Defeasible Deontic Robot Control Based on Extended Vector Annotated Logic Programming

摘要

We have already proposed an annotated logic program called an EVALPSN (Extended Vector Annotated Logic Program with Strong Negation) to deal with defeasible deontic reasoning. In this paper, we propose a defeasible deontic action control system for a virtual robot based on EVALPSN. We suppose a beetle robot who is traveling a maze with three kinds of obstacles and has some different kinds of sensors to detect the obstacles. If some sensor values are input to the robot control, the next action that the robot should do is computed by the EVALPSN programming system.