融合改进Dijkstra算法和MPC的单舵轮机器人路径规划  

Path Planning of Single-Wheeled Robot Based on Improved Dijkstra Algorithm and Model Prediction

作　　者：郭建 曾志豪 黄熙航 谢瑞伦 杨朋 王家淳 GUO Jian;ZENG Zhihao;HUANG Xihang;XIE Ruilun;YANG Peng;WANG Jiachun(School of Mechanical Engineering,Guangzhou City University of Technology,Guangzhou 510800,China)

机构地区：[1]广州城市理工学院机械工程学院,广州510800

出　　处：《组合机床与自动化加工技术》2024年第6期18-23,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：2022广东省普通高校特色创新人才类项目(2022KTSCX185);广州城市理工学院2022年度校级青年科研基金项目(K0222005);2023年广东省科技创新战略专项资金项目(pdjh2023b0778);广州城市理工学院2023年度校级科研基金项目(K0223002)。

摘　　要：为提高单舵轮机器人在路径规划性能和路径跟踪精度,提出一种融合改进Dijkstra算法和MPC的路径规划算法。首先,对比Dijkstra算法和A^(*)算法两种方法获得的路径规划轨迹长度;随后,改进Dijkstra算法,替换与障碍物碰撞的子路径段,并对路径进行平滑优化处理;再者,建立单舵轮机器人数学模型,基于优化后的MPC算法,分别以直线和圆弧路径跟踪进行仿真。仿真实验结果表明,跟踪直线速度从1.19 m/s平稳到达1.91 m/s,角加速度为1 rad/s^(2),跟踪圆弧速度从1.19 m/s增加至2 m/s,角速度平稳变化;最后,通过样机实验结果表明,MPC算法能够使得单舵轮移动机器人运动高度贴合改进的Dijkstra算法路径规划的轨迹,且行驶平稳。In order to improve the path planning performance and path tracking accuracy of single steering wheel robot,a path planning algorithm combining improved Dijkstra algorithm and MPC is proposed.Firstly,the path planning trajectory length obtained by Dijkstra algorithm and A^(*) algorithm is compared.Then,the Dijkstra algorithm is improved to replace the sub-path segment that collides with the obstacle,and the path is smoothed and optimized.Furthermore,the mathematical model of the single-wheeled robot is established.Based on the optimized MPC algorithm,the straight line and arc path tracking are simulated respectively.The simulation results show that the tracking linear velocity increases from 1.19 m/s to 1.91 m/s,the angular acceleration is 1 rad/s^(2),the tracking arc velocity increases from 1.19 m/s to 2 m/s,and the angular velocity changes smoothly.Finally,the experimental results of the prototype show that the MPC algorithm can make the motion of the single steering wheel mobile robot highly fit the trajectory of the improved Dijkstra algorithm path planning,and the driving is stable.

关 键 词：单舵轮机器人 DIJKSTRA算法 模型预测控制轨迹跟踪 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]