Abstract
Based on the utilitarian navigation concept, the paper introduces a recurrent neural network for the search of sensory sources by a mobile robot. First, a utility function for sensory-based search is defined and a dynamic optimization process is obtained. Next, a bio-inspired neural model of sensory-motor coordination is proposed. The paper analyzes the proposed motor neural circuit in more detail, using a dynamic model of the respective motor neurons. Experimental results confirm the viability of the recurrent neural model for implementing sensory-based search by a mobile robot. © Springer-Verlag Berlin Heidelberg 2003.
Cite
CITATION STYLE
Maravall, D., De Lope, J., & Patricio, M. Á. (2003). A recurrent neural network for robotic sensory-based search. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2687, 153–160. https://doi.org/10.1007/3-540-44869-1_20
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
