Dogfight Search (DoS)

This paper proposes a metaphor-free metaheuristic algorithm called Dogfight Search for global optimization problems.

Dogfight Search (DoS) is a non-metaphorical meta-heuristic algorithm inspired by dogfighting behavior. Drawing inspiration from dogfighting dynamics, DoS features a structural framework distinct from traditional algorithms. Its search mechanism is designed based on the requirements of optimization problems, incorporating kinematic equations. In the original paper, DoS was compared against 7 advanced competitors and three SOTA algorithms on CEC2017 and CEC2022 benchmark test functions. Results demonstrate that DoS achieves significantly superior optimization performance compared to both advanced competitors and the evaluated SOTA algorithms. Furthermore, the original paper presents experimental results comparing DoS with advanced competitors on 10 constrained engineering optimization problems (maximum dimension of 30, including 91 inequality constraints) and mountainous terrain path planning problems. All experiments consistently demonstrate that DoS outperforms advanced competitors in both constraint handling capability and optimization accuracy.