An MPI Implementation of the Fast Messy Genetic Algorithm

The fast messy genetic algorithm fmGA belongs to a class of algorithms inspired by the principles of evolution known appropriately as evolutionary algorithms EAs These techniques operate by applying biologically inspired operators such as recom bination mutation and selection to a population of individuals EAs are frequently applied as optimum seeking techniques by way of analogy to the principle of survival of the ttest In contrast to many EAs the fmGA consists of several evolutionary phases each with distinct characteristics of local global computation These are ex plained in the paper Parallel computational experiments are performed to determine the e ectiveness ef ciency and scaled e ciency of fmGA implementations for several parallel distributed environments The underlying communication libraries are based on the Message Pass ing Interface MPI as well as an NX based Intel Paragon implementation Each imple mentation is evaluated with respect to two selection strategies and two recombination strategies The optimization problem for these experiments is the minimization of the CHARMM energy model of the pentapeptide Met Enkephalin With this application the use of the MPI constructs permit e ective and e cient execution on a variety of platforms