing the transformation from numerical calculation to decision output.

[0085] In some embodiments, the resource loading status is analyzed to identify the fragmentable and idle resources occupied by low-priority tasks; the task execution progress of low-priority tasks is obtained; the interruption penalty is calculated based on the task execution progress; the resource preemptibility score is determined based on the interruption penalty; and the set of preemptible resources per unit time is determined based on the resource preemptibility score, according to the fragmentable and idle resources.

[0086] Low-priority tasks can be tasks with lower priority.

[0087] Fragmentable resources can be resource types that can be divided or partially released from the resources currently occupied by low-priority tasks.

[0088] Idle resources can be resources that are not currently being used by any task.

[0089] Task execution progress can be the current completion status of low-priority tasks.

[0090] Interruption loss can be a measure of the combined loss in terms of schedule and resources when a task is interrupted.

[0091] Resource preemptibility score can be a rating value that represents the feasibility and risk level of a resource being preemptible.

[0092] Specifically, the system analyzes the resource loading status, identifies resources currently occupied by low-priority tasks, and filters out those resources that can be divided or partially released as shardable resources (such as a CPU core that can be divided into multiple time slices). At the same time, it identifies resources that are not currently occupied by any tasks as idle resources (such as idle network bandwidth or idle executors).

[0093] The task execution progress of low-priority tasks is read directly from the task execution tracking system. Based on the task execution progress value, an interruption loss is generated using a proportional rule (a linear proportional relationship used to convert the task execution progress value into an interruption loss). The interruption loss is directly proportional to the task execution progress value; for example, the higher the task execution progress value, the greater the interruption loss, indicating that the invested resources may be wasted.

[0094] Based on the interruption loss, a resource preemptibility score is generated using a preset mapping rule (used to convert the interruption loss into a resource preemptibility score). The resource preemptibility score is inversely proportional to the interruption loss; for example, a higher interruption loss corresponds to a lower resource preemptibility score, indicating a greater risk of preemption. Priority is given to shardable resources with high resource preemptibility scores (e.g., a high score indicates easier preemption); secondly, all idle resources are directly included; finally, these resources are combined to form a set of preemptible resources per unit time.

[0095] This solution analyzes resource utilization status, identifies the fragmentable and idle resources occupied by low-priority tasks, and avoids blind operations in resource-congestion scenarios. It obtains the execution progress of low-priority tasks and introduces real-time feedback on task progress to ensure that subsequent interruption loss calculations are based on actual task status rather than static presets, thus supporting dynamic decision-making. Based on task execution progress, interruption losses are calculated to ensure that resource preemption decisions consider the actual cost of task interruptions, avoiding resource waste caused by unreasonable interruptions. Based on interruption losses, a resource preemptibility score is determined to dynamically assess the feasibility of resource preemption, thereby optimizing resource management and reducing scheduling risks. Based on the resource preemptibility score, and according to fragmentable and idle resources, the set of preemptible resources per unit time is determined, enabling high-priority tasks to acquire the resources they need immediately, thereby improving responsiveness and alleviating resource congestion problems.

[0096] In some embodiments, the task type is analyzed to identify whether the task is associated with a preset high-value target; if associated with a high-value target, the target value weight coefficient is obtained; the real-time operating status is analyzed to determine the robot's remaining power; the control task set is analyzed to determine the robot's control operation set; the expected energy consumption of the task is determined based on the control operation set; the probability of task completion is calculated based on the expected energy consumption and remaining power; and the real-time priority of each control task is determined based on the interaction urgency, environmental state influence value, target value weight coefficient, and the probability of task completion.

[0097] Pre-defined high-value targets can be predefined high-importance targets, stored in the server beforehand, and invoked when needed.

[0098] The target value weighting coefficient can be a numerical value that represents the importance of a pre-set high-value target.

[0099] The remaining power can be the robot's current available energy value.

[0100] A set of control operations can be a set of basic execution units within a set of control tasks.

[0101] The estimated energy consumption of a task can be the total energy required to perform the control task.

[0102] The probability of task completion can be the probability value of successful task execution under the constraint of remaining battery power.

