A Novel Cooperative Hybrid Metaheuristic Optimization Method Based on Collective Intelligence

Abstract

This paper studies a cooperative hybrid
metaheuristic that combines the Dragonfly Algorithm
(DA), the Firefly Algorithm (FA) and Cuckoo Search (CS)
in a multi-population setting with fuzzy-based parameter
adaptation. Each algorithm keeps its own update rules,
while a shared global best solution allows information
exchange among subpopulations. Type-1, Interval Type
2 and General Type-2 fuzzy controllers adjust key
parameters during the run so that exploration and
exploitation are modified according to the current
progress of the search. The hybrid is tested on a set of
standard continuous benchmark functions. The
cooperative model without fuzzy logic already improves
over the individual algorithms in most cases, and the
fuzzy variants further reduce the mean error and make
convergence more stable. In particular, the General
Type-2 configuration is often the best performer. These
results indicate that combining complementary swarm
behaviors with fuzzy uncertainty handling is a practical
way to implement dynamic parameter control in
hybrid metaheuristics.