New Ant Colony Algorithm for Continuous Function Optimization in 2D and 3D Search Spaces

Y.H.F. Jbara (Jordan)


Ant Algorithms, 3D Spaces, Continuous


This paper proposes a novel ant-based approach for solving continuous function optimization in 2D and 3D search spaces. The proposed approach called 2D-3D Continuous Ant Colony Approach (2D-3D-CACA), and is able to handle continuous 3D real word spaces applications. Novel concepts of this algorithm that distinguish it from the other heuristics are: (1) the inclusion of a dynamic discretization representation in order to change from discrete nature to continuous one. The paper briefly proposes the benefits and necessity of applying the dynamic discretization representation to resolve 3D application representation issues. (2) The use of a novel simulated annealing-based local search approach for local optimization with the aim of improving the overall performance. The proposed approach consists of two phases, that is, the global phase, and the local phase. The global phase provides a high quality starting solutions while the local phase operates on these quality solutions to obtain higher quality solutions. By iterating these two phases, global optimum obtained. Extensive experimental results show that the proposed algorithm performs very well against other continuous ant-related approaches found in the literature on a set of four well-known benchmark algebraic problems and one 3D real world application.

Important Links:

Go Back