Active control of smart structures containing a large number of actuators andsensors presents a tradeoff between increased system performance and the cost and
bulk of the required hardware and computational resources. A technique called multiplexed
control offers advantages when software and hardware resources are scarce
and performance specifications call for a large number of actuators and sensors. In
structural control applications, in particular those using smart materials, it is often
desirable to increase the number of actuators to enhance controllability.
The focus of this research is to demonstrate real-time multiplexing on the hardware
side of the actively controlled structure. Multiplexing effectively reduces the
number of power units by sharing them among a large number of actuators according
to a switching schedule. Multiplexing introduces periodicity in the closed-loop plant,
requiring the use of periodic linear systems theory to tune as optimal quadratic regulator.
In this thesis a multiplexed implementation is developed to control the three
actuators mounted on single smart beam, where the control inputs are updated sequentially
and cyclically instead of simultaneously, thus allowing the use of a single
power amplifier. Focus was placed on the application of multiplexed control theory to
a smart structure, including experimental procedures to obtain a plant model using
system identification tools, and tuning of a discrete-time periodic quadratic regulator.
An observer was also implemented. Simulations and real-time implementation
validate the approach.