Multi-level composite robot scheduling method and system and storage medium

Abstract

The invention discloses a multi-level composite robot scheduling method. The method comprises the following steps: S1, obtaining task information and collecting state information of an online robot; s2, the robots are divided into a plurality of formations according to the positions and available resources of the robots, and then tasks are distributed to the formations; s3, the robots in the formations generate target paths through distributed path planning and execute tasks, and local dynamic adjustment is carried out according to needs during execution; not only are communication and computing power bottlenecks of centralized scheduling effectively broken through, but also the stability, the intelligent level and the engineering availability of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of robotics, and more specifically, to a multi-level composite robot scheduling method, a robot, and a storage medium. Background Technology

[0002] In multi-robot scheduling, high communication load and centralized computing bottlenecks are encountered. Currently, there are two main solutions for multi-robot scheduling: centralized and fully distributed. The centralized scheduling method, adopted by most existing industrial-grade multi-robot systems, relies on a central controller to obtain real-time status information such as the position, speed, and task progress of all robots. It centrally calculates the task allocation and path planning results for each robot before issuing them for execution. While it has advantages in path optimality and task scheduling efficiency, it suffers from technical bottlenecks such as a surge in computational load as the system scales up due to centralized computing resources, high communication pressure requiring continuous high-frequency communication with all robots, and insufficient robustness due to the impact of central node failure on the entire system. The fully distributed scheduling method, through a decentralized mechanism, allows individual robots to complete path generation and task execution through local communication and collaborative decision-making within their perception range. While this method has low communication load and flexible deployment, it lacks global optimization capabilities, making it prone to local optima due to individual robots making decisions based on local information. Its task allocation efficiency is lower than centralized methods, and it cannot perform optimal resource scheduling based on global load. Summary of the Invention

[0003] To address the above problems, this invention provides a multi-level composite robot scheduling method, comprising the following steps:

[0004] S1. Obtain task information and collect status information of the online robot;

[0005] S2. Divide the robots into multiple squads based on their location and available resources, and then assign tasks to each squad.

[0006] S3. Robots in each formation generate target paths and execute tasks through distributed path planning, making local dynamic adjustments as needed during execution.

[0007] As a preferred technical solution, the status information of the online robot in S1 includes the robot's location and available resources, and the status of each robot is represented as follows:

[0008] [x i E i ]

[0009] Where x represents the robot's location on the map, and E represents the currently available resources, which are related to the remaining battery power and the remaining time to complete the task.

[0010] As a preferred technical solution, the step of dividing the robots into formations in S2 includes optimizing the first objective function to divide the robots into K formations, each formation being denoted as KK. The first objective function is as follows:

[0011]

[0012] Among them, R k Let r represent the set of robots in the k-th formation. i μ represents the serial number of each robot. k E represents the assembly center of the k-th formation. i λ represents the available resources for the robot, and λ is the weighting coefficient.

[0013] As a preferred technical solution, the task allocation step in S2 includes: obtaining a task set T = {t1, t2, ..., t}. n Define task assignment variable d. n,k The tasks are assigned to each formation by optimizing the second objective function, which is as follows:

[0014]

[0015] Each task t in the task set n The state can be represented as: [x t,n ,L n ], x t L represents the target location of the task, L represents the task load, and β is the weighting coefficient, a task allocation variable.

[0016] As a preferred technical solution, in the distributed path planning to generate the target path, the objective function for optimizing the robot path within the formation is:

[0017]

[0018]

[0019] Among them, h(E) i ) represents the weights related to available resources, and p represents the target point of the task. Indicates the relationship with robot r i The set of neighboring robots, where γ is the weighting coefficient and φ is a function negatively correlated with the distance between two adjacent robots.

[0020] As a preferred technical solution, the path optimization is achieved through distributed gradient descent, including: the robot obtains the state of neighboring robots through local communication; and iteratively solves and updates the path according to the distributed gradient descent method until two adjacent paths converge to a preset threshold.

[0021] As a preferred technical solution, the distributed gradient descent method iteratively solves for the update path, as shown in the following formula:

[0022]

[0023] Where t represents the number of iterations, Let α represent the gradient, and let α represent the gradient descent rate.

[0024] As a preferred technical solution, the S3 process includes on-demand local dynamic adjustments, including:

[0025] Each robot moves to the target point according to the optimized path and performs the task;

[0026] If obstacles or robot malfunctions are encountered during execution, local path replanning is initiated, and the task is re-executed within the current formation. If the task cannot be completed, it is reported to the robot central scheduling system for reassignment.

[0027] As a preferred technical solution, the present invention also provides a multi-level composite robot scheduling system, comprising:

[0028] The central scheduling module is used to acquire task information, collect robot status, divide into formations, and assign tasks.

[0029] The distributed planning module is used to generate target paths and execute tasks;

[0030] The dynamic adjustment module is used to make local dynamic adjustments as needed during task execution.

[0031] As a preferred technical solution, the present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program implements the above-described method when executed by a processor.

