Human Center of Mass Trajectory Models Using Nonlinear Model Predictive Control

Joos, Alexander; Hoffmann, Marian; Stein, Thorsten

doi:10.1016/j.ifacol.2018.11.569

Cited by 2 publications

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, for instance, the actions of a person are defined as a set of probabilities which influences the control, as shown in [14] or [15]. Another option to avoid the writing of the knowledge consists of using a computer to analyze the data or also using a group of people instead of a single expert, for instance to learn how a person looks for a destination in an unknown environment with obstacles using a simulated environment, in [16], or the motion of the body to avoid an obstacle, in [17]. To avoid the writing of the knowledge by an expert we propose to use a Neural Network, but we could use another tool, such as a Support Vector Machine.…”

Section: Introductionmentioning

confidence: 99%

Performance evaluation of a Time Scale Controller

Antolines

2022

DYNA

View full text Add to dashboard Cite

Regardless of the advances in intelligent control the analysis and use of the human capacity to control are far from exhausted. For instance, industrial applications could be too fast or too slow for a human to control. The proposed solution in this paper starts by scaling the model of the system in time, so that it results comfortable to control. The control actions of the human are learned by a Neural Network, which is blind to the changes in the time scale, thus the Neural Network controls the scaled model and the real plant as well. The Time Scaling Controller is evaluated by controlling the angular position of a motor and the result is contrasted with a fuzzy controller and a piecewise linear controller. Time Scaling Control resulted better than the other two controllers because it has the lowest effort and the highest effectiveness among the three controllers.

show abstract