[0103] Specifically, based on the task type, the system queries a pre-defined association rule database (which stores the mapping relationship between task types and pre-defined high-value goals, such as emergency obstacle avoidance → safety goal, user assistance → service goal) built by the database management system to identify whether the current task is associated with a pre-defined high-value goal. If it is associated with a high-value goal, the system retrieves the corresponding target value weight coefficient from a pre-defined weight coefficient table built with expert experience (which stores the mapping relationship between pre-defined high-value goals and their corresponding target value weight coefficients).

[0104] Analyze the real-time operating status, obtain the current battery level reading, and determine the robot's remaining battery power. Traverse the control task set, and for each control task, extract the set of robot control operations it contains. For example, for a path guidance task, the set of control operations includes operations such as calculating the path, driving the actuators, and updating the position.

[0105] Based on the set of control operations, query the preset energy consumption model database (which stores the energy consumption values ​​of different types of control operations) constructed from the test data, extract the energy consumption value of each operation (the baseline energy consumption value of a single control operation (such as calculating a path or driving an actuator)); then sum up the energy consumption values ​​of all operations to obtain the expected energy consumption of the task.

[0106] Compare the estimated energy consumption of the task with the remaining power; based on the comparison results, determine the probability of task completion. For example, if the estimated energy consumption of the task is less than or equal to the remaining power, the probability of task completion is "complete"; if the estimated energy consumption of the task is greater than the remaining power, the probability of task completion is "incomplete".

[0107] The interaction urgency, environmental state influence value, target value weight coefficient, and task completion probability are used as input parameters. Then, a weighted aggregation rule (a weighted summation rule used to determine the real-time priority of each control task) is used to sum the input parameters to determine the real-time priority of each control task.

[0108] This solution analyzes task types and identifies whether tasks are associated with preset high-value objectives, addressing the issue of a single dimension in priority calculation and avoiding reliance solely on static task types for priority allocation. If associated with high-value objectives, it obtains the objective value weight coefficient, strengthening multi-dimensional priority evaluation and dynamically adjusting priority decisions based on task value. It analyzes real-time operating status to determine the robot's remaining battery power, enhancing dynamic response capabilities. Addressing insufficient real-time response capabilities, it adjusts priorities by integrating the robot's own status. It analyzes the control task set to determine the robot's control operation set, supporting efficient resource management and indirectly alleviating resource congestion and inefficient management. Based on the control operation set, it determines the expected energy consumption of tasks, optimizing resource allocation and avoiding waste. Based on the expected energy consumption and remaining battery power, it calculates the probability of task completion, addressing the lack of a dynamic feedback loop and improving decision-making accuracy and efficiency. Based on interaction urgency, environmental state influence, objective value weight coefficient, and task completion probability, it determines the real-time priority of each control task, achieving efficient and non-blocking resource management.

[0109] In some embodiments, the task execution progress is analyzed to determine the percentage of work completed; the value decay curve corresponding to the task type is obtained; the progress loss factor is calculated based on the percentage of work completed and the value decay curve; historical task data is obtained, and the historical resource occupancy level is determined based on the historical task data; the resource sunk cost is determined based on the historical resource occupancy level; and the progress loss factor and the resource sunk cost are weighted and summed to determine the interruption loss.

[0110] The percentage of work completed can be the proportion of the total work currently completed for a task.

[0111] A value decay curve can be a curve used to describe the decay law of task value as the progress changes.

[0112] The schedule loss factor can be the value loss caused by the unfinished remaining work when a task is interrupted.

[0113] Historical task data can be task execution records over a past period of time, which can be obtained based on experience or defined manually.

[0114] Historical resource consumption level can represent the average resource consumption level of different task types during historical execution.

[0115] Sunk costs can be the amount of resources that have been consumed but cannot be recovered when a task is interrupted.

[0116] Specifically, the task execution progress is analyzed, and the ratio of the amount of work already executed (the number of work units actually completed from the start of the task to the current moment) to the total amount of work (the total number of all work units that the task needs to complete) is extracted; then the ratio is converted into a percentage to obtain the percentage of work completed.

