Electrical and Computer Engineering ETDs

Publication Date

Fall 7-31-2017

Abstract

The purpose of this thesis is to show the development of an aerial testbed based on the Robot Operating System (ROS). Such a testbed provides flexibility to control heterogenous vehicles, since the robots are able to simply communication with each other on the High Level (HL) control side. ROS runs on an embedded computer on-board each quadrotor. This eliminates the need of a Ground Base Station, since the complete HL control runs on-board the Unmanned Aerial Vehicle (UAV).

The architecture of the system is explained throughout the thesis with detailed explanations of the specific hardware and software used for the system. The implementation on two different quadrotor models is documented and shows that even though they have different components, they can be controlled similarly by the framework. The user is able to control every unit of the testbed with position, velocity and/or acceleration data. To show this independency, control architectures are shown and implemented. Extensive tests verify their effectiveness. The flexibility of the proposed aerial testbed is demonstrated by implementing several applications that require high-performance control.

Additionally, a framework for a flying inverted pendulum on a quadrotor using robust hybrid control is presented. The goal is to have a universal controller which is able to swing-up and balance an off-centered pendulum that is attached to the UAV linearly and rotationally. The complete dynamic model is derived and a control strategy is presented. The performance of the controller is demonstrated using realistic simulation studies. The realization in the testbed is documented with modifications that were made to the quadrotor to attach the pendulum. First flight tests are conducted and are presented.

The possibilities of using a ROS based framework is shown at every step. It has many advantages for implementation purposes, especially in a heterogeneous robotic environment with many agents. Real-time data of the robot is provided by ROS topics and can be used at any point in the system. The control architecture has been validated and verified with different practical tests, which also allowed improving the system by tuning the specific control parameters.

Keywords

ROS, quadrotor

Document Type

Thesis

Language

English

Degree Name

Electrical Engineering

Level of Degree

Masters

Department Name

Electrical and Computer Engineering

First Committee Member (Chair)

Rafael Fierro

Second Committee Member

Francesco Sorrentino

Third Committee Member

Svetlana Poroseva

Share

COinS