Evolution of event and delay controlled neuronal network for locomotion

Fine neurocontrol of complex dynamical systems such as locomotion of robot requires feedback for the estimation of deviations from the target behavior. Artificial neural networks is a powerful tool for locomotion control but traditional architectures lack explicit mechanisms for the evaluation of ac...

Full description

Saved in:
Bibliographic Details
Published inProceedings of the International Conference on Genetic and Evolutionary Methods (GEM) p. 1
Main Authors Shirshova, Maria, Burtsev, Mikhail
Format Conference Proceeding
LanguageEnglish
Published Athens The Steering Committee of The World Congress in Computer Science, Computer Engineering and Applied Computing (WorldComp) 01.01.2014
Online AccessGet full text

Cover

More Information
Summary:Fine neurocontrol of complex dynamical systems such as locomotion of robot requires feedback for the estimation of deviations from the target behavior. Artificial neural networks is a powerful tool for locomotion control but traditional architectures lack explicit mechanisms for the evaluation of actions results. To address this issue we propose a novel model of formal neuron capable of simple assessment of the timing as well as results of its' own activity. This event and delay controlled (ECD) model of neuron was used to evolve neurocontrollers for robot locomotion. Suggested ECD-model had significantly better performance in our simulations compared to traditional neural networks architectures.