文章目录
8.2. 系统函数 H(z)
8.2.1. 系统函数 H(z) 定义
- 定义:
F ( z ) → H ( z ) x = 0 → Y f ( z ) F(z) \to \boxed{\underset{x=0}{H(z)}} \to Y_f(z) F(z)→x=0H(z)→Yf(z)
H ( z ) = Y z s ( z ) F ( z ) H(z) = \frac{Y_{zs}(z)}{F(z)} H(z)=F(z)Yzs(z)
- 物理含义:
h
(
k
)
↔
H
(
z
)
=
∑
k
=
0
∞
h
(
k
)
z
−
k
h(k) \leftrightarrow H(z) = \sum^{\infty}_{k=0} h(k) z^{-k}
h(k)↔H(z)=k=0∑∞h(k)z−k
H
(
z
)
=
Z
[
h
(
k
)
]
H(z) = \mathcal{Z}[h(k)]
H(z)=Z[h(k)]
-
计算方法:
- H ( z ) = Y z s ( z ) F ( z ) H(z) = \frac{Y_{zs}(z)}{F(z)} H(z)=F(z)Yzs(z)
- H ( z ) = Z [ h ( k ) ] H(z) = \mathcal{Z}[h(k)] H(z)=Z[h(k)]
- 由系统差分方程求 H ( z ) H(z) H(z)
-
应用:
- 求 y z s ( k ) = Z [ Y z s ( z ) ] , Y z s ( z ) = H ( z ) F ( z ) y_{zs}(k) = \mathcal{Z}[Y_{zs}(z)],\; Y_{zs}(z)= H(z)F(z) yzs(k)=Z[Yzs(z)],Yzs(z)=H(z)F(z) ;
- 求 h ( z ) = Z − 1 [ H ( z ) ] h(z)= \mathcal{Z}^{-1}[H(z)] h(z)=Z−1[H(z)] ;
- 求 f ( k ) = Z − 1 [ F ( z ) ] , F ( z ) = Y z s ( z ) H ( z ) f(k) = \mathcal{Z}^{-1}[F(z)], \; F(z) = \frac{Y_{zs}(z)}{H(z)} f(k)=Z−1[F(z)],F(z)=H(z)Yzs(z) ;
- 表示系统特性:频率特性、稳定性等。
-
分解:
f ( k ) = 1 2 π j ∮ c F ( z ) z z k d z , − ∞ < k < ∞ f({\color{red}k}) = \frac{1}{2\pi j} \oint_c \frac{F(z)}{z} z^{ {\color{red}k}} dz, \; -\infty < k < \infty f(k)=2πj1∮czF(z)zkdz,−∞<k<∞
- 基本信号 z k z^k zk :
z 0 k → h ( k ) → z 0 k ⋅ H ( z 0 ) z_0^k \to \boxed{h(k)} \to z_0^k \cdot H(z_0) z0k→h(k)→z0k⋅H(z0)
f ( k ) → h ( k ) → y f ( k ) f(k) \to \boxed{h(k)} \to y_f(k) f(k)→h(k)→yf(k)
- 任意信号:
1
2
π
j
F
(
z
)
z
⋅
z
k
→
1
2
π
j
F
(
z
)
z
⋅
z
k
H
(
z
)
\frac{1}{2\pi j} \frac{F(z)}{z}\cdot z^k \to \frac{1}{2\pi j} \frac{F(z)}{z} \cdot z^k H(z)
2πj1zF(z)⋅zk→2πj1zF(z)⋅zkH(z)
∮
c
1
2
π
j
F
(
z
)
z
⋅
z
k
d
z
→
∮
c
1
2
π
j
F
(
z
)
z
⋅
z
k
H
(
z
)
d
z
\oint_c\frac{1}{2\pi j} \frac{F(z)}{z}\cdot z^k dz \to \oint_c\frac{1}{2\pi j} \frac{F(z)}{z} \cdot z^k H(z) dz
∮c2πj1zF(z)⋅zkdz→∮c2πj1zF(z)⋅zkH(z)dz
∮
c
1
2
π
j
F
(
z
)
z
⋅
z
k
d
z
→
∮
c
1
2
π
j
F
(
z
)
⋅
H
(
z
)
z
⋅
z
k
d
z
\oint_c\frac{1}{2\pi j} \frac{F(z)}{z}\cdot z^k dz \to \oint_c\frac{1}{2\pi j} \frac{ {\color{blue}F(z)\cdot H(z)} }{z} \cdot z^k dz
∮c2πj1zF(z)⋅zkdz→∮c2πj1zF(z)⋅H(z)⋅zkdz
Y
f
(
z
)
=
F
(
z
)
⋅
H
(
z
)
Y_f(z) = F(z) \cdot H(z)
Yf(z)=F(z)⋅H(z)
- 回顾 转换成时域
f ( k ) ⋆ h ( k ) ↔ F ( z ) ⋅ H ( z ) f(k)\star h(k) \leftrightarrow F(z) \cdot H(z) f(k)⋆h(k)↔F(z)⋅H(z)
8.2.2. 系统特性
离散系统的零点与极点
- 类比
