Inhaltsverzeichnis

Wichtige mathematische Anwendungen

Hier werden die mathematischen Anwendungen beschrieben, die nicht als Funktionen in Taschenrechnern und Computerprogrammen zu finden sind. Diese Liste wird laufend ergänzt.

Lineare Gleichungssysteme

Hat man drei Gleichungen in der Form \[\begin{align} b_1 &= a_{11} \cdot x + a_{12} \cdot y + a_{13} \cdot z \\ b_2 &= a_{21} \cdot x + a_{22} \cdot y + a_{23} \cdot z \\ b_3 &= a_{31} \cdot x + a_{32} \cdot y + a_{33} \cdot z \end{align}\tag{1}\]

so kann man diese leicht als Matrix und Vektor darstellen: \[\left(\begin{matrix} b_1 \\ b_2 \\ b_3 \end{matrix}\right) = \left(\begin{array}{ccc} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{array}\right) \cdot \left(\begin{matrix} x \\ y \\ z \end{matrix}\right)\tag{2}\]

Die Multiplikation erfolgt mit der folgenden graphischen Rechenregel:

Abb. 1: Drehmatrix um die x - Achse

Vektoren

Aus den drei Gleichungen \[\begin{align} x &= r\cdot \cos(\beta)\cdot \cos(\alpha)\\ y &= r\cdot \cos(\beta)\cdot \sin(\alpha)\\ z &= r\cdot \sin(\beta)\end{align}\tag{3}\]

formt man den Vektor \[\vec{r} = r \left(\begin{matrix} \cos(\beta)\cdot \cos(\alpha)\\ \cos(\beta)\cdot \sin(\alpha)\\ \sin(\beta)\end{matrix}\right) = r \vec{e}\tag{4}\]

Skalarprodukt (Punktprodukt)

$$\vec{v} \cdot \vec{u} = v_1 \cdot u_1 + v_2 \cdot u_2 + + v_3 \cdot u_3\tag{5}$$

Rechenregeln: $$\vec{v} \cdot \vec{u} = \vec{u} \cdot \vec{v}\tag{6}$$ $$\vec{u} \cdot \vec{u} = |u|^2\tag{7}$$ $$\vec{u} \cdot \vec{v} = |\vec{u}||\vec{v}| \cos(\theta)\tag{8}$$

senkrecht $\theta = 90^\circ$: $$\vec{u} \cdot \vec{v} = 0\tag{9}$$

parallel $\theta = 0^\circ$: $$\vec{u} \cdot \vec{v} = |\vec{u}||\vec{v}|\tag{10}$$

Vektorprodukt (Kreuzprodukt)

$$\vec{v} \times \vec{u} = \left(\begin{array} vv_2 \cdot u_3 - v_3 \cdot u_2 \\ v_3 \cdot u_1 - u_3 \cdot v_1 \\ v_1 \cdot u_2 - u_1 \cdot v_2 \end{array}\right)\tag{11}$$

Rechenregeln: $$\vec{v} \times \vec{u} = - \vec{u} \times \vec{v}\tag{12}$$ $$\vec{u} \times \vec{u} = 0\tag{13}$$ $$\vec{u} \times \vec{v} = |\vec{u}||\vec{v}| \sin(\theta)\tag{14}$$

senkrecht $\theta = 90^\circ$: $$\vec{u} \times \vec{v} = |\vec{u}||\vec{v}|\tag{15}$$

parallel $\theta = 0^\circ$: $$\vec{u} \times \vec{v} = 0\tag{16}$$

Matrizen

Drehmatrizen

Abb. 2: passive Drehung

Ein Vektor wird mit Hilfe einer Drehmatrix in eine neues Koordinatensystem (i.e. neuer Vektor) gedreht (Abb. 1). \[\begin{split} & x = \cos(\alpha) \\ & y = \sin(\alpha) \\ & z = 0 \end{split} \Rightarrow\vec{r}(\alpha,0) = \left( \begin{matrix} x \\ y \\ z \end{matrix} \right) = \left( \begin{matrix} \cos(\alpha) \\ \sin(\alpha) \\ 0 \\ \end{matrix} \right) = \vec{e}(\alpha,0)\tag{17}\]

Nach einer Drehung mit dem Winkel $\varphi$ um die z - Achse erhält man: \[\vec{r} = \left( \begin{matrix} x' \\ y' \\ z' \end{matrix} \right) = \left( \begin{matrix} \cos(\alpha') \\ \sin(\alpha') \\ z \end{matrix} \right) = \left( \begin{matrix} \cos(\alpha - \varphi) \\ \sin(\alpha - \varphi) \\ z \end{matrix} \right) = \left( \begin{array}{lll} + x \ \cos(\varphi) & + y \ \sin(\varphi) & + z \ 0 \\ - x \ \sin(\varphi) & + y \ \cos(\varphi) & + z \ 0 \\ + x \ 0 & + y \ 0 & + z \ 1 \end{array} \right)\tag{18}\]

\[\Rightarrow \vec{r}'(\alpha',0) = \left( \begin{matrix} x' \\ y' \\ z' \end{matrix} \right) = \left( \begin{array}{ccc} + \cos(\varphi) & + \sin(\varphi) & 0 \\ - \sin(\varphi) & + \cos(\varphi) & 0 \\ 0 & 0 & 1 \end{array} \right) \left( \begin{matrix} x \\ y \\ z \end{matrix} \right) = R_3(\varphi) \vec{r}(\alpha,0)\tag{19}\]

