← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9201.c · 生成模型:GPT-6

观测矩阵不一定直接测到状态

理解等级:中级

线性观测z=Hx+v可能只看到状态的某些组合,创新应为z−Hx_prior,增益还需考虑H与测量协方差。原书本章只提供简化融合直觉,没有完整写出所有预测和协方差更新,不能据此宣称滤波器已实现完整。

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

      where Hkxk is the measurement prediction. In our example, Hk was just the identity matrix; in a robot position estimation problem, Hk is a function that would predict how a robot would observe a shift in position through a sensor. As you can see, all the weighting based on variances is done in the Kalman gain K.

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第15章《不确定性与误差传播》(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_15.xhtml) - 结论依据:摘录中关于观测矩阵不一定直接测到状态的定义、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:where Hkxk is the measurement prediction. In our example, Hk was just the identity matrix; in a robot position estimation problem, Hk is a function that would predict how a robot would observe a shift in position through a sensor. As you can see, all the weighting based on variances is done in the Kalman gain K.