系统工程与电子技术 ›› 2024, Vol. 46 ›› Issue (10): 3473-3483.doi: 10.12305/j.issn.1001-506X.2024.10.24

• 制导、导航与控制 • 上一篇    

动态拓扑下四旋翼无人机集群蜂拥控制

殷雅萱, 张安, 毕文豪, 杨盼, 黄湛钧   

  1. 西北工业大学航空学院, 陕西 西安 710072
  • 收稿日期:2023-06-29 出版日期:2024-09-25 发布日期:2024-10-22
  • 通讯作者: 毕文豪
  • 作者简介:殷雅萱(1999—), 女, 硕士研究生, 主要研究方向为无人机集群协同控制
    张安(1962—), 男, 教授, 博士, 主要研究方向为航空武器火力控制技术
    毕文豪(1986—), 男, 副研究员, 博士, 主要研究方向为火力指挥与控制
    杨盼(1996—), 女, 博士研究生, 主要研究方向为多智能体协同控制
    黄湛钧(1989—), 男, 副教授, 博士, 主要研究方向为机电系统、航电系统
  • 基金资助:
    国家自然科学基金(62073267);航空科学基金(201905053001)

Flocking control for quadrotor unmanned aerial vehicle swarm with dynamic topology

Yaxuan YIN, An ZHANG, Wenhao BI, Pan YANG, Zhanjun HUANG   

  1. School of Aeronautics, Northwestern Polytechnical University, Xi'an 710072, China
  • Received:2023-06-29 Online:2024-09-25 Published:2024-10-22
  • Contact: Wenhao BI

摘要:

针对动态拓扑下无严格构型约束的无人机集群协同控制问题, 基于四旋翼无人机(quadrotor unmanned aerial vehicle, QUAV)内外环串级控制思想, 提出一种QUAV集群蜂拥控制策略。引入卡尔曼一致性滤波(Kalman-consensus filter, KCF)算法对带有噪声的通信数据进行融合, 实现对变速度领导者状态的精确估计; 考虑无人机集群拓扑的动态变化及可扩展性需求, 设计基于KCF的蜂拥控制算法以实现无人机集群位置控制, 利用李雅普诺夫稳定性定理证明算法的稳定性; 基于大脑情感学习(brain emotional learning, BEL)模型设计姿态控制器, 实现了QUAV的姿态控制。通过仿真实验验证了控制算法的有效性。

关键词: 无人机集群, 蜂拥控制, 动态拓扑, 卡尔曼一致性滤波, 大脑情感学习

Abstract:

In view of the cooperative control problem of unmanned aerial vehicle (UAV) swarm without strict configuration constraints under dynamic topology, a flocking control strategy of quadrotor UAV (QUAV) swarm is proposed on the basis of cascade control idea of the inner-outer loop of QUAV. The Kalman-consensus filter (KCF) algorithm is utilized to fuse the communication data with noise, which realizes the accurate estimation of the state of leader with varying velocity. Considering the dynamic variation and scalability requirements of UAV swarm, a flocking control algorithm based on the KCF is designed to realize the position control for the UAV swarm. The stability of the proposed algorithm is proved by the Lyapunov stability theory. An attitude controller is designed for QUAV based on brain emotional learning (BEL) model, which enables the pose control for QUAV. Simulation results prove the validity of control algorithm.

Key words: unmanned aerial vehicle (UAV) swarm, flocking control, dynamic topology, Kalman-consensus filter (KCF), brain emotional learning (BEL)

中图分类号: