Trolley robot task scheduling method and device, electronic equipment and storage medium
By using an auction-based approach based on performance calibration factors and congestion prediction time, the problems of inaccurate task prediction and local congestion caused by the performance degradation of the robot were solved, thus achieving optimized task allocation and improved efficiency.
Patent Information
- Application Number
- CN202511743525.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-25
- Publication Date
- 2026-02-17
AI Technical Summary
Existing swarm intelligence task scheduling methods fail to effectively consider inaccurate task predictions caused by the performance degradation of small robots and local traffic congestion, resulting in uneven task allocation and affecting overall logistics efficiency and system stability.
By calculating the target task completion time based on the performance calibration factor and combining it with the additional time predicted by congestion, the task is allocated to the robot with the lowest bid through an auction, thereby achieving optimized task allocation and avoiding potential traffic congestion.
This improved the efficiency of task scheduling for the robot car, ensured more accurate task allocation, avoided local traffic congestion, and enhanced the system's operational efficiency and stability.
Smart Images
Figure CN121543976A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of task scheduling for small robots, and more specifically, to a task scheduling method, apparatus, electronic device, and storage medium for small robots. Background Technology
[0002] In distributed swarm robot systems, key performance parameters of the robots, such as speed, acceleration, and sensor accuracy, degrade non-uniformly due to wear and aging during long-term operation. However, existing swarm intelligence task scheduling methods typically estimate task completion times and perform task load balancing based on the initial, standardized performance parameters of the robots at the time of manufacture. This estimation method based on standardized performance parameters deviates significantly from the actual performance state of the robots, resulting in actual task completion times far exceeding expectations.
[0003] This prediction bias not only prevents the system from achieving truly even task load distribution, causing some robots to work for excessively long periods and become inefficient, but also further triggers localized dynamic congestion on critical paths within the warehouse. When tasks are unevenly distributed, some robots may be assigned too many tasks, causing them to operate for extended periods in specific areas, thus creating traffic bottlenecks and forcing other robots to wait frequently or detour. This congestion severely reduces overall logistics efficiency and system predictability, impacting the operational efficiency and stability of the entire distributed swarm of robot systems.
[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention
[0005] The purpose of this application is to provide a task scheduling method, device, electronic device, and storage medium for a small vehicle robot. By calculating the completion time of the target task based on a performance calibration factor and the predicted additional time due to congestion, the target task is allocated to the small vehicle robot with the lowest bid in an auction format, thus obtaining the task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and local traffic congestion caused by the performance degradation of small vehicle robots in existing task scheduling methods. It not only considers the performance differences of the small vehicle robots themselves, but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of small vehicle robots.
[0006] Firstly, this application provides a task scheduling method for a small robot, used for task scheduling of a small robot, including the following steps: Obtain historical key performance data for each robot; Based on the aforementioned historical key performance data, the performance calibration factors for each of the aforementioned small robot vehicles were determined. When acquiring the target task, the task information of the target task and the performance calibration factor are input into the preset task completion time calculation model to calculate the completion time of each of the small robots to complete the target task; Based on the completion time and the predicted additional time for congestion, the target task is assigned to the robot with the lowest bid in an auction format, resulting in a task scheduling result.
[0007] The task scheduling method for small robots provided in this application can schedule tasks for small robots. By calculating the completion time of the target task based on the performance calibration factor and the predicted additional time of congestion, the target task is allocated to the small robot with the lowest bid in an auction format, thus obtaining the task scheduling result. This method solves the technical problems of inaccurate task prediction, uneven task allocation, and local traffic congestion caused by the performance degradation of small robots in existing task scheduling methods. It not only considers the performance differences of the small robots themselves, but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of small robots.
[0008] Optionally, based on the historical key performance information, the performance calibration factor for each robot is calculated, including: Obtain the factory performance index information of each of the aforementioned small robots; Based on the ratio of the historical key performance data to the factory performance index information, the performance calibration factor of each of the robot cars is calculated.
[0009] Optionally, based on the ratio of the historical key performance data to the factory performance index information, a performance calibration factor for each of the robot vehicles is calculated, including: Based on the ratio of the historical key performance data to the factory performance index information, multiple individual performance calibration factors for each of the robot cars are calculated. Set corresponding weights for the multiple individual performance calibration factors; The performance calibration factor of each of the robot cars is calculated based on the multiple individual performance calibration factors and their corresponding weights.
[0010] The task scheduling method for small robots provided in this application can perform task scheduling for small robots. By introducing multiple individual performance calibration factors and their weights, it can more meticulously evaluate the degradation of various performance aspects of the small robot and perform weighted calculations based on their importance, so that the final performance calibration factors can more comprehensively and accurately reflect the overall performance status of the small robot.
[0011] Optionally, corresponding weights are set for the multiple individual performance calibration factors, including: Set corresponding initial weights for the multiple individual performance calibration factors; Based on the difference between the historical key performance data and the factory performance index information, the initial weights are adjusted to obtain the weights corresponding to the multiple individual performance calibration factors.
[0012] Optionally, when acquiring the target task, the task information of the target task and the performance calibration factor are input into a preset task completion time calculation model to calculate the completion time of each of the robot vehicles for the target task, including: When acquiring a target task, acquire the task information of the target task; The task information and the performance calibration factor of each of the small robots are input into a preset task completion time calculation model to calculate the execution time of each of the small robots in performing the target task. The execution time of each of the robot cars is added to the current task queue time of each of the robot cars to calculate the completion time of each robot car in completing the target task.
