Date of Award

December 2013

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Engineering

First Advisor

Anoop K. Dhingra

Committee Members

Ronald Perez, Ilya Avdeev, Habibollah Tabatabai, Istvan Lauko

Keywords

Active Control, Control, Optimization, Structural, Topology

Abstract

The design and performance of complex engineering systems often depends on several conflicting objectives which, in many cases, cannot be represented as a single measure of performance. This thesis presents a multi-objective formulation for a comprehensive treatment of the structural and topological considerations in the design of actively controlled structures.

The dissertation addresses three main problems. The first problem deals with optimum placement of actuators in actively controlled structures. The purpose of control is to reduce the vibrations when the structure is subjected to a disturbance. In order to mitigate the structural vibrations as quickly as possible, it is necessary to place the actuators at locations such that their ability to control the vibrations is maximized. Since the actuator locations are discrete (0-1) variables, a genetic algorithm based approach is used to solve the resulting optimization problem.

The second problem this dissertation addresses is the multi-objective design of actively controlled structures. Structural weight, controller performance index and energy dissipated by the actuators are considered as the objective functions. It is assumed that a hierarchical structure exist between the actuator placement and structural-control design objective functions with the actuator placement problem considered being more important. The resulting multi-objective optimization problem is solved using Stackelberg game and cooperative game theory approaches. The exchange of information between different levels of the multi-level problem is done by constructing the rational reaction set of follower solution using design of experiments and response surface methods.

The third problem addressed in this dissertation is the optimization of structural topology in the context of structural/control system design. Despite the recognition that an optimization of topology can significantly improve structural performance, most of the work in design of actively controlled structures has been done with structures of a known topology. The combined topology and sizing optimization of actively controlled structures is also considered in this thesis. The approach presented involves the determination of optimum topology followed by a sizing and control system optimization of the optimum topology. Using two numerical examples, it is shown that a simultaneous consideration of topological, control and structural aspects yields solutions that outperform designs when topological considerations are neglected.

Share

COinS