Evolution of central pattern generators for bipedal walking in a real-time physics environment

Author

Reil, Torsten ; Husbands, Phil

Author_Institution

Dept. of Zoology, Oxford Univ., UK

Volume

6

Issue

2

fYear

2002

fDate

4/1/2002 12:00:00 AM

Firstpage

159

Lastpage

168

Abstract

We describe an evolutionary approach to the control problem of bipedal walking. Using a full rigid-body simulation of a biped, it was possible to evolve recurrent neural networks that controlled stable straight-line walking on a planar surface. No proprioceptive information was necessary in order to achieve this task. Furthermore, simple sensory input to locate a sound source was integrated to achieve directional walking. To our knowledge, this is the first work that demonstrates the application of evolutionary optimization to 3D physically simulated biped locomotion

Keywords

acoustic transducers; direction-of-arrival estimation; evolutionary computation; legged locomotion; neurocontrollers; optimal control; recurrent neural nets; stability; 3D physically simulated biped locomotion; bipedal walking; central pattern generator evolution; directional walking; evolutionary algorithms; evolutionary control problem; evolutionary optimization; evolutionary robotics; planar surface; real-time physics environment; recurrent neural network evolution; rigid-body simulation; sensory input; sound source location; stable straight-line walking; Animation; Computer graphics; Evolution (biology); Evolutionary computation; Humans; Legged locomotion; Physics; Recurrent neural networks; Robot sensing systems; Zoology;

fLanguage

English

Journal_Title

Evolutionary Computation, IEEE Transactions on

Publisher

ieee

ISSN

1089-778X

Type

jour

DOI

10.1109/4235.996015

Filename

996015