Abstract

Traditional control methods require extensive tuning or a derivation of a system model making them increasingly antiquated for use on new, more complex systems. Sliding Mode Control has emerged as a more effective alternative as a control method that can directly handle nonlinear systems with increased robustness while guaranteeing stability. However, it is still limited by the need for a system model for the derivation of the controller form. This work proposes a new model-free control method based on Sliding Mode Control referred to as Model-Free Sliding Mode Control where the form of the controller is only dependent on system order, state measurements, and previous control inputs. Lyapunov’s stability theorem is used to ensure global asymptotic stability and a boundary layer is incorporated to reduce chattering. The model-free properties of the controller are enabled by a least-squares online parameter estimation method used to estimate the control input influence gain matrix of the system directly. The estimation process is essential as previous work was limited by assuming the bounds of the control input influence gain matrix are known or the assumption it was unitary. The estimation method also incorporated exponential forgetting to only include updated data for parameter estimation, increasing the speed of convergence. Another addition was a bounded gain forgetting factor to ensure that the magnitude of the control input influence gain was upper bounded. The performance of this controller was simulated on various example systems to test its performance. These included single-input, single-output and multi-input, multi-output first and second order systems. Principally, the controller was implemented to control a lateral-directional state space model of an aircraft with a shaped input characterizing aircraft roll and yaw dynamics. The controller proved to exhibit outstanding tracking performance, convergence of estimated parameters, smooth and acceptable control input, and increased robustness to parameter uncertainty. Therefore, the controller was proven to be a feasible method to control the lateral and directional dynamics of an aircraft.

Library of Congress Subject Headings

Sliding mode control; Drone aircraft--Control systems; Lyapunov stability

Publication Date

8-1-2023

Document Type

Thesis

Student Type

Graduate

Degree Name

Mechanical Engineering (MS)

Department, Program, or Center

Mechanical Engineering (KGCOE)

Advisor

Jason Kolodziej

Advisor/Committee Member

Kathleen Lamkin-Kennard

Advisor/Committee Member

Sarilyn Ivancic

Campus

RIT – Main Campus

Plan Codes

MECE-MS

Download

Share

COinS