Title :
Evolutionary computing systems for obstacle avoidance
Author :
Bostan, I. ; Franti, E. ; Dascalu, M. ; Goschin, S.
Author_Institution :
Univ. of Pitesti, Romania
fDate :
28 Sept.-2 Oct. 2003
Abstract :
In this paper we present two neuronal network architectures to perform the task of obstacle avoidance for a little mobile robot named Khepera, which has 8 infrared proximity detectors and two independent motors. The first one used Hebbian learning for weight adaptation and the second one use evolutionary weight selection. The experiments are only simulated.
Keywords :
Hebbian learning; mobile robots; neural nets; Hebbian learning; evolutionary computing systems; evolutionary weight selection; infrared proximity detectors; little mobile robot; neuronal network architectures; obstacle avoidance; weight adaptation; Biological neural networks; Control systems; Electronic mail; Genetic mutations; Infrared detectors; Intelligent robots; Mobile robots; Neural networks; Robot control; Testing;
Conference_Titel :
Semiconductor Conference, 2003. CAS 2003. International
Print_ISBN :
0-7803-7821-0
DOI :
10.1109/SMICND.2003.1252464
