The Bridge - Issue 1, 2023 - 31

Graduate Research SP
Dulip Madurasinghe
Zeta Iota
Clemson University, Ph.D. Student in Electrical Engineering
RESEARCH TOPIC
AI Based Power System Monitoring and Control Applications
My research is focused on applying intelligent algorithms to improve the power system monitoring and control. I have
progressed through my research from the power system branch event identification to substation configuration identification
to proposing a novel power system transmission network topology processing (TNTP) at energy control center (ECC). All three
of these steps are to establish the
foundation to propose novel real-time
applications for power system control
at the ECC. The research has already
produced several publications. I'm
motivated to propose a meaningful
energy management system (EMS)
application based on an artificial
intelligence (AI) algorithm named
real-time intelligent security constraint
dynamic power flow approach.
Left to Right: Distributed Intelligent Branch Event Identification Approach, Substation configuration identification,
Transmission Network Topology Processing based on Substation Configuration Identification and Branch Event Identification
LEARN MORE
https://scholar.google.com/citations?user=5YkcqEoAAAAJ&hl=en&oi=ao
CONTACT
www.linkedin.com/in/dtmadurasinghe/
Jared Miller
Gamma Beta
Northeastern University, Ph.D. Student in Electrical Engineering
RESEARCH TOPIC
Convex Optimization, Nonlinear Systems, Control
Peak estimation is the problem of bounding extreme values of a state function along system
trajectories. Examples of peak estimation include finding the maximum speed of a car, height of an
aircraft, and proportion of infected people in an epidemic model. Peak estimates may be employed to
analyze system behavior, evaluate controller performance, and design policy interventions. These peak
estimation problems are finite-dimensional but nonconvex when optimizing over the initial condition/
time that could generate the peak value along a trajectory. Measure-theoretic and algebraic methods
may be applied to approximate these nonconvex problems by a sequence of convergent (under mild
conditions) convex optimization problems in increasing complexity and accuracy.
Jared Miller has extended these measure-based peak estimation methods to extract near-optimal
trajectories, to incorporate uncertainty in dynamics, and to quantify the safety of trajectories by
bounding the distance of the closest approach to an unsafe set. Further research in this topic includes
peak estimation and optimal control for systems with time delays (e.g., the incubation periods). Other
work he has performed includes the data-driven stabilization and control of systems with input and
measurement noise and the exploitation of sparse structure to improve the approximation quality of
large-scale semidefinite programs. He is a member of the Robust Systems Laboratory at Northeastern
University and was in part supported by a Chateaubriand Fellowship.
LEARN MORE
https://jarmill.github.io/
TLIGHT
No trajectory (cyan) starting
from an initial set (blackcircle)
following dynamics
with uncertainty reaches a
defined distance (red contour)
surrounding the unsafe set (red
half-circle).
CONTACT
www.linkedin.com/in/jared-f-miller
HKN.ORG
31

https://scholar.google.com/citations?user=5YkcqEoAAAAJ&hl=en&oi=ao https://www.linkedin.com/in/dtmadurasinghe/ https://jarmill.github.io/ http://www.linkedin.com/in/jared-f-miller https://hkn.ieee.org/

The Bridge - Issue 1, 2023

Table of Contents for the Digital Edition of The Bridge - Issue 1, 2023