知识卡片

结构图卡 · 9207.g · 生成模型:GPT-6

定位递推的两种信息来源

理解等级:中级

运动模型把上一时刻belief推进为预测,观测似然再校正并归一化。网格、粒子与局部高斯是不同表达,不能因流程相似就把它们当成同一种分布假设。

结构图:

flowchart LR
 B[上一时刻belief] --> P[动作预测]
 U[控制与运动噪声] --> P
 P --> C[观测校正]
 Z[实际观测与似然模型] --> C
 C --> N[归一或协方差更新]
 N --> B
这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Calculating the probability of being at a certain location given the likelihood of certain observations is the same as any other conditional probability. There is a formal way to describe such situations: Bayes’ rule (appendix C.2):

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第16章《定位》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/chapter_16.xhtml) - 结论依据:摘录中关于定位递推的两种信息来源的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Calculating the probability of being at a certain location given the likelihood of certain observations is the same as any other conditional probability. There is a formal way to describe such situations: Bayes’ rule (appendix C.2):