A Framework for Dynamical Distributed Flocking Control in Dense Environments
Ziqing Zhou,Chun Ouyang,Linqiang Hu,Yi Xie,Yuning Chen,Zhongxue Gan
DOI: https://doi.org/10.1016/j.eswa.2023.122694
IF: 8.5
2024-01-01
Expert Systems with Applications
Abstract:The bio-inspired flocking model has been widely utilized in self-organized swarms. However, existing potential field methods fail to guarantee safe and orderly swarm movement in dense environments featuring both static and dynamic obstacles. In this study, we propose an innovative, dynamically optimized flocking control framework that integrates an improved and tunable self-propelled flocking model with an enhanced dynamic particle swarm optimization (DPSO) algorithm. Our model incorporates pigeon-inspired obstacle gap selection based on a probability approach to optimize steering decisions. Moreover, we introduce the DPSO algorithm, which combines a collaboration-based particle swarm optimizer with fractional order velocity incorporated history-guided estimation. This hybrid algorithm dynamically determines the optimal set of control parameters to maintain the desired swarm state. Through numerical comparison experiments with state-of-the-art models, we demonstrate that our approach enhances the safety, synchronization, and efficiency of the swarm even within a limited visual field, without prior environmental information. These features bring the simulations closer to real-world experiments.