CURRENT RESEARCH AREAS
1.Research on Very Small-Scale Servo Systems
Sensing and actuation
devices on length scales of millimeter and smaller will be considered
in large arrays, and communication and hierarchical control issues will be
treated. Nonlinear problems in MEMS devices such as snap-through instabilities,
hysteresis, thermal conditioning etc. will be attacked from the viewpoint
of local control.
2. Modeling for Sensing and Control: Diverse fundamental physical
models of magnetoelastic and piezoelastic materials will be treated from
rigorous, modern, mathematical viewpoints. Methods for reduction of nonlinear
partial differential equations into low order ordinary differential equations
capturing the essential physics will be developed. Methods for stability analysis,
and feedback stabilization will be developed. Problems involving frequency
dependent hysteresis will be attacked from rigorous
coupled with experimental characterization. The tools created here will
be useful in the design of millimeter scale and larger scale actuators and
3. Systems of Embedded Micro-actuators for the Control of Flow over
Airfoils and in Arrays of Microvalves
Models of fluid flow over lifting
surfaces will be used to capture sufficient detail to control flow
separation. Hybrid control (combining gross pitch motions with coordinated
motions of microactuators) strategies will be developed. A modular ducted
flow experiment will serve as a testbed. The needed MEMS sensor/actuator
arrays will be fabricated using the Smart-MUMPS process at MCNC. This
will enable integration of control electronics with actuators.
4. Issues in the Control of Fluids on Small Length Scales
of fluid transport based on (i) direct boundary actuation, (ii) thermal
effects exploiting bubbles, (iii) electro-osmotic flows, (iv) streaming
based on rectification of oscillatory flows, (v) electrocapillarity, and
(vi) electro/magneto-rheological effects will be investigated. Fundamental
mathematical analysis will be carried out.
5. The Communications Theory of Very Large-Scale Device Networks
Quantized actuator concepts needing low local bandwidth will be investigated
in the formation of effective sensor-actuator networks. Tradeoffs between
network bandwidth and control performance will be analyzed.
6. Numerical Methods and CAD
Numerical integrators for geometrically
nonlinear models will be developed (especially for magnetostrictive materials).
Tools for systematic model reduction will be developed and implemented in
software based on scripting languages. Field visualization algorithms and
tools will be developed, to support actuator design and optimization.
Send comments, etc. to email@example.com.