Heterogeneous robot group cooperation system and method based on dynamic role allocation and collaborative planning

Abstract

The invention discloses a heterogeneous robot group collaboration system and method based on dynamic role allocation and collaborative planning, and relates to the technical field of robot control and collaboration, and the system comprises a cloud collaboration management platform, a plurality of heterogeneous robot agents, and a distributed task allocation and coordination algorithm and a group collaboration path planning module which run between the cloud collaboration management platform and the heterogeneous robot agents; information islands of the heterogeneous robots are broken through a unified communication protocol, and efficient communication, intelligent task allocation and conflict-free group cooperation of the heterogeneous robots in a dynamic environment are realized in combination with a task bidding allocation mechanism of dynamic context sensing and a path planning scheme of space-time reservation. According to the invention, the problems of low communication efficiency, rigid distribution and planning conflict in the prior art are effectively solved, and key technical support is provided for complex scene applications such as intelligent logistics and intelligent security and protection.

Description

Technical Field

[0001] This invention relates to the technical field of robot control and collaboration, specifically to a heterogeneous robot swarm collaboration system and method based on dynamic role allocation and collaborative planning. Background Technology

[0002] With the continuous development of robotics technology, single-function robots have been widely used in specific scenarios such as cleaning, security, and logistics. However, tasks in complex scenarios usually have multi-dimensional requirements, such as the need to simultaneously achieve functions such as movement, perception, manipulation, and transportation, which a single type of robot cannot complete independently and efficiently.

[0003] Existing research on multi-robot systems mainly focuses on formation control or simple task allocation for homogeneous robots. For heterogeneous robot systems, the following technical bottlenecks exist: 1. Inconsistent communication protocols: Differences in communication protocols between different robot platforms lead to high data transmission latency and low efficiency, affecting information exchange and collaborative decision-making.

[0004] 2. Rigid task allocation: Traditional task allocation algorithms (such as auction algorithms and market mechanisms) are based on predefined static robot capability models, which cannot adapt to the dynamic changes in robot status (such as battery level and fault) and task context (such as task urgency and item category), making it difficult to achieve globally optimal allocation.

[0005] 3. Path planning conflicts: In dynamic environments, independent path planning by each robot can easily lead to path conflicts, resource competition (such as the struggle for right-of-way in narrow passages) and inefficiency, lacking a collaborative planning mechanism that considers the overall effectiveness of the group. Summary of the Invention

[0006] The purpose of this invention is to provide a heterogeneous robot group collaboration system and method based on dynamic role allocation and collaborative planning, so as to overcome the above-mentioned defects in the prior art.

[0007] A heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning includes a cloud-based collaborative management platform, multiple heterogeneous robot agents, and a unified communication protocol connecting the cloud-based collaborative management platform and the heterogeneous robot agents. The cloud-based collaborative management platform includes a global dynamic map module and a task pool management module. The global dynamic map module is used to receive and integrate local perception information uploaded by all robots, construct and update a unified global semantic map that includes static obstacles, dynamic objects, task points, and the real-time position and status of robots. The task pool management module is used to receive, create, and manage task requests submitted by human users, triggered by system presets, or actively discovered and reported by robots. Each heterogeneous robot agent has a unified communication interface and a local decision-making and execution module.

[0008] Preferably, it also includes a distributed task allocation and coordination algorithm, which runs between the cloud-based collaborative management platform and various heterogeneous robot agents, with the core being a dynamic role allocation mechanism; Preferably, the dynamic role allocation mechanism includes: when a new task is generated, the cloud-based collaborative management platform broadcasts a task announcement to the robot group, including a task description, required capabilities, target location, and priority; after receiving the announcement, the robot calculates its bid value based on its real-time status and local environmental information; the cloud-based collaborative management platform collects all bid values, selects the robot with the optimal comprehensive bid value based on an improved contract network protocol or a greedy algorithm, assigns the task, and notifies other robots; and supports robots actively rejecting unsuitable tasks and reallocation when tasks fail.

[0009] Preferably, it also includes a group collaborative path planning module, which runs on a cloud-based collaborative management platform.

[0010] Preferably, the group collaborative path planning module includes: a cloud-based collaborative management platform maintaining a spatiotemporal conflict prediction map to predict potential spatiotemporal conflict points for all robot planned paths; when performing global path planning for a single robot, the predicted trajectories of other robots are used as dynamic obstacles; for robots that need to pass through the same narrow area, a right-of-way management mechanism based on spatiotemporal reservation is introduced, requiring the robot to reserve exclusive access to specific spatial resources for a certain period of time from the cloud-based collaborative management platform.