[0013] Optionally, based on the completion time and combined with the predicted congestion forecast time, the target task is allocated to the robot with the lowest bid in an auction format to obtain the task scheduling result, including: Obtain the execution path information and execution location information of the planned tasks for each of the aforementioned small robots; Generate the planned paths for each of the aforementioned small robots to complete the target task; Based on the execution path information, the execution location information, and the planned path, the congestion situation of each of the small robots is detected within the execution time period corresponding to the planned path, and the additional congestion prediction time of each of the small robots is obtained. According to the auction format, the target task is assigned to the robot with the lowest bid based on the completion time and the congestion prediction additional time, thus obtaining the task scheduling result.
[0014] The task scheduling method for small robots provided in this application can perform task scheduling for small robots. By introducing congestion prediction additional time, the method considers the robot's own time to complete the task and the congestion prediction additional time when allocating tasks, and uses an auction method for allocation. This enables the task allocation results to effectively avoid potential congestion risks, thereby optimizing the overall task scheduling efficiency and avoiding local traffic bottlenecks.
[0015] Optionally, based on the execution path information, the execution location information, and the planned path, the congestion situation of each of the robot cars within the execution time period corresponding to the planned path is detected, and the congestion prediction additional time for each of the robot cars is obtained, including: Based on the execution path information and the planned path, the path overlap of each of the small robots within the execution time period corresponding to the planned path is detected, and the spatial congestion time cost of each of the small robots with respect to the target task is determined. Based on the execution location information and the planned path, the queuing situation of each of the robot cars when it arrives at the task start position and task target position corresponding to the target task is detected, and the queuing congestion time cost of each of the robot cars for the target task is determined. The spatial congestion time cost and the queuing congestion time cost are added together to determine the additional congestion prediction time for each of the small robots.
[0016] Secondly, this application provides a task scheduling device for a small robot, used for task scheduling of a small robot, including: The acquisition module is used to acquire historical key performance data for each robot. The determination module is used to determine the performance calibration factor for each of the robot cars based on the historical key performance data. The calculation module is used to input the task information of the target task and the performance calibration factor into a preset task completion time calculation model when the target task is acquired, and to calculate the completion time of each of the robot cars to complete the target task. The scheduling module is used to allocate the target task to the robot with the lowest bid based on the completion time and the predicted additional time of congestion, in an auction format, to obtain the task scheduling result.
[0017] This robot task scheduling device, based on the target task completion time calculated using a performance calibration factor and the predicted congestion time, allocates the target task to the robot with the lowest bid in an auction format, thus obtaining the task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and local traffic congestion caused by performance degradation of existing robot task scheduling methods. It not only considers the performance differences of the robots themselves, but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of robot robots.
[0018] Thirdly, this application provides an electronic device including a processor and a memory, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, it runs the steps in the robot task scheduling method described above.
[0019] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the robot task scheduling method described above.
[0020] Beneficial effects: The task scheduling method, apparatus, electronic device, and storage medium for small vehicles provided in this application allocate the target task to the small vehicle robot with the lowest bid by calculating the completion time of the target task based on the performance calibration factor and the predicted additional time of congestion prediction in an auction manner, thereby obtaining the task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and local traffic congestion caused by the performance degradation of small vehicles in existing task scheduling methods. It not only considers the performance differences of the small vehicles themselves, but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of small vehicles. Attached Figure Description
[0021] Figure 1 A flowchart of a robot task scheduling method provided in an embodiment of this application.
[0022] Figure 2 This is a schematic diagram of the structure of the robot task scheduling device provided in the embodiments of this application.
[0023] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.
[0024] Labeling Explanation: 1. Acquisition Module; 2. Determination Module; 3. Calculation Module; 4. Scheduling Module; 301. Processor; 302. Memory; 303. Communication Bus. Detailed Implementation
[0025] 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 a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.
[0026] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.
[0027] Please refer to Figure 1 , Figure 1 This application discloses a task scheduling method for a small robot, which is used for task scheduling of a small robot, including: Step S1: Obtain historical key performance data for each robot car; Step S2: Based on historical key performance data, determine the performance calibration factor for each robot car. Step S3: When acquiring the target task, input the task information and performance calibration factor of the target task into the preset task completion time calculation model to calculate the completion time of each robot car to complete the target task. Step S4: Based on the completion time and the predicted additional time of congestion, the target task is assigned to the robot with the lowest bid in an auction format to obtain the task scheduling result.
[0028] This new task scheduling method for small robots calculates the completion time of the target task based on a performance calibration factor, and calculates the additional time due to predicted congestion. Then, using an auction format, it allocates the target task to the robot with the lowest bid, resulting in a task scheduling outcome. This method addresses the technical problems of existing task scheduling methods, such as inaccurate task prediction, uneven task allocation, and localized traffic congestion caused by robot performance degradation. It not only considers the performance differences of the robots themselves but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of small robots.
[0029] Specifically, in step S1, historical key performance data of each robot is acquired. The historical key performance data includes various key performance data recorded by the robot during its past operation (i.e., the historical records of various key performance data), such as average driving speed, acceleration time, battery life, picking time (referring to the time to pick up goods from the task start position and put them on the robot) and placing time (referring to the time to place goods from the robot to the task target position). This data can be acquired through the data acquisition module installed on the robot, or by periodically conducting performance tests on the robot and recording its performance under standard test conditions as historical key performance data.
