Home /Research /2A1-D10 A Neural Network Controller Designed for a Rough Terrain Mobile Robot : 2nd Report: Effects on the controller property caused by the process of teaching data generation
LEARNING

2A1-D10 A Neural Network Controller Designed for a Rough Terrain Mobile Robot : 2nd Report: Effects on the controller property caused by the process of teaching data generation

Masanori Sato, Atushi Kanda, Kazuo Ishii

Year
2007
Citations
4
Access
Open access

Abstract

In the previous researches, a 6-wheeled mobile robot employing the linkage mechanism, "Zaurus", has been developed to extend maneuverability of wheel type mobile robots. Neural network and PID controllers are introduced as the control system for Zaurus, and their performance are compared. In this research, neural network and PID controllers are redesigned, because the experimental results using Zaurus are different from the results of simulation. The results of the simulation and experiments using Zaurus have improved by introducing the models of motor driver.

Keywords

TerrainMobile robotArtificial neural networkControl engineeringController (irrigation)PID controllerProperty (philosophy)RobotProcess (computing)Computer science

Related papers

Browse all LEARNING papers