Simulation and Inference in Electrical Power Networks

Event Sponsor: 
Mathematics and Computer Science Division Seminar
Start Date: 
Jan 24 2017 - 10:00am
Building/Room: 
Building 240/Room 4301
Location: 
Argonne National Laboratory
Speaker(s): 
Daniel Maldonado
Speaker(s) Title: 
Postdoc Interviewee, MCS
Host: 
Emil Constantinescu

The increasing penetration of Distributed Generation and the introduction of policies to allow customer participation in the electricity market is challenging the way that Electric Power Networks are operated. Whereas a good amount of research has been produced in order to address the economic operation of the power system and the demand-supply problem, more work is required to develop frameworks that permit power systems operators to keep performance objectives of quality, reliability and stability.
 
We are also seeing interest for the modernization of the grid, particularly regarding communications, supervisory control and data acquisition (SCADA) systems. Harnessing high density measurements, high speed communications and new computational capacity can provide a solution to manage the increasing complexity in the electrical power system.
 
In this talk I will discuss some experiences using PETSc as a simulation platform for Electrical Power Networks. Additionally, I will talk about PETSc's new class for expressing unstructured network problems: DMNetwork. I will talk about my experience using DMNetwork to create an application to simulate large-scale water distribution networks.
 
Finally, I will discuss the importance of load modelling and how statistical methods can improve our understanding of the grid dynamics. In particular, I will present recent work regarding the application of Monte-Carlo techniques to state and parameter estimation of induction motor load using high density measurements.