[0030] Specifically, in step S2, based on historical key performance information, the performance calibration factors for each robot are determined, including: Obtain the factory performance specifications of each robot car; The performance calibration factor for each robot is calculated based on the ratio of historical key performance data to factory performance indicators.
[0031] In step S2, by introducing the robot's factory performance specifications and comparing them with historical key performance data, the degradation or change in the robot's actual performance can be assessed more accurately. The factory performance specifications can be understood as the various performance parameters set at the time of manufacture or obtained through testing, such as maximum speed and maximum battery life. These specifications typically represent the theoretical performance of the robot under ideal conditions. This factory performance specification information can be obtained from the manufacturer's specifications or through benchmark testing of newly manufactured robots.
[0032] For example, if the actual average speed of a robot is 90% of its factory maximum speed, its speed performance calibration factor might be 0.9. This ratio calculation yields a quantitative indicator reflecting the robot's current true performance status.
[0033] Specifically, in step S2, based on the ratio of historical key performance data to factory performance index information, the performance calibration factor for each robot is calculated, including: Based on the ratio of historical key performance data to factory performance index information, multiple individual performance calibration factors for each robot are calculated. Set corresponding weights for multiple individual performance calibration factors; The performance calibration factor for each robot is calculated based on multiple individual performance calibration factors and their corresponding weights.
[0034] In step S2, for different key performance dimensions of the robot, such as driving speed and battery life, the ratios of its historical key performance data to its corresponding factory performance indicators are calculated. These ratios each represent the calibration status of the robot in a specific performance aspect.
[0035] The calculation method for the individual performance calibration factor is as follows: ; in, For the single performance calibration factor of the key performance data x of the robot i (where the subscript is... Key performance data x for the robot i (such as driving speed, acceleration, battery life, pickup time, and delivery time). This is the recent average of the key performance data x of the robot i (e.g., the average of the key performance data x over the past 10 executions). The key performance data x of the small robot i is the factory performance index; The aging weight factor is set within the range of [0,1]. It can be set according to the time difference between the latest acquisition time of the key performance data x and the current time. The smaller the time difference, the larger the value of the aging weight factor.
[0036] Specifically, in step S2, corresponding weights are set for multiple individual performance calibration factors, including: Set corresponding initial weights for multiple individual performance calibration factors; Based on the difference between historical key performance data and factory performance indicators, the initial weights are adjusted to obtain the weights corresponding to multiple individual performance calibration factors.
[0037] In step S2, the initial weights are preset weight values for each individual performance calibration factor, based on general experience or design specifications, for the initial use of the robot or when sufficient historical data is lacking. These initial weights aim to provide a benchmark to ensure that preliminary performance evaluation can be performed in the early stages of the robot's operation.
[0038] By comparing historical key performance data accumulated during the actual operation of the robot with the performance indicators set at the time of manufacture, the deviation between the actual and expected performance of each indicator is quantified. For example, if the historical data of a certain performance indicator (such as battery life) has decreased significantly compared with the factory specifications, it can be assumed that the weight of that indicator's impact on the overall performance should be increased accordingly to more accurately reflect its current state. Conversely, if the performance of a certain performance indicator remains stable at the factory specifications, its weight can be appropriately reduced or kept unchanged.
[0039] In step S2, the performance calibration factor for each robot is calculated based on multiple individual performance calibration factors and their corresponding weights using weighted summation or other weighted aggregation algorithms. For example, each individual performance calibration factor can be multiplied by its corresponding weight, and then all weighted individual performance calibration factors can be summed to obtain the final comprehensive performance calibration factor. The purpose is to integrate multi-dimensional performance evaluation results into a unified indicator that comprehensively reflects the current performance status of the robot.
[0040] Specifically, in step S3, when acquiring the target task, the task information and performance calibration factor of the target task are input into a preset task completion time calculation model to calculate the completion time of each robot to complete the target task, including: When acquiring the target task, acquire the task information of the target task; Input the task information and the performance calibration factors of each robot into the preset task completion time calculation model to calculate the execution time of each robot in performing the target task; The execution time of each robot is added to the current task queue time of each robot to calculate the completion time of each robot in completing the target task.
[0041] In step S3, when acquiring the target task, it is first necessary to obtain the task information of the target task. The task information may include, but is not limited to, task type, task priority, task start position, task target position, type and quantity of materials to be processed, etc. This information is the basis for calculating the task execution time.
[0042] The task information and performance calibration factors of each robot are input into a preset task completion time calculation model to calculate the execution time of each robot in performing the target task. The specific task completion time calculation model is as follows: ; in, The execution time for robot i to perform target task j; Let be the path distance from robot i to the starting position of target task j; The factory performance specifications for the driving speed of the robot i; For the performance calibration factor of the small robot i; The factory performance indicators for the pickup time of robot i; Let be the path distance from the starting position of task j to the target position. This refers to the factory performance indicators for the delivery time of the robot i.
[0043] By considering the robot's existing task load (i.e., the current task queue time, which refers to the remaining execution time of the robot's existing tasks and the queuing time of assigned tasks), the execution time of the new task is added to the remaining execution time of existing tasks and the queuing time of assigned tasks to obtain a comprehensive completion time prediction. This method avoids the limitation of only considering task execution time while ignoring the robot's actual working state, enabling task scheduling decisions to be based on more realistic time costs. The current task queue time refers to the total estimated time occupied by tasks that the robot has received but has not yet started executing or is currently executing.
