Constraint Based World Modeling for Multi Agent Systems in Dynamic Environments

Abstract

Mobile autonomous robotics is a young and complex field of research. Since the world is uncertain and since robots can only gain partial information about it, probabilistic navigation algorithms became popular whenever a robot has to localize itself or surrounding objects. Furthermore, cooperative exploration and localization approaches have become very relevant lately, as robots begin to act not just alone but in groups. Within my thesis I analyze, how information can be exchanged between robots in order to improve their world model. Therefore I examine how communication of spatial percept-relations can help to improve the accuracy of the world model, in particular when the robots are poorly self-localized. First, percept-relations are being used to increase the modeling accuracy in static situations, later the approach is extended to moving objects. After focussing on suitable sensory data for communication, in the second part I present a Bayesian modeling approach, using constraint satisfaction techniques for complex belief functions. Constraint based localization methods will be analyzed in order to have a group of robots efficiently localized and to model their environment. The presented algorithms were implemented and tested within the RoboCup Standard Platform League (SPL).