Virtual autonom robot

The main goal of this paper is to extend the potential based guiding (PBG) model to a more general form that can be approximated by a common type neuro-fuzzy algorithm. The extended model (VFB) eliminates the strongly alternating behavior of PBG. As a result of virtual training of VFB based on different guiding-style description models, some examples are given, which show the effectiveness of VFB. These examples also show the advantages of using virtual environment, COSMOS for training of different guiding-style description robot models.