[0117] Based on the task type (e.g., path guidance, data collection), a pre-defined value decay curve database (which stores the mapping relationship between task types and corresponding value decay curves) is queried, and the value decay curve corresponding to the task type is extracted. The percentage of completed work is input into the value decay curve to query the task value corresponding to the current progress (the value of the task at the current execution progress); then, based on the interruption scenario assumption (e.g., the task is interrupted before completion), the decay loss of task value at the interruption point is calculated (the amount of value loss caused by the interruption); and then the decay loss is directly used as the progress loss factor.

[0118] Access the historical database (which stores the execution records of historical tasks) built from historical task execution data, retrieve matching historical task data based on task type; parse the historical task data to extract resource usage information (such as CPU utilization and memory usage) during the execution of historical tasks; then calculate the average resource usage level (such as average CPU utilization and average memory usage) of historical tasks as the historical resource usage level.

[0119] Historical resource occupancy levels are directly mapped to sunk costs. The schedule loss factor and sunk costs are weighted and summed using pre-defined weighting coefficients (used to determine interruption losses) constructed with professional knowledge.

[0120] This solution analyzes task execution progress, determines the percentage of completed work, and ensures that interruption loss assessment is based on actual execution status, avoiding subjective estimation. It obtains the value decay curve corresponding to each task type, ensuring that interruption loss assessment considers the value characteristics of each task type. Based on the percentage of completed work and the value decay curve, it calculates the schedule loss factor, quantifying the schedule value decay caused by interruption, providing a schedule-dimensional loss indicator for interruption loss calculation. It acquires historical task data, determines historical resource occupancy levels, assesses the amount of resources potentially wasted during interruptions, and compensates for the neglect of sunk resource costs. Based on historical resource occupancy levels, it determines sunk resource costs, quantifies the potential waste of consumed resources, incorporates interruption loss into the resource cost dimension, and addresses the problem of interruption decisions not considering sunk resource costs. Finally, it weights and sums the schedule loss factor and sunk resource costs to determine interruption losses, optimizes resource preemption mechanisms, and improves efficiency.

[0121] In some embodiments, the basic preemption cost is determined based on the interruption loss; the critical protection weight is determined by analyzing the task type; the resource stress coefficient is determined based on the resource carrying status; and a resource preemptibility score is generated based on the basic preemption cost, the critical protection weight, and the resource stress coefficient.

[0122] Basic preemption cost can represent the combined loss in terms of schedule and resources when a task is interrupted.

[0123] Criticality protection weight can be the degree of protection for the criticality of a task.

[0124] The resource scarcity coefficient can be considered as the current level of resource scarcity.

[0125] Specifically, the interruption loss is directly assigned to the base preemption cost (i.e., base preemption cost = interruption loss). The task type of the current task (e.g., emergency obstacle avoidance, data acquisition) is analyzed, and a preset criticality weight mapping table (storing the mapping relationship between task types and criticality protection weights, used to query the criticality protection weights corresponding to the task type) is queried based on domain expert experience. The corresponding criticality protection weights are then extracted. For example, a high weight indicates strong task criticality, high protection requirements, and amplified preemption cost; a low weight indicates weak task criticality, low protection requirements, and reduced preemption cost.

[0126] The quotient of the calculated resource load status and the preset maximum resource capacity value (maximum available capacity of resources, such as maximum CPU computing power and maximum network bandwidth) set according to hardware specifications is used as the resource stress coefficient. For example, when resources are scarce (such as high CPU utilization), the resource stress coefficient increases, indicating that resources are scarce and the cost of preemption increases; when resources are normal, the resource stress coefficient is small, indicating that resources are abundant and the cost of preemption decreases.

[0127] Calculate the product of basic preemption cost, criticality protection weight, and resource scarcity coefficient; use the product directly as the resource preemptibility score. For example, a higher resource preemptibility score indicates a higher preemption cost (high task criticality, resource scarcity, or significant interruption loss), making the task less suitable for preemption; a lower resource preemptibility score indicates a lower preemption cost (weak task criticality, abundant resources, or minimal interruption loss), making the task easier to preempt.

