Ganesh Hegde

Development of a Rappture based Generic Optimization Engine

Development of a Rappture based Generic Optimization Engine Ganesh Hegde, Michael McLennan, Gerhard Klimeck, George Howlett

Objective/Problem:

  • To Build a Generic Optimization engine within Rappture Framework.
  • To Enable Automatic optimization for all Rappture based nanoHUB.org tools.

Approach:

  • Use existing software for basic GA framework.
  • Modify Rappture framework to include the optmization API.
  • Use API to make Rappture talk to PGAPack.

Results:

  • Demonstrated sequential optimization of Quantum dot absorption plot.
  • Parallelization enables solving of sequentially intractable problems in orders of magnitude lesser time.

Desired Impact:

  • First generic parallel optimization package with programmable fitness functions.
  • Plug any simulation tool into the Rappture framework in order to solve optimization problems on the nanoHUB.

Powerpoint slide as ppt, pdf, or as image below.