Chaos embedded opposition based learning for gravitational search algorithm

Due to its robust search mechanism, Gravitational search algorithm (GSA) has achieved a lot of popularity in different research communities. However, stagnation reduces its searchability towards global optima for rigid and complex multi-modal problems. This paper proposes a GSA variant that incorporates chaos-embedded opposition-based learning into the basic GSA for the stagnation-free search. Additionally, a sine-cosine based chaotic gravitational constant is introduced to balance the trade-off between exploration and exploitation capabilities more effectively. The proposed variant is tested over 23 classical benchmark problems, 15 test problems of CEC 2015 test suite, and 15 test problems of CEC 2014 test suite. Different graphical, as well as empirical analyses, reveal the superiority of the proposed algorithm over conventional meta-heuristics and most recent GSA variants.

Automated summary

This paper proposes an improved gravitational search algorithm that incorporates chaos-embedded opposition-based learning and a sine-cosine based chaotic gravitational constant for stagnation-free search of global optima. Experimental results demonstrate the superiority of the proposed algorithm over conventional meta-heuristics and recent GSA variants in various benchmark and test problems.