摘要
针对目前移动机器人全局路径规划算法中存在规划效率低、路径质量低等问题,提出一种新的基于梯度优化的路径规划算法。首先使用欧氏符号距离场方法表示已知的二维地图环境,构建从障碍物中心向外梯度下降的距离场,以获得机器人到障碍物中心的距离信息来进行碰撞惩罚代价的计算;其次将路径规划问题描述为优化问题,通过引入目标距离信息和障碍物距离信息,设计了一个新的路径规划代价函数,并使用梯度下降方法优化该函数得到一条最优路径;最后通过仿真对比实验表明,所提算法在规划时间、路径长度和最大转向角方面均具有一定的优越性,并通过搭建基于ROS的移动机器人实验平台进行实际环境中的路径规划实验,验证了所提算法的有效性和可行性。
In view of the low planning efficiency and poor path quality in the current global path planning algorithm for mobile robots,a new path planning algorithm based on gradient optimization is proposed.The Euclidean signed distance field(ESDF)method is used to represent the known two⁃dimensional map environment,and a distance field with the gradient descent from the center of the obstacle towards outside is constructed to obtain the distance information from the robot to the center of the obstacle and then calculate the collision penalty cost.Then,the path planning problem is described as an optimization problem,and a new path planning cost function is designed by introducing the object distance information and obstacle distance information.This function is optimized by the gradient descent method to obtain an optimal path.Simulation experiments show that the proposed algorithm has certain advantages in terms of planning time,path length and maximum steering angle.Path planning experiments were performed in the real environment by building a ROS⁃based mobile robot experiment platform.Its effectiveness and feasibility are verified.
作者
杨帆
李玮
严天宏
YANG Fan;LI Wei;YAN Tianhong(College of Mechanical and Electrical Engineering,China Jiliang University,Hangzhou 310018,China)
出处
《现代电子技术》
2023年第21期99-104,共6页
Modern Electronics Technique
基金
浙江省自然科学基金项目(LTGG23E090002)。
