Search NASAโŒ• Search

DOE OSTI ยท 1231244

Hybrid Optimization Parallel Search PACKage

Abstract

HOPSPACK is open source software for solving optimization problems without derivatives. Application problems may have a fully nonlinear objective function, bound constraints, and linear and nonlinear constraints. Problem variables may be continuous, integer-valued, or a mixture of both. The software provides a framework that supports any derivative-free type of solver algorithm. Through the framework, solvers request parallel function evaluation, which may use MPI (multiple machines) or multithreading (multiple processors/cores on one machine). The framework provides a Cache and Pending Cache of saved evaluations that reduces execution time and facilitates restarts. Solvers can dynamically create other algorithms to solve subproblems, a useful technique for handling multiple start points and integer-valued variables. HOPSPACK ships with the Generating Set Search (GSS) algorithm, developed at Sandia as part of the APPSPACK open source software project.

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

2009-11-10. Hybrid Optimization Parallel Search PACKage. https://www.osti.gov/biblio/1231244

Cite the original work for its findings. Save a collection to share your selection of sources.