HSF: The iOpt’s Framework to Easily Design Metaheuristic Methods

Abstract

The Heuristic Search Framework (HSF) is aJava object-oriented framework allowing to easily implement single solution algorithms such as Local Search, population-based algorithms such as Genetic Algorithms, and hybrid methods being a combination of the two. The main idea in HSF is to break down any of these heuristic algorithms into a plurality of constituent parts. Thereafter, a user can use this library of parts to build existing or new algorithms. The main motivation behind HSF is to provide a “well-designed” framework dedicated to heuristic methods in order to offer representation of existing methods and to retain flexibility to build new ones. In addition, the use of the infra-structure of the framework avoid the need to re-implement parts that have already been incorporated in HSF and reduces the code necessary to extend existing components.