An Improvement Of Hybrid Whale Optimization Algorithm


Creative Commons License

Danacı M. , Alızada B.

EUROASIA JOURNAL OF MATHEMATICS-ENGINEERING NATURAL & MEDICAL SCIENCES, cilt.2, ss.60-68, 2019 (Diğer Kurumların Hakemli Dergileri)

  • Cilt numarası: 2 Konu: 7
  • Basım Tarihi: 2019
  • Dergi Adı: EUROASIA JOURNAL OF MATHEMATICS-ENGINEERING NATURAL & MEDICAL SCIENCES
  • Sayfa Sayıları: ss.60-68

Özet

The difficulty in solving engineering problems creates difficulties in the selection of the methods to be

used. Nature-inspired herd intelligence-based meta-heuristic optimization techniques have recently

become the most popular algorithms for solving such problems. In this work, a new hybrid algorithm

model has been developed to adapt to various problems. The developed models were adapted to 23

Benchmark test problems in the literature and compared with meta-heuristic algorithms. The

algorithms aim to balance the optimization processes of exploration and exploitation. In the

development of a meta-heuristic algorithm, it is very difficult to achieve a balance due to its stochastic

structure. In this study, the new hybrid model improved by Multi-Verse Optimization (MVO) on the

Sine Cosine Whale Optimization Algorithm (SCWOA) hybrid model, which is available in the

literature, has increased the success of test problems. Although the SCWOA hybrid balances

exploitation and exploration, the MVSCWOA (Multi-Verse Sine Cosine Whale Optimization

Algorithm) hybrid algorithm, which was modified by modifying MVO's wormhole existence

probability (WEP) and traveling distance rate (TDR), has succeeded in improving this balance further.

WEP is used instead of 𝒓𝟏 parameter, which determines the update direction in SCA, and TDR is used

in place of 𝒂𝟐 (varies between -1 and -2) used in the update of l, which is the inter-element

multiplication parameter in WOA. The results obtained from the newly developed hybrid model have

shown that it makes the search and exploitation feature more effective by showing better results than

SCWOA, WOA, SCA, and MVO. MVSCWOA was successful in test problems.

Keywords: Benchmark, Sine Cosine Algorithm, Whale Optimization Algorithm, Multi-Verse

Optimization