移动机器人
机器人
避碰
控制器(灌溉)
控制工程
计算机科学
方案(数学)
机器人控制
碰撞
控制理论(社会学)
工程类
控制(管理)
人工智能
数学
数学分析
计算机安全
农学
生物
作者
Guodong Wang,Xiangyu Wang,Shihua Li
出处
期刊:IEEE Transactions on Automation Science and Engineering
[Institute of Electrical and Electronics Engineers]
日期:2022-12-19
卷期号:21 (1): 382-393
被引量:8
标识
DOI:10.1109/tase.2022.3228397
摘要
This paper investigates distributed formation control of multi-mobile robot systems with collision avoidance. A novel guidance module based formation control scheme is established, which consists of two main parts. In the first part, on the communication network of the multi-mobile robot system, a distributed guidance module is constructed from some predesigned virtual dynamics scattered in the robots' feedback loops. By using some proper distributed formation control methods and distributed observer techniques, some formation references are generated by the guidance module. In the second part, by taking these references as the tracking references, some tracking controllers are designed and assigned to the robots such that the robots' positions track their respective references asymptotically. A "two-layer constraint mechanism" is presented in the above controller design to limit both the formation references and the robots' tracking errors such that robots' collision avoidance is guaranteed. In this way, the multi-mobile robot system completes the desired collision-free formation control task. Compared with existing results, the method proposed in this paper has a better plug-and-play function and wider application scope, especially when the practical robots have tight encapsulations such that their pre-equipped tracking controllers cannot be arbitrarily redesigned. Moreover, under the proposed scheme, the asymptotic formation convergence is achieved by the robots without extra limitations on robots' initial states and the utilization of global information. Some simulations and an experiment are given to validate the effects of the proposed methods. Note to Practitioners—This paper is motivated by controller encapsulation characteristics of practical mobile robot products. To complete various formation tasks, most of the related existing formation control methods require robots' controllers to be continually updated and redesigned. However, this purpose is sometimes unachievable because actual mature robot products may have tight controller packages for trade secrets and intellectual property, or engineers want to retain robots' well-adjusted default performances. To overcome this practical limitation, this paper provides a novel "guidance module based formation control scheme". Under this scheme, engineers only need to put most efforts on a guidance module design to generate desired real-time desired tracking references for robots. A new "two-layer constraint mechanism" is proposed and applied here to guarantee collision avoidance among the robots when using the proposed scheme. With this scheme, engineers can establish required formation control strategies without robots' pre-equipped controllers changed and the original control performances abandoned, as long as the robots have some basic capability of tracking given trajectories. Theoretical analysis and experiments on this scheme are all shown. In future research, based on the proposed scheme, we will continue to develop application-oriented formation control methods to improve the anti-disturbance performance and robustness of the multi-mobile robot systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI