Partial differential equation

Japanese: 偏微分方程式 - へんびぶんほうていしき(英語表記)partial differential equation
Partial differential equation

A relational equation between two or more independent variables and an unknown function and its partial derivative is called a partial differential equation. The highest order of the partial derivatives contained in the relational equation is called the order of the equation. A partial differential equation that is a first-order equation with respect to the unknown function and each of its partial derivatives is said to be linear, and if it is not, it is said to be nonlinear. A partial differential equation that is a first-order equation with respect to the highest-order partial derivative is said to be quasi-linear. When the independent variables are x and y , a first-order quasi-linear equation is a ( x , y , u ) u x + b ( x , y , u ) u y = c ( x , y , u ) (1)
where u is the unknown function,

where a , b , and c are known functions. Similarly, the second-order linear equation is a ( x , y ) u xx +2 b ( x , y ) u xy
+ c ( x , y ) uyy
= d ( x , y ) u x + e ( x , y ) y
+ f ( x , y ) u + g ( x , y ) (2)
where a , b , c , etc. are known functions,

A function that satisfies a partial differential equation identically is called a solution to that equation, and finding the solution to a partial differential equation is called solving the equation.

Consider the first-order quasi-linear equation (1). The problem of finding a solution to (1) that is equal to a given function at each point on the curve P in the x - y plane is called the general initial value problem for (1). For (1), consider the simultaneous ordinary differential equations x ′= a ( x , y , u ), with unknown functions x , y , and u .
y ′= b ( x , y , u ),
u ′= c ( x , y , u )
is called the characteristic differential equation of (1), and the curve in ( x , y , u ) space given by its solution is called the characteristic curve of (1). The surface u = u ( x , y ) represented by the solution of (1) is equal to the surface generated by the characteristic curve. Therefore, if the projection of the characteristic curve onto the x - y plane intersects the initial curve P without being tangent, the solution of the general initial value problem exists uniquely in the neighborhood of P , and the solution can be obtained by solving the characteristic differential equation. A general initial value problem for a nonlinear first-order partial differential equation, which is not necessarily quasi-linear, can also be reduced to an initial value problem for an ordinary differential equation under appropriate conditions on the initial function and initial curve.

The generalized initial value problem for a second-order linear equation (2) is the problem of finding a solution such that at each point on a given curve P, a given function is equal, including its normal derivatives.

Now, let the equation of curve P be ( x , y )=0. If,
Q ()≡ a ( x , y ) x 2 +2 b ( x , y ) x y + c ( x , y ) y 2 =0
If Q ()≠0 always holds, then P is called the characteristic curve of (2). If the initial curve, initial function, and coefficient functions a , b , and c are analytical and the initial curve is noncharacteristic, then there exists a unique real analytical solution of (2) in the vicinity of P. This also holds for higher-order nonlinear partial differential equations (Cauchy–Kowalewski theorem).

For the wave equation u xx - u yy = 0, the curve y = 0 is non-characteristic. In this case, a solution exists even if the initial function is not analytical, and the solution depends continuously on the initial function.

Laplace's equation u xx + u yy =0
For y = 0, the solution is not continuously dependent on the initial function, and no solution may exist if the initial function is not analytic. In general, when a problem is solved by imposing additional conditions on a partial differential equation, if a solution exists uniquely for the additional conditions and if the solution depends continuously on the additional conditions, the problem is said to be well-formed.

In equation (2), "if ac - b 2 > 0, it is elliptic; if ac - b 2 = 0, it is parabolic; if ac - b 2 < 0, it is hyperbolic." For hyperbolic equations, initial value problems are appropriate. For elliptic equations, boundary value problems are appropriate. Second-order linear partial differential equations are important in mathematical physics and have been studied in detail.

[Yoshikazu Kobayashi]

[Reference] | Boundary value problems | Heat equation | Wave equation | Laplace equation

Source: Shogakukan Encyclopedia Nipponica About Encyclopedia Nipponica Information | Legend

Japanese:

2個以上の独立変数と未知関数およびその偏導関数の間の関係式を偏微分方程式という。関係式に含まれる偏導関数の最高階数をその方程式の階数という。未知関数とその各偏導関数について一次式であるような偏微分方程式は線形であるといわれ、そうでないとき非線形であるといわれる。また、最高階の偏導関数について一次式であるような偏微分方程式は準線形であるといわれる。独立変数がxyのとき、一階準線形方程式は
  a(x,y,u)ux+b(x,y,u)uy=c(x,y,u)(1)
と書くことができる。ここでuは未知関数で、