Die Drehungen um die x (R$_1$), y (R$_2$) und die z - Achse (R$_3$): \[R_1(\varphi) = \left( \begin{array}{ccc} 1 & 0 & 0 \\ 0 & + \cos(\varphi) & + \sin(\varphi) \\ 0 & - \sin(\varphi) & + \cos(\varphi) \end{array} \right) \tag{20}\]

\[R_2(\varphi) = \left( \begin{array}{ccc} + \cos(\varphi) & 0 & - \sin(\varphi) \\ 0 & 1 & 0 \\ + \sin(\varphi) & 0 & + \cos(\varphi) \end{array} \right) \tag{21}\]

\[R_3(\varphi) = \left( \begin{array}{ccc} + \cos(\varphi) & + \sin(\varphi) & 0 \\ - \sin(\varphi) & + \cos(\varphi) & 0 \\ 0 & 0 & 1 \end{array} \right)\tag{22}\]

Determinante

Die Determinante ist der skalare Wert einer Matrix. Die Regel von Sarrus berechnet die Determinante (Abb. 2):

Abb. 3: Regel von Sarrus

\[\det(\mathbf{M}) = \left| \begin{array}{ccc} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{array} \right| = \begin{array}{c} + a_{11} \ a_{22} \ a_{33} + a_{12} a_{23} \ a_{31} + a_{13} \ a_{21} \ a_{32} \\ - a_{31} \ a_{22} \ a_{13} - a_{32} \ a_{23} \ a_{11} - a_{33} \ a_{21} \ a_{12} \end{array}\tag{23}\]

Invertierte Matrix

Die Invertierung (Umkehrung) der Matrix M erfolgt mit Hilfe der Cramerschen Regel: Die Bildung einer Unterdeterminante N$_{ik}$ durch das Herausstreichen der i - ten Spalte und k - ten Zeile. \[\det(N_{ik}) = (-1)^{i+k} \mathbf{M} \quad und \quad b_{ik} = \frac{\det(N_{ik})}{\det(M)}\tag{24}\]

det($N_{ik}$) ist eine Streich- oder Unterdeterminante von M. Sie wird aus M gebildet, wenn die i - te Zeile und die k - te Spalte gestrichen wurde. Bei einer 3 $\times$ 3 Matrix sind es neun Unterdeterminanten. Die invertierte Matrix ist dann: \[M^{-1} = \frac{1}{\det(M)} \left(\begin{array}{ccc} b_{11} & b_{21} & b_{13} \\ b_{12} & b_{22} & b_{32} \\ b_{31} & b_{23} & b_{33} \end{array}\right)\tag{25}\]

Für die oben genannten Drehmatrizen gilt $R^{-1}(\varphi) = R(-\varphi)$.

Transponierte Matrix

Hat man eine Matrix M \[\mathbf{M} = \left( \begin{array}{ccc} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{array} \right)\tag{26}\]

so ist die transponierte Matrix $M^t$: \[\mathbf{M}^t = \left( \begin{array}{ccc} a_{11} & a_{21} & a_{31} \\ a_{12} & a_{22} & a_{32} \\ a_{13} & a_{23} & a_{33} \end{array} \right)\tag{27}\]

Es vertauschen sich alle Elemente bis auf die Diagonalelemente gegenseitig. Überträgt man die transponierte Matrix $M^t$ auf die Drehmatrix $R(\varphi)$, so zeigt sich, daß $R^t(\varphi)$ = $R(-\varphi)$ gilt.

Interpolation

Taylorpolynom

Das Taylorpolynom (auch Taylorreihe/-entwicklung) ist die beliebteste Anwendung zur Näherung von Funktionen an der Stelle $r_0$ mit $|\vec{r} - \vec{r}_0| \ll 1$. \[f_{r_0}(\vec{r}) = f(\vec{r}_0) + f'(\vec{r}_0) \cdot (\vec{r} - \vec{r}_0) + \frac{1}{2} \cdot f''(\vec{r}_0) \cdot (\vec{r} - \vec{r}_0)^2 + \cdots = \sum_{k = 0}^{\infty} \frac{1}{k!} \cdot \frac{d^k}{d\vec{r}_0^k} f(\vec{r}_0) \cdot(\vec{r} - \vec{r}_0)^k\tag{28}\]