[0128] This solution determines the basic preemption cost based on the interruption loss, avoiding additional conversion or computational overhead and ensuring consistency between preemption cost assessment and interruption loss, thereby simplifying the data flow for scheduling decisions. It analyzes task types to determine criticality protection weights, ensuring that highly critical tasks receive higher weights in the scoring, thus reducing their likelihood of being preempted. Based on resource utilization status, it determines the resource stress coefficient, quantifying the degree of resource stress and reflecting overall resource availability. Based on the basic preemption cost, criticality protection weights, and resource stress coefficient, it generates a resource preemptibility score, thereby quantifying the preemptibility of tasks.

[0129] In some embodiments, the robot's current acceleration is determined based on the real-time operating status; the distance difference between the obstacle distance and the safe distance threshold is calculated; the minimum safe braking distance is calculated based on the current acceleration and the distance difference; and when the obstacle distance is less than the minimum safe braking distance, it is determined that there is an operational impact.

[0130] The current acceleration can be the instantaneous rate of change of the robot's motion.

[0131] The minimum safe braking distance can be the minimum distance required for the robot to stop safely under the current acceleration and distance difference conditions.

[0132] Specifically, the instantaneous rate of change of the robot's motion is extracted from the real-time operating status and used as the current acceleration (representing how fast the robot's motion changes). The distance difference between the obstacle distance and the safe distance threshold (representing the minimum distance boundary for safe robot operation) is calculated through a subtraction operation (obstacle distance minus the safe distance threshold represents the real-time deviation of the environmental safety margin).

[0133] Based on the current acceleration and distance difference, preset calculation rules are set according to dynamic characteristics and safety requirements (used to generate the minimum safe braking distance) to obtain the minimum safe braking distance. For example, when the current acceleration increases (indicating enhanced deceleration capability), the minimum safe braking distance decreases (reflecting a shorter braking requirement); when the distance difference decreases (indicating reduced safety margin), the minimum safe braking distance further strengthens the space requirements for emergency braking.

[0134] The distance to the obstacle is compared with the minimum safe braking distance; when the distance to the obstacle is less than the minimum safe braking distance, it is determined that there is an operational impact (indicating that the robot faces a collision risk or operation interruption).

[0135] This solution determines the robot's current acceleration based on its real-time operating status, providing real-time motion data and ensuring safety assessments are based on the latest information. It calculates the distance difference between the obstacle and the safe distance threshold, providing crucial input for calculating the minimum safe braking distance and effectively assessing whether the obstacle's position poses a potential risk. Based on the current acceleration and distance difference, the minimum safe braking distance is calculated, ensuring that it considers both real-time motion and the relative position of the obstacle. When the obstacle distance is less than the minimum safe braking distance, it is determined that there is an operational impact, ensuring timely response in high-risk scenarios and preventing safety accidents caused by delays.

[0136] Figure 3 A schematic diagram of a state-aware robot scheduling system provided in one embodiment of this application is shown below. Figure 3 As shown, the state-aware robot scheduling system 300 of this embodiment includes: a task analysis module 301, a state analysis module 302, a priority determination module 303, a resource determination module 304, and a scheduling module 305.

[0137] Task analysis module 301 is used to acquire a set of control tasks; analyze the set of control tasks, and determine the task type; The status analysis module 302 is used to acquire environmental perception data and robot operation data in real time; analyze the operation data, and determine the real-time operation status; The priority determination module 303 is used to determine the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type. The resource determination module 304 is used to acquire the resource carrying status in real time and determine the set of preemptible resources per unit time based on the resource carrying status. The scheduling module 305 is used to allocate the set of preemptible resources according to the real-time priority, and schedule the robot to operate according to the allocation result.

[0138] Optionally, when the priority determination module 303 determines the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type, it is configured to: acquire and analyze user input to determine the interaction urgency; analyze the environmental perception data to determine the obstacle distance and movement trajectory; determine the environmental state impact based on the real-time operating status, the obstacle distance, and the movement trajectory; and determine the real-time priority of each control task based on the interaction urgency, the real-time operating status, the environmental state impact, and the task type.

