RPL LEARN : Extending an Autonomous Robot Control Language to Perform Experience-based Learning
Abstract
In this paper, we extend the autonomous robot control and plan language RPL with constructs for specifying experiences , control tasks, learning systems and their param-eterization, and exploration strategies. Using these constructs , the learning problems can be represented explicitly and transparently and become executable. With the extended language we rationally reconstruct parts of the AG-ILO autonomous robot soccer controllers and show the feasibility and advantages of our approach.
Origin : Files produced by the author(s)
Loading...