①线性特性
②(因果信号的)延时特性
③尺度变换
④复频移性
⑤时域微积分
⑥时/频域的卷积定理
1 线性性质
{ f 1 ( t ) ↔ X 1 ( s ) , f 2 ( t ) ↔ X 2 ( s ) a f 1 ( t ) + b f 2 ( t ) ↔ a X 1 ( s ) + b X 2 ( s ) \begin{cases} f_1(t)\leftrightarrow X_1(s),f_2(t)\leftrightarrow X_2(s)\\ af_1(t)+bf_2(t)\leftrightarrow aX_1(s)+bX_2(s) \end{cases} {f1(t)↔X1(s),f2(t)↔X2(s)af1(t)+bf2(t)↔aX1(s)+bX2(s)
2 延时特性(特别:针对因果信号)
- 因果信号:有因才有果——系统没有 输入信号激励 的时候,必不可能有 输出信号响应。若系统在 t = 0 t=0 t=0 时接入输入信号激励,在 t < 0 t<0 t<0 时段必不能存在激励 f ( t ) ≠ 0 f(t)\neq 0 f(t)=0 的情况
- 这就意味着,使用这个性质必须是右边信号(即 f ( t ) f(t) f(t) 中必须包含如 u ( t ) u(t) u(t) 类的阶跃信号)
- 同时时移只能是单方向的延时(右移),即必须保证实常数 t 0 > 0 t_0>0 t0>0
- 其中,奥本海默《信号与系统》一书给出:
{ f ( t ) ↔ X ( s ) f ( t − t 0 ) ↔ e − s t 0 X ( s ) \begin{cases} f(t)\leftrightarrow X(s)\\ f(t-t_0)\leftrightarrow e^{-st_0}X(s)\\ \end{cases} {f(t)↔X(s)f(t−t0)↔e−st0X(s)
- 推导:
拟设定 f ( t ) f(t) f(t) 的定义域 t > 0 t>0 t>0
L [ f ( t ) ] = X ( s ) = ∫ 0 + ∞ f ( t ) ⋅ e − s t d t L[f(t)]=X(s)=\int_{0}^{+\infty}f(t)\cdot e^{-st}dt L[f(t)]=X(s)=∫0+∞f(t)⋅e−stdt
则 f ( t − t 0 ) f(t-t_0) f(t−t0) 的定义域 t > t 0 t>t_0 t>t0
L [ f ( t − t 0 ) ] = ∫ t 0 + ∞ f ( t − t 0 ) ⋅ e − s t d t = ∫ t 0 + ∞ f ( t − t 0 ) ⋅ e − s ( t − t 0 ) ⋅ e − s t 0 d ( t − t 0 ) L[f(t-t_0)]=\int_{t_0}^{+\infty}f(t-t_0)\cdot e^{-st}dt=\int_{t_0}^{+\infty}f(t-t_0)\cdot e^{-s(t-t_0)}\cdot e^{-st_0}d(t-t_0) L[f(t−t0)]=∫t0+∞f(t−t0)⋅e−stdt=∫t0+∞f(t−t0)⋅e−s(t−t0)⋅e−st0d(t−t0)
令 m = t − t 0 m=t-t_0 m=t−t0,则:
L [ f ( t − t 0 ) ] = e − s t 0 ∫ 0 + ∞ f ( m ) ⋅ e − s m d m L[f(t-t_0)]=e^{-st_0}\int_{0}^{+\infty}f(m)\cdot e^{-sm}dm L[f(t−t0)]=e−st0∫0+∞f(m)⋅e−smdm
用上面 L [ f ( t ) ] L[f(t)] L[f(t)] 中的公式,将 m m m 替换 t t t,即 ∫ 0 + ∞ f ( t ) ⋅ e − s t d t = ∫ 0 + ∞ f ( m ) ⋅ e − s m d m \int_{0}^{+\infty}f(t)\cdot e^{-st}dt=\int_{0}^{+\infty}f(m)\cdot e^{-sm}dm ∫0+∞f(t)⋅e−stdt=∫0+∞f(m)⋅e−smdm:
L [ f ( t − t 0 ) ] = e − s t 0 X ( s ) L[f(t-t_0)]=e^{-st_0}X(s) L[f(t−t0)]=e−st0X(s)
傅里叶变换有个与这个很像的性质叫时移性质:
{ f ( t ) ↔ F ( j w ) f ( t ± t 0 ) ↔ F ( j w ) e ± j w t 0 \begin{cases} f(t)\leftrightarrow F(jw)\\ f(t\pm t_0)\leftrightarrow F(jw)e^{\pm jwt_0}\\ \end{cases} {f(t)↔F(jw)f(t±t0)↔F(jw)e±jwt0
- 推导过程:
F ( j w ) = ∫ − ∞ + ∞ e − j w t f ( t ) d t F(jw)=\int^{+\infty}_{-\infty}e^{-jwt}f(t)dt F(jw)=∫−∞+∞e−jwtf(t)dt
F T [ f ( t ± t 0 ) ] = e ± j w t 0 ∫ − ∞ + ∞ f ( t ± t 0 ) e − j w ( t ± t 0 ) d ( t ± t 0 ) FT[f(t\pm t_0)]=e^{\pm jwt_0}\int^{+\infty}_{-\infty}f(t\pm t_0)e^{-jw(t\pm t_0)}d(t\pm t_0) FT[f(t±t0)]=e±jwt0∫−∞+∞f(t±t0)e−jw(t±t0)d(t±t0)
因为 ∫ − ∞ + ∞ f ( m ) e − j w m d m = ∫ − ∞ + ∞ e − j w t f ( t ) d t \int^{+\infty}_{-\infty}f(m)e^{-jwm}dm=\int^{+\infty}_{-\infty}e^{-jwt}f(t)dt ∫−∞+∞f(m)e−jwmdm=∫−∞+∞e−jwtf(t)dt,所以:
F T [ f ( t ± t 0 ) ] = F ( j w ) e ± j w t 0 FT[f(t\pm t_0)]=F(jw)e^{\pm jwt_0} FT[f(t±t0)]=F(jw)e±jwt0
例、求 f ( t ) = ( t − t 0 ) u ( t − t 0 ) f(t)=(t-t_0)u(t-t_0) f(t)=(t−t0)u(t−t0) 的拉普拉斯变换
其中:
y
(
t
)
=
t
u
(
t
)
↔
1
s
2
y(t)=tu(t)\leftrightarrow \frac{1}{s^2}
y(t)=tu(t)↔s21
则
y
(
t
−
t
0
)
=
(
t
−
t
0
)
u
(
t
−
t
0
)
↔
1
s
2
e
−
t
0
s
y(t-t_0)=(t-t_0)u(t-t_0)\leftrightarrow \frac{1}{s^2}e^{-t_0s}
y(t−t0)=(t−t0)u(t−t0)↔s21e−t0s
-
f
1
(
t
)
=
t
−
t
0
f_1(t)=t-t_0
f1(t)=t−t0
由于单边拉氏变换,得:
X 1 ( s ) = L [ ( t − t 0 ) u ( t ) ] = X 2 X_1(s)=L[(t-t_0)u(t)]=X_2 X1(s)=L[(t−t0)u(t)]=X2 -
f
2
(
t
)
=
(
t
−
t
0
)
u
(
t
)
f_2(t)=(t-t_0)u(t)
f2(t)=(t−t0)u(t)
X 2 ( s ) = L [ f 2 ( t ) ] = L [ t u ( t ) ] − L [ t 0 u ( t ) ] X_2(s)=L[f_2(t)]=L[tu(t)]-L[t_0u(t)] X2(s)=L[f2(t)]=L[tu(t)]−L[t0u(t)]
其中 L [ t u ( t ) ] = 1 s 2 L[tu(t)]=\frac{1}{s^2} L[tu(t)]=s21, L [ t 0 u ( t ) ] = t 0 L [ u ( t ) ] = t 0 s L[t_0u(t)]=t_0L[u(t)]=\frac{t_0}{s} L[t0u(t)]=t0L[u(t)]=st0,则:
X 2 ( s ) = 1 s 2 − t 0 s X_2(s)=\frac{1}{s^2}-\frac{t_0}{s} X2(s)=s21−st0 -
f
3
(
t
)
=
t
u
(
t
−
t
0
)
f_3(t)=tu(t-t_0)
f3(t)=tu(t−t0)
X 3 ( s ) = [ ( t − t 0 ) + t 0 ] u ( t − t 0 ) X_3(s)=[(t-t_0)+t_0]u(t-t_0) X3(s)=[(t−t0)+t0]u(t−t0)
其中 L [ ( t − t 0 ) u ( t − t 0 ) ] = 1 s 2 e − t 0 s L[(t-t_0)u(t-t_0)]=\frac{1}{s^2}e^{-t_0s} L[(t−t0)u(t−t0)]=s21e−t0s, L [ t 0 u ( t − t 0 ) ] = t 0 e − t 0 s L[t_0u(t-t_0)]=t_0e^{-t_0s} L[t0u(t−t0)]=t0e−t0s
X 3 ( s ) = 1 s 2 e − t 0 s + t 0 e − t 0 s X_3(s)=\frac{1}{s^2}e^{-t_0s}+t_0e^{-t_0s} X3(s)=s21e−t0s+t0e−t0s
3 尺度变换
{ f ( t ) ↔ X ( s ) f ( a t ) ↔ 1 a X ( s a ) , a > 0 \begin{cases} f(t)\leftrightarrow X(s)\\ f(at)\leftrightarrow \frac{1}{a}X(\frac{s}{a}),a>0\\ \end{cases} {f(t)↔X(s)f(at)↔a1X(as),a>0
4 复频移性