[0139] Optionally, when the priority determination module 303 determines the impact of the environmental state based on the real-time operating status, the obstacle distance, and the motion trajectory, it is configured to: obtain a preset safe distance threshold based on the real-time operating status; compare the safe distance threshold with the obstacle distance to determine whether there is an operating impact; if there is no operating impact, analyze the real-time operating status to determine the robot's motion direction; analyze the motion trajectory to determine the real-time motion angle between the obstacle's motion direction and the robot's motion direction; and determine the impact of the environmental state based on the real-time motion angle.

[0140] Optionally, when the priority determination module 303 determines the environmental state impact based on the real-time motion angle, it is used to: analyze the real-time motion angle; when the real-time motion angle is less than 90 degrees, determine the relative velocity of the obstacle based on the motion trajectory; acquire obstacle size data; calculate the effective collision cross section based on the obstacle size data and the obstacle distance; calculate the dynamic threat coefficient based on the obstacle relative velocity and the effective collision cross section; and map the dynamic threat coefficient to the environmental state impact value.

[0141] Optionally, when the resource determination module 304 determines the set of preemptible resources per unit time based on the resource carrying status, it is used to: analyze the resource carrying status, identify the fragmentable resources and idle resources occupied by low-priority tasks; obtain the task execution progress of low-priority tasks; calculate the interruption loss based on the task execution progress; determine the resource preemptibility score based on the interruption loss; and determine the set of preemptible resources per unit time based on the resource preemptibility score, the fragmentable resources, and the idle resources.

[0142] Optionally, when the priority determination module 303 determines the real-time priority of each control task based on the interaction urgency, the real-time operating status, the environmental state influence, and the task type, it is configured to: analyze the task type to identify whether the task is associated with a preset high-value target; if associated with a high-value target, obtain the target value weight coefficient; parse the real-time operating status to determine the robot's remaining battery power; analyze the control task set to determine the robot's control operation set; determine the expected energy consumption of the task based on the control operation set; calculate the task completion probability based on the expected energy consumption and the remaining battery power; and determine each control task based on the interaction urgency, the environmental state influence value, the target value weight coefficient, and the task completion probability.

[0143] Optionally, when the resource determination module 304 calculates the interruption loss based on the task execution progress, it is configured to: analyze the task execution progress and determine the percentage of completed work; obtain the value decay curve corresponding to the task type; calculate the progress loss factor based on the percentage of completed work and the value decay curve; obtain historical task data and determine the historical resource occupancy level based on the historical task data; determine the resource sunk cost based on the historical resource occupancy level; and weightedly sum the progress loss factor and the resource sunk cost to determine the interruption loss.

[0144] Optionally, when the resource determination module 304 determines the resource preemptibility score based on the interruption loss, it is used to: determine the basic preemption cost based on the interruption loss; analyze the task type and determine the critical protection weight; determine the resource tension coefficient based on the resource carrying status; and generate the resource preemptibility score based on the basic preemption cost, the critical protection weight, and the resource tension coefficient.

[0145] Optionally, when the priority determination module 303 compares the safe distance threshold and the obstacle distance to determine whether there is an operational impact, it is used to: determine the robot's current acceleration based on the real-time operating status; calculate the distance difference between the obstacle distance and the safe distance threshold; calculate the minimum safe braking distance based on the current acceleration and the distance difference; and determine that there is an operational impact when the obstacle distance is less than the minimum safe braking distance.

