A propagating interface model strategy for global trajectory planning among moving obstacles

Abstract

A global trajectory planning method which employs a collision-trend index and a propagating interface model to perform mobile robot navigation is presented in this paper. To simplify the mathematical representation and geometrical approximation, all the objects in the workspace are modeled as ellipses. Using a series of geometrical transformations between the ellipses, which represent the mobile robot and obstacles, the computational complexity of collision detection in trajectory planning can be reduced tremendously. To keep the front propagating in the normal direction with positive propagating speed, the collision-trend index plays an important role in determining the propagating speed for a front over workspace. The index is obtained by mapping the geometrical relationship between the ellipses into the profile of a Gaussian distribution. Several simulations to demonstrate the performance of the proposed method are given. The results reveal that the proposed method is always able to generate an optimal collision-free trajectory for a mobile robot navigating in an environment with dynamic and static obstacles.