The Mobile Robot Map Building And Path Planning

Improving robot intelligence is an important topic in robot field. With the background, this paper focuses on robot map building and path planning in unknown environment. In the unknown environment, robot can not get the environment information in advance, so it is hardly to make the right plan and decision. In order to improve robot intelligence and carrying out task effective, a environment map that robot can understand is build. Map building is the first problem to be solved in this paper. The second problem to be solved is path planning that make robot quickly and with least time to complete task. The main research contents and results are as follows:(1) The robot kinematics model is established. And the robot spatial position presentation and coordinate transformation are analyzed in this paper. The sensor model is also build to test their performance. There are laser sensor, infrared sensor, anti-collision sensors, gyroscopes, encoders and so on in the sensor system. These model of the robot is the basis of building map and planning path.(2) Present a algorithm of local map generating global map. To get the indoor map, we used LMS100laser sensor of SICK to get the environment information, extracted environmental characteristics and matched local map and global map. With the map information, robot uses extended kalman filter (UKF) algorithm to locate.(3) Introduce a algorithm of multi-sensor data fusion and its structure. The robot carries many sensors, and uses the multi-sensor data fusion algorithm to avoid obstacle. The robot finds the door to self-location, and uses the method of point to point to get the optimal path.At last, experiments are carried in our laboratory. These results show the method get satisfactorily effect.