ROBUST ACTUATOR FORCE ANALYSIS OF A HEAVY-DUTY MANIPULATOR USING GMM/GMR

Miao Li, Xiaoping Zhang, and Wenyu Yang

Keywords

Grasping force optimization, uncertainty, Gaussian mixture model, Gaussian mixture regression

Abstract

Uncertainty is inevitable in modelling of physical world, including sensor noise, control errors and inaccurate models of the environment. In robotic grasping force analysis (GFA), the uncertainties mainly come from the location of contact points and the coefficient of friction. Previous researches of uncertainty analysis are mainly focusing on of robotic grasp synthesis and planning, while little work has been done from the aspect of grasping force. In this paper, a point contact model with uncertainty is proposed to investigate the effect of uncertainty in GFA, which was solved using Monte Carlo simulation. Moreover, to predict the minimal actuator force in an industrial heavy-duty manipulator, a Gaussian mixture model (GMM) is proposed to describe the relationship between the object configuration and the minimal actuator force.

Important Links:

Go Back