各种傅里叶

连续时间信号

傅里叶级数 CFS:

正变换:$F_n = \frac{1}{T} \int_0 ^T f(t) e^{-jn \omega_0 t} \mathrm{d} t$
基频:$F_0 = a_0 = \frac{1}{T}\int_0^T f(t) \mathrm{d} t$
逆变换:$f(t) = \sum_{n=-\infty}^{\infty}F_n e^{jn\omega_0 t}$

傅里叶变换 CFT:

$\text{正变换:}F(\omega) = \int_{-\infty}^\infty f(t) e^{-j\omega t} \mathrm{d}t$
$\text{逆变换:}f(t) = \frac{1}{2\pi}\int_{-\infty}^\infty F(\omega) e^{j\omega t}\mathrm{d} \omega$

常用变换公式:

$$\begin{equation}\begin{aligned}
\mathrm{rect}(\frac{t}{T}) = u(t + \frac{T}{2}) - u(t - \frac{T}{2}) &\leftrightarrow T \mathrm{sinc}(\frac{T\omega}{2}) \\
e^{-\alpha t} &\leftrightarrow \frac{1}{\alpha + j\omega} \\
\delta(t) &\leftrightarrow 1 \\
\delta^\prime (t) &\leftrightarrow j\omega \\
1 &\leftrightarrow 2\pi \delta(\omega) \\
\sum_{k=-\infty}^{\infty}\delta(t-kT) &\leftrightarrow \frac{2\pi}{T}\sum_{k=-\infty}^{\infty}\delta(\omega-k\frac{2\pi}{T}) \\
\mathrm{sgn}(t) = u(t) - u(- t) &\leftrightarrow \frac{2}{j\omega} \\
u(t) &\leftrightarrow \pi \delta(\omega) + \frac{1}{j\omega} \\
\cos(\omega_0 t) &\leftrightarrow \pi[\delta(\omega - \omega_0) + \delta(\omega + \omega_0)] \\
\sin(\omega_0 t) &\leftrightarrow \frac{\pi}{j}[\delta(\omega - \omega_0) - \delta(\omega + \omega_0)]
\end{aligned}\end{equation}$$

CFT 性质:

$$\begin{equation}\begin{aligned}
F(t) &\leftrightarrow 2\pi f(-\omega) \\
f(at) &\leftrightarrow \frac{1}{|a|}F(\frac{\omega}{a}) \\
f_1(t) \ast f_2(t) &\leftrightarrow F_1(\omega) \cdot F_2(\omega)
\\ f_1(t) \cdot f_2(t) &\leftrightarrow \frac{1}{2\pi}[F_1(\omega) \ast F_2(\omega)]
\end{aligned}\end{equation}$$

ps: 这一块容易出现卷积积分的求解,

一些简化计算的方法:

利用时移性质:

若 $x_1(t) \ast x_2(t) = v(t)$,
则:$x_1(t - t_1) \ast x_2(t - t_2) = v(t - t_1 - t_2)$


:若 $x(t) = u(t + 4) - u(t + 2) + u(t - 2) - u(t - 4)$,求 $y(t) = x(t)\ast x(t)$,$x(t)$ 图像:

:$x(t)$ 是 $u(t)$ 的线性组合,所以是 LTI 信号。可以用时移性质简化计算:
$$\begin{align}
y(t) &= x(t)\ast x(t) \\
&= [u(t + 4) - u(t + 2) + u(t - 2) - u(t - 4)] \ast [u(t + 4) - u(t + 2) + u(t - 2) - u(t - 4)] \\
&= u(t + 4) \ast u(t + 4) - u(t + 2) \ast u(t + 4) + u(t - 2) \ast u(t + 4) + \cdots \quad\quad \text{(卷积的结合律)} \\
&= (t + 8)u(t + 8) - (t + 6)u(t + 6) + (t + 2)u(t + 2) + \cdots\quad(\text{应用 }u(t)\ast u(t) = tu(t)) \\
\end{align}
$$
最后得到结果:
$y(t) = (t + 8)u(t + 8) - 2(t + 6)u(t + 6) + (t+ 4)u(t + 4) + 2(t + 2)u(t + 2) - 4tu(t) + 2(t - 2)u(t - 2) + (t - 4)u(t - 4) - 2(t - 6)u(t - 6) + (t - 8)u(t - 8)$


此外还有常用的:$\mathrm{rect}(t)\ast \mathrm{rect}(t)=(t+1)u(t+1)-2tu(t)+(t-1)u(t-1)$

利用微、积分性质求解:

$$\begin{align}
f_1(t) \ast f_2(t) &= \frac{\mathrm{d}}{\mathrm{d}t} \int_{-\infty}^t [f_1(\tau) \ast f_2(\tau)]\mathrm{d}\tau \\
&= f_1 ^\prime (t) \ast \int_{-\infty}^t f_2(\tau)\mathrm{d} \tau
\end{align}$$
解题时选两个信号中一阶导数简单的那个求导,就可以间接解题。

离散时间信号

##傅里叶级数 DFS:

正变换:$c_k = \frac{1}{N}\sum_{n=}x[n]e^{-jk\frac{2\pi}{N}n}$ ,
逆变换:$x[n] = \sum_{k=}c_k e^{jk\frac{2\pi}{N}n}$