[0044] In an alternative embodiment, key performance data can be used when performing calculations using a task completion time calculation model. The individual performance calibration factor corresponding to the key performance data (i.e., speed, pickup time, and delivery time) is used instead of the performance calibration factor in the calculation, thus adjusting the task completion time calculation model to... ,in, Let i be the single performance calibration factor for the travel speed of robot i. For robot i, a single performance calibration factor for pickup time. The single performance calibration factor for the delivery time of robot i is used; thus, steps S2 and S3 are adjusted to: based on historical key performance data, determine the single performance calibration factor corresponding to each key performance data of each robot car; when acquiring the target task, input the task information of the target task and the single performance calibration factor corresponding to each key performance data into the preset task completion time calculation model to calculate the completion time of each robot car to complete the target task.
[0045] Specifically, in step S4, based on the completion time and combined with the predicted congestion time, the target task is assigned to the robot with the lowest bid in an auction format, resulting in the task scheduling result, including: Obtain the execution path information and execution location information of the planned tasks for each robot; Generate the planned paths for each robot car to complete the target task; Based on the execution path information, execution location information, and planned path, the congestion situation of each robot is detected within the execution time period corresponding to the planned path, and the additional congestion prediction time for each robot is obtained. Using an auction format, with the completion time and congestion prediction time as bids, the target task is assigned to the robot with the lowest bid, resulting in the task scheduling outcome.
[0046] In step S4, by introducing a prediction of congestion for the robotic vehicles and quantifying it as a congestion prediction time, the task allocation auction mechanism considers not only the performance of the robotic vehicles themselves and the current task queue, but also the external environmental obstacles they may encounter when executing new tasks. By acquiring the path and location information of planned tasks and generating a planned path for the target task (which can be calculated using existing path planning algorithms such as A* algorithm, RRT (Rapid Exploratory Random Tree) algorithm, or Dijkstra's algorithm based on the current location of the robotic vehicles, the task's starting location to the target location, and map information), it is possible to anticipate potential temporal or spatial conflicts between different robotic vehicles when executing the target task. These conflicts are converted into quantifiable congestion prediction time, which is then combined with the inherent time for the robotic vehicle to complete the task as a bid, making the task allocation decision more comprehensive and accurate. The target task is assigned to the robotic vehicle with the lowest bid (the minimum sum of completion time and congestion prediction time), resulting in the final task scheduling result. This mechanism ensures that tasks are assigned to robots that are not only highly efficient themselves but also least affected by congestion in the current environment, thus avoiding a decline in overall efficiency due to localized congestion.
[0047] The planned task execution path information refers to the predetermined travel route of each robot currently assigned a task, while the execution location information refers to the expected location of each robot at a specific point in time. This information is the basic data for predicting potential congestion.
[0048] Specifically, in step S4, based on the execution path information, execution location information, and planned path, the congestion situation of each robot is detected within the execution time period corresponding to the planned path, and the congestion prediction additional time for each robot is obtained, including: Based on the execution path information and the planned path, the path overlap of each robot is detected within the execution time period corresponding to the planned path, and the spatial congestion time cost of each robot for the target task is determined. Based on the execution location information and the planned path, the queuing situation of each robot car when it arrives at the task start position and task target position corresponding to the target task is detected, and the queuing congestion time cost of each robot car for the target task is determined. The additional time for congestion prediction for each robot is determined by adding the time cost of spatial congestion and the time cost of queuing congestion.
[0049] In step S4, spatial congestion time cost refers to the time delay caused by the overlap between the planned paths of each robot and the planned paths of other robots during the execution of the target task, which may lead to mutual avoidance or waiting. Path overlap can be detected by analyzing the temporal and spatial intersections between the planned paths of each robot and the execution paths of other robots in the environment. The spatial congestion time cost is quantified based on the degree and duration of the overlap (i.e., for each overlapping path in the planned path, the passage order is arranged according to task priority or task generation order, the queuing time of the corresponding robot on each overlapping path is determined, and the spatial congestion time cost of the corresponding robot is obtained by summing them). Its purpose is to assess the delays caused by dynamic conflicts that the robot may encounter during its movement.
[0050] Queue congestion time cost refers to the time delay incurred when a robot arrives at a designated location for a task (task start and task target locations) because that location is already occupied by other robots or requires waiting for resources (such as loading / unloading points). Specifically, queuing can be determined by detecting the estimated arrival time of each robot at the task start and task target locations and the occupancy status of those locations. The queue duration can then be estimated based on the pickup or unloading times of other robots in the queue to calculate the queue congestion time cost. Its purpose is to assess the delays caused by static waiting that robots may encounter at the task endpoint or critical nodes.
[0051] By subdividing congestion prediction time into spatial congestion time cost and queuing congestion time cost, we can more comprehensively and accurately capture various congestion situations that the robot may encounter during task execution. Spatial congestion time cost focuses on the dynamic conflicts of the robot on its movement path, while queuing congestion time cost focuses on the static waiting of the robot at a specific task location. By calculating and superimposing these two costs separately, a more refined and accurate congestion prediction model can be formed, thus providing a more reliable basis for subsequent task allocation. This detailed analysis helps avoid prediction bias caused by a single congestion assessment dimension, making the task scheduling results closer to actual operation.
