-
- Research is being conducted in the design of efficient coding
and modulation schemes for reliable high-speed transmission of data
over band-limited channels and for high-density packing of digital
information on magnetic storage media.
Faculty:
Costello
-
-
The modern theory of robust feedback systems
can be based upon the mathematical theory of
H-infinity equilibria, upon the methods of
stochastic control, or upon the approaches of
dynamic, noncooperative games. Research in
this area at Notre Dame has placed a focus upon
the second viewpoint, specifically Nonlinear
Risk-Sensitive Control based upon the cumulants
of traditional cost functions. In the area of
robust stability, work upon the use of reliability
methods to influence the probability of instability
is in progress.
Faculty:
Sain, Spencer
-
-
The protection of the civil infrastructure from the
hazards of earthquakes, winds, and seas---and perhaps
from unfriendly activities as well---depends upon the
use of the latest technological means of actuation and
sensing, together with the most appropriate nonlinear,
reliable, and robust control algorithms. The work at
Notre Dame is a joint venture of the Departments of
Civil Engineering and Geological Sciences and of Electrical
Engineering, which operate the
Structural Dynamics and
Control/Earthquake Engineering Laboratory.
Current
emphasis is upon nonlinear modelling and control using
actuators which have long shelf-life, require small power
and energy, and do not insert energy into the structure
during operation. Such actuators will operate successfully
even when the earthquake or other excitation damages external
power sources.
Faculty:
Spencer, Sain
-
-
Investigations in system theory included the application of
tensor algebra to nonlinear optimal and robust control, the
study of possibilities for defining and designing the transient
response of multi-input, multi-output systems, the algebraic
theory of zeros for multi-input, multi-output, systems, and the
modelling and control issues associated with various forms of
hysteresis in circuits and systems. Applications have included
gas turbine engine control, pressure modelling in internal
conbustion engines, protection of buildings from earthquakes,
and hybrid models for autonomous control systems.
Faculty:
Sain
-
-
The present project addresses the modeling of very general
dynamical systems which are suitable in the qualitative analysis of such
systems, with an emphasis on Lyapunov stability. The models of the
dynamical systems are sufficiently general to accommodate hybrid dynamical
systems and discrete event systems. Examples of specific classes of
problems that have been considered thus far include switching systems,
manufacturing systems with priority batch processing, computer network load
balancing, and the like.
Faculty:
Michel
-
-
The present project addresses the effects of parameter
perturbations on the stability properties and on the location of equilibria
in dynamical systems with applications to power systems and recurrent
neural networks. In the case of the robust control of power systems, the
motivation for this work is parameter sensitivity while in the case of
artificial neural networks, the motivation of this work is the existence or
nonexistence of memories in neural nets subjected to parameter
perturbations, the invariance of the stability properties of memories in
neural nets subjected to parameter perturbations, and the accuracy of the
stable memories of neural nets subjected to parameter perturbations.
Faculty:
Michel
-
-
The present project investigates the effects of time delays on the
qualitative behavior of a class of systems described by ordinary
differential equations (resp., a class of delay equations). The motivation
of the study is to investigate the effects of time delays incurred during
the implementation process of integrated circuits and artificial neural
networks.
Faculty:
Michel
-
-
The objective of the present work is to study the qualitative
effects of the implementation process in digital feedback control systems.
Specific issues addressed include quantization effects, overflow
nonlinearity effects, types of arithmetic used in a controller, and the
like.
Faculty:
Michel
-
- Intelligent control refers to a large set of
control methods used to control highly complex
dynamical systems. Current research projects include
robust control, fault detection (FDI), neural network
controllers, gain-scheduling, and multiple-model switching
control.
Faculty:
Antsaklis,
Lemmon
-
- Hybrid dynamical systems (HDS) are systems containing
both symbolic and continuous variables; a typical example is that of
a continuous system supervised by a digital computer.
Hybrid systems are found in manufacturing, process control, and
communication networks. Models for hybrid systems are derived by
combining models of discrete event systems (DES) with those of continuous-state
systems. Current research projects are developing
a formal systems theory for HDS, developing
computationally efficient methods for controller synthesis,
and studying the extraction of DES models for
the continuous state systems.
Faculty:
Antsaklis,
Lemmon
Discrete Event Systems
- Discrete event systems (DES) are dynamical systems which evolve over
an alphabet of symbols. These systems are very useful in
the high-level supervision of complex dynamical processes
found in flexible manufacturing, process control, and data
communication networks. Current research projects
include investigating the use of structural invariants in the
synthesis of Petri net controllers, and the use of adaptive logical
DES control.
Faculty:
Antsaklis,
Lemmon
-
-
The effects of QoS in networks on feedback control applications is
investigated for local as well as heterogeneous Wide Area Networks. In
particular congestion control mechanisms and their impact on networked
feedback
control systems is of prime interest. Other applications of interest are
in the area of distributed sensor/actuator networks, resource management,
end to end network modeling and stability of networked control systems.