Ein Rechenbeispiel sind die trigonometrischen Funktionen und die e - Funktion: \[\begin{split} e^{\pm\varphi} &= 1 \pm \frac{\varphi}{1!} - \frac{\varphi^2}{2!} \mp \frac{\varphi^3}{3!} + \frac{\varphi^4}{4!} \pm \frac{\varphi^5}{5!} - ...= \sum_{n=0}^\infty \frac{\varphi^n}{n!} \\ \cos(\varphi) &= 1 - \frac{\varphi^2}{2!} + \frac{\varphi^4}{4!} - ... = \sum_{n=0}^\infty \frac{(-1)^n \varphi^{2n}}{(2n)!} \\ \sin(\varphi) &= \varphi - \frac{\varphi^3}{3!} + \frac{\varphi^5}{5!} - ... = \sum_{n=0}^\infty \frac{(-1)^n \varphi^{2n+1}}{(2n+1)!} \end{split}\tag{29}\]

Stirling-Interpolation

$x_1 = x_0 - 2\cdot h$ $y_1 = f(x_0 - 2\cdot h)$
$\nabla^{1} f\left(x_0 - \frac{3}{2}\cdot h\right)$
$x_2 = x_0 - h$ $y_2 = f(x_0 - h)$ $\nabla^{2} f(x_0 - h)$
$x = x_0 - t\cdot h$ $\nabla^{1} f\left(x_0 - \frac{1}{2}\cdot h\right)$ $\nabla^{3} f\left(x_0 - \frac{1}{2}\cdot h\right)$
$x_3 = x_0$ $y_3 = f(x_0)$ $\delta^{2} f(x_0)$ $\delta^{4} f(x_0)$
$x = x_0 + t\cdot h$ $\Delta^{1} f\left(x_0 + \frac{1}{2}\cdot h\right)$ $\Delta^{3} f\left(x_0 + \frac{1}{2}\cdot h\right)$
$x_4 = x_0 + h$ $y_4 = f(x_0 + h)$ $\Delta^{2} f(x_0 + h)$
$\Delta^{1} f\left(x_0 + \frac{3}{2}\cdot h\right)$
$x_5 = x_0 + 2\cdot h$ $y_5 = f(x_0 + 2\cdot h)$

$\Delta f(x_0) = f(x_0 + t \ h) - f(x_0)$ bilden die Vorwärtsdifferenzen und $\nabla f(x_0) = f(x_0) - f(x_0 - t \ h)$ sind die Rückwärtsdifferenzen. Die Hochzahlen bilden die Ordnungen; es sind keine Potenzen. Das gleiche gilt auch für die Zentraldifferenzen $\delta f(x_0) = f(x_0 + t \ h) - f(x_0 - t \ h)$.

Die entsprechende Interpolationsgleichung lautet $$f(x_0 \pm t h) = f(x_0) \pm t \cdot \delta^1 f(x_0) + \frac{t^2}{2} \cdot \delta^2 f(x_0) \pm \frac{(t + 1) t (t - 1)}{6} \cdot \delta^3 f(x_0) + \frac{(t + 1) t^2 (t - 1)}{24} \cdot \delta^4 f(x_0) \pm \cdots\tag{30}$$

mit \[\begin{align} \delta^1 f(x_0) &= \frac{1}{2} (\nabla^1 f(x_0 - \frac{1}{2} h) + \Delta^1 f(x_0 + \frac{1}{2} h)) \\ \delta^3 f(x_0) &= \frac{1}{2} (\nabla^3 f(x_0 - \frac{1}{2} h) + \Delta^3 f(x_0 + \frac{1}{2} h)) \end{align}\tag{31}\]

Eine Anwendungsmöglichkeit ist die Bestimmung der mittleren Bahnelemente aus den oskulierenden Bahnelementen. Diese sind jedoch nur innerhalb des Interpolationszeitraums gültig.

weitere Interpolationsverfahren

Die Interpolation hat aufgrund ihres Umfangs eine eigene Seite erhalten.

Komplexe Zahlen

Komplexe Zahlen entstehen, wenn Quadratwurzeln negative Argumente bei der Berechnung erhalten. Diese werden dann mit $i = \sqrt{-1}$ dargestellt. Eine komplexe Zahl $z$ setzt sich aus einem imaginären Anteil $b$ und einem realen Anteil $a$ zusammen: $$z = a + i\cdot b \qquad\text{mit}\qquad i = \sqrt{-1}\tag{32}$$

Für $i$ gilt:

Tritt die Variante $$z^{*} = a - i\cdot b\tag{33}$$

auf, so ist diese konjungiert komplex zu $z$.

$$\text{real:} \qquad (a + b)(a - b) = a^2 + ab - ba - b^2 = a^2 - b^2\tag{34}$$ $$\text{komplex:} \qquad (a + ib)(a - ib) = a^2 + a ib - ib a - i^2b^2 = a^2 + b^2\tag{35}$$

Es gilt $z\cdot z^{*} = a^2 + b^2$, eine reelle Zahl.

Abb. 4: komplexe Zahlen

Eine anschauliche Darstellung der komplexen Zahlen im Koordinatenkreuz. $z$ kann als ein um den Ursprung kreisenden Radius vorgestellt werden. $a$, $b$ und $z$ bilden die Seiten eines rechtwinkeligen Dreiecks. Die konjugiert komplexe Form ist die Spiegelung an der reellen Achse.