であり、abcは既知関数である。同様に、二階線形方程式は
  a(x,y)uxx+2b(x,y)uxy
    +c(x,y)uyy
   =d(x,y)ux+e(x,y)y
    +f(x,y)u+g(x,y)     (2)
と書くことができる。ここでabcなどは既知関数で、

などである。偏微分方程式を恒等的に満たす関数をその方程式の解といい、偏微分方程式の解を求めることをその方程式を解くという。

 一階準線形方程式(1)を考える。x-y平面上の曲線P上の各点で与えられた関数に等しい(1)の解を求める問題を、(1)に対する一般初期値問題という。(1)に対し、xyuを未知関数とする連立常微分方程式
  x′=a(x,y,u),
  y′=b(x,y,u),
  u′=c(x,y,u)
を(1)の特性微分方程式といい、その解で与えられる(x,y,u)空間の曲線を(1)の特性曲線という。(1)の解が表す曲面u=ux,y)は特性曲線により生成される曲面に等しい。したがって、特性曲線のx-y平面への射影が初期曲線Pに接しないで交わるならば、一般初期値問題の解はPの近傍で一意的に存在し、解は特性微分方程式を解くことにより得られる。かならずしも準線形とは限らない非線形一階偏微分方程式に対する一般初期値問題も、初期関数と初期曲線に対する適当な条件の下で、常微分方程式の初期値問題に帰着される。

 二階線形方程式(2)に対する一般初期値問題は、与えられた曲線P上の各点で与えられた関数に、その法線導関数を含めて等しくなるような解を求める問題である。

 いま、曲線Pの方程式を(x,y)=0とする。もし、
  Q()≡a(x,y)x2+2b(x,y)xy+c(x,y)y2=0
が成り立つならばPを(2)の特性曲線といい、またつねにQ()≠0が成り立つならばPを非特性な曲線という。初期曲線、初期関数、係数関数abcなどが解析的で、初期曲線が非特性ならばPの近傍で(2)の実解析的な解が一意的に存在する。これは高階非線形偏微分方程式に対しても成り立つ(コーシー‐コワレフスキーの定理)。

波動方程式uxx-uyy=0に対し、曲線y=0は非特性である。この場合、初期関数が解析的でなくとも解が存在し、解は初期関数に連続的に依存する。

ラプラスの方程式
  uxx+uyy=0
に対し、y=0は非特性であるが、解は初期関数に連続的に依存しないし、また初期関数が解析的でないとき解が存在しないことがある。一般に偏微分方程式に対し付加条件を課して解を求める問題について、付加条件に対し解が一意的に存在して、さらに解が付加条件に連続的に依存するとき、この問題は適切であるといわれる。

 方程式(2)において「ac-b2>0ならば楕円(だえん)型、ac-b2=0ならば放物型、ac-b2<0ならば双曲型である」という。双曲型方程式に対しては初期値問題が適切である。楕円型方程式に対しては境界値問題が適切である。二階線形偏微分方程式は数理物理学において重要であり、詳しく調べられている。

[小林良和]

[参照項目] | 境界値問題 | 熱方程式 | 波動方程式 | ラプラス方程式

出典 小学館 日本大百科全書(ニッポニカ)日本大百科全書(ニッポニカ)について 情報 | 凡例

<<:  Penfield, Wilder-Graves

>>:  Partial Differentiation

Recommend

Influenza bacillus - Influenza bacillus

...Short bacilli are called short bacilli, and lo...

"Complete medical texts"

…The Asaino family was a prominent family in Saka...

Ippei Okamoto

Manga artist. Born in Hakodate, Hokkaido. Grew up...

Aluminothermy process

...A method of reducing metal oxides using the la...

personal social services

...However, from the late 1960s onwards, the prob...

yajé (English spelling) yaje

...However, several plants are known from the Old...

Haute-Volta (English spelling)

The former name of Burkina Faso, a landlocked cou...

Differential geometry

In the classical sense, differential geometry is ...

homestead

…This is almost the same as in Germany. (1) Farms...

Bank for International Settlements

…Abbreviation for Bank for International Settleme...

Rapatea paludosa (English spelling) Rapatea paludosa

… [Mitsuru Hotta]... *Some of the terminology tha...

Ichthyornis

…Archaeopteryx [Appearance of new birds] No bird ...

Rose moss (English spelling) rosemoss

…It is also called 'Hiderisou' (Sunshine ...

spike-wave stupor

…Recently, it has become relatively easy to treat...

Kushida Shrine (Fukuoka) - Kushida Shrine

…Several yearly gyoji were in charge of the town ...