Faculty:
Bauer
-
- Detection and estimation schemes with improved performance
over conventional ones are developed and investigated for
one- and multi-dimensional signals.
Applications of these schemes to areas such as array signal processing
(radar, sonar), acoustic signal separation, medical signal processing,
and image processing are being explored.
In the context of system identification and adaptive
signal processing, closely related research investigates the
topics of recursive parameter estimation using membership-set technique,
blind identification methods, eigen-decomposition methods, etc.
Faculty:
Huang
Liu
-
-
Research focuses on fundamental aspects of m-D digital filters and
systems. Topics such as stability, response behavior, the effects
of various types of singularities, distributed m-D processes,
synchronization of m-D processes on a computer network, continuous to
discrete
system transformations and properties of nonlinear m-D systems
are investigated.
Faculty:
Bauer
-
- A variety of problems associated with digital processing
of two- and three-dimensional visual data are under study.
Emphasis is currently on applications in data
compression for image sequences, real-time video data processing,
tomographic image reconstruction from sparse data and image restoration.
The design and analysis of multidimensional linear and nonlinear filters,
artificial neural networks, varying frame rate allocation techniques,
and multimensional detection and estimation all
contribute to the goal of matching application requirements with
realistic capacities of image transmission and processing systems.
Faculty:
Huang,
Sauer,
Stevenson
-
-
Research focuses on the development of statistical
image models for use in a wide variety of imaging
related inverse problems.
These problems include tomography, image and video
enhancement, inverse halftoning, post-processing of
compressed image/video data amoung others.
Interest has also focused on the computational issues
in the use of these models in real applications.
Faculty:
Sauer,
Stevenson
-
- Several theoretical projects dealing with the inbestigation
of quantum transport and hot electron effects in nanometer
semiconductor devices are under way.
They deal with such basic phenomena as the Aharonov-Bohm effect,
localization and universal conductance fluctuations in nanostructures,
transport through laterally confined structures, quantum interference
devices, resonant tunneling, and also include the more applied aspects
of hot electron engineeringt and wavefunction engineering in quantum
well structures.
Some theoretical projects deal with the investigation of the optical
properties of quantum confined systems.
Faculty:
Bernstein,
Lent,
Merz,
Porod,
Snider
-
- Several projects dealing with nanofabrication using
electron-beam-lithography are currently under way.
They also include investigation of the ultimate resolution
limits of e-beam resists, effects of electron backscattering
in pattern definition, and associated problems.
The nanolithography facility can delineate feature sizes
approaching 100 Angstroms.
The nanostructures fabricated in this facility are tested in our
measurement laboratories at cryogenic and room
temperatures within an excellent noise-isolated environment.
Current projects include the study of quantum interference devices
and phenomena, spin effects in InGaAs/AlGaAs
and InGaAs/InAlAs heterostructures, MBE-grown HEMTs, single-electron tunneling,
oxide breakdown mechanisms, and the performance of ultrashort-gate MESFETs.
Faculty:
Bernstein,
Merz,
Snider
-
- The influence of hot carrier effects on the
long-term reliability of MOS devices is investigated.
In particular, oxide breakdown
phenomena are studied using Monte Carlo simulation techniques.
Faculty:
Bernstein,
Lent,
Porod
-
- As part of the Cold Weather Transit Technology Program, a
$13-million porgram sponsored by the Department of Transportation,
experimental and theoretical investigations are conducted on the coupling
of radio-frequency energy to ice-coated metal substrates.
The goal is more effective deicing techniques.
The work has resulted in full-scale demonstrations of radio-frequency
and water-jet deicing systems for transit power rails at speeds in excess
of 20 mph.
Faculty:
Berry
-
- This project is studying the velocities of erodent particles
as they rebound from a metal surface during impact erosion.
It includes computer modeling of the impact problem and measurements
of the rebound parameters using a novel velocity measurement technique.
Faculty:
Kosel
-
- Our optoelectronic materials and device research is focused on
understanding and developing new applications of the III-V compound
semiconductor oxidation process discovered by Dallesasse and Holonyak
(University of Illinois, 1990) for optical and optoelectronic device
integration. Presently, native-oxide-based waveguides and integrated
optics components are being designed, fabricated and characterized.
Related studies of the material and optical properties of III-V
semiconductor oxides are also underway. Applications include high-power
master oscillator power amplifier (MOPA) laser devices. New materials and
device structures for incorporating luminescent rare-earth dopants into
compound semiconductors are also under investigation.
Faculty:
Hall
-
- MEMs research is an exciting field where semiconductor processing
technology is used to fabricate mechanical structures. These devices can
be extremely small, and can be mass-produced, which greatly decreases their
cost. At Notre Dame we are working on several projects, including sensors
which will measure the force of an aerosol particle hitting a surface, and
pumps which will move tiny amounts of fluids in a "Chemistry Lab on a
Chip".
Faculty:
Bernstein,
Snider
