Particle Filter-based State Estimation in a Competitive and Uncertain Environment

In this paper we present the application of particle filters for state estimation on a humanoid robot. These filters are used for self-localization and ball tracking in a competitive soccer scenario using robots with limited perceptual and processing capabilities. Some extensions have been applied to the basic algorithms to adapt them to the special needs of this domain which is a subject of high noise and dynamic changes. Different experiments have been carried out to confirm the applicability and the precision of the approach.