Optimization

The idea of Evolutionary Optimization Algorithms (EA) is inspired by the evolution processes that take place in nature, where sucessfull species must adapt continuously to the changing enviroment by creating new individuals and by selectio of the fittest. This natural optimization technique can easily be transfered to mathematical optimization problems. Theobjective function F on a search space S can be interpreted as the environment and each vector v is a possible individuum in this environment. The fitness of these mathematical individuals is given by the objective function F(v). The EA searches for the optimal solution, by starting with an initial set of vectors called "population". In each following iteration (generation) a new set of vectors is created from the old population by modifying the fittest of the old vectors. For the new population of the next generation the fittest among the newly created and old vectors are selected.

I started working with Evolutionary Algorithms during the diploma thesis for my stdy of buissnes and administration (Betriebswirtschaftslehre) at the comprehensive Unniversity in Hagen. Here I developed an object oriented C++-libary for EAs.

Computer simulation is an important tool in modern product development. New prototypes of products are designed, improved and tested using computers before a first real prototype is build. In this process different geometries, materials and process parameters are changed by hand until a sufficient solution was found. DesParO provides the possibility to optimise these design parameters using the computer. DesParO is an optimisation toolbox especially design for industrial simulation. The toolbox contains a collection of efficient algorthims for the computer based optimisation and can easily be adopted to any industrial simulation code. Consulting by the GMD institute SCAI ensures an efficient use of the toolbox. DesParO was especially designed for computational expensive simulation codes and supports parallel computation.