|
| 工业机器人群体协作的边缘计算任务分配算法研究 |
| Research on Edge Computing Task Allocation Algorithms for Collaborative Industrial Robot Swarms |
| 投稿时间：2025-07-27 |
| DOI：10.16018/j.cnki.cn32-1650/n.202602006 |
| 中文关键词: 边缘计算 工业机器人 多机器人协同 任务分配 路径规划 |
| 英文关键词: edge computing industrial robot multi-robot collaboration task allocation path planning |
| 基金项目:2024年安徽省高校质量工程项目(2024xqjd048)；2023年安徽省高校质量工程项目(2023jyxm1671)；2022年安徽省高校自然科学研究项目(2022AH053008) |
|
| 摘要点击次数: 2 |
| 全文下载次数: 1 |
| 中文摘要: |
| 针对工业机器人群体协作在实时性与资源分配效率方面所面临的挑战,提出一种基于边缘计算的协同任务分配与路径规划方法。在任务分配层面,采用改进合同网协议与动态负载均衡策略,以应对设备异构与任务突变;在路径规划层面,融合时空地图与强化学习避障策略,确保运动安全;同时,设计边缘-云端协同机制,通过智能卸载决策与断网应急方案保障系统可靠性;最后,通过6台UR5e机械臂与3台边缘服务器搭建的测试平台,在装配与分拣场景中进行验证。结果表明,相较于基于区块链的方案,该方法使系统响应延迟降低38.74%,边缘节点带宽消耗减少61.94%。 |
| 英文摘要: |
| To address the challenges in real-time performance and resource allocation efficiency faced by collaborative industrial robot swarms, this paper proposes an edge computing-based approach for cooperative task allocation and path planning. At the task allocation level, an improved Contract Net Protocol combined with a dynamic load balancing strategy is employed to handle device heterogeneity and sudden task changes. At the path planning level, a safety-assured motion strategy is achieved by integrating spatiotemporal maps with reinforcement learning-based obstacle avoidance. Additionally, an edge-cloud collaboration mechanism is designed, incorporating intelligent task offloading decisions and network-disconnection contingency plans to ensure system reliability.A testbed comprising six UR5e robotic arms and three edge servers was built to validate the proposed method in assembly and sorting scenarios. Experimental results demonstrate that, compared to blockchain-based approaches, the proposed method reduces system response latency by 38.74% and decreases bandwidth consumption at edge nodes by 61.94%. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|