Dynamic Path Planning Of Autonomous Mobile Robot In Complex Indoor Scenarios

Because of its powerful function and widespread application, wheeled mobile robot has become a pop research in the field of robotics. It attracts more and more scholars' attention. Path planning and obstacle avoidance are key technologies in the research area of robot, which are prerequisites for other high-level task. To build a path planner with real-time and effectiveness characters, this paper carries out the research of path planning on a nonholonomic wheeled robot in complex indoor scenarios. In this way, the wheeled mobile robot could adapt to the changing environment, and the path planning and dynamic obstacle avoidance capabilities are also improved. The robot can complete the autonomous navigation tasks in a dynamic environment, which lays a good foundation for further applications.Watching a number of moving objects on a robot in motion is one of the key technologies in the field of robot's application. It is a precondition of achieving path planning and obstacle avoidance in a dynamic environment. Based on a LRF sensor, the dynamic multi-target tracking program is verified effective to track a number of dynamic targets. During the research, the dynamic objects in consideration are human targets. Based on the geometric features of human body and the NNJPDA arithmetic, this program ensures the effective tracking of human targets, which can be used in dynamic path planning.In this paper, a path planning scheme in a dynamic partly known environment is proposed. Combined with the dynamic objects tracking method above, a local oriented solution based on dynamic range histogram is put forward, which takes the moving trend and dynamic factors into consideration. On the global level, this paper uses a dynamic distance transform method to achieve the subgoals' update. This method could reflect the change of environment quickly and make its decision. The simulation and experiment results in dynamic environment verify that the method is effective and practical.In the last part of this paper, some academic discussions about dynamic path planning are showed. In order to consider the effects of dynamic objects on the model level, some exploratory research has been carried out, which mainly about map building and environment representation. More work is still needed in further research.