当前位置: 首页 > news >正文

全微分证明 链式法则 乘法法则 除法法则

z = f ( u ( x , y ) , v ( x , y ) ) z = f(u(x, y), v(x, y)) z=f(u(x,y),v(x,y))

全微分(Total Differential)、链式法则(Chain Rule),以及它在涉及乘法法则除法法则下的应用。


🧠 一、函数背景

你给出的函数是一个复合函数

  • 外层函数: z = f ( u , v ) z = f(u, v) z=f(u,v)
  • 内层函数: u = u ( x , y ) , v = v ( x , y ) u = u(x, y),\quad v = v(x, y) u=u(x,y),v=v(x,y)

也就是说, z z z x , y x, y x,y间接函数


📘 二、链式法则(Chain Rule)

对于复合函数 z = f ( u ( x , y ) , v ( x , y ) ) z = f(u(x,y), v(x,y)) z=f(u(x,y),v(x,y)),我们求其相对于 x x x y y y 的偏导时,应用链式法则:

∂ z ∂ x = ∂ f ∂ u ⋅ ∂ u ∂ x + ∂ f ∂ v ⋅ ∂ v ∂ x \frac{\partial z}{\partial x} = \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial x} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial x} xz=ufxu+vfxv

∂ z ∂ y = ∂ f ∂ u ⋅ ∂ u ∂ y + ∂ f ∂ v ⋅ ∂ v ∂ y \frac{\partial z}{\partial y} = \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial y} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial y} yz=ufyu+vfyv


📗 三、全微分公式(Total Differential)

全微分 d z dz dz 是表示函数小变化的线性近似。由链式法则导出:

d z = ∂ z ∂ x d x + ∂ z ∂ y d y dz = \frac{\partial z}{\partial x} dx + \frac{\partial z}{\partial y} dy dz=xzdx+yzdy

代入链式法则的结果:

d z = ( ∂ f ∂ u ⋅ ∂ u ∂ x + ∂ f ∂ v ⋅ ∂ v ∂ x ) d x + ( ∂ f ∂ u ⋅ ∂ u ∂ y + ∂ f ∂ v ⋅ ∂ v ∂ y ) d y dz = \left( \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial x} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial x} \right) dx + \left( \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial y} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial y} \right) dy dz=(ufxu+vfxv)dx+(ufyu+vfyv)dy

也可以整理成:

d z = ∂ f ∂ u ⋅ d u + ∂ f ∂ v ⋅ d v dz = \frac{\partial f}{\partial u} \cdot du + \frac{\partial f}{\partial v} \cdot dv dz=ufdu+vfdv

其中: d u = ∂ u ∂ x d x + ∂ u ∂ y d y , d v = ∂ v ∂ x d x + ∂ v ∂ y d y \text{其中:}\quad du = \frac{\partial u}{\partial x} dx + \frac{\partial u}{\partial y} dy,\quad dv = \frac{\partial v}{\partial x} dx + \frac{\partial v}{\partial y} dy 其中:du=xudx+yudy,dv=xvdx+yvdy



📘 全微分证明

根据全微分定义:

d z = ∂ z ∂ x d x + ∂ z ∂ y d y dz = \frac{\partial z}{\partial x} dx + \frac{\partial z}{\partial y} dy dz=xzdx+yzdy

把上面链式法则中的偏导数代入:

d z = ( ∂ f ∂ u ⋅ ∂ u ∂ x + ∂ f ∂ v ⋅ ∂ v ∂ x ) d x + ( ∂ f ∂ u ⋅ ∂ u ∂ y + ∂ f ∂ v ⋅ ∂ v ∂ y ) d y dz = \left( \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial x} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial x} \right) dx + \left( \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial y} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial y} \right) dy dz=(ufxu+vfxv)dx+(ufyu+vfyv)dy