[0011] Preferably, the unified communication interface follows a preset high-efficiency communication protocol, defines a standardized message format to transmit robot status, perception data and task-related notifications, and supports publish / subscribe mode; the local decision and execution module is responsible for the robot's motion control, task execution and local obstacle avoidance.

[0012] The heterogeneous robot swarm collaboration method based on dynamic role allocation and collaborative planning is applied to the aforementioned heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning, including the following steps: S1, System Initialization: The cloud-based collaborative management platform starts the global dynamic map module and the task pool management module, and each heterogeneous robot agent starts the unified communication interface and the local decision-making and execution module to establish a communication connection between the cloud and the robots; S2, Task Acquisition and Release: The task pool management module acquires task requests and creates tasks, and the cloud-based collaborative management platform broadcasts task announcements to the robot swarm through the unified communication protocol; S3, Dynamic Task Bidding and Allocation: After receiving the task announcement, the robots calculate the bidding value and feed it back to the cloud. The cloud collects the bidding values ​​and selects the best robot to grant the task; S4, Group Collaborative Path Planning: The cloud-based collaborative management platform plans a conflict-free global path for the robots executing the task based on the spatiotemporal conflict prediction map and the spatiotemporal reservation mechanism; S5, Task Execution and Monitoring: The robots execute tasks according to the planned path and report their status and task progress to the cloud in real time. The cloud monitors the robot status and task execution through the global dynamic map module. If the task fails, a reassignment process is triggered.

[0013] Preferably, in step S3, when the robot calculates the bid value, it will calculate based on its real-time status and local environmental information. The real-time status includes the distance to the task target point, the current battery level, whether it is idle, and whether it has the functions required for the task. The local environmental information includes the congestion situation on the path leading to the task target.

[0014] Preferably, in step S4, when planning the global path, the cloud-based collaborative management platform first predicts the spatiotemporal conflict points of each robot's planned path, then incorporates the predicted trajectories of other robots as dynamic obstacles into the path planning, and finally manages the spatiotemporal reservations of robots that need to pass through narrow areas.

[0015] The beneficial effects achieved by this invention are as follows: 1. This application establishes a "common language" for heterogeneous robot groups through a unified communication protocol, breaks down information silos, and enables high-dimensional, semantic information (real-time maps, object poses, task instructions) to flow seamlessly and efficiently among robots, laying a data foundation for group intelligent decision-making, realizing a leap from "individual perception" to "group cognition", and achieving efficient and deep group collaborative perception.

[0016] 2. Enhance the intelligence and robustness of task allocation: A dynamic context-aware task bidding and allocation mechanism overcomes the shortcomings of traditional static allocation models. The robot dynamically bids based on its real-time state and environmental context, upgrading the task allocation result from merely achievable to globally optimal. Simultaneously, this mechanism is flexible, adapting to unexpected situations such as robot malfunctions and insufficient power, automatically triggering task reallocation, significantly enhancing system efficiency and robustness.

[0017] 3. Ensure conflict-free and efficient group path planning: Introduce a collaborative path planning scheme based on spatiotemporal reservation, which improves conflict resolution from passive real-time collision avoidance to proactive and forward-looking planning. The system predicts and resolves potential path conflicts at the global level and intelligently allocates spatiotemporal resources through the reservation mechanism, fundamentally avoiding deadlock and congestion. This enables the robot group to work smoothly and efficiently in parallel in complex environments, significantly improving the overall task throughput. Attached Figure Description

[0018] Figure 1 This is an architecture diagram of the cloud collaborative management platform of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims and drawings of this application are intended to cover non-exclusive inclusion.

[0021] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of the phrase "embodiment" in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] like Figure 1 As shown, this invention provides a heterogeneous robot group collaboration system and method based on dynamic role assignment and collaborative planning, wherein: I. Heterogeneous Robot Swarm Collaboration System The system includes a cloud-based collaborative management platform, multiple heterogeneous robot agents, and a unified communication protocol connecting the two. It is also equipped with a distributed task allocation and coordination algorithm and a group collaborative path planning module.

[0023] The cloud-based collaborative management platform, acting as the system's brain, is responsible for macro-level information fusion and decision-making. It includes a global dynamic map module and a task pool management module. The global dynamic map module receives and integrates local perception information uploaded by all robots (such as SLAM maps and object detection results), constructs and updates a unified global semantic map in real time, and labels static obstacles, dynamic objects, task points, and the robot's real-time position and status. The task pool management module receives, creates, and manages task requests. Task sources include submissions from human users, system-preset triggers, and proactive reporting by robots (such as a security robot reporting a "stain discovery" event).

