| 引用本文: | 肖凡,杨庆凯,周勃,方浩.面向平均区域覆盖的多机器人分布式控制[J].控制理论与应用,2023,40(3):441~449.[点击复制] |
| XIAO Fan,YANG Qing-kai,ZHOU Bo,FANG Hao.Distributed even coverage control of multi-robot systems[J].Control Theory and Technology,2023,40(3):441~449.[点击复制] |
|
| 面向平均区域覆盖的多机器人分布式控制 |
| Distributed even coverage control of multi-robot systems |
| 摘要点击 2512 全文点击 615 投稿时间：2021-12-15 修订日期：2023-02-21 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2022.11226 |
| 2023,40(3):441-449 |
| 中文关键词 多智能体系统 覆盖控制 维诺图 Lloyd算法 分布式控制 控制系统 |
| 英文关键词 multi agent system coverage control Voronoi diagram Lloyd algorithm distributed control control systems |
| 基金项目 国家自然科学基金项目(61903035, U1913602, 61903036)资助. |
|
| 中文摘要 |
| 随着通信技术、传感技术和控制技术的发展, 多机器人系统因其良好的鲁棒性, 灵活性和可扩展性, 在理论
研究和工程应用中展现出广阔的前景. 区域覆盖是多机器人系统典型应用之一, 目前多采用维诺图分割覆盖区域
并使用Lloyd算法控制机器人前往维诺图细胞中心. 然而传统Lloyd算法存在不平衡问题, 即机器人覆盖区域面积大
小不一, 这降低了多机器人协作效率. 针对平均区域覆盖问题, 本文提出了一种改进的Lloyd算法, 将维诺图中各细
胞面积方差引入Lloyd算法, 相应地设计了基于梯度下降法的分布式控制器. 本文方法降低了维诺图中各细胞面积
的方差, 改善了Lloyd算法的平衡性, 能够实现整个区域面积更为平均的划分与机器人对该区域的覆盖. 数值仿真与
无人机实物实验均验证了改进算法的有效性. |
| 英文摘要 |
| With the development of communication technology, sensing technology and control technology, multi-robot
system shows broad prospects in theoretical research and engineering application due to its good robustness, flexibility
and scalability. Area coverage is one of the typical applications of multi-robot system. At present, the mainstream is to
divide the coverage area using Voronoi diagram and control the robot to go to the cell center by Lloyd algorithm. However,
the traditional Lloyd algorithm has imbalance problem, that is, the coverage area of robots is different, which reduces
the efficiency of multi-robot cooperation. To solve the even coverage problem, this paper improves the Lloyd algorithm
by introducing the cell area variance in the Voronoi diagram into the Lloyd algorithm, and correspondingly deriving the
controller based on the gradient descent method. This method reduces the variance of each cell area in the Voronoi diagram,
which improves the balance of the Lloyd algorithm, and can realize a more even division of the whole area and the coverage
of the robot. The effectiveness of the improved algorithm is verified by numerical simulation and UAV physical experiment. |
|
|
|
|
|