.png){kind=icon}
Return
Stagnation-avoidance framework with hybrid dynamics evolutionary algorithms for constrained and combinatorial optimization problems
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2026.116632.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper proposes a novel four-algorithm hybrid framework called the Red Fire Meerkat Naked (RFMN) algorithm, integrating the Meerkat Optimization Algorithm (MOA), Red Panda Optimizer (RPO), Fire Hawks Optimizer (FHO), and Naked Mole Rat Algorithm (NMRA). The algorithm uses selected operators from the four methods within a fixed four-stage and employs a fractional iteration strategy to alternate between the exploration and exploitation phases. In addition, a stagnation avoidance mechanism and five adaptive mutation operators are used to enhance its performance and robustness. For discrete problems, a binary variant (bRFMN) is also developed using the S-shaped transfer function. An analysis is performed to check the suitability of the RFMN algorithm using classical benchmarks, such as the CEC 2014, CEC 2019, and CEC 2022 benchmarks, six constraints engineering design cases, and 15 CEC 2020 constraint problems. In addition to that, we are using ten feature selection datasets to test the bRFMN algorithm. Results demonstrate that RFMN and bRFMN outperform or remain competitive with leading algorithms such as SHADE, JADE, LSHADE-SPACMA, LX-TLA, SASS, HPSO, EnMODE, multi-strategy chameleon algorithm (MCSA), co-evolutionary PSO (CPSO), quadratic binary harris hawk optimizer (QBHHO), and binary multiverse optimizer (BMVO) across diverse continuous and combinatorial optimization problems.
Keywords:
Multi-hybrid algorithms
CEC benchmark problems
Adaptive properties
Engineering design problems
Feature selection
Journal
K
IF:
7.6
Papers:
1.3W
Citations:
4.5W
Organization
View more organizations
Cited Papers
No cited papers available