[0052] For example, suppose robot A is assigned a target task that requires it to move from position P1 to position P2. After generating the planned path for robot A, its spatial congestion time cost is first detected. For instance, if robot A's planned path overlaps with robot B's planned path within the corresponding execution time period, the spatial congestion time cost that robot A might incur due to avoiding robot B is calculated based on the size and duration of the overlap, for example, 10 seconds. Next, the queuing congestion situation when robot A reaches the target task position P2 is detected. For instance, if robot C is expected to still occupy the loading / unloading point at P2 when robot A arrives, causing robot A to wait, the queuing congestion time cost that robot A might incur is calculated based on robot C's expected occupancy time, for example, 20 seconds. Ultimately, by adding the 10-second spatial congestion time cost to the 20-second queuing congestion time cost, we arrive at a congestion prediction additional time of 30 seconds for robot A to complete the target task. This more accurate congestion prediction additional time will be used in the task allocation auction bid, making the task allocation result more reasonable and efficient.
[0053] As shown above, this robot task scheduling method obtains historical key performance data for each robot, determines performance calibration factors for each robot based on this data, and inputs the task information and performance calibration factors into a preset task completion time calculation model when acquiring a target task. The method calculates the completion time for each robot to complete the target task. Based on the completion time and the predicted congestion time, the target task is allocated to the robot with the lowest bid in an auction format, resulting in a task scheduling result. Therefore, by using the target task completion time calculated based on the performance calibration factors and the predicted congestion time, the target task is allocated to the robot with the lowest bid in an auction format, resulting in a task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and localized traffic congestion caused by robot performance degradation in existing robot task scheduling methods. It not only considers the performance differences of the robots themselves but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of robot robots.
[0054] refer to Figure 2 This application provides a task scheduling device for a small robot, used for task scheduling of a small robot, including: Module 1 is used to acquire historical key performance data for each robot car. Module 2 is used to determine the performance calibration factors for each robot based on historical key performance data. Calculation module 3 is used to input the task information and performance calibration factor of the target task into the preset task completion time calculation model when the target task is acquired, and to calculate the completion time of each robot car to complete the target task. The scheduling module 4 is used to allocate the target task to the robot with the lowest bid based on the completion time and the predicted additional time of congestion, in an auction format, so as to obtain the task scheduling result.
[0055] This robot task scheduling device, based on the target task completion time calculated using a performance calibration factor and the predicted congestion time, allocates the target task to the robot with the lowest bid in an auction format, thus obtaining the task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and local traffic congestion caused by performance degradation of existing robot task scheduling methods. It not only considers the performance differences of the robots themselves, but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of robot robots.
[0056] Specifically, when module 1 is executed, it acquires historical key performance data of each robot. The historical key performance data includes various key performance data recorded by the robot during its past operation (i.e., the historical records of various key performance data), such as average driving speed, acceleration time, battery life, picking time (referring to the time to pick up goods from the task start position and put them on the robot) and placing time (referring to the time to place goods from the robot to the task target position). This data can be acquired through the data acquisition module installed on the robot, or by periodically conducting performance tests on the robot and recording its performance under standard test conditions as historical key performance data.
[0057] Specifically, when determining the performance calibration factors for each robot based on historical key performance information, module 2 executes the following: Obtain the factory performance specifications of each robot car; The performance calibration factor for each robot is calculated based on the ratio of historical key performance data to factory performance indicators.
[0058] When Module 2 is executed, it incorporates the robot's factory performance specifications and compares them with historical key performance data to more accurately assess the degradation or change in the robot's actual performance. The factory performance specifications refer to the various performance parameters set at the time of manufacture or obtained through testing, such as maximum speed and maximum battery life. These specifications typically represent the robot's theoretical performance under ideal conditions. This factory performance specification information can be obtained from the manufacturer's specifications or through benchmark testing of newly manufactured robots.
[0059] For example, if the actual average speed of a robot is 90% of its factory maximum speed, its speed performance calibration factor might be 0.9. This ratio calculation yields a quantitative indicator reflecting the robot's current true performance status.
[0060] Specifically, when determining the performance calibration factor of each robot car based on the ratio of historical key performance data to factory performance index information, module 2 executes the following: Based on the ratio of historical key performance data to factory performance index information, multiple individual performance calibration factors for each robot are calculated. Set corresponding weights for multiple individual performance calibration factors; The performance calibration factor for each robot is calculated based on multiple individual performance calibration factors and their corresponding weights.
[0061] When module 2 is executed, it calculates the ratio of historical key performance data to corresponding factory performance indicators for different key performance dimensions of the robot, such as driving speed and battery life. These ratios represent the calibration status of the robot in specific performance aspects.
[0062] The calculation method for the individual performance calibration factor is as follows: ; in, For the single performance calibration factor of the key performance data x of the robot i (where the subscript is... Key performance data x for the robot i (such as driving speed, acceleration, battery life, pickup time, and delivery time). This is the recent average of the key performance data x of the robot i (e.g., the average of the key performance data x over the past 10 executions). The key performance data x of the small robot i is the factory performance index; The aging weight factor is set within the range of [0,1]. It can be set according to the time difference between the latest acquisition time of the key performance data x and the current time. The smaller the time difference, the larger the value of the aging weight factor.
[0063] Specifically, when determining the weights for multiple individual performance calibration factors, module 2 performs the following: Set corresponding initial weights for multiple individual performance calibration factors; Based on the difference between historical key performance data and factory performance indicators, the initial weights are adjusted to obtain the weights corresponding to multiple individual performance calibration factors.
