Robot Trajectory Generation in Joint Space
Abstract
In order to generate desired robot trajectory, this chapter proposes a different method from the previous works like inverse kinematics and dynamic time warping. The classic hidden Markov model (HMM) is modified such that it is more feasible to generate trajectory in joint space. We introduce a new auxiliary output in HMM to help the training process. The Lloyd's algorithm is modified for HMM, such that it can solve the problems in joint space learning. The proposed method is validated with a two-link planar robot and a four degree-of-freedom (4-DoF) exoskeleton robot.
Keywords
Joint space; Tarjectory generation; Hidden Markov model
10.1 Codebook and key-points generation ...
Get PID Control with Intelligent Compensation for Exoskeleton Robots now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
