Improvement on obstacle avoiding ability based on laser range finder

Mobile robot works in complicated environment have to deal with all kinds of obstacles, some obstacles may leave only a narrow way, and some obstacles moves quickly or arbitrarily, such as a child. The robot must improve its ability to quickly sensing the obstacles and tracking their state, and then avoid these obstacle and move to the original goal at the same time by motion planning. To improve the obstacle avoiding ability, a three layers obstacle avoiding policy is adopted to deal with the obstacle in different range. The dangerous obstacles are determined and the sub-goal based avoiding policy is presented. To improve the responding ability to moving object, a tracking model is set up, which takes into account the influence of moving platform. The robot may anticipate the collision possibility, collision time and position from tracking algorithm, and then three obstacle avoiding policy is presented according to the collision estimation. The experiment shows good moving object avoiding result.