[0064] When Module 2 is executed, the initial weights are preset weight values for each individual performance calibration factor, based on general experience or design specifications, for the initial use of the robot or when sufficient historical data is lacking. These initial weights aim to provide a benchmark to ensure that preliminary performance evaluation can be performed at the beginning of the robot's operation.
[0065] By comparing historical key performance data accumulated during the actual operation of the robot with the performance indicators set at the time of manufacture, the deviation between the actual and expected performance of each indicator is quantified. For example, if the historical data of a certain performance indicator (such as battery life) has decreased significantly compared with the factory specifications, it can be assumed that the weight of that indicator's impact on the overall performance should be increased accordingly to more accurately reflect its current state. Conversely, if the performance of a certain performance indicator remains stable at the factory specifications, its weight can be appropriately reduced or kept unchanged.
[0066] When module 2 is executed, it calculates the performance calibration factor for each robot by using weighted summation or other weighted aggregation algorithms, based on multiple individual performance calibration factors and their corresponding weights. For example, each individual performance calibration factor can be multiplied by its corresponding weight, and then all weighted individual performance calibration factors can be summed to obtain the final comprehensive performance calibration factor. The purpose is to integrate multi-dimensional performance evaluation results into a unified indicator that can comprehensively reflect the current performance status of the robot.
[0067] Specifically, when acquiring the target task, the calculation module 3 inputs the task information and performance calibration factor of the target task into the preset task completion time calculation model. When the completion time of each robot completing the target task is calculated, the following steps are executed: When acquiring the target task, acquire the task information of the target task; Input the task information and the performance calibration factors of each robot into the preset task completion time calculation model to calculate the execution time of each robot in performing the target task; The execution time of each robot is added to the current task queue time of each robot to calculate the completion time of each robot in completing the target task.
[0068] When the calculation module 3 executes, it first needs to obtain the task information of the target task. The task information may include, but is not limited to, task type, task priority, task start position, task target position, and the type and quantity of materials to be processed. This information is the basis for calculating the task execution time.
[0069] The task information and performance calibration factors of each robot are input into a preset task completion time calculation model to calculate the execution time of each robot in performing the target task. The specific task completion time calculation model is as follows: ; in, The execution time for robot i to perform target task j; Let be the path distance from robot i to the starting position of target task j; The factory performance specifications for the driving speed of the robot i; For the performance calibration factor of the small robot i; The factory performance indicators for the pickup time of robot i; Let be the path distance from the starting position of task j to the target position. This refers to the factory performance indicators for the delivery time of the robot i.
[0070] By considering the robot's existing task load (i.e., the current task queue time, which refers to the remaining execution time of the robot's existing tasks and the queuing time of assigned tasks), the execution time of the new task is added to the remaining execution time of existing tasks and the queuing time of assigned tasks to obtain a comprehensive completion time prediction. This method avoids the limitation of only considering task execution time while ignoring the robot's actual working state, enabling task scheduling decisions to be based on more realistic time costs. The current task queue time refers to the total estimated time occupied by tasks that the robot has received but has not yet started executing or is currently executing.
[0071] In an optional embodiment, when calculating using the task completion time calculation model, the individual performance calibration factor corresponding to a single key performance data point (i.e., speed, pickup time, and delivery time) can be used instead of the performance calibration factor for calculation, thereby adjusting the task completion time calculation model to... ,in, Let i be the single performance calibration factor for the travel speed of robot i. For robot i, a single performance calibration factor for pickup time. The single performance calibration factor for the delivery time of robot i is used; thus, steps S2 and S3 are adjusted to: based on historical key performance data, determine the single performance calibration factor corresponding to each key performance data of each robot car; when acquiring the target task, input the task information of the target task and the single performance calibration factor corresponding to each key performance data into the preset task completion time calculation model to calculate the completion time of each robot car to complete the target task.
[0072] Specifically, when scheduling module 4 obtains the task scheduling result by allocating the target task to the robot with the lowest bid based on the completion time and the predicted congestion time, in an auction format, it executes: Obtain the execution path information and execution location information of the planned tasks for each robot; Generate the planned paths for each robot car to complete the target task; Based on the execution path information, execution location information, and planned path, the congestion situation of each robot is detected within the execution time period corresponding to the planned path, and the additional congestion prediction time for each robot is obtained. Using an auction format, with the completion time and congestion prediction time as bids, the target task is assigned to the robot with the lowest bid, resulting in the task scheduling outcome.
[0073] During execution, scheduling module 4 incorporates predictions of congestion for the robotic vehicles, quantifying these predictions as congestion prediction time. This ensures that the task allocation auction mechanism considers not only the performance of the robotic vehicles themselves and the current task queue but also potential external environmental obstacles when executing new tasks. By acquiring the path and location information of planned tasks and generating a planned path for the target task (based on the current location of the robotic vehicles, the task's starting and ending points, and map information, using existing path planning algorithms such as A*, RRT (Rapid Exploratory Random Tree), or Dijkstra's algorithm to calculate the optimal planned path), it can anticipate potential temporal or spatial conflicts between different robotic vehicles when executing the target task. These conflicts are converted into quantifiable congestion prediction time, which is then combined with the inherent time for each robot to complete the task as a bid, making task allocation decisions more comprehensive and accurate. The target task is assigned to the robotic vehicle with the lowest bid (the minimum sum of completion time and congestion prediction time), resulting in the final task scheduling outcome. This mechanism ensures that tasks are assigned to robots that are not only highly efficient themselves but also least affected by congestion in the current environment, thus avoiding a decline in overall efficiency due to localized congestion.
