A Novel Human-Based Meta-Heuristic Algorithm: Dragon Boat Optimization

Xiang Li, Long Lan, Husam Lahza, Shaowu Yang, Shuihua Wang‎, Wenjing Yang, Hengzhu Liu, Yudong ZhangarXiv 2023

Abstract

Inspired by the traditional team sport of Dragon Boat Racing, this paper proposes a novel human-based meta-heuristic algorithm called Dragon Boat Optimization (DBO). The algorithm models the unique behaviors and interactions of crew members by introducing social psychology mechanisms such as social loafing and social incentive. By adjusting state updating strategies based on crew behavior and environmental factors, DBO maintains high performance efficiency. Experimental results across 29 mathematical optimization problems and two structural design problems demonstrate that DBO is highly competitive with state-of-the-art meta-heuristic algorithms and conventional methods.

Introduction

In the field of meta-heuristic optimization, researchers continuously seek nature-inspired strategies to solve complex, high-dimensional problems. While biological models like ant colonies or particle swarms are common, Dragon Boat Optimization (DBO) draws from a 2,000-year-old human tradition. DBO shifts the focus from simple biological instincts to complex social psychology and team coordination, offering a fresh perspective on collective intelligence.

Problem Statement

Modern engineering often involves "intricate problems" characterized by non-linear, high-dimensional landscapes riddled with local optima. Traditional swarm-based models often treat agents as identical units, which can limit their ability to model the specialized collaboration found in human teams. DBO addresses this gap by assigning distinct roles and modeling psychological states that dictate the search for a global optimum.

Methodology

DBO translates the mechanics of a boat race into four primary mathematical search strategies:

Role Differentiation

Unlike homogeneous swarms, DBO assigns specific roles to its agents to simulate a real crew. The Drummer regulates the global pace and acceleration, the Paddlers serve as the primary search agents providing propulsion, and the Steersman maintains the overall course.

The specialized roles within a dragon boat crew: Drummer, Paddlers, and Steersman.

Social Psychology and Kinematics

The core of DBO lies in its Social Behavior Factor ($\psi$), which incorporates:

  • Social Loafing: Agents may exert less effort in unconstrained search spaces where the demand for precision is lower.
  • Social Incentive: In difficult or constrained spaces, increased "pressure" motivates agents to refine their search.

Kinematics modeling further enhances the search through an Acceleration Factor and an Attenuation Factor, the latter of which models the natural decay of physical performance over time. Additionally, an Imbalance Rate accounts for external noise, such as water surface fluctuations.

Modeling the optimal paddle entry angle $\theta_B$ to maximize propulsion and maintain team synchronization.

Experimental Evaluation

DBO was rigorously tested against 29 benchmark functions, including unimodal functions for convergence speed, multimodal functions for global search capability, and hybrid composition functions for complex landscapes. It was also applied to real-world structural design tasks: tension/compression spring design and pressure vessel design.

Results and Discussion

Experimental results indicate that DBO often outperforms established algorithms such as the Whale Optimization Algorithm (WOA) and Grey Wolf Optimizer (GWO).

  • Convergence: DBO showed superior stability and convergence, reaching absolute minima on several complex multimodal functions where others failed.
  • Engineering Application: In the pressure vessel design problem, DBO achieved an optimum cost of 5977.2759, surpassing nearly 15 other established algorithms including PSO and GA.
  • Robustness: DBO exhibited lower standard deviation across multiple runs, suggesting high reliability and less sensitivity to initial conditions.

Conclusion

By integrating human social psychology into a mathematical framework, Dragon Boat Optimization provides a robust and efficient tool for complex engineering optimization. Future research will focus on adapting DBO for discrete optimization problems and multi-objective landscapes, continuing to draw lessons from human coordination for modern computational challenges.

Summary updated

Abstract

Dragon Boat Racing, a popular aquatic folklore team sport, is traditionally held during the Dragon Boat Festival. Inspired by this event, we propose a novel human-based meta-heuristic algorithm called dragon boat optimization (DBO) in this paper. It models the unique behaviors of each crew member on the dragon boat during the race by introducing social psychology mechanisms (social loafing, social incentive). Throughout this process, the focus is on the interaction and collaboration among the crew members, as well as their decision-making in different situations. During each iteration, DBO implements different state updating strategies. By modelling the crew’s behavior and adjusting the state updating strategies, DBO is able to maintain high-performance efficiency. We have tested the DBO algorithm with 29 mathematical optimization problems and 2 structural design problems. The experimental results demonstrate that DBO is competitive with state-of-the-art meta-heuristic algorithms as well as conventional methods.
BibTeX
@misc{li2023novel,
  author = {Xiang Li and Long Lan and Husam Lahza and Shaowu Yang and Shuihua Wang‎ and Wenjing Yang and Hengzhu Liu and Yudong Zhang},
  title = {{A Novel Human-Based Meta-Heuristic Algorithm: Dragon Boat Optimization}},
  year = {2023},
  eprint = {2311.15539},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2311.15539},
}
Figure