Evolving Drivers for TORCS using On-Line Neuroevolution

We applied on-line neuroevolution to evolve non-player characters for
The Open Racing Car Simulator. While previous approaches allowed on-line
learning with performance improvements during each generation,
our approach enables a finer grained on-line learning with performance
improvements within each lap. We tested our approach on three tracks
using two methods of on-line neuroevolution (NEAT and rtNEAT) combined
with four evaluation strategies (epsilon-greedy,
epsilon-greedy-improved. softmax, & interval-based) taken from the
literature. We compared the eight resulting configurations on several
driving tasks involving (i) the learning of a driving behavior for a
specific track, (ii) its adaptation to a new track, and (iii) the
generalization capability to unknown tracks. The results we present show
that our approach can successfully evolve drivers from scratch
and can also be used to transfer evolved knowledge to other tracks.
Overall, our results suggest that the approach performs significantly
better when coupled with on-line NEAT and also indicate that
epsilon-greedy-improved and softmax are generally better than the other
evaluation strategies.

This entry was posted in Technical reports. Bookmark the permalink.