[0024] Heterogeneous Robot Agents: Each robot acts as an independent intelligent agent, possessing a unified communication interface and a local decision-making and execution module. The unified communication interface follows a pre-defined high-efficiency communication protocol, defining standardized message formats for transmitting robot status (position, battery level, speed), perception data (local map, detected objects), and task requests / responses / completion notifications, etc., and supports publish / subscribe modes to ensure efficient and targeted information distribution; the local decision-making and execution module is responsible for the robot's motion control, task execution, and local obstacle avoidance.

[0025] Distributed task allocation and coordination algorithm: This algorithm operates between the cloud platform and the robots, with a core dynamic role allocation mechanism. When a new task is generated, the cloud platform broadcasts a task announcement to the group, containing metadata such as task description, required capabilities, target location, and priority. After receiving the announcement, the robots calculate their bid value (quantifying the cost or suitability of executing the task) based on their real-time status (distance from the target point, current battery level, idle status, and availability of the required functions) and local environmental information (congestion on the path to the target). After collecting all bid values, the cloud platform selects the robot with the optimal comprehensive bid value based on an improved contract network protocol or a greedy algorithm, assigns the task, and notifies the other robots. This process allows robots to actively reject unsuitable tasks and can trigger a reassignment when a task fails.

[0026] Group collaborative path planning module: running on a cloud platform, the cloud platform maintains a spatiotemporal conflict prediction map, predicting potential spatiotemporal conflict points for all robot planned paths; when performing global path planning for a single robot, in addition to considering static obstacles, the predicted trajectories of other robots are also considered as dynamic obstacles; for robots that need to pass through the same narrow area, a right-of-way management mechanism based on spatiotemporal reservation is introduced, whereby robots need to reserve exclusive access to specific spatial resources for a certain period of time from the cloud, fundamentally avoiding deadlock and collisions.

[0027] II. Heterogeneous Robot Swarm Collaboration Method This method is applied to the aforementioned heterogeneous robot swarm collaboration system, and the specific steps are as follows: S1. System Initialization: The cloud-based collaborative management platform starts the global dynamic map module and task pool management module, and each heterogeneous robot agent starts the unified communication interface and local decision-making and execution module to establish a stable communication connection between the cloud and the robot.

[0028] S2. Task Acquisition and Release: The task pool management module acquires various task requests and creates tasks. The cloud-based collaborative management platform broadcasts task announcements containing key task information to the robot group through a unified communication protocol.

[0029] S3. Dynamic Task Bidding and Allocation: After receiving a task announcement, the robot calculates its bid value based on its real-time status and local environmental information, and feeds the bid value back to the cloud. The cloud-based collaborative management platform collects all bid values, selects the robot with the best comprehensive bid value based on the improved contract network protocol or greedy algorithm, assigns the task to it, and notifies other robots at the same time.

[0030] S4. Group Collaborative Path Planning: The cloud-based collaborative management platform predicts the spatiotemporal conflict points of each robot's planned path through a spatiotemporal conflict prediction map, and uses the predicted trajectories of other robots as dynamic obstacles to initially plan the path for the robot performing the task. For robots that need to pass through the same narrow area, the platform allows the robot to reserve a time slot for exclusive use of space resources based on the spatiotemporal reservation mechanism, ultimately generating a conflict-free global path.

[0031] S5. Task Execution and Monitoring: The robot executes tasks according to the planned global path, and reports its status (location, battery level, etc.) and task progress to the cloud collaborative management platform in real time during the process. The cloud monitors the robot's status and task execution in real time through the global dynamic map module. If robot failure or task execution failure occurs, the task reassignment process is immediately triggered to ensure the task is completed smoothly.

[0032] In summary, this application achieves efficient and in-depth group collaborative perception: through a unified communication protocol, a standard "common language" is established for heterogeneous robot groups, breaking down information silos and enabling the efficient flow of high-dimensional semantic information, thus laying a solid data foundation for group intelligent decision-making.

[0033] It improves the intelligence level and robustness of task allocation: the task bidding and allocation mechanism based on dynamic context awareness overcomes the drawbacks of the traditional static allocation model, makes the task allocation result globally optimal, and can adaptively cope with sudden situations such as robot failure and insufficient power, with extremely strong elasticity.

[0034] It ensures global conflict-free and high-efficiency group path planning: By adopting a collaborative path planning scheme based on spatiotemporal reservation, conflict resolution is upgraded from passive real-time collision avoidance to proactive forward planning, which fundamentally avoids deadlock and congestion and significantly improves the overall task throughput.