H ( z ) = B ( z ) A ( z ) = b m z m + b m − 1 z m − 1 + ⋯ + b 1 z + b 0 z n + a n − 1 z n − 1 + ⋯ + a 1 z + a 0 = b m ( z − ζ 1 ) ( z − ζ 2 ) ⋯ ( z − ζ m ) ( z − P 1 ) ( z − P 2 ) ⋯ ( z − P n ) = b m ∏ j = 1 m ( z − ζ j ) ∏ i = 1 n ( z − P i ) , m ≤ n \begin{aligned}H(z) & = \displaystyle \frac{B(z)}{A(z)}\\ & = \displaystyle \frac{b_m z^m + b_{m-1}z^{m-1} + \cdots + b_1 z + b_0}{z^n + a_{n-1}z^{n-1} + \cdots +a_1z+a_0}\\ & = \frac{b_m(z-\zeta_1)(z-\zeta_2)\cdots(z-\zeta_m)}{(z-P_1)(z-P_2)\cdots(z-P_n)} \\ &= \frac{b_m \prod^{m}_{j=1}(z-\zeta_j)}{\prod^{n}_{i=1}(z-P_i)}, \; m\leq n \end{aligned} H(z)=A(z)B(z)=zn+an−1zn−1+⋯+a1z+a0bmzm+bm−1zm−1+⋯+b1z+b0=(z−P1)(z−P2)⋯(z−Pn)bm(z−ζ1)(z−ζ2)⋯(z−ζm)=∏i=1n(z−Pi)bm∏j=1m(z−ζj),m≤n
- H ( z ) H(z) H(z) 的零点:
ζ i , i = 1 , 2 , ⋯ , m \zeta_i, \; i =1,2,\cdots, m ζi,i=1,2,⋯,m
- H ( z ) H(z) H(z) 的极点:
P i , i = 1 , 2 , ⋯ , m P_i, \; i =1,2,\cdots, m Pi,i=1,2,⋯,m
- 零/极点的种类:
- 实数、
- 复数 (复数零、极点必共轭 )
- 一阶、二阶及二阶以上极点
零、极点与h(k)的关系
-
极点在单位圆内
-
在实轴上:
- 一阶极点:
A z z − a ↔ A a k ↓ ε ( k ) , ∣ a ∣ < 1 \frac{Az}{z-a}\leftrightarrow A {\color{blue}a^k\downarrow} \varepsilon(k),\; \lvert a \rvert <1 z−aAz↔Aak↓ε(k),∣a∣<1 - 二阶极点:
A z ( z − a ) 2 ↔ A k a k − 1 ↓ ε ( k ) , ∣ a ∣ < 1 \frac{Az}{(z-a)^2}\leftrightarrow A{\color{blue}ka^{k-1}\downarrow} \varepsilon(k),\; \lvert a \rvert <1 (z−a)2Az↔Akak−1↓ε(k),∣a∣<1
- 一阶极点:
-
不在实轴上:
- 一阶极点:
A z z − r e j β + A ∗ z z − r e − j β ↔ 2 ∣ A ∣ r k ↓ cos ( β k + θ ) ε ( k ) , r < 1 \frac{Az}{z-r e^{j\beta}} + \frac{A^*z}{z-re^{-j\beta}}\leftrightarrow 2\lvert A \rvert {\color{blue}r^k\downarrow} \cos(\beta k + \theta) \varepsilon(k),\; r<1 z−rejβAz+z−re−jβA∗z↔2∣A∣rk↓cos(βk+θ)ε(k),r<1 - 二阶极点:
A z ( z − r e j β ) 2 + A ∗ z ( z − r e − j β ) 2 ↔ 2 ∣ A ∣ r k − 1 ↓ cos ( β ( k − 1 ) + θ ) ε ( k ) , r < 1 \frac{Az}{(z-r e^{j\beta})^2} + \frac{A^*z}{(z-re^{-j\beta})^2}\leftrightarrow 2\lvert A \rvert {\color{blue}r^{k-1}\downarrow} \cos(\beta (k-1) + \theta) \varepsilon(k),\; r<1 (z−rejβ)2Az+(z−re−jβ)2A∗z↔2∣A∣rk−1↓cos(β(k−1)+θ)ε(k),r<1
- 一阶极点:
-
结论: 对应 h ( k ) h(k) h(k)按指数规律衰减;
-
-
极点在单位圆上
-
在实轴上:
- 一阶极点:
A z z ± 1 ↔ A ( ± 1 k ) ε ( k ) \frac{Az}{z\pm 1}\leftrightarrow A {\color{blue}(\pm 1^k)} \varepsilon(k) z±1Az↔A(±1k)ε(k) - 二阶极点:
A z ( z ± 1 ) 2 ↔ A k ↑ ( ± 1 k − 1 ) ε ( k ) \frac{Az}{(z\pm 1)^2}\leftrightarrow A {\color{red}k\uparrow}{\color{blue}(\pm 1^{k-1})} \varepsilon(k) (z±1)2Az↔Ak↑(±1k−1)ε(k)
- 一阶极点:
-
不在实轴上:
- 一阶极点:
A z z − r e j β + A ∗ z z − r e − j β ↔ 2 ∣ A ∣ cos ( β k + θ ) ε ( k ) , r < 1 \frac{Az}{z-r e^{j\beta}} + \frac{A^*z}{z-re^{-j\beta}}\leftrightarrow 2\lvert A \rvert \cos(\beta k + \theta) \varepsilon(k),\; r<1 z−rejβAz+z−re−jβA∗z↔2∣A∣cos(βk+θ)ε(k),r<1 - 二阶极点:
A z ( z − r e j β ) 2 + A ∗ z ( z − r e − j β ) 2 ↔ 2 ∣ A ∣ k ↑ cos ( β ( k − 1 ) + θ ) ε ( k ) , r < 1 \frac{Az}{(z-r e^{j\beta})^2} