报告题目：Cooperative Multi-Robot Systems

报告时间：2020年12月8日上午9:00-10:00

报告地点：腾讯会议 981 882 535

报告人：Ya-jun Pan 教授

报告人简介：

Ya-Jun Pan is currently a Professor in the Department of Mechanical Engineering at Dalhousie University, Canada. Her research interests are robust nonlinear control, cyber physical systems, intelligent transportation systems, haptics and collaborative multiple robotic systems. Currently she serves as Associate Editors for IEEE Transactions on Industrial Electronics, IEEE Transactions on Cybernetics and ASME/IEEE Transactions on Mechatronics, Vice President-Atlantic of CSME, Canada and has served as AEs for several other journals and conferences. She is a Fellow of ASME, a Senior Member of IEEE, a Member of CSME and a registered Professional Engineer in Nova Scotia, Canada.

报告内容简介：

Research on cooperative multi-robot systems in the Advanced Control and Mechatronics lab will be introduced. Specifically, the presentation will focus on the consensus problem of a team of multiple quadcopter systems. Multiple-quadcopter systems have various civilian and military applications, such as forest fire monitoring and load transportation. However, since multiple-quadcopter systems are networked control systems (NCSs), they suffer from network-induced constraints, such as time delays and packet losses. Consensus, which is a basic coordination problem, is often desired for the group in achieving tasks. The objective of this work is to develop novel distributed consensus algorithms for multiple-quadcopter systems over two types of communication delays: uniform constant delays and asynchronous time-varying delays. The quadcopter system is simplified into four decoupled subsystems such that it can be studied in a multi-agent system (MAS) scale. The interactions among quadcopters are modeled using algebraic graph theory. The consensus problem is then converted to a stability analysis problem by defining the consensus error dynamics. Sufficient conditions for stabilizing controller design are developed based on Lyapunov's method and linear matrix inequality (LMI) techniques for both cases. Finally, extensive MATLAB simulations are carried out for both cases to verify the proposed algorithms. Discussions are given regarding the feasibility and effectiveness of the proposed controllers under various conditions. Real-time hardware implementation will be briefly discussed as well.