Navigation Of The Autonomous Mobile Robot

This thesis is based on the key technique project of"Intelligent Information Processing and Navigation of lunar rover"in National 863 Project, and puts emphasis on navigation of lunar rover.Current robot control architectures are analyzed in detail, a hybrid control architecture based on function behavior interaction is proposed, which endows the robot with quick reactive speed, also with high-level intelligence. According to the control architecture proposed in this thesis, the autonomous mobile robot–AMR-ITL-I is designed and developed, and kinematics model is studied.Basic behaviors of the navigation system are realized based on fuzzy logic method, and the input is the data gain using vision sensor, and the robot navigation based on monocular vision system is realized. Visual navigation algorithms are applied to automated aviation package storage system, validity and practicability of visual navigation have been proved.Dead reckoning algorithm of relative coordinate by odometer and a visual servoing control strategy monitoring dynamic multi-object with just one PTZ(Pan-Tilt-Zoom) camera are proposed in order to increase the precision of navigation, by which the field of vision of the camera is widened and reliability of objects in environment are also enhanced. The model monitoring objects based on finite state machine is put forward. Decision-making table about switch conditions between states of the finite state machine is built, minimal set for monitoring objects rules is achieved by using reduction algorithm of rough set theory. Monitoring dynamic multi-objects is achieved based on the minimal set. The rationality , validity , superiority of the approach have been proved by the experiments accomplished using autonomous mobile robots to finish certain tasks.