NFDDE: A novelty-hybrid-fitness driving differential evolution algorithm

• DOI: 10.1016/j.ins.2021.07.082
• Publication Date: 2021-07-31T03:07:05Z
• Authors: Xuewen Xia; Lei Tong; Yinglong Zhang; Xing Xu; Honghe Yang; Ling Gui; Yuanxiang Li; Kangshun Li
• Keywords: Differential Evolution; Test suite

In differential evolution algorithm (DE), it is a widely accepted method that selecting individuals with higher fitness to generate mutant vector. this case, the population under fitness-based driving force. Although force beneficial for exploitation, sacrifices performance on exploration. paper, novelty-hybrid-fitness introduced trade off contradictions between exploration and exploitation of DE. new proposed DE, named as NFDDE, both novelty values are considered when choosing create vectors. addition, two adaptive scaling factors adjust weights novelty-based force, respectively, then distinct properties forces can be effectively utilized. At last, save computational resources, some lower deleted has converged certain extent. The comprehensive NFDDE extensively evaluated by comparisons other 9 state-of-art DE variants based CEC2017 test suite. newly strategies involved parameters further confirmed set experiments.