Motion Planning And Control Of Tele-autonomous Mobile Robot

In recent years, mobile robots, especially autonomous mobile robot, have become a very active research field of robot technology. The extents of their intelligence constantly improve, while have more and more strong autonomy. However, in lots of applications, with the current level of the relevant supporting technologies, the mobile robot is also very difficult to have the ability to achieve complete autonomy, and in many cases require the participation of human. So, to combine the teleoperation technology and autonomous system while improving the level of robot intelligence, and then create a tele-autonomous system, characterized the integration of human intelligence and machine intelligence, has great significance to enhance the robot adaptability and the ability to achieve mission operations, broaden the scope of application of the mobile robot. This dissertation does some research on the motion planning and control of tele-autonomous mobile robot.First of all, based on human-machine cooperation concept, a three-layer hierarchical control architecture for tele-autonomous mobile robot is proposed. In the architecture, a deliberation-reaction hybrid control strategy is adopted, and combining with the upper layer of teleoperation system, namely monitoring assistance layer. With the teleoperation-based human-machine interaction, the operator can take part in robot's navigation and control process, thus the human-machine intelligence integration is implemented.Secondly, intelligent mobile robot and remote reality, based on Omni-Directional Vision, is combined together to design the teleoperation system of mobile robot. Then, the augmented reality technology is used in the remote-reality-based teleoperation in order to enhance the capacity of human-computer interaction and overcome the impact of communication delay. Besides, make research on digital image stabilization technology for robot surveillance, a fast and robust image motion estimation algorithm is designed. Thirdly, for the motion planning of tele-autonomous mobile robot in uncertain dynamic environments, a human-robot cooperative motion planning and control method is proposed based on multi-constrained environment modeling. This method uses adaptive dynamic window, and dynamics constrain and tracking control problems are considered. Remote operator can cooperate mobile robot using shared control, and effective wall following action and urgent collision avoiding action are designed. Simulation results show that the environment modeling and real-time motion planning method has very good environmental adaptability and obstacle avoidance capabilities, access to a better security and accessibility, but the exercise of dynamic collision avoidance of moving obstacles less than ideal. So, in order to deal with the problem of dynamic collision avoidance, the velocity obstacles method is used. For the existing problems of the past velocity obstacles methods used for dynamic collision-avoidance, corresponding improvement measures have been worked out. In the velocity change space, based on collision-avoidance behavior dynamics, proposes the dynamic window method of multi-step reachable velocity change region, and a new evaluation function is designed for the dynamic collision-avoidance planning. Through simulation experiments, the feasibility and superiority of the above-mentioned methods are verified respectively.Finally, the path following control technology of mobile robot is researched. According to the description of robot oriented positioning control, a novel path following fuzzy controller which based on human driving behavior has been designed. As the robot's direction error and orientation error relative to oriented goal are hardly coupled, and can not be controlled in the same time only by the velocity control vector, so a position-orientation alternate control method is presented. Simulation experimental results show that the method is effective, can be able to achieve fast and accurate path following.