Information-Driven Path Planning
Information-Driven Path Planning
复制标题
DOI：
10.1007/s43154-021-00045-6
复制
发表时间：
2021-04
期刊：
影响因子：
--
通讯作者：
Shi Bai;Tixiao Shan;Fanfei Chen;Lantao Liu;Brendan Englot
中科院分区：
文献类型：
--
作者：
Shi Bai;Tixiao Shan;Fanfei Chen;Lantao Liu;Brendan Englot
基于机器人的环境监测时代产生了许多有趣的研究领域。一个关键的挑战是，机器人平台及其操作通常受到能量、时间或移动距离的限制，这反过来又限制了可以收集的测量数据的数量。因此，需要对路径进行规划，以便在满足给定预算约束的情况下，最大限度地收集关于未知环境的信息，这被称为信息规划问题。这篇综述讨论了致力于信息驱动路径规划的文献，介绍了关键的算法构建块以及突出的挑战。机器学习方法已被引入解决信息驱动的路径规划问题，提高了效率和鲁棒性。本综述从环境建模和监测的信息规划的基本构建块开始，然后与机器学习集成，强调如何使用机器学习来提高机器人技术中信息路径规划的鲁棒性和效率。
Purpose of ReviewThe era of robotics-based environmental monitoring has given rise to many interesting areas of research. A key challenge is that robotic platforms and their operations are typically constrained in ways that limit their energy, time, or travel distance, which in turn limits the number of measurements that can be collected. Therefore, paths need to be planned to maximize the information gathered about an unknown environment while satisfying the given budget constraint, which is known as the informative planning problem. This review discusses the literature dedicated to information-driven path planning, introducing the key algorithmic building blocks as well as the outstanding challenges.Recent FindingsMachine learning approaches have been introduced to solve the information-driven path planning problem, improving both efficiency and robustness.SummaryThis review started with the fundamental building blocks of informative planning for environment modeling and monitoring, followed by integration with machine learning, emphasizing how machine learning can be used to improve the robustness and efficiency of informative path planning in robotics.