Bio-inspired ModelofRobotAdaptive Learning and Mapping
Alejandra Barrera Ramírez, Weitzenfeld Ridel · 2006
alfredogitam.mx Abstract -Inthis paperwepresent amodeldesigned onthe basis oftheneurophysiology oftherathippocampus tocontrol thenavigation ofa realrobot. Themodelallows therobotto learnrewardlocations dynamically moved in different environments, tobuildatopological map,andtoreturnhome autonomously. We describe robotexperimentation results from ourtests ina T-maze, an8-armradial mazeandanextended maze. IndexTerms-Affordances, mapping, pathintegration, rat hippocampus, reinforcement learning.