整理得:

d z = ∂ f ∂ u ⋅ ( ∂ u ∂ x d x + ∂ u ∂ y d y ) + ∂ f ∂ v ⋅ ( ∂ v ∂ x d x + ∂ v ∂ y d y ) dz = \frac{\partial f}{\partial u} \cdot \left( \frac{\partial u}{\partial x} dx + \frac{\partial u}{\partial y} dy \right) + \frac{\partial f}{\partial v} \cdot \left( \frac{\partial v}{\partial x} dx + \frac{\partial v}{\partial y} dy \right) dz=uf(xudx+yudy)+vf(xvdx+yvdy)

注意:

d u = ∂ u ∂ x d x + ∂ u ∂ y d y , d v = ∂ v ∂ x d x + ∂ v ∂ y d y du = \frac{\partial u}{\partial x} dx + \frac{\partial u}{\partial y} dy,\quad dv = \frac{\partial v}{\partial x} dx + \frac{\partial v}{\partial y} dy du=xudx+yudy,dv=xvdx+yvdy

所以:

d z = ∂ f ∂ u ⋅ d u + ∂ f ∂ v ⋅ d v \boxed{dz = \frac{\partial f}{\partial u} \cdot du + \frac{\partial f}{\partial v} \cdot dv} dz=ufdu+vfdv

全微分公式得证。


📗 三、乘法法则的推导(复合乘积函数)

设:

z = f ( u , v ) = u ( x , y ) ⋅ v ( x , y ) z = f(u, v) = u(x,y) \cdot v(x,y) z=f(u,v)=u(x,y)v(x,y)

直接对 x x x 求偏导:

∂ z ∂ x = ∂ ( u v ) ∂ x = ∂ u ∂ x ⋅ v + u ⋅ ∂ v ∂ x \frac{\partial z}{\partial x} = \frac{\partial (uv)}{\partial x} = \frac{\partial u}{\partial x} \cdot v + u \cdot \frac{\partial v}{\partial x} xz=x(uv)=xuv+uxv

这其实也是链式法则的一个特例,因为:

  • f ( u , v ) = u ⋅ v f(u,v) = u \cdot v f(u,v)=uv
  • 所以 ∂ f ∂ u = v \frac{\partial f}{\partial u} = v uf=v, ∂ f ∂ v = u \frac{\partial f}{\partial v} = u vf=u

所以由链式法则:

∂ z ∂ x = ∂ f ∂ u ⋅ ∂ u ∂ x + ∂ f ∂ v ⋅ ∂ v ∂ x = v ⋅ ∂ u ∂ x + u ⋅ ∂ v ∂ x \frac{\partial z}{\partial x} = \frac{\partial f}{\partial u} \cdot \frac{\partial u}{\partial x} + \frac{\partial f}{\partial v} \cdot \frac{\partial v}{\partial x} = v \cdot \frac{\partial u}{\partial x} + u \cdot \frac{\partial v}{\partial x} xz=ufxu+vfxv=vxu+uxv

全微分:

d z = v ⋅ d u + u ⋅ d v dz = v \cdot du + u \cdot dv dz=vdu+udv

乘法法则得证。


📕 四、除法法则的推导

设:

z = f ( u , v ) = u ( x , y ) v ( x , y ) z = f(u, v) = \frac{u(x,y)}{v(x,y)} z=f(u,v)=v(x,y)u(x,y)

x x x 求偏导:

使用商法则(Quotient Rule):

∂ z ∂ x = v ⋅ ∂ u ∂ x − u ⋅ ∂ v ∂ x v 2 \frac{\partial z}{\partial x} = \frac{v \cdot \frac{\partial u}{\partial x} - u \cdot \frac{\partial v}{\partial x}}{v^2} xz=v2vxuuxv

这个可以从链式法则推导:

  • f ( u , v ) = u v ⇒ ∂ f ∂ u = 1 v , ∂ f ∂ v = − u v 2 f(u,v) = \frac{u}{v} \Rightarrow \frac{\partial f}{\partial u} = \frac{1}{v},\quad \frac{\partial f}{\partial v} = -\frac{u}{v^2} f(u,v)=vuuf=v1,vf=v2u
  • 所以:

∂ z ∂ x = 1 v ⋅ ∂ u ∂ x − u v 2 ⋅ ∂ v ∂ x \frac{\partial z}{\partial x} = \frac{1}{v} \cdot \frac{\partial u}{\partial x} - \frac{u}{v^2} \cdot \frac{\partial v}{\partial x} xz=v1xuv2uxv

全微分:

d z = 1 v ⋅ d u − u v 2 ⋅ d v = v ⋅ d u − u ⋅ d v v 2 dz = \frac{1}{v} \cdot du - \frac{u}{v^2} \cdot dv = \frac{v \cdot du - u \cdot dv}{v^2} dz=v1duv2udv=v2vduudv

除法法则得证。


✅ 总结表格

类型表达式导数/微分公式
链式法则 z = f ( u ( x , y ) , v ( x , y ) ) z = f(u(x,y), v(x,y)) z=f(u(x,y),v(x,y)) ∂ z ∂ x = ∂ f ∂ u ∂ u ∂ x + ∂ f ∂ v ∂ v ∂ x \frac{\partial z}{\partial x} = \frac{\partial f}{\partial u}\frac{\partial u}{\partial x} + \frac{\partial f}{\partial v}\frac{\partial v}{\partial x} xz=ufxu+vfxv
全微分同上 d z = ∂ f ∂ u ⋅ d u + ∂ f ∂ v ⋅ d v dz = \frac{\partial f}{\partial u} \cdot du + \frac{\partial f}{\partial v} \cdot dv dz=ufdu+vfdv
乘法法则 z = u ⋅ v z = u \cdot v z=uv d z = v ⋅ d u + u ⋅ d v dz = v \cdot du + u \cdot dv dz=vdu+udv
除法法则 z = u v z = \frac{u}{v} z=vu d z = v ⋅ d u − u ⋅ d v v 2 dz = \frac{v \cdot du - u \cdot dv}{v^2} dz=v2vduudv

http://www.xdnf.cn/news/960823.html

相关文章:

  • 基于正点原子阿波罗F429开发板的LWIP应用(6)——SNTP功能和lwiperf测速
  • 第一章 空间解析几何与向量代数 ~ 空间直角坐标系
  • 【Fifty Project - D35】
  • 在线学堂-第二章媒资管理模块上
  • 高效清理C盘
  • Quick BI 自定义组件开发 -- 第一篇 Lifecycle 接口的定义
  • esp_image: invalid segment length 0xffffffff
  • MySQL自定义函数零基础学习教程
  • FastAPI 与 JWT 身份验证:保护你的 API
  • SpringBoot配置最新的AI版本加入Maven的配置方式
  • CDBench论文精读
  • 树莓派4B, ubuntu20.04, 安装Ros Noetic[踩坑记录]
  • 当拼音文字遇上回文:英语中的诗意镜像与文化密码
  • Profinet转CAN网关如何实现profinet与can协议互转
  • 如何通过API接口获取淘宝商品列表?操作详解
  • Quick BI 自定义组件开发 -- 第二篇 添加 echart 组件,开发图表
  • Spring AMQP
  • 打造高效能技术组织的逆向法则
  • 解读新交规中关于“电动自行车能否在快车道骑行”的核心问题
  • Shellshock漏洞与永恒之蓝(WannaCry)勒索病毒深度分析
  • [大A量化专栏] 看盘界面设置(未完待续)
  • Linux进程信号(一)
  • AI Bot到底是真助手,还是又一个流量收割伎俩?
  • 软件功能测试有哪些类型?软件测评机构
  • CppCon 2015 学习:The Importance of Being const
  • 鸠摩搜书官网入口,免费电子书小说在线搜索下载网站
  • 火山 RTC 引擎10 ----远端视频 转网易视频格式
  • 镜头景深的影响因素有哪些
  • 【西门子杯工业嵌入式-7-OLED】
  • 高防CDN是什么?和传统CDN有什么区别?