知识卡片
贝叶斯滤波交替预测与校正
预测把先前belief经运动模型推进,校正再乘当前观测似然并归一化,使全部历史通过当前belief递推保留。动作与观测应按时间顺序处理,不能原地覆盖旧分布同时计算所有新格;归一和为零还提示模型或异常问题。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 定位
原书出处 · 1 处
- · 查看原书依据
This final equation is remarkable because it allows us to perform a belief update for a given state by incorporating a sensor measurement and/or a motion prediction based on an action we took. With this formulation, we can define an algorithm for belief updates that takes our current belief, an array of action and observation data, and the set of states that comprise the state
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:摘录中关于贝叶斯滤波交替预测与校正的定义、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:This final equation is remarkable because it allows us to perform a belief update for a given state by incorporating a sensor measurement and/or a motion prediction based on an action we took. With this formulation, we can define an algorithm for belief updates that takes our current belief, an array of action and observation data, and the set of states that comprise the state