Combining trajectory control and position control for autonomous mobile robot navigation

Abstract

In the present paper, we described the system architecture of the autonomous mobile robot B21 from RWI focussing on the double control loop implementing the kinematic control together with the underlying drive control. As a highly accurate drive control is taken for granted only the kinematic model is considered for trajectory and position control. In order to manage successfully demanding tasks in dynamic environments, a motion strategy combining trajectory control and position control is suggested. Two different examples where this combination is required are shown. When a self-localization error occurs, the starting point of the reference trajectory is far away from the actual position of the mobile robot. In such a case, to avoid re-planning a motion control strategy is implemented such that the robot finds again the starting point of the reference trajectory without crossing large parts of the unknown environment. When an unexpected obstacle appears on the planned path a collision avoidance manoeuvre is initiated. This manoeuvre also combines trajectory control and position control. Experimental results for both cases show the performance of the presented motion strategies. The used trajectory and position controllers are described.