Constrained Optimization Methods In The Design Problems Of Thermal-Hydraulic Systems

Seminarium Zakładu Energetyki Jądrowej i Analiz Środowiska
Prelegent i afiliacja: 
dr inż. Maciej Kulig
wt., 2019-04-02 11:30
CYFRONET (bud. 39), sala 172 (III piętro)

The subject of the presentation is related to the research project conducted
a long time ago at the Institute of Nuclear Research (IBJ) and Atomic Energy (IEA), Świerk, devoted to modeling and optimization of thermal hydraulic systems/installations. Specific features of engineering problems in the design of thermal hydraulic systems will be discussed, illustrated by an example of
a design problem considered within the framework of IEA projects. Mathematical formulation of the related nonlinear optimization problems will be given, difficulties and challenges associated with practical applications will be discussed. Overview of the numerical methods available at that time will be made, and research work on the development/implementation of the most suitable analytical tools will be described, including the results of numerical experiments with the optimization solver developed within the framework of this project.

Significant advances in the field of nonlinear optimization achieved since that time and future prospects will be summarized. Finally, a brief overview will be provided of the recent algorithms considered relevant for the subject of NLP optimization that are available in the public domain.

The author believes that effective and reliable NLP solvers should be one of the relevant elements of the analytical, High Performance Computing (HPC) tools being made available at NCBJ and CIŚ, in particular. Resuming the research/ implementation work in this field at NCBJ/CIŚ would be desirable, consistently with the NCBJ/CIŚ profile and its mission to support a vast variety of analytical services, not only in the area of energy and power industry, but also in many other fields.

PDF icon seminarium_uz3_02-04-2019.pdf595.99 KB
Plik seminarium_uz3_02-04-2019.docx17.37 KB