An Efficient Sampling-based Method for Online Informative Path Planning in Unknown Environments
Abstract
new RRT*- inspired online informative path planning algorithm,continuously expands a single tree of candidate trajectories and rewires nodes to maintain the tree and refine intermediate
propose a novel TSDF-based 3D reconstruction gain and cost-utility formulation
Introduction
传统的基于采样的在线informative path plan,通常是把采样的轨迹存储到树中,但是只会执行第一步,这样有两个弊端
- 树的其他部分都被丢弃了,因为只有最好的branch 是被执行的,这样的话会造成计算量特别大;
- 由于倾向于选择局部最优情况,会导致路径非最优,或者陷入到dead-end中;
为了解决这两个问题,采用的方法是:根据节点的用途重新布线,这样未执行的节点及其子树将保持活动状态。 结合自适应更新策略,可以遍历不断增长的树并对其进行维护