@article{Cully:2016:10.1162/EVCO_a_00143,
author = {Cully, A and Mouret, J-B},
doi = {10.1162/EVCO_a_00143},
journal = {Evolutionary Computation},
pages = {59--88},
title = {Evolving a behavioral repertoire for a walking robot},
url = {http://dx.doi.org/10.1162/EVCO_a_00143},
volume = {24},
year = {2016}
}

Download

TY  - JOUR
AB  - Numerous algorithms have been proposed to allow legged robots to learn to walk.However,  most  of  these  algorithms  are  devised  to  learn  walking  in  a  straight  line,which is not sufficient to accomplish any real-world mission. Here we introduce theTransferability-based  Behavioral  Repertoire  Evolution  algorithm  (TBR-Evolution),  anovel evolutionary algorithm that simultaneously discovers several hundreds of simplewalking controllers, one for each possible direction. By taking advantage of solutionsthat are usually discarded by evolutionary processes, TBR-Evolution is substantiallyfaster than independently evolving each controller. Our technique relies on two meth-ods: (1) novelty search with local competition, which searches for both high-performingand diverse solutions, and (2) the transferability approach, which combines simulationsand real tests to evolve controllers for a physical robot. We evaluate this new techniqueon a hexapod robot. Results show that with only a few dozen short experiments per-formed on the robot, the algorithm learns a repertoire of controllers that allows therobot to reach every point in its reachable space. Overall, TBR-Evolution introduceda  new  kind  of  learning  algorithm  that  simultaneously  optimizes  all  the  achievablebehaviors of a robot.
AU  - Cully,A
AU  - Mouret,J-B
DO  - 10.1162/EVCO_a_00143
EP  - 88
PY  - 2016///
SN  - 1063-6560
SP  - 59
TI  - Evolving a behavioral repertoire for a walking robot
T2  - Evolutionary Computation
UR  - http://dx.doi.org/10.1162/EVCO_a_00143
UR  - http://hdl.handle.net/10044/1/50402
VL  - 24
ER  - 

Download