Intelligent navigation for an autonomous mobile robot

Abstract

The authors present a navigational algorithm for solving the problem of collision-free path planning and real-time control of an autonomous mobile robot in an environment cluttered with moving obstacles. The proposed approach is based on geometric representation/multiobjective A* search and the path smoothing/steering control techniques. An attempt was made to introduce a geometric structure into a paradigm that lacked any previous structure, as well as a multiobject search technique. The proposed navigational system has the following three-level structure: identifier, goal selector, and adapter. The authors are currently implementing the algorithm on a Sun 4 artificial intelligence workstation.<>