mirror of
https://github.com/tu-darmstadt-informatik/stosigsysfs.git
synced 2025-12-13 10:25:49 +00:00
Konvergenz
This commit is contained in:
parent
a368522ce5
commit
14585b90d0
BIN
document.pdf
BIN
document.pdf
Binary file not shown.
68
document.tex
68
document.tex
@ -151,6 +151,9 @@ Seien $X_i$ Zufallsvariablen
|
||||
\E\left[\sum\limits_{i=0}^n X_i\right]=\sum\limits_{i=0}^n \E[X_i]
|
||||
\end{equation}
|
||||
\section{Varianz}
|
||||
|
||||
|
||||
|
||||
\subsection{Berechnung der Varianz}
|
||||
\begin{equation}
|
||||
\Var(X)=\E(X^2)-\E(X)^2
|
||||
@ -163,6 +166,38 @@ Seien $X_i$ Zufallsvariablen
|
||||
\begin{equation}
|
||||
\Var\left[\sum\limits_{i=0}^n X_i\right]=\sum\limits_{i=0}^n \Var[X_i]
|
||||
\end{equation}
|
||||
|
||||
|
||||
|
||||
\section{Konvergenz}
|
||||
Es wird eine Konvergenz von Zufallsvariablen $X_k$ mit $k=0,1,2 \dots$ betrachtet:
|
||||
\subsection{Konvergenz mit Wahrscheinlichkeit eins (Convergence with probability one)} \label{conv:one}
|
||||
\begin{equation}
|
||||
P \left(\lim_{k\rightarrow\infty} |X_k-X|=0\right)=1
|
||||
\end{equation}
|
||||
|
||||
\subsection{Konvergenz im ``Mean Square Sense''} \label{conv:mss}
|
||||
\begin{equation}
|
||||
\lim_{k\rightarrow\infty} \E\left[|X_k-X|^2\right]=0
|
||||
\end{equation}
|
||||
|
||||
\subsection{Convergence in Pobability} \label{conv:prob}
|
||||
\begin{equation}
|
||||
\lim_{k\rightarrow\infty}P \left( |X_k-X|>\epsilon\right)=0
|
||||
\end{equation}
|
||||
|
||||
\subsection{Convergence in Distribution} \label{conv:dist}
|
||||
\begin{equation}
|
||||
\lim_{k\rightarrow\infty}F_{X_k} (x)=F_X(x) \quad \text{Für alle stetigen punkte $x$ aus } F_X
|
||||
\end{equation}
|
||||
|
||||
\subsection{Gewichtung der Konvergenzen}
|
||||
\begin{itemize}
|
||||
\item Convergence with probability 1 (\ref{conv:one}) implies convergence in probability (\ref{conv:prob})
|
||||
\item Convergence with probability 1 (\ref{conv:one}) implies convergence in the MSS (\ref{conv:mss}), provided second order moments exist.
|
||||
\item Convergence in the MSS (\ref{conv:mss}) implies convergence in probability (\ref{conv:prob}).
|
||||
\item Convergence in probability (\ref{conv:prob}) implies convergence in distribution (\ref{conv:dist}).
|
||||
\end{itemize}
|
||||
\chapter{Discrete-Time-Fourier-Transformation}
|
||||
|
||||
\section{Abtastung}
|
||||
@ -430,6 +465,39 @@ C_{XX}(e^{j \omega})=\sum\limits_{n=-\infty}^\infty c_{xx}(n)e^{-j\omega n}
|
||||
\end{equation}
|
||||
|
||||
|
||||
\subsubsection{Eigenschaften des Spektrums}
|
||||
\begin{enumerate}
|
||||
\item Wenn $\sum_n |c_{XX}(n)|<\infty$, dann existiert $C_{XX}$ und ist begrenzt und stetig
|
||||
\item $C_{XX}$ ist Real, $2\pi$-Periodisch und $C_{XX}\geq0$
|
||||
\item \begin{equation}
|
||||
c_{XX}(n)=\frac{1}{2\pi}\int_{-\pi}^{\pi}C_{XX}(e^{j \omega})e^{j \omega n}d\omega
|
||||
\end{equation}
|
||||
\end{enumerate}
|
||||
|
||||
\subsection{Kreuzspektrum zweier gemeinsam stationärer Zufallsprozesse}
|
||||
Ist $X(n)$ und $Y(n)$ \emph{gemeinsam stationär} (\ref{jointstationary}), dann ist das Kreuzspektrum definiert durch
|
||||
\begin{equation}
|
||||
C_{XY}(e^{j \omega})=\sum\limits_{n=-\infty}^\infty c_{XY}(n)e^{-j \omega n}
|
||||
\end{equation}
|
||||
|
||||
\subsubsection{Eigenschaften der Kreuzspektrums}
|
||||
Das Spektrum eines Realen Zufallsprozesses ist komplett im Intervall $[0,\pi]$ bestimmt
|
||||
\begin{subequations}
|
||||
\begin{align}
|
||||
C_{XY}(e^{j \omega})&=C_{YX}(e^{j \omega})^* \\
|
||||
c_{XY}(n) &= \frac{1}{2\pi}\int\limits_{-\pi}^\pi C_{XY}(e^{j \omega})e^{j \omega n} d\omega \\
|
||||
\text{Wenn } &X(n),Y(n) \in \mathbb{R} \text{ dann} \notag\\
|
||||
C_{XX}(e^{j \omega}) &= C_{XX}(e^{-j \omega})\\
|
||||
C_{XY}(e^{j \omega}) =C_{XY}(e^{-j \omega})^*&=C_{YX}(e^{-j \omega})=C_{YX}(e^{j \omega})^*
|
||||
\end{align}
|
||||
\end{subequations}
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
\chapter{Sonstiges}
|
||||
\section{Spezielle Funktionen}
|
||||
\subsection{Gaussian white noise process}
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user