Home News & Events Research Seminars Research Seminar: How Linear Algebra and Optimization Is Useful for Power Industry – Dr. Riadh Zorgati

Research Seminar: How Linear Algebra and Optimization Is Useful for Power Industry – Dr. Riadh Zorgati

 About the Seminar

In this talk, we illustrate how linear algebra and optimization, with continuous or binary variables, is useful for power industry, for example for improving maintenance of power plants or for energy management. Optimization under constraints is a key approach for energy management where we aim at minimizing costs of production while respecting operating constraints of power plants. After a brief introduction to power systems, we present how optimization problems related to energy management are designed and review some useful optimization schemes like linear programming, with real and binary variables, quadratic programming and price decomposition. We will discuss some examples illustrate this approach.

About the Speaker
Dr. Riadh Zorgati is a research engineer at EDF, Research & Development Division since 1990. He was a project manager and a researcher on signal processing and inverse problems applied to nondestructive testing of nuclear power plants during 8 years. He joined the department OSIRIS (Optimization, Simulations, Risks & Statistics) in 2000 where he was a project manager on weekly electricity management. His current research activity is related to optimization under uncertainty and inverse problems and their application to energy management with an emphasis on probabilistic, robust and conic optimization methods. His research activities have been awarded in 1994 and 2010. Riadh ZORGATI received his PhD degree in Physical Sciences from the Université de Paris-Diderot in 1990. He taught Optimization and its applications to power systems at the Université de Paris-Orsay from 2005 to 2016. He was Associate Director of Research at the French national research center (CNRS), from 2003 to 2008 and carried out research on inverse methods applied to non-destructive evaluation.

Recent Events

[Research Seminar] Foundation of Mixture of Experts in Complex and Massive AI Models
10:00- 11:00 AM, July 26, 2024

[Research Seminar] Foundation of Mixture of Experts in Complex and Massive AI Models

About the Talk: Since the release of the original Transformer model, extensive efforts have been devoted to scaling up the model complexities to take advantage of massive datasets and advanced computing resources. To go beyond simply increasing the network depth and width, Sparse Mixture-of-experts (SMoE) has risen as an appealing solution for scaling Large Language […]

[Research Seminar] Scalable Uncertainty: AI Needs to Know What it Doesn’t Know
2:00 - 4:00 PM, July 15, 2024

[Research Seminar] Scalable Uncertainty: AI Needs to Know What it Doesn’t Know

About the Talk: Large models fuel impressive capabilities.  However, they currently do not know what they do not know, which is critical to the important challenges of exploration and alignment.  Knowing what is not known is essential, for example, to gathering informative data.  In this talk, I will discuss the need for scalable uncertainty and […]