Safety is one of the primary concerns before autonomous robots can be widely accepted by people. To address the safety problem of autonomous navigation, obstacle avoidance continues to be an active research field and an important topic in robotics. Another concern is controlling multi-robot systems to achieve and maintain formation and heading consensus without collision.;A novel switched-system approach is proposed to address obstacle avoidance. The proposed approach is rooted in the classic problem of avoiding static circular objects, but is well suited to large, non-circular objects and dynamic obstacles. For circular obstacles, we define a switching surface as a local subset of ;Two decentralized approaches are developed for formation and heading consensus control of the networked multi-robot systems without collision. The key feature of these approaches are a virtual robot system, which is derived by adding a set of relative translation vectors to the coordinate of real robots. Thus, the real and the virtual multi-robot systems are coupled to each other. A novel continuous and smooth control approach transforms the formation and heading consensus problem to a pose regulation problem. If the system is holonomic, traditional consensus on the virtual graph can stabilize the real system to the desired formation. Another novel switched-system approach results in the real multi-robot system moving to the desired formation and achieving heading consensus as the virtual robots reach pose consensus. The proposed control schemes are proved globally asymptotically stable and asymptotically stable, respectively. Extensive simulation and experiments have been conducted to demonstrate the effectiveness of the proposed approaches. These experiments extend the theoretical development by introducing a teleoperated quadrotor as a leader robot of the heterogeneous multi-robot systems. The same control law works for the extended system, with no modifications.
展开▼
