Learning both a world model and navigation paths in an unknown environment

This paper presents a new method for mobile robot navigation in an unknown world. The partigame learning approach is used to construct a world model, and to learn to navigate from a starting position to a known goal region. T he navigation architecture then integrates a new approach, based on the app lication of the Fuzzy ART neural architecture, for on-line map building fro m actual sensor data. This: leads to an improved world model, which is then used to introduce a predictive on-line trajectory filtering method resulti ng in a new and more effective navigation approach. It is assumed that the robot knows its own current world location, is able to perform sensor-based obstacle detection (not avoidance), and straight-line motions. Simulation and real-robot results obtained with a Nomad 200 mobile robot will be prese nted demonstrating the effectiveness of the discussed methods. Quantitative results will demonstrate (I) the exploration and planning improvements of the new navigation approach, and (2) that the constructed world model (orig inal or improved) is general purpose in the sense that its usefulness is no t restricted to be used on learning a particular path, but is valuable for learning paths with different (Start,Goal) pairs.