当前位置:网站首页>微商的差商近似
微商的差商近似
2022-07-27 10:59:00 【char~lie】
有限差分法是用变量离散的、含有有限个未知数的差分方程近似替代连续变量的微分方程,因此首要任务是构建合理的差分格式,使解能够保持原问题的性质,并且具有相当的精度。
为了表明精确度,下面用一元函数做例子,令
y = f ( x ) y=f(x) y=f(x)
在x轴上按h为间隔取点,f(xi+1)可以用Taylor级数表示为:
f ( x i + 1 ) = f ( x i ) + h 1 ! ∂ f ( x ) ∂ x ∣ x = x i + h 2 2 ! ∂ 2 f ( x ) ∂ x 2 ∣ x = x i + . . . f(x_{i+1})=f(x_i)+\frac h {1!}\frac {\partial f(x)} {\partial x}|_{x=x_i}+\frac {h^2} {2!}\frac {\partial^2 f(x)} {\partial x^2}|_{x=x_i}+... f(xi+1)=f(xi)+1!h∂x∂f(x)∣x=xi+2!h2∂x2∂2f(x)∣x=xi+...
f ( x i − 1 ) = f ( x i ) − h 1 ! ∂ f ( x ) ∂ x ∣ x = x i − h 2 2 ! ∂ 2 f ( x ) ∂ x 2 ∣ x = x i − . . . f(x_{i-1})=f(x_i)-\frac h {1!}\frac {\partial f(x)} {\partial x}|_{x=x_i}-\frac {h^2} {2!}\frac {\partial^2 f(x)} {\partial x^2}|_{x=x_i}-... f(xi−1)=f(xi)−1!h∂x∂f(x)∣x=xi−2!h2∂x2∂2f(x)∣x=xi−...
由此可得:
f ( x i + 1 ) − f ( x i ) h = ∂ f ( x ) ∂ x ∣ x = x i + h 2 ! ∂ 2 f ( x ) ∂ x 2 ∣ x = x i + . . = ∂ f ( x ) ∂ x ∣ x = x i + O ( h ) \frac {f(x_{i+1})-f(x_i)}{h}=\frac {\partial f(x)} {\partial x}|_{x=x_i}+\frac {h} {2!}\frac {\partial^2 f(x)} {\partial x^2}|_{x=x_i}+..=\frac {\partial f(x)} {\partial x}|_{x=x_i}+O(h) hf(xi+1)−f(xi)=∂x∂f(x)∣x=xi+2!h∂x2∂2f(x)∣x=xi+..=∂x∂f(x)∣x=xi+O(h)
f ( x i ) − f ( x i − 1 ) h = ∂ f ( x ) ∂ x ∣ x = x i − h 2 ! ∂ 2 f ( x ) ∂ x 2 ∣ x = x i − . . = ∂ f ( x ) ∂ x ∣ x = x i − O ( h ) \frac {f(x_{i})-f(x_{i-1})}{h}=\frac {\partial f(x)} {\partial x}|_{x=x_i}-\frac {h} {2!}\frac {\partial^2 f(x)} {\partial x^2}|_{x=x_i}-..=\frac {\partial f(x)} {\partial x}|_{x=x_i}-O(h) hf(xi)−f(xi−1)=∂x∂f(x)∣x=xi−2!h∂x2∂2f(x)∣x=xi−..=∂x∂f(x)∣x=xi−O(h)
f ( x i + 1 ) − f ( x i ) h ( 1 ) \frac {f(x_{i+1})-f(x_i)}{h} (1) hf(xi+1)−f(xi)(1)
f ( x i ) − f ( x i − 1 ) h ( 2 ) \frac {f(x_{i})-f(x_{i-1})}{h} (2) hf(xi)−f(xi−1)(2)
(1)式叫前向差商,(2)式叫后向差商。
f ( x i + 1 ) − f ( x i − 1 ) 2 h = ∂ f ( x ) ∂ x ∣ x = x i + h 2 3 ! ∂ 3 f ( x ) ∂ x 3 ∣ x = x i + h 4 5 ! ∂ 5 f ( x ) ∂ x 5 ∣ x = x i + . . = ∂ f ( x ) ∂ x ∣ x = x i + O ( h 2 ) \frac {f(x_{i+1})-f(x_{i-1})}{2h}=\frac {\partial f(x)} {\partial x}|_{x=x_i}+\frac {h^2} {3!}\frac {\partial^3 f(x)} {\partial x^3}|_{x=x_i}+\frac {h^4} {5!}\frac {\partial^5 f(x)} {\partial x^5}|_{x=x_i}+..=\frac {\partial f(x)} {\partial x}|_{x=x_i}+O(h^2) 2hf(xi+1)−f(xi−1)=∂x∂f(x)∣x=xi+3!h2∂x3∂3f(x)∣x=xi+5!h4∂x5∂5f(x)∣x=xi+..=∂x∂f(x)∣x=xi+O(h2)
f ( x i + 1 ) − f ( x i − 1 ) 2 h ( 3 ) \frac {f(x_{i+1})-f(x_{i-1})}{2h}(3) 2hf(xi+1)−f(xi−1)(3)
(3)式叫做中心差商,这个是微商的二阶近似。
边栏推荐
- The C programming language 2nd -- Notes -- 6.7
- "My" bug collection (Reprinted)
- Analysis of distributed database and cache double write consistency scheme (Reprint)
- C programming language (2nd Edition) -- Reading Notes -- 1.5
- 第10章 枚举类与注解
- State compression DP acwing 91. shortest Hamilton path
- Moveit2 - 5. Scenario Planning
- TapNet: Multivariate Time Series Classification with Attentional Prototypical Network
- Error while unzipping file in win10: unable to create symbolic link. You may need to run WinRAR with administrator privileges. The client does not have the required privileges
- 局域网SDN硬核技术内幕 25 展望未来——RDMA(下)
猜你喜欢

为什么选择智能电视?

Game theory acwing 893. Set Nim game

Remember an experience of using canvas to make the banner streamer effect of Tencent cloud homepage

树形DP AcWing 285. 没有上司的舞会

Moveit2 - 4. robot model and robot state

Knapsack model acwing 1024. Packing problem

Error encountered in adding quick open option to right-click menu:

Game theory acwing 894. Split Nim game

Find the combination number acwing 886. find the combination number II

你真的会写二分查找吗——变种二分查找
随机推荐
第10章 枚举类与注解
Moveit2 - 5. Scenario Planning
高斯消元 AcWing 883. 高斯消元解线性方程组
Maker harmony OS application development training notes 01
Solutions to errors in tensorflow operation
Knapsack model acwing 1024. Packing problem
Find the combination number acwing 886. find the combination number II
LAN SDN technology hard core insider 11 the key of cloud convergence CP -- hierarchical port binding
LAN SDN hard core technology insider 23 looking forward to the future - RDMA (Part 1)
Error encountered in adding quick open option to right-click menu:
EfficientNet
[shader realizes shake random shaking effect _shader effect Chapter 10]
The longest ascending subsequence model acwing 1016. The sum of the largest ascending subsequence
力扣——10. 正则表达式匹配
Properties file
Raw socket grabs packets, and packets on some ports cannot be caught
torch‘ has no attribute ‘inference_mode‘
数字三角形模型 AcWing 1018. 最低通行费
局域网SDN技术硬核内幕 11 云网融合CP的关键——层次化端口绑定
Digital triangle model acwing 1015. Picking flowers