Back to search

Clemson University

Deep Reinforcement Learning of Variable Impedance Control for Object-Picking Tasks

Abstract

dc:description.abstract

The increasing deployment of robots in industries with varying tasks has accelerated the development of various control frameworks, enabling robots to replace humans in repetitive, exhaustive, and hazardous jobs. One critical aspect is the robots' interaction with their environment, particularly in unknown object-picking tasks, which involve intricate object weight estimations and calculations when lifting objects. In this study, a unique control framework is proposed to modulate the force exerted by a manipulator for lifting an unknown object, eliminating the need for feedback from a force/torque sensor. The framework utilizes a variable impedance controller to generate the required force, and an admittance controller models the robot's motion as a mass-spring-damper system. The combined framework mimics a human hand guiding a robot arm, where the force generated by the variable impedance controller pulls the robot to the desired position. The distance to the desired position, stiffness, and damping parameters influence the variable impedance force generated. The stiffness and damping parameters are uniquely tailored for specific object masses and require learning. Here, deep reinforcement learning is employed to learn the stiffness parameter, enabling the framework to lift objects of unknown mass effectively. The effectiveness of the proposed control framework is demonstrated through training and testing in the ROS Gazebo simulator, employing a UR5 manipulator. The trained model exhibits the ability to lift objects with unknown masses to predetermined positions, showcasing the framework's practical applicability and potential in diverse industrial settings.

Degree

thesis:*
Name thesis:degree_name
Master of Science (MS)
Level thesis:degree_level
Thesis
Discipline thesis:degree_discipline
Mechanical Engineering

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Lunia, Akshit
Contributors dc:contributor
  • Yue Wang
  • Phanindra Tallapragada
  • John Wagner

Subjects

dc:subject × 8

Identifiers

dc:identifier.*
Repository record dc:identifier
https://open.clemson.edu/all_theses/4257

Chain of custody

source
Harvested from
Clemson University
Base URL
open.clemson.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Lunia, Akshit. Deep Reinforcement Learning of Variable Impedance Control for Object-Picking Tasks. Thesis thesis, https://open.clemson.edu/all_theses/4257