[0074] The planned task execution path information refers to the predetermined travel route of each robot currently assigned a task, while the execution location information refers to the expected location of each robot at a specific point in time. This information is the basic data for predicting potential congestion.
[0075] Specifically, when scheduling module 4 detects the congestion situation of each robot within the execution time period corresponding to the planned path based on the execution path information, execution location information, and planned path, and obtains the congestion prediction additional time for each robot, it executes: Based on the execution path information and the planned path, the path overlap of each robot is detected within the execution time period corresponding to the planned path, and the spatial congestion time cost of each robot for the target task is determined. Based on the execution location information and the planned path, the queuing situation of each robot car when it arrives at the task start position and task target position corresponding to the target task is detected, and the queuing congestion time cost of each robot car for the target task is determined. The additional time for congestion prediction for each robot is determined by adding the time cost of spatial congestion and the time cost of queuing congestion.
[0076] When scheduling module 4 is executed, spatial congestion time cost refers to the time delay caused by overlapping execution paths of other robots' planned tasks during the execution of the target task. This overlap may lead to mutual avoidance or waiting. Path overlap can be detected by analyzing the temporal and spatial intersections between each robot's planned path and the execution paths of other robots in the environment. The spatial congestion time cost is quantified based on the degree and duration of overlap (i.e., for each overlapping path in the planned path, the passage order is arranged according to task priority or task generation order, the queuing time of the corresponding robot on each overlapping path is determined, and the spatial congestion time cost of the corresponding robot is obtained by summing them). Its purpose is to assess the delays caused by dynamic conflicts that the robot may encounter during movement.
[0077] Queue congestion time cost refers to the time delay incurred when a robot arrives at a designated location for a task (task start and task target locations) because that location is already occupied by other robots or requires waiting for resources (such as loading / unloading points). Specifically, queuing can be determined by detecting the estimated arrival time of each robot at the task start and task target locations and the occupancy status of those locations. The queue duration can then be estimated based on the pickup or unloading times of other robots in the queue to calculate the queue congestion time cost. Its purpose is to assess the delays caused by static waiting that robots may encounter at the task endpoint or critical nodes.
[0078] By subdividing congestion prediction time into spatial congestion time cost and queuing congestion time cost, we can more comprehensively and accurately capture various congestion situations that the robot may encounter during task execution. Spatial congestion time cost focuses on the dynamic conflicts of the robot on its movement path, while queuing congestion time cost focuses on the static waiting of the robot at a specific task location. By calculating and superimposing these two costs separately, a more refined and accurate congestion prediction model can be formed, thus providing a more reliable basis for subsequent task allocation. This detailed analysis helps avoid prediction bias caused by a single congestion assessment dimension, making the task scheduling results closer to actual operation.
[0079] For example, suppose robot A is assigned a target task that requires it to move from position P1 to position P2. After generating the planned path for robot A, its spatial congestion time cost is first detected. For instance, if robot A's planned path overlaps with robot B's planned path within the corresponding execution time period, the spatial congestion time cost that robot A might incur due to avoiding robot B is calculated based on the size and duration of the overlap, for example, 10 seconds. Next, the queuing congestion situation when robot A reaches the target task position P2 is detected. For instance, if robot C is expected to still occupy the loading / unloading point at P2 when robot A arrives, causing robot A to wait, the queuing congestion time cost that robot A might incur is calculated based on robot C's expected occupancy time, for example, 20 seconds. Ultimately, by adding the 10-second spatial congestion time cost to the 20-second queuing congestion time cost, we arrive at a congestion prediction additional time of 30 seconds for robot A to complete the target task. This more accurate congestion prediction additional time will be used in the task allocation auction bid, making the task allocation result more reasonable and efficient.
[0080] As shown above, this robot task scheduling device acquires historical key performance data of each robot and determines its performance calibration factor based on this data. When acquiring a target task, the device inputs the task information and performance calibration factor into a preset task completion time calculation model to calculate the completion time of each robot. Based on the completion time and the predicted congestion time, the device allocates the target task to the robot with the lowest bid in an auction format, thus obtaining the task scheduling result. Therefore, by using the target task completion time calculated based on the performance calibration factor and the predicted congestion time, the device allocates the target task to the robot with the lowest bid in an auction format, thus obtaining the task scheduling result. This solves the technical problems of inaccurate task prediction, uneven task allocation, and localized traffic congestion caused by performance degradation in existing robot task scheduling methods. It not only considers the performance differences of the robots themselves but also predicts and avoids potential traffic congestion, thereby achieving optimized task allocation and improving the task scheduling efficiency of robot tasks.
[0081] Please refer to Figure 3 , Figure 3This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device includes a processor 301 and a memory 302. The processor 301 and the memory 302 are interconnected and communicate with each other via a communication bus 303 and / or other connection mechanisms (not shown). The memory 302 stores a computer program executable by the processor 301. When the electronic device is running, the processor 301 executes the computer program to perform the robot task scheduling method in any optional implementation of the above embodiments, to achieve the following functions: acquiring historical key performance data of each robot; determining the performance calibration factor of each robot based on the historical key performance data; when acquiring a target task, inputting the task information and performance calibration factor of the target task into a preset task completion time calculation model to calculate the completion time of each robot in completing the target task; based on the completion time and combined with the predicted congestion prediction additional time, allocating the target task to the robot with the lowest bid in an auction format to obtain the task scheduling result.
[0082] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it executes the robot task scheduling method in any optional implementation of the above embodiments to achieve the following functions: acquiring historical key performance data of each robot; determining the performance calibration factor of each robot based on the historical key performance data; when acquiring a target task, inputting the task information and performance calibration factor of the target task into a preset task completion time calculation model to calculate the completion time of each robot to complete the target task; and, based on the completion time and combined with the predicted congestion prediction additional time, allocating the target task to the robot with the lowest bid in an auction format to obtain the task scheduling result. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.
[0083] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.
[0084] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.
[0085] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.
[0086] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.
[0087] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.
Claims
1. A task scheduling method for a small robot, used for task scheduling of a small robot, characterized in that, Including the following steps: Obtain historical key performance data for each robot; Based on the aforementioned historical key performance data, the performance calibration factors for each of the aforementioned small robot vehicles were determined. When acquiring the target task, the task information of the target task and the performance calibration factor are input into the preset task completion time calculation model to calculate the completion time of each of the small robots to complete the target task; Based on the completion time and the predicted additional time for congestion, the target task is assigned to the robot with the lowest bid in an auction format, resulting in a task scheduling result.
2. The task scheduling method for a small robot according to claim 1, characterized in that, Based on the aforementioned historical key performance information, the performance calibration factors for each robot are calculated, including: Obtain the factory performance index information of each of the aforementioned small robots; Based on the ratio of the historical key performance data to the factory performance index information, the performance calibration factor of each of the robot cars is calculated.
3. The task scheduling method for a small robot according to claim 2, characterized in that, Based on the ratio of the historical key performance data to the factory performance index information, the performance calibration factor for each of the robot vehicles is calculated, including: Based on the ratio of the historical key performance data to the factory performance index information, multiple individual performance calibration factors for each of the robot cars are calculated. Set corresponding weights for the multiple individual performance calibration factors; The performance calibration factor of each of the robot cars is calculated based on the multiple individual performance calibration factors and their corresponding weights.
4. The task scheduling method for a small robot according to claim 3, characterized in that, Assign corresponding weights to multiple individual performance calibration factors, including: Set corresponding initial weights for the multiple individual performance calibration factors; Based on the difference between the historical key performance data and the factory performance index information, the initial weights are adjusted to obtain the weights corresponding to the multiple individual performance calibration factors.
5. The task scheduling method for a small robot according to claim 1, characterized in that, When acquiring the target task, the task information of the target task and the performance calibration factor are input into a preset task completion time calculation model to calculate the completion time of each of the robot vehicles for the target task, including: When acquiring a target task, acquire the task information of the target task; The task information and the performance calibration factor of each of the small robots are input into a preset task completion time calculation model to calculate the execution time of each of the small robots in performing the target task. The execution time of each of the robot cars is added to the current task queue time of each of the robot cars to calculate the completion time of each robot car in completing the target task.
6. The task scheduling method for a small robot according to claim 1, characterized in that, Based on the completion time and the predicted congestion time, the target task is assigned to the robot with the lowest bid in an auction format, resulting in a task scheduling outcome, including: Obtain the execution path information and execution location information of the planned tasks for each of the aforementioned small robots; Generate the planned paths for each of the aforementioned small robots to complete the target task; Based on the execution path information, the execution location information, and the planned path, the congestion situation of each of the small robots is detected within the execution time period corresponding to the planned path, and the additional congestion prediction time of each of the small robots is obtained. According to the auction format, the target task is assigned to the robot with the lowest bid based on the completion time and the congestion prediction additional time, thus obtaining the task scheduling result.
7. The task scheduling method for a small robot according to claim 6, characterized in that, Based on the execution path information, the execution location information, and the planned path, the congestion situation of each of the robot cars is detected within the execution time period corresponding to the planned path, and the congestion prediction additional time for each of the robot cars is obtained, including: Based on the execution path information and the planned path, the path overlap of each of the small robots within the execution time period corresponding to the planned path is detected, and the spatial congestion time cost of each of the small robots with respect to the target task is determined. Based on the execution location information and the planned path, the queuing situation of each of the robot cars when it arrives at the task start position and task target position corresponding to the target task is detected, and the queuing congestion time cost of each of the robot cars for the target task is determined. The spatial congestion time cost and the queuing congestion time cost are added together to determine the additional congestion prediction time for each of the small robots.
8. A task scheduling device for a small robot, used for task scheduling of a small robot, characterized in that, include: The acquisition module is used to acquire historical key performance data for each robot. The determination module is used to determine the performance calibration factor for each of the robot cars based on the historical key performance data. The calculation module is used to input the task information of the target task and the performance calibration factor into a preset task completion time calculation model when the target task is acquired, and to calculate the completion time of each of the robot cars to complete the target task. The scheduling module is used to allocate the target task to the robot with the lowest bid based on the completion time and the predicted additional time of congestion, in an auction format, to obtain the task scheduling result.
9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing a computer program executable by the processor, and when the processor executes the computer program, it performs the steps in the robot task scheduling method as described in any one of claims 1-7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it performs the steps in the robot task scheduling method as described in any one of claims 1-7.