Feshedilebilir mantıksal programlama ile tartışma tabanlı sahne yorumlama

In an agent system that needs to operate in the real world, the problem of maintaining a consistent world model in the face of unreliable, incomplete and inconsistent sensory data should be solved. In this thesis, an approach is presented that addresses the problem of constructing a consistent world model through inconsistent sensor readings by applying an argumentation-based scene interpretation framework in order to accurately model and represent the observations and beliefs of an agent. The proposed approach is based on temporal and probabilistic defeasible logic programming for reasoning. The performance of the proposed approach is evaluated on simulation experiments in the Stage Robot Simulator. It is also shown that the approach is applicable to real world scenarios with an autonomous Pioneer 3-AT robot.