CN-122022251-A - Robot scheduling method

CN122022251ACN 122022251 ACN122022251 ACN 122022251ACN-122022251-A

Abstract

The application discloses a robot scheduling method, which comprises the steps of firstly obtaining a target task, determining candidate robots, then calculating task cost representing the resource consumption of the task to be executed by the candidate robots by combining the attribute information of the target task and the candidate robots, and finally selecting the target robot to execute the task according to the task cost. The scheduling decision is fully brought into the real-time specific state of the candidate robot terminal side, the resource consumption is accurately quantified through the task cost, the task is effectively prevented from being distributed to the robots with overhigh resource consumption or insufficient suitability, and the operation efficiency and the resource utilization rate of the scheduling system are remarkably improved. Meanwhile, the clear step flow and quantitative evaluation ensure that decision making is scientific and operable, the task requirements of different scenes can be adapted, the task execution stability and reliability are improved, and clear basis is provided for the optimization of the follow-up scheduling strategy.

Inventors

LI GUANGSHENG
GONG HANYUE
ZHI TAO

Assignees

北京云迹科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251224

Claims (10)

1. A method of robotic scheduling, the method comprising: Acquiring a target task; Determining candidate robots according to the target tasks; determining task costs corresponding to the candidate robots according to the target tasks and attribute information corresponding to the candidate robots, wherein the task costs corresponding to the candidate robots are used for representing resource conditions required to be paid by the candidate robots for executing the target tasks; Determining target robots according to task costs corresponding to the candidate robots; and executing the target task by using the target robot.
2. The method according to claim 1, wherein the target task includes task start point position information, task end point position information, task content type, and task execution time limit; The attribute information corresponding to the candidate robot comprises a position attribute, an operation attribute, an energy attribute and a configuration attribute, wherein the position attribute is a real-time positioning coordinate of the candidate robot, the operation attribute is a current task execution state of the candidate robot, the energy attribute is a residual electric quantity percentage and a unit distance power consumption rate of the candidate robot, and the configuration attribute is hardware form, load capacity and functional module configuration of the candidate robot.
3. The robot scheduling method of claim 2, wherein the determining a candidate robot from the target task comprises: Screening candidate robots with configuration attributes matched with the task content types from registered robots according to the task content types and the configuration attributes of the registered robots; determining a task execution range according to the task starting point position information and the task ending point position information; determining a robot screening area according to the task execution range; And taking the candidate robot with the real-time positioning coordinate corresponding to the position attribute in the robot screening area as a candidate robot.
4. The robot scheduling method according to claim 2, wherein the determining, for each candidate robot, a task cost corresponding to the candidate robot according to the target task and attribute information corresponding to the robot, includes: Determining the position cost of each candidate robot based on the position attribute of the candidate robot, the task starting point position information and the task ending point position information of the target task, wherein the position cost represents the forward path degree and the total moving distance of the candidate robot for executing the target task; Determining the operation cost of the candidate robot based on the operation attribute of the candidate robot, wherein the operation cost represents whether the candidate robot needs to interrupt the current task to accept the target task or not; determining energy costs of the candidate robots based on the energy attributes of the candidate robots, wherein the energy costs represent total electric quantity consumption of the candidate robots for completing the target tasks; determining configuration cost of the candidate robot based on the configuration attribute of the candidate robot, wherein the configuration cost characterizes the matching degree of the candidate robot and the task content type; and inputting the position cost, the operation cost, the energy agent and the configuration cost into a preset cost calculation model to obtain task cost corresponding to the candidate robot.
5. The robot scheduling method of claim 4, wherein determining the position cost of the robot based on the position attribute of the candidate robot and the task start point position information and the task end point position information of the target task comprises: Determining a theoretical running path of the candidate robot for executing the target task according to the real-time positioning coordinates of the candidate robot, the task starting point position coordinates and the task ending point position coordinates of the target task; Judging whether the theoretical driving path is along the current driving path of the candidate robot, if so, overlapping a preset along-path rewarding coefficient, and if not, overlapping a preset detour punishment coefficient; calculating the actual mileage of the theoretical travel path, and if the actual mileage relates to the movement of the cross-floor, calculating the cross-floor sub-cost according to the product of the preset unit distance cost of the cross-floor and the cross-floor mileage, wherein the cross-floor sub-cost does not exceed the preset maximum threshold value of the cross-floor; Calculating according to the product of a preset flat layer unit distance cost and a flat layer mileage related to the actual mileage, and if the flat layer mileage is smaller than a preset flat layer minimum mileage threshold value, calculating a flat layer running sub-cost according to the preset flat layer minimum mileage threshold value, wherein the flat layer running sub-cost does not exceed the preset flat layer maximum threshold value; If the candidate robot needs to back to the current upper cabin and then execute the target task, acquiring the driving distance of the back to the current upper cabin, calculating the back sub-price of the upper cabin according to the product of the preset back single distance cost and the driving distance of the back to the current upper cabin, and if the driving distance of the back to the current upper cabin is smaller than a preset back minimum distance threshold, calculating according to the preset back minimum distance threshold, wherein the back sub-price of the upper cabin does not exceed the preset back maximum threshold; The position cost is the sum of the cross-floor sub-cost, the flat-layer driving sub-cost and the upper cabin back sub-cost which are adjusted based on the preset forward road rewarding coefficient or the preset detour punishment coefficient.
6. The robot scheduling method according to claim 4, wherein the determining the operation cost of the candidate robot based on the operation attribute of the candidate robot specifically includes: Detecting the operation attribute of the candidate robot and judging whether the candidate robot is in an idle state or a busy state currently; If the operation cost is in an idle state, the operation cost is a preset idle basic cost; The method comprises the steps of obtaining a target task, extracting the residual execution time length, the residual driving distance and the task priority of the current task of a candidate robot, calculating a time length interrupt sub-cost according to the product of a preset interrupt time length weight coefficient and the residual execution time length, calculating a mileage interrupt sub-cost according to the product of a preset interrupt mileage weight coefficient and the residual driving distance, determining a preset high priority punishment coefficient if the task priority of the current task is higher than the task priority of the target task, determining a preset low priority punishment coefficient if the task priority of the current task is lower than the task priority of the target task, and determining an operation cost according to the time length interrupt sub-cost, the mileage interrupt sub-cost and the preset high priority punishment coefficient or the preset low priority punishment coefficient, wherein the operation cost does not exceed a preset operation maximum threshold.
7. The robot scheduling method according to claim 4, wherein the determining the energy cost of the candidate robot based on the energy attribute of the candidate robot specifically comprises: extracting energy attributes of the candidate robots, wherein the energy attributes comprise current residual electric quantity percentage, unit distance power consumption rate and charging efficiency parameters; Determining the basic total power consumption for completing the target task according to the theoretical driving path mileage of the target task and the power consumption rate per unit distance; if the current remaining capacity percentage is lower than a preset safe electric capacity threshold, the total power consumption of the base is weighted and amplified according to a preset low electric capacity coefficient, and if the current remaining capacity percentage is higher than a preset sufficient electric capacity threshold, the total power consumption of the base is weighted and reduced according to a preset high electric capacity coefficient; If the candidate robot needs to return to a charging point after executing the target task, calculating the total amount of additional power consumption returned to the charging point, wherein the total amount of additional power consumption=the returned mileage×the power consumption rate per unit distance; And determining the energy cost of the candidate robot according to the basic power consumption total amount, the additional power consumption total amount and the preset low power coefficient or the preset high power coefficient, wherein the energy cost does not exceed a preset energy maximum threshold.
8. The robot scheduling method according to claim 4, wherein the determining the configuration cost of the candidate robot based on the configuration attribute of the candidate robot specifically includes: Extracting configuration attributes of the candidate robot, including hardware form, load capacity, function module type and adapter interface specification; Analyzing the task content type of the target task, and determining a minimum load threshold value, a necessary functional module and interface requirements required by the target task; If the load capacity of the candidate robot is lower than a minimum load threshold, the configuration cost is a preset load non-standard punishment cost; if the load capacity is higher than a minimum load threshold, calculating load sub-cost according to the load capacity, the minimum load threshold and a preset load redundancy reward coefficient; Determining the number of the missing necessary functional modules of the candidate robot, and calculating the functional sub-cost according to the number of the missing necessary functional modules and the penalty cost of a preset functional missing unit; if the specification of the adaptation interface of the candidate robot is inconsistent with the task requirement, determining the preset interface adaptation punishment cost, and if so, determining the interface adaptation rewarding coefficient; And determining the configuration cost of the candidate robot according to the load sub-cost, the function sub-cost and the interface adaptation adjustment coefficient, wherein the configuration cost does not exceed a preset configuration maximum threshold.
9. The method for scheduling robots according to claim 4, wherein the inputting the position cost, the operation cost, the energy cost and the configuration cost into a preset cost calculation model to obtain task costs corresponding to the candidate robots specifically includes: acquiring a cost weight parameter pre-configured by a cloud scheduling platform, wherein the cost weight parameter comprises a position weight coefficient, an operation weight coefficient, an energy weight coefficient and a configuration weight coefficient, and the sum of the weight coefficients is 1; Calculating the task cost according to the following formula of task cost = position cost x position weight coefficient + operation cost x operation weight coefficient + energy cost x energy weight coefficient + configuration cost x configuration weight coefficient; if the calculated task cost is lower than a preset minimum cost threshold, correcting according to the preset minimum cost threshold, and if the calculated task cost is higher than a preset maximum cost threshold, correcting according to the preset maximum cost threshold; And outputting the corrected result as the task cost corresponding to the candidate robot.
10. The robot scheduling method according to claim 1, wherein the determining the target robot according to the task cost corresponding to each candidate robot comprises: Setting a task cost effective threshold and a feedback time window; Collecting task costs of all candidate robots in the feedback time window; screening candidate robots with task cost lower than the effective threshold value to form a qualified robot set; And selecting a robot with the minimum task cost from the qualified robot set, and determining the robot as a target robot.

