An Improved Hybrid Algorithm for Numerical Optimization
Prabir Kumar Jena, D. Chandrasekhar Rao, Pradipta Kumar Das · 2018 International Conference on Inventive Research in Computing Applications (ICIRCA) · 2018
Design of a robust and reliable optimization algorithm to solve various real world optimization problems now becomes a major concern for the research community. In this perspective, this paper presents a hybrid DEEKH algorithm by combining differential evolution (DE) with an enhanced version of krill herd (EKH) in a new conceptual fashion. The motion of KH individuals are mostly influenced by its neighbors. The control parameters in enhanced KH (EKH) are updated dynamically to avoid the krill individuals being trapped at local optima by preserving the basic structure of KH algorithm. The DE algorithm has the potential of better exploration of problem search domain. In DEEKH, the solution is further improved by improving the exploitative nature of EKH through sharing of information during the search process. The potential of the proposed DEEKH is validated using six standard benchmark functions. From simulation study, it is verified that the proposed DEEKH provides superior results and highly viable with the existing state of the art.