机器人学-估计与学习-第三周
Posted gdut-gordon
tags:
篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了机器人学-估计与学习-第三周相关的知识,希望对你有一定的参考价值。
3.1 robotic mapping
3.1.1 Introduction
3.1.2 Introduction to mapping
Map and Mapping
(1) map is a spatial model of robot‘s environment
(2) mapping is a process for building a map
(3) consideration for mapping
- map representation
- avalable sensors
- purpose of mapping
Type of map:
(1) metric map : a localization is represented as a coordinate
(2) topological : Locations are represented as nodes and their connectivity as arcs. Only the connectivity between nodes matter. Graph representation is useful for path planning.
(3) semantic map : semantic map is a map with labels.
What make it challengin?
- Noisy measurement in local coordinate
- Motion involved
- Change over time
3.2 occupancy grid mapping
3.2.1 occupancy grid map
3.2.2 log-odd update
在贝叶斯框架下根据观测模型直接更新每个珊格的占据概率是困难的,引入一个新的变量Odd,赢率,其含义为某件事发生的概率与不发生的概率之比。
取对数,获得odd的更新方法,即每次观测后都更新先验地图,获得的后验地图将作为下一时刻的先验地图,以此方式不断迭代。
例题:
3.2.3 handling range sensor
3.2.4 2D occupancy grid mapping
3.3 3D mapping
以上是关于机器人学-估计与学习-第三周的主要内容,如果未能解决你的问题,请参考以下文章
2019-2020-1 学期 2427 《网络空间安全导论》 第三周学习总结
2019-2020-1学期 20202429《网络空间安全专业导论》第三周学习总结