Abstract

This paper investigates the benefits of using non-dimension analysis to develop a control law for a flexible electro-mechanical system. The system that is analyzed consists of a DC motor connected to a load inertia through a set of gears. A state space system model is derived using LaGrange's equation and then non-dimensionalized using a linear transformation. The resulting system model reveals the system character more clearly through the resulting dimensionless parameters. The parameters highlight the interaction between system properties and motor constants and demonstrate the benefits of a concurrent mechatronics design process. Open-loop behavior is analyzed and an optimal value for these paramaters can be found by varying the gear ratio. Once the best possible gear ratio is determined, a PID control law is developed and the closed loop performance is analyzed. With the optimal gear ratio, the power required to control the system is minimized. Also, dynamic inversion is applied to control the system. Dynamic inversion requires a square "B" matrix in the state space model. A new method to apply dynamic inversion to a system with a non-square "B" matrix is demonstrated. To make the matrix invertible, a linear transform is applied to the state space model. A Linear-Quadratic Regulator (LQR) design method is applied to find the transformation matrix values that will make the "B" matrix invertible. The power consumption of this control law is also minimized when the system contains the optimal gear ratio.

Library of Congress Subject Headings

Electromechanical devices--Dynamics; Mechatronics; H2 control

Publication Date

2004

Document Type

Thesis

Student Type

Graduate

Degree Name

Mechanical Engineering (MS)

Department, Program, or Center

Mechanical Engineering (KGCOE)

Advisor

Agamemnon Crassidis

Advisor/Committee Member

Mark Kemski

Advisor/Committee Member

Josef Torok

Comments

Physical copy available from RIT's Wallace Library at TK153 .W55 2004

Campus

RIT – Main Campus

Download

Share

COinS