摘要
基于拟态物理(artificial physics)法研究多无人机系统的集群构型生成及重构控制问题.在改进传统的拟态物理法基础上定义新的引力和斥力,使质点模型多无人机系统能够生成均匀分布的标准构型,并借助LaSalle不变集理论分析了集群控制律的稳定性;在此基础上,进一步讨论了序号无关的集群构型局部极值的改进策略.探索了一般性集群构型和标准构型之间的可逆且唯一的双射变换,利用双射变换提出了一般性集群构型生成的集群控制算法,并进而讨论了集群队形的收缩与扩张、集群构型切换和集群避障等集群重构问题.为验证提出的质点模型集群构型生成和重构算法,论文基于ROS(机器人操作系统)及Gazebo仿真器构建了多四旋翼无人机的协同仿真平台,分析验证了构型生成和重构算法的有效性.最后,将设计的基于拟态物理法的集群控制律扩展到多固定翼无人机系统,并搭建了基于X-Plane飞行模拟器和无人机自驾仪的多无人机协同控制半实物仿真环境,验证集群构型生成控制律的有效性.
This paper investigates the flocking and reconfiguration control problem of multiple unmanned aerial vehicles（UAVs） systemby using a modified artificial physics（AP） method.Firstly,in order to dirve the multi-UAV system with double-intergrator dynamics to form the uniform distribued standard formation,we redefine the attractive and repulsive forces in the traditional artificial physics method,and analyze the stability of the flocking control method using the LaSalle invariance principle accordingly.Immediately,we develop an improved strategy regards to the local optimal solution of the flocking problem in the serial number independent case.Then,we explore the reversible and exclusive bijective transformation between the standard formation and an arbitrary formation.Using the bijective transformation,we propose a flocking control method to form the arbitrary formation,and futher extend to the flocking reconfiguration problem,such as flocking contraction and expansion,formation switching and flocking with obstacle avoidence.In order to verify the effectiveness of the proposed flocking and reconfiguration control method,we construct a multiple quadrotors coodination simulation platform based on the robot operating system（ROS） and Gazebo simulator.The simulation results demonstrate the effectiveness of the flocking and reconfiguration control method.Finally,we expand our artificial physics based flocking control method to the multiple fixed-wing UAVs system,and the effectiveness is verified in a semi-physical simulation environment for multi-UAV coodination by using X-Plane flight simulaor and the UAV autopilot.
出处
《中国科学：技术科学》
EI
CSCD
北大核心
2017年第3期266-285,共20页
Scientia Sinica(Technologica)
基金
国家自然科学基金(批准号:61403406)资助项目
关键词
无人机集群
拟态物理法
双射变换
队形生成
队形重构
UAV flocking
artificial physics
bijective transformation
flocking control
reconfiguration