[0032] Compared with the prior art, the beneficial effects of this invention are:

[0033] The multi-level composite robot scheduling method provided by this invention significantly reduces the system's dependence on a central node by distributing path optimization tasks across local formations. Even if some communication links are interrupted, local tasks can still be completed successfully. By introducing a distributed optimization algorithm, robots achieve collaborative decision-making based on neighbor perception, which helps improve the overall coordination and flexibility of task execution. The regional formation mechanism and local autonomous optimization make the system easier to deploy in sites of different sizes or complex structures, exhibiting good versatility and maintainability. This invention not only effectively overcomes the communication and computing power bottlenecks of centralized scheduling but also improves the system's stability, intelligence level, and engineering usability. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating the multi-level composite robot scheduling method of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0037] It should be noted that the terms "first" and "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0038] Figure 1 This is a flowchart of a multi-level composite robot scheduling method provided in an embodiment of the present invention. The method is executed by a central scheduling system for the robot, such as... Figure 1 As shown, it includes the following steps:

[0039] S1. Obtain task information and collect status information of the online robot;

[0040] S2. Divide the robots into multiple squads based on their location and available resources, and then assign tasks to each squad.

[0041] S3. Robots in each formation generate target paths and execute tasks through distributed path planning, making local dynamic adjustments as needed during execution.

[0042] Compared to existing technologies, the multi-level composite robot scheduling method provided by this invention proposes a multi-level scheduling method that combines centralized task partitioning with distributed path optimization. Through a "centralized-distributed" hybrid architecture, it effectively overcomes several shortcomings of existing technologies. Specifically: First, the central scheduling system is only responsible for the initial decomposition of the overall task and the allocation of tasks to each formation according to region, without needing to collect and process the status information of all robots in real time; the path planning task is completed collaboratively by robots within each formation, thereby reducing the computational complexity of the central system. Second, by adopting a locally distributed path planning strategy, robots only need to communicate with neighboring individuals to complete path collaborative optimization, thereby reducing communication bandwidth dependence. Third, each formation has independent path optimization capabilities; after a task change, only local adjustments are needed, without global recalculation.

[0043] In a preferred embodiment, the status information of the online robot in S1 includes the robot's location and available resources, and the status of each robot is represented as follows:

[0044] [x i E i ]

[0045] Where x represents the robot's location on the map, and E represents the currently available resources, which are related to the remaining battery power and the remaining time to complete the task.

[0046] It should be noted that the lower the remaining battery power, the longer the remaining time to complete the task, and the less available resources there are.

[0047] In a preferred embodiment, the step of dividing the robots into formations in S2 includes optimizing a first objective function to divide the robots into K formations, each formation denoted as KK. The first objective function is as follows:

[0048]

[0049] Among them, R k Let r represent the set of robots in the k-th formation. i μ represents the serial number of each robot. k E represents the assembly center of the k-th formation. i λ represents the available resources for the robot, and λ is the weighting coefficient.

[0050] In a preferred embodiment, the task allocation step in S2 includes: obtaining a task set T = {t1, t2, ..., t}. n Define task assignment variable d. n,k The tasks are assigned to each formation by optimizing the second objective function, which is as follows:

[0051]

[0052] Each task t in the task set n The state can be represented as: [x t，n L n ], x t L represents the target location of the task, L represents the task load, and β is the weighting coefficient, a task allocation variable.

[0053] It should be noted that the workload of a task can be described by the time required to complete the task.

[0054] In a preferred embodiment, the objective function for optimizing the robot path within the formation in the distributed path planning to generate the target path is:

[0055]

[0056] Where h represents a function negatively correlated with available resources, and p represents the target point of the task. Indicates the relationship with robot r i The set of neighboring robots, where γ is the weighting coefficient and φ is a function negatively correlated with the distance between two adjacent robots.

[0057] In a more preferred embodiment, the function h, which is negatively correlated with available resources, is defined as follows:

[0058]

[0059] In a more preferred embodiment, the function φ, which is negatively correlated with the distance between two adjacent robots, is defined as follows:

[0060]

[0061] Here, ∈ is a user-defined parameter.

[0062] In a preferred embodiment, the path optimization is achieved using a distributed gradient descent method, including: robot i acquiring the state x of neighboring robot j through local communication. j The updated path is solved iteratively using the distributed gradient descent method. The above communication and gradient update are repeated until two adjacent paths converge to a preset threshold.

[0063] In a preferred embodiment, the distributed gradient descent method iteratively solves for the update path, as shown in the following formula:

[0064]

[0065] Where t represents the number of iterations, Let α represent the gradient, and let α represent the gradient descent rate.

[0066] In a preferred embodiment, the on-demand local dynamic adjustments performed during execution in step S3 include:

[0067] Each robot moves to the target point according to the optimized path and performs the task;

[0068] If obstacles or robot malfunctions are encountered during execution, local path replanning is initiated, and the task is re-executed within the current formation. If the task cannot be completed, it is reported to the robot central scheduling system for reassignment.