{ f ( t ) ↔ X ( s ) f ( t ) e ± s 0 t ↔ X ( s ∓ s 0 ) \begin{cases} f(t)\leftrightarrow X(s)\\ f(t)e^{\pm s_0t}\leftrightarrow X(s\mp s_0)\\ \end{cases} {f(t)↔X(s)f(t)e±s0t↔X(s∓s0)
5 时域微积分特性
5-1 时域微分特性
{
f
(
t
)
↔
X
(
s
)
d
d
t
[
f
(
t
)
]
↔
s
X
(
s
)
−
f
(
0
−
)
\begin{cases} f(t)\leftrightarrow X(s)\\ \frac{d}{dt}[f(t)]\leftrightarrow sX(s)-f(0_{-})\\ \end{cases}
{f(t)↔X(s)dtd[f(t)]↔sX(s)−f(0−)
拓展:
d
n
d
t
n
[
f
(
t
)
]
↔
s
n
X
(
s
)
−
∑
m
=
0
n
−
1
s
n
−
1
−
m
f
(
m
)
(
0
−
)
\frac{d^n}{dt^n}[f(t)]\leftrightarrow s^nX(s)-\sum_{m=0}^{n-1}s^{n-1-m}f^{(m)}(0_{-})
dtndn[f(t)]↔snX(s)−m=0∑n−1sn−1−mf(m)(0−)
居多使用的是:
f
′
′
(
t
)
↔
s
2
X
(
s
)
−
s
f
(
0
−
)
−
f
′
(
0
−
)
f''(t)\leftrightarrow s^2X(s)-sf(0_{-})-f'(0_{-})
f′′(t)↔s2X(s)−sf(0−)−f′(0−)
- 提到微分容易想起电容电感的VCR方程:
对电容 C C C: i C = C d u C d t i_C=C\frac{du_C}{dt} iC=CdtduC 【电容电流是其电压对时间的微分】
对电感 L L L: u L = L d i L d t u_L=L\frac{di_L}{dt} uL=LdtdiL 【电感电压是其电流对时间的微分】
{ i C ↔ C [ s U C ( s ) − u C ( 0 − ) ] u L ↔ L [ s I L ( s ) − i L ( 0 − ) ] \begin{cases} i_C\leftrightarrow C[sU_C{(s)}-u_C(0_{-})]\\ u_L\leftrightarrow L[sI_L(s)-i_L(0_{-})]\\ \end{cases} {iC↔C[sUC(s)−uC(0−)]uL↔L[sIL(s)−iL(0−)]
- 例、
u
S
(
t
)
=
δ
(
t
)
u_S(t)=\delta(t)
uS(t)=δ(t)
U S ( s ) = 1 U_S(s)=1 US(s)=1, Z R = R = 1 Z_R=R=1 ZR=R=1, Z C ( s ) = 1 s C = 1 s Z_C(s)=\frac{1}{sC}=\frac{1}{s} ZC(s)=sC1=s1, Z L ( s ) = s L Z_L(s)=sL ZL(s)=sL
H ( s ) = Y ( s ) X ( s ) = U C ( s ) U S ( s ) = Z C ( s ) Z C ( s ) + Z R ( s ) = 1 s 1 s + 1 = 1 s + 1 H(s)=\frac{Y(s)}{X(s)}=\frac{U_C(s)}{U_S(s)}=\frac{Z_C(s)}{Z_C(s)+Z_R(s)}=\frac{\frac{1}{s}}{\frac{1}{s}+1}=\frac{1}{s+1} H(s)=X(s)Y(s)=US(s)UC(s)=ZC(s)+ZR(s)ZC(s)=s1+1s1=s+11
d d t [ u C ( t ) ] + u C ( t ) = u S ( t ) = δ ( t ) \frac{d}{dt}[u_C(t)]+u_C(t)=u_S(t)=\delta(t) dtd[uC(t)]+uC(t)=uS(t)=δ(t)
对上边微分方程两边进行拉氏变换:
s U C ( s ) − u C ( 0 − ) + U C ( s ) = 1 sU_C(s)-u_C(0_{-})+U_C(s)=1 sUC(s)−uC(0−)+UC(s)=1
U C ( s ) = 1 + u C ( 0 − ) ( s + 1 ) = 2 s + 1 U_C(s)=\frac{1+u_C(0_{-})}{(s+1)}=\frac{2}{s+1} UC(s)=(s+1)1+uC(0−)=s+12
u C ( t ) = 2 e − t u ( t ) u_C(t)=2e^{-t}u(t) uC(t)=2e−tu(t)
5-1 时域积分特性
{ f ( t ) ↔ X ( s ) ∫ 0 − t [ f ( τ ) d τ ] ↔ X ( s ) s \begin{cases} f(t)\leftrightarrow X(s)\\ \int_{0_{-}}^{t}[f(\tau)d\tau]\leftrightarrow \frac{X(s)}{s}\\ \end{cases} {f(t)↔X(s)∫0−t[f(τ)dτ]↔sX(s)
6 时/频域的卷积定理
与傅里叶变换性质一致