Date of Award

Spring 2026

Degree Type

Honors College Thesis

Academic Program

Ocean Engineering BS

Department

Ocean Science and Technology

First Advisor

Dr. Gero Nootz

Advisor Department

Ocean Science and Technology

Abstract

Proportional integral derivative (PID) controllers are used for precise position and orientation control in systems such as autonomous underwater vehicles (AUVs). This project supports the University of Southern Mississippi’s (USM) Robotics Club’s RoboSub AUV effort by developing, troubleshooting, and manually tuning PID controllers to characterize tracking performance and settling time across systems of increasing complexity. Initially, the project hypothesized that tracking performance would be reduced and settling times would increase as system complexity advanced from one degree-of-freedom (DOF) to two DOF. However, prior research was found that suggests that for small disturbances around an equilibrium state, separate PID-controlled DOFs can follow the assumption of being independent linear control loops. This would mean that each PID controller could be analyzed and tuned as if it were operating by itself, without the need to account for interactions from other controllers. This notion raised the possibility that the initial hypothesis might not hold true if the 2-DOF system continues to behave linearly and controllers do not couple during testing.

A 1-DOF PID controller for depth and a 2-DOF PID controller for depth and roll underwent manual tuning and were evaluated through settling time and tracking tests. Settling time for the 1-DOF PID controlled system averaged 0.68s and 0.79s for downward and upward jumps respectively, while the 2-DOF PID controlled system averaged 1.32s and 1.52s. Tracking performance revealed that both systems performed very similarly, implying that the depth and roll controllers for the 2-DOF system remained decoupled and behaved in an approximately linear manner. In spite of this, the settling time performance showed that some coupling is present between the two controllers as inputs led to a nonlinear response not captured by the decoupled assumption. Manual tuning produced stable and reproducible results, but as system complexity increases toward a fully free-floating 6-DOF AUV, the validity of assuming that controllers are linear and decoupled will need to be retested for added DOFs before more reliable tuning can be implemented.

Download

Share

COinS