[0069] Traditional centralized scheduling, as the number of robots increases, requires the central system to monitor all states and calculate the optimal trajectory in real time, leading to a surge in computing and communication pressure. This invention significantly reduces the central computing pressure by combining task-level optimization of central scheduling with distributed optimization within the formation. Simultaneously, it incorporates available resource items (such as load and power) into the objective function, enabling more refined task allocation and path planning. Furthermore, this invention employs a local communication strategy, eliminating network-wide broadcasting and allowing state exchange only between robots within the formation, significantly reducing communication load. Moreover, if one formation fails, other formations can still operate normally, demonstrating strong system fault tolerance.

[0070] The present invention also provides a multi-level composite robot scheduling system, comprising:

[0071] The central scheduling module is used to acquire task information, collect robot status, divide into formations, and assign tasks.

[0072] The distributed planning module is used to generate target paths and execute tasks;

[0073] The dynamic adjustment module is used to make local dynamic adjustments as needed during task execution.

[0074] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the above-described multi-level composite robot scheduling method.

[0075] It should be understood that the computer-readable storage medium is any data storage device capable of storing data or programs that can subsequently be read by a computer system. Examples of computer-readable storage media include read-only memory, random access memory, CD-ROM, HDD, DVD, magnetic tape, and optical data storage devices. Computer-readable storage media can also be distributed across network-coupled computer systems, enabling computer-readable code to be stored and executed in a distributed manner. The program code contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wireless, wired, optical fiber, radio frequency (RF), or any suitable combination thereof.

[0076] In some implementations, the computer-readable storage medium may be non-transitory.

[0077] In summary, this invention significantly reduces the system's dependence on a central node by distributing path optimization tasks across local formations. Even if some communication links are interrupted, local tasks can still be completed successfully. By introducing distributed optimization algorithms, robots can achieve collaborative decision-making based on neighbor perception, which helps improve the overall coordination and flexibility of task execution. The regional formation mechanism and local autonomous optimization make the system easier to deploy in sites of different sizes or complex structures, exhibiting good versatility and maintainability. This invention not only effectively overcomes the communication and computing power bottlenecks of centralized scheduling but also improves the system's stability, intelligence level, and engineering usability.

[0078] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0079] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A multi-level composite robot scheduling method, executed by a central scheduling system for the robot, characterized in that, Includes the following steps: S1. Obtain task information and collect status information of the online robot; S2. Divide the robots into multiple squads based on their location and available resources, and then assign tasks to each squad. S3. Robots in each formation generate target paths and execute tasks through distributed path planning, making local dynamic adjustments as needed during execution.

2. The multi-level composite robot scheduling method according to claim 1, characterized in that, The status information of the online robot in S1 includes the robot's location and available resources. The status of each robot is represented as follows: [x i ,AND i ] Where x represents the robot's location on the map, and E represents the currently available resources, which are related to the remaining battery power and the remaining time to complete the task.

3. The multi-level composite robot scheduling method according to claim 2, characterized in that, The step of dividing the robot into formations in S2 includes optimizing the first objective function to divide the robot into K formations, each formation being denoted as K. The first objective function is as follows: Among them, R k Let r represent the set of robots in the k-th formation. i μ represents the serial number of each robot. k E represents the assembly center of the k-th formation. i λ represents the available resources for the robot, and λ is the weighting coefficient.

4. The multi-level composite robot scheduling method according to claim 3, characterized in that, The task allocation steps in S2 include: obtaining the task set T = {t1, t2, ..., t}. n Define task assignment variable d. n,k The tasks are assigned to each formation by optimizing the second objective function, which is as follows: Each task t in the task set n The state can be represented as: [x t,n ,L n ], x t L represents the target location of the task, L represents the task load, and β is the weighting coefficient, a task allocation variable.

5. The multi-level composite robot scheduling method according to claim 4, characterized in that, In the distributed path planning that generates the target path, the objective function for optimizing the robot path within the formation is: Among them, h(E) i ) represents the weights related to available resources, and p represents the target point of the task. Indicates the relationship with robot r i The set of neighboring robots, where γ is the weighting coefficient and φ is a function negatively correlated with the distance between two adjacent robots.

6. The multi-level composite robot scheduling method according to claim 5, characterized in that, The path optimization is achieved through distributed gradient descent, including: the robot obtains the state of neighboring robots through local communication; and iteratively solves and updates the path according to the distributed gradient descent method until two adjacent paths converge to a preset threshold.

7. The multi-level composite robot scheduling method according to claim 6, characterized in that, The distributed gradient descent method iteratively solves for the update path, and the formula is as follows: Where t represents the number of iterations, Let α represent the gradient, and let α represent the gradient descent rate.

8. The multi-level composite robot scheduling method according to claim 7, characterized in that, The S3 process involves on-demand local dynamic adjustments, including: Each robot moves to the target point according to the optimized path and performs the task; If obstacles or robot malfunctions are encountered during execution, local path replanning is initiated, and the task is re-executed within the current formation. If the task cannot be completed, it is reported to the robot central scheduling system for reassignment.

9. A multi-level composite robot scheduling system, characterized in that, include: The central scheduling module is used to acquire task information, collect robot status, divide into formations, and assign tasks. The distributed planning module is used to generate target paths and execute tasks; The dynamic adjustment module is used to make local dynamic adjustments as needed during task execution.

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 implements the multi-level composite robot scheduling method according to any one of claims 1-8.

Images