Implementing Parallel Metaheuristic Optimization Framework Using Metaprogramming and Design Patterns

In the present paper we introduce an approach to implementing parallel metaheuristic optimization frameworks which is used in the design of the framework HeO. This experimental cross-platform framework is a collection of popular optimization methods implemented in C++ as algorithmic skeletons. The key feature of the discussed approach is the wide usage of metaprogramming and design patterns which allow to increase the reusability of the code and ease the process of hybrid algorithms construction for the end-user. We consider framework structure and implementation details and provide the results of numerical experiments for some well-known optimization problems.