Robotic trajectory control employing anyanet neural architecture
Loading...
Authors
Simons, Anya L.
Issue Date
1992-12
Type
Electronic thesis
Thesis
Thesis
Language
en_US
Keywords
Engineering science
Alternative Title
Abstract
Popular and novel neural network architecture and training approaches are combined with conventional trajectory smoothing and obstacle avoidance techniques to demonstrate the feasibility of an industrial robot that can safely, reliably, and effectively interact with humans, avoid unforeseen moving obstacles in a stable fashion, and maintain integrity of its payload. An appreciation is sought for the level of hardware and process integration required for the implementation of a successful comprehensive automated manufacturing package. Extensions of the methods demonstrated are discussed.
Description
School of Engineering
December 1992
December 1992
Full Citation
Publisher
Rensselaer Polytechnic Institute, Troy, NY