Cooperative Manipulation of Polygonal Objects with Nonholonomic Wheeled Mobile Robots

Document Type

Student Presentation

Presentation Date



College of Engineering


Department of Mechanical & Biomedical Engineering

Faculty Sponsor

Dr. Aykut C. Satici


This work presents a framework in which a group of nonholonomic wheeled robotic manipulators are cooperatively utilized to manipulate a polygonal object in the plane SE(2). In this framework, the robots are assumed to contact the object without friction, applying forces only in the normal direction. The desired object pose is achieved by finding a controller through online optimization, also called model predictive control. Since the system is highly nonlinear for online implementation, we linearize the equations of motion at the current state and solve the model predictive control problem for the linearized system. We then apply the resulting controller to the original nonlinear system. The optimization problem is set up so that the solution respects the contact constraints and is such that contact between the robots and the object ensures force-closure throughout the manipulation. We evaluate the method through a simulation of four circular wheeled robotic manipulators in contact with a hexagonal object.

This document is currently not available here.