Skip to content

Latest commit

 

History

History
16 lines (16 loc) · 926 Bytes

Conjugate Direction Method.md

File metadata and controls

16 lines (16 loc) · 926 Bytes

![[Pasted image 20240525185636.png]] 用于纯二次问题

Conjugate directions

![[Pasted image 20240525190749.png]]

Conjugate Direction Theorem

![[Pasted image 20240525194242.png]]

The Conjugate Gradient Method

![[Pasted image 20240527001429.png]] 共轭梯度法是一种共轭方向法,通过选择连续方向向量作为方法进展过程中获得的连续梯度的共轭版本而获得。因此,方向不是预先指定的,而是在迭代的每个步骤中依次确定的。在步骤 $k$ 处,评估当前的负梯度向量,并将先前方向向量的线性组合添加到其中,以获得要沿其移动的新共轭方向向量。

The Conjugate Gradient Algorithm

![[Pasted image 20240527001448.png]]

Conjugate Gradient Theorem

![[Pasted image 20240527001509.png]] ![[Pasted image 20240618202010.png]]

Fletcher-Reeves Conjugate Gradient Method

![[Pasted image 20240527001614.png]]