Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Nonlinear Global Optimization Using Curdling Algorithm

Software ·
OSTI ID:1230340
An algorithm for performing curdling optimization which is a derivative-free, grid-refinement approach to nonlinear optimization was developed and implemented in software. This approach overcomes a number of deficiencies in existing approaches. Most notably, it finds extremal regions rather than only single external extremal points. The program is interactive and collects information on control parameters and constraints using menus. For up to four dimensions, function convergence is displayed graphically. Because the algorithm does not compute derivatives, gradients or vectors, it is numerically stable. It can find all the roots of a polynomial in one pass. It is an inherently parallel algorithm. Constraints are handled as being initially fuzzy, but become tighter with each iteration.
Short Name / Acronym:
OPTIMIZE; 001077MLTPL00
Version:
00
Programming Language(s):
Medium: X; OS: UNIX
Research Organization:
Argonne National Laboratory (ANL), Argonne, IL (United States)
Sponsoring Organization:
DOE/ER
DOE Contract Number:
P
OSTI ID:
1230340
Country of Origin:
United States

Similar Records

Nonlinear Global Optimization Using Curdling Algorithm in Mathematica Environmet
Software · Tue Aug 05 00:00:00 EDT 1997 · OSTI ID:1230398

New advances in the GOP algorithm for global optimization
Conference · Fri Dec 30 23:00:00 EST 1994 · OSTI ID:36016

An adaptive, multivariate partitioning algorithm for global optimization of nonconvex programs
Journal Article · Sun Jan 13 19:00:00 EST 2019 · Journal of Global Optimization · OSTI ID:1514953

Related Subjects