Date of Award

11-2013

Document Type

Thesis - Open Access

Degree Name

Master of Science in Mechanical Engineering

Department

Mechanical Engineering

Committee Chair

Hever Y. Moncayo, Ph.D.

First Committee Member

Richard Prazenica, Ph.D.

Second Committee Member

Yan Tang, Ph.D.

Abstract

In practical applications, an Unmanned Aerial System's (UAS) baseline performance is dictated by how well it can follow a given trajectory with limited stress on the actuators. However, these can be insufficient performance metrics when the UAS is allowed to adapt to an unpredicted external influence such as turbulence or actuation failure, while maintaining a satisfactory baseline performance.

In this thesis, different control laws based on the formation flight geometry problem, nonlinear dynamic inversion and an artificial immune system adaptive mechanism , are implemented in hardware-in-the-loop as a precursor for in-flight testing. These controllers are compared based on three performance metrics: trajectory following, control activity and computer task execution time. The controllers chosen for comparison are: Basic Proportional-Integral-Derivative (PID), Outer loop Non-Linear Dynamic Inversion (NLDI), Extended NLDI, and the previous three controllers augmented with an AIS for a total of six controllers. The Extended NLDI augmented with the AIS outperformed all of the other algorithms under falure conditions on a global scale.

Share

COinS