← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9133.c · 生成模型:GPT-6

阻尼最小二乘约束关节步幅

理解等级:高级

对e=目标−当前,可最小化||Jδq−e||²+λ²||δq||²,得到δq=(JᵀJ+λ²I)⁻¹Jᵀe。λ>0抑制弱奇异方向的大步幅,代价是允许残差。书中3.42漏平方、3.43末项写J⁺均不应直接作为该目标的标准解。

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

      In this case, we do not only minimize the error but also the joint velocities. The minimization problem then becomes: where λ is a constant. One can show that the resulting controller that achieves this has the form This is known as the damped least-squares method. Problems that can arise when taking this approach include the existence of local minima and singularities of the mechanism,

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》第03章《运动学》(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_3.xhtml) - 结论依据:摘录中关于阻尼最小二乘约束关节步幅的定义、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:In this case, we do not only minimize the error but also the joint velocities. The minimization problem then becomes: where λ is a constant. One can show that the resulting controller that achieves this has the form This is known as the damped least-squares method. Problems that can arise when taking this approach include the existence of local minima and singularities of the mechanism,