[0146] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A state-aware based robot dispatching method, characterized in that, include: Acquire the control task set; Analyze the control task set to determine the task type; Real-time acquisition of environmental perception data and robot operation data; Analyze the operational data to determine the real-time operational status; The real-time priority of each control task is determined based on the real-time operating status, the environmental perception data, and the task type. Real-time acquisition of resource carrying status; and determination of the set of preemptible resources per unit time based on the resource carrying status. Based on the real-time priority, the set of preemptible resources is allocated, and the robot is scheduled to operate according to the allocation result; The step of determining the set of preemptible resources per unit time based on the resource carrying status includes: Analyze resource utilization status to identify the fragmentable and idle resources occupied by low-priority tasks; Get the task execution progress of low-priority tasks; Calculate the interruption loss based on the task execution progress; Based on the interruption loss, a resource preemptibility score is determined; Based on the resource preemptibility score, the set of preemptible resources per unit time is determined according to the shardable resources and the idle resources. The calculation of interruption loss based on the task execution progress includes: Analyze the task execution progress to determine the percentage of work completed; Obtain the value decay curve corresponding to the task type; Calculate the schedule loss factor based on the percentage of completed work and the value decay curve; Obtain historical task data, and determine the historical resource usage level based on the historical task data; Determine the sunk cost of resources based on the historical resource utilization levels. The interruption loss is determined by weighted summation of the schedule loss factor and the sunk cost of resources. The determination of resource preemptibility score based on the interruption loss includes: Determine the basic preemption cost based on the aforementioned interruption loss; Analyze the task types to determine critical protection weights; Based on the resource carrying capacity status, determine the resource scarcity coefficient; Based on the basic preemption cost, the critical protection weight, and the resource scarcity coefficient, a resource preemptibility score is generated.

2. The method of claim 1, wherein, The step of determining the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type includes: Acquire and analyze user input to determine the urgency of the interaction; Analyze the environmental perception data to determine the distance and trajectory of obstacles; The environmental conditions are determined based on the real-time operating status, the distance to the obstacle, and the movement trajectory. The real-time priority of each control task is determined based on the urgency of the interaction, the real-time running status, the impact of the environmental status, and the task type.

3. The method of claim 2, wherein, The step of determining the environmental state influence based on the real-time operating status, the obstacle distance, and the movement trajectory includes: Based on the real-time operating status, a preset safe distance threshold is obtained; By comparing the safe distance threshold with the distance to the obstacle, it is determined whether there is any operational impact at present; If there are no operational impacts at present, analyze the real-time operating status to determine the robot's direction of movement; Analyze the motion trajectory to determine the real-time motion angle between the obstacle's motion direction and the robot's motion direction; The influence of environmental conditions is determined based on the real-time motion angle.

4. The method of claim 3, wherein, The step of determining the influence of environmental conditions based on the real-time motion angle includes: Analyze the real-time motion angle; when the real-time motion angle is less than 90 degrees, determine the relative velocity of the obstacle based on the motion trajectory. Obtain obstacle size data; calculate the effective collision cross section based on the obstacle size data and the obstacle distance; Calculate the dynamic threat coefficient based on the relative velocity of the obstacle and the effective collision cross section; The dynamic threat coefficient is mapped to the environmental state impact value.

5. The method of claim 2, wherein, The step of determining the real-time priority of each control task based on the interaction urgency, the real-time operating status, the environmental status impact, and the task type includes: Analyze the task type to identify whether the task is associated with a preset high-value goal; If high-value targets are associated, obtain the target value weight coefficient; Analyze the real-time operating status to determine the robot's remaining battery power; Analyze the control task set to determine the robot's control operation set; Based on the set of control operations, the estimated energy consumption of the task is determined; Calculate the probability that the task can be completed based on the estimated energy consumption of the task and the remaining power. Each control task is determined based on the interaction urgency, the environmental state influence value, the target value weight coefficient, and the task completion probability.

6. The method of claim 3, wherein, The step of comparing the safe distance threshold with the obstacle distance to determine whether there is an operational impact includes: Based on the real-time operating status, determine the robot's current acceleration; Calculate the distance difference between the obstacle distance and the safe distance threshold; Calculate the minimum safe braking distance based on the current acceleration and the distance difference; When the distance to an obstacle is less than the minimum safe braking distance, it is determined that there is an operational impact.

7. A state-aware based robotic dispatching system, characterized in that, The method applied to any one of claims 1-6 includes: The task analysis module is used to acquire a set of control tasks; analyze the set of control tasks, and determine the task type. The status analysis module is used to acquire environmental perception data and robot operation data in real time; analyze the operation data to determine the real-time operating status; The priority determination module is used to determine the real-time priority of each control task based on the real-time operating status, the environmental perception data, and the task type. The resource determination module is used to acquire the resource carrying status in real time and determine the set of preemptible resources per unit time based on the resource carrying status. The scheduling module is used to allocate the set of preemptible resources according to the real-time priority, and schedule the robot to operate according to the allocation result.

Images