Skip to main content

移动机器人编队控制

状态反馈编队控制算法与 V-REP 仿真,结合仅基于激光雷达输入的去中心化神经网络实现编队控制。

Wheeled mobile robots in precise triangular formation in a lab setting

Project Overview

This project investigates distance-based formation control of a multi-robot system. A state-feedback formation control algorithm is designed for a three-robot system to maintain a precise geometric formation using relative position measurements.

Key Features

  • Formation Control Theory: Designed and proved stability for a distance-based formation control algorithm using Lyapunov theory.
  • V-REP/CoppeliaSim Simulation: Implemented a simulation environment featuring three differential-drive wheeled robots.
  • Decentralized DNN Integration: Integrated control policies with a decentralized Deep Neural Network (DNN) that uses LiDAR ranges directly as input to generate smooth control commands while maintaining formation.

项目概览

该项目研究了多机器人系统基于距离的编队控制。为三机器人系统设计了状态反馈编队控制算法,以利用相对位置测量值保持精确的几何编队。

核心特色

  • 编队控制理论: 利用李雅普诺夫理论设计了基于距离的编队控制算法并证明了其稳定性。
  • V-REP/CoppeliaSim 仿真: 实现了一个包含三个差速驱动轮式机器人的仿真环境。
  • 去中心化 DNN 集成: 将控制策略与去中心化深度神经网络 (DNN) 集成,直接利用激光雷达测距数据作为输入,在保持编队的同时生成平滑的控制指令。

相关发表论文

  1. CCDC Distance-based formation control of a three-robot system
    Zhuo Chen, Chao Jiang, Yi Guo · Chinese Control And Decision Conference (CCDC) , 2019
  2. AIM Learning decentralized control policies for multi-robot formation
    Chao Jiang, Zhuo Chen, Yi Guo · IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM) , 2019
  3. JCD Multi-robot formation control: a comparison between model-based and learning-based methods
    Chao Jiang, Zhuo Chen, Yi Guo · Journal of Control and Decision , 2020