Description

Robot scheduling method Technical Field The application relates to the technical field of robots, in particular to a robot scheduling method. Background In a multi-robot cooperation scene, the traditional robot scheduling method generally only makes a centralized decision by a cloud scheduling platform according to the basic requirements of a target task and part of basic information of robots, and determines a robot for executing the task, and the resource condition required by the robot for executing the target task is not quantified by combining the attribute information corresponding to the target task and the robots for each robot, so that the real-time specific state of a terminal side of the robot cannot be fully considered, the suitability of the robot for executing the target task is difficult to accurately evaluate, the problem of unreasonable task allocation easily occurs, the situation that the resource consumption is too high and the execution efficiency is low exists when part of the robots execute the task, and finally the operation efficiency and the resource utilization rate of the whole scheduling system are influenced. Disclosure of Invention The application provides a robot scheduling method, which aims to effectively solve the problems that the suitability of a robot for executing a target task is difficult to accurately evaluate, the task allocation is unreasonable, the situation that the resource consumption is too high and the execution efficiency is low when part of robots execute the task is caused, and finally the operation efficiency and the resource utilization rate of an overall scheduling system are influenced. In a first aspect, the present application provides a robot scheduling method, the method comprising: Acquiring a target task; Determining candidate robots according to the target tasks; determining task costs corresponding to the candidate robots according to the target tasks and attribute information corresponding to the candidate robots, wherein the task costs corresponding to the candidate robots are used for representing resource conditions required to be paid by the candidate robots for executing the target tasks; Determining target robots according to task costs corresponding to the candidate robots; and executing the target task by using the target robot. In a second aspect, the present application provides a robot scheduling device, the device comprising: The first unit is used for acquiring a target task; A second unit for determining candidate robots according to the target task; the third unit is used for determining task cost corresponding to each candidate robot according to the target task and attribute information corresponding to the candidate robot, wherein the task cost corresponding to the candidate robot is used for representing resource conditions required by the candidate robot to execute the target task; a fourth unit, configured to determine a target robot according to task costs corresponding to each candidate robot; and a fifth unit for executing the target task by using the target robot. In a third aspect, the present application provides a readable medium comprising execution instructions which, when executed by a processor of an electronic device, perform the method according to any of the first aspects. In a fourth aspect, the present application provides an electronic device comprising a processor and a memory storing execution instructions, the processor performing the method according to any one of the first aspects when executing the execution instructions stored in the memory. According to the robot scheduling method, the target tasks are acquired firstly, then the candidate robots are determined according to the target tasks, then for each candidate robot, the task cost for representing the resource condition required by the execution of the target tasks is calculated by combining the target tasks and the attribute information corresponding to the candidate robots, finally the target robots are determined based on the task cost of each candidate robot and the target tasks are executed by utilizing the target robots, so that the real-time specific state of the terminal side of the candidate robots can be fully brought into the scheduling decision process, the resource consumption level of the candidate robots for executing the target tasks is accurately quantized through the task cost, the situation that the target tasks are distributed to the candidate robots with high resource consumption or insufficient suitability is effectively avoided, the running efficiency and the resource utilization rate of the whole scheduling system are obviously improved, meanwhile, the method ensures the scientificity and the operability of scheduling decisions through an explicit step flow and a quantized cost evaluation mode, can adapt to the task requirements under different scenes, the stability and the reliability of the execution of the target task