Cable driven parallel robots (CDPRs) have been widely used in precision Industries. However, an error occurs in repetitive motion of the CDPR because of the elongation and recovery of cable. Also, it is difficult to apply the numerical cable model because of its complexity. In this paper, we use the frequency based H-RNN method for accurate kinematics. H-RNN predicted more accurately compared with LSTM and RNN algorithms.
Volume Subject Area:Advanced Simulation in Science and Engineering
Topics:Artificial neural networks, Cables, Calibration, Robots, Algorithms, Elongation, Errors, Kinematics
This content is only available via PDF.
Copyright © 2018 by ASME