启发式自适应步长优化Informed-RRT*算法
Informed-RRT*Algorithm Based on Heuristic Adaptive Step Size Optimization
葛启兴 1章伟 1陈浩 1胡陟 1谢贵亮1
作者信息
- 1. 上海工程技术大学机器人智能控制实验室,上海 201620
- 折叠
摘要
Informed-RRT*算法是解决全局路径规划问题常用的算法.当处理狭窄环境时,Informed-RRT*算法往往容易陷入局部最优解,而在复杂环境中路径规划的成本又往往过高.为了解决这些问题,提出了一种基于启发式自适应步长的采样策略,以改进Informed-RRT*算法的不足之处.通过在随机节点周围扩展采样点集来计算启发式值,选择最优节点并按照其生长方向进行扩张.通过计算最优节点与最近节点的距离,确定下一次采样的步长.这使得机器人能够更好地适应二维和三维环境中的狭窄区域和复杂环境.将改进的算法在二维和三维环境中进行仿真验证,实验结果表明了该算法的有效性和鲁棒性较为优异.
Abstract
The Informed-RRT*algorithm is a commonly used algorithm for solving global path planning problems.When dealing with narrow environments,the Informed-RRT*algorithm often gets trapped in local optima,while the cost of path planning in complex environments tends to be excessively high.To address these issues,a sampling strategy based on heuristic adaptive step size is proposed to improve the limitations of the Informed-RRT*algorithm.Firstly,the heuristic value is computed by expanding the sample node set around random points,and then the optimal node is selected and expanded along its growth direction.Secondly,the distance between the optimal node and the nearest node is calculated to determine the step size for the next sampling.This enables the robot to better adapt to narrow areas and complex environments in both 2D and 3D scenarios.Finally,the improved algorithm is validated through simulation in both 2D and 3D scenarios.The experimental results demonstrate the remarkable effectiveness and robustness of the algorithm.
关键词
路径规划/启发式自适应步长/Informed-RRT*/三维场景Key words
path planning/heuristic adaptive step size/Informed-RRT*/3D scenarios引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据