面向未知环境的快速安全自主轨迹生成  

Fast and Safe Autonomous Trajectory Generation in Unknown Environments

作　　者：李海 江涛 苏晓杰[1] 付文豪 LI Hai;JIANG Tao;SUXiao-jie;FU Wen-hao(School of Automation,Chongqing University,Chongqing 400044,China)

机构地区：[1]重庆大学自动化学院,重庆400044

出　　处：《控制工程》2021年第11期2153-2157,共5页Control Engineering of China

摘　　要：为了实现在未知环境中移动机器人的最优路径规划,并具有快速安全的特性,提出了一种基于空间凸分解和改进区间时间分配的路径规划算法。本算法借助三次贝赛尔曲线来描述路径,同时提出了混合整数二次规划来优化路径。快速性是通过对轨迹的每个区间施加特别的时间分配方案来实现的。在每个重新规划步骤开始时,在自由已知空间中始终有一个可行、安全的备用轨迹,从而确保安全性。仿真实验环境基于Gazebo平台搭建,并基于自主装配的移动机器人进行物理实验,实验结果表明,本算法适用于未知环境下的移动机器人路径规划,且可以有效地进行快速安全的轨迹规划。In order to achieve fast, safe and autonomous optimal path planning for mobile robots in unknown environments, a path planning algorithm based on spatial convex decomposition and a special interval time allocation scheme is proposed. This algorithm uses cubic Bezier curve to describe the path, and at the same time mixed integer quadratic programming(MIQP) is proposed to optimize the path. The fastness is achieved by applying a special time allocation scheme to each interval of the trajectory. At the beginning of each replanning step, there is always a feasible and safe alternate trajectory in the free-known space to ensure safety.The simulation experiment environment is built on the Gazebo platform, and the physical experiments are carried out based on autonomously assembled mobile robots. The experimental results show that this algorithm is suitable for path planning of mobile robots in unknown environments, and can effectively carry out fast and safe trajectory planning.

关 键 词：未知环境 局部规划 凸分解 贝塞尔曲线 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]