[0035] The embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A heterogeneous robot swarm collaboration system based on dynamic role assignment and collaborative planning, characterized in that: The system includes a cloud-based collaborative management platform, multiple heterogeneous robot agents, and a unified communication protocol connecting the cloud-based collaborative management platform and the heterogeneous robot agents. The cloud-based collaborative management platform includes a global dynamic map module and a task pool management module. The global dynamic map module is used to receive and integrate local perception information uploaded by all robots, construct and update a unified global semantic map that includes static obstacles, dynamic objects, task points, and the real-time position and status of the robots. The task pool management module is used to receive, create, and manage task requests submitted by human users, triggered by system presets, or actively discovered and reported by robots. Each heterogeneous robot agent has a unified communication interface and a local decision-making and execution module.

2. The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning according to claim 1, characterized in that: It also includes a distributed task allocation and coordination algorithm, which runs between the cloud-based collaborative management platform and various heterogeneous robot agents, with a core dynamic role allocation mechanism.

3. The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning according to claim 2, characterized in that: The dynamic role allocation mechanism includes: when a new task is generated, the cloud-based collaborative management platform broadcasts a task announcement to the robot group, including the task description, required capabilities, target location, and priority; after receiving the announcement, the robot calculates its bid value based on its real-time status and local environmental information; the cloud-based collaborative management platform collects all bid values, selects the robot with the best comprehensive bid value based on the improved contract network protocol or greedy algorithm, assigns the task, and notifies other robots; and supports robots actively rejecting unsuitable tasks and reassignment when tasks fail.

4. The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning according to claim 1, characterized in that: It also includes a group collaborative path planning module, which runs on a cloud-based collaborative management platform.

5. The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning according to claim 4, characterized in that: The group collaborative path planning module includes: a cloud-based collaborative management platform that maintains a spatiotemporal conflict prediction map to predict potential spatiotemporal conflict points for all robot planned paths; when performing global path planning for a single robot, the predicted trajectories of other robots are used as dynamic obstacles; for robots that need to pass through the same narrow area, a right-of-way management mechanism based on spatiotemporal reservation is introduced, requiring the robot to reserve exclusive access to specific spatial resources for a certain period of time from the cloud-based collaborative management platform.

6. The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning according to claim 1, characterized in that: The unified communication interface follows a preset high-efficiency communication protocol, defines a standardized message format to transmit robot status, perception data and task-related notifications, and supports publish / subscribe mode; the local decision and execution module is responsible for the robot's motion control, task execution and local obstacle avoidance.

7. A heterogeneous robot group collaboration method based on dynamic role assignment and collaborative planning, characterized in that: The heterogeneous robot swarm collaboration system based on dynamic role allocation and collaborative planning, as described in any one of claims 1-6, comprises the following steps: S1, System initialization: The cloud-based collaborative management platform starts the global dynamic map module and the task pool management module, and each heterogeneous robot agent starts the unified communication interface and the local decision-making and execution module to establish a communication connection between the cloud and the robots; S2, Task acquisition and release: The task pool management module acquires task requests and creates tasks, and the cloud-based collaborative management platform broadcasts task announcements to the robot swarm through a unified communication protocol; S3, Dynamic task bidding and allocation: After receiving the task announcement, the robot calculates the bidding value and feeds it back to the cloud. The cloud collects the bidding values ​​and selects the best robot to grant the task; S4, Swarm collaborative path planning: The cloud-based collaborative management platform plans a conflict-free global path for the robots performing the task based on a spatiotemporal conflict prediction map and a spatiotemporal reservation mechanism; S5, Task execution and monitoring: The robot executes the task according to the planned path and feeds back the status and task progress to the cloud in real time. The cloud monitors the robot status and task execution through the global dynamic map module. If the task fails, a reassignment process is triggered.

8. The heterogeneous robot group collaboration method based on dynamic role assignment and collaborative planning according to claim 7, characterized in that: In step S3, when the robot calculates the bid value, it will perform the calculation based on its real-time status and local environmental information. The real-time status includes the distance to the task target point, the current battery level, whether it is idle, and whether it has the functions required for the task. The local environmental information includes the congestion situation on the path leading to the task target.

9. The heterogeneous robot group collaboration method based on dynamic role assignment and collaborative planning according to claim 7, characterized in that: In step S4, when planning the global path, the cloud-based collaborative management platform first predicts the spatiotemporal conflict points of each robot's planned path, then incorporates the predicted trajectories of other robots as dynamic obstacles into the path planning, and finally manages the spatiotemporal reservations of robots that need to pass through narrow areas.

Images