首页|融合改进Dijkstra算法和动态窗口法的移动机器人路径规划

融合改进Dijkstra算法和动态窗口法的移动机器人路径规划

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
为解决移动机器人在智能制造车间的全局路径规划和局部动态避障问题,提出一种融合改进的Dijkstra算法和改进的DWA算法,对传统Dijkstra算法的路径进行平滑优化,使得路径轨迹更加平滑,动态改变DWA算法中速度评价权重函数,提高避障效率.仿真结果表明,改进Dijkstra算法路径平滑优化后,平均路程缩短比例为 0.65%,平均偏航角震荡次数减少了 67.70%,改进后的DWA算法运行路程缩小9.68%,路径转折次数降低了 33%,运行时间缩短 3.88%.基于改进的Dijkstra算法和改进的DWA算法提出一种融合算法,仿真和样机实验结果表明:面对静态、动态障碍物,机器人运行线速度平缓,轨迹光滑,角速度波动明显,证明机器人运动稳定,实时调整方位,具有良好的避障能力.并且多次机器人循环定点实验中机器人纵向(X轴方向)平均误差≤30 mm,横向(Y轴)平均误差≤30 mm,定位精度满足工业需求.
Mobile Robot Path Planning Based on Improved Dijkstra Algorithm and Dynamic Window Method
To solve the global path planning and local dynamic obstacle avoidance problems of mobile ro-bots in intelligent manufacturing workshops,an improved Dijkstra algorithm fused with improved DWA al-gorithm is proposed.The path of the traditional Dijkstra algorithm is smoothed and optimised to make the path trajectory smoother,and the speed evaluation weight function in the DWA algorithm is dynamically changed to improve the obstacle avoidance efficiency.Simulation results show that after the path smoothing optimization of the improved Dijkstra algorithm,the average distance reduction ratio is 0.65%,the average number of yaw angle oscillations is reduced by 67.70%,the running distance of the improved DWA algo-rithm is reduced by 9.68%,the number of path turns is reduced by 33%and the running time is shortened by 3.88%.Based on the improved Dijkstra algorithm and the improved DWA algorithm,a fusion algorithm is proposed.Simulation and prototype experimental results show that the robot runs with smooth linear ve-locity,smooth trajectory and obvious angular velocity fluctuation in the face of static and dynamic obsta-cles,proving that the robot has stable motion,real-time orientation adjustment and good obstacle avoidance capability.In addition,the average error of the robot's longitudinal direction(X-axis direction)is≤30 mm and the average error of the robot's lateral direction(Y-axis)is≤30 mm in several robot cyclic fixing ex-periments,and the positioning accuracy meets industrial requirements.

Dijkstra algorithmdynamic window methodmobile robot path planningreal-time obstacle a-voidance

郭建、杨朋、曾志豪、黄颖驹、王家淳

展开 >

广州城市理工学院机械工程学院,广州 510800

Dijkstra算法 动态窗口法 移动机器人路径规划 实时避障

广东省普通高等学校特色创新人才类项目(2022)广州城市理工学院校级青年科研基金(2022)广东省科技创新战略专项(2023)广州城市理工学院校级科研基金(2023)

2022KTSCX185K0222005pdjh2023b0778K0223002

2024

10.13462/j.cnki.mmtamt.2024.03.008

组合机床与自动化加工技术
大连组合机床研究所 中国机械工程学会生产工程分会

组合机床与自动化加工技术

CSTPCD北大核心
影响因子：0.671
ISSN：1001-2265
年,卷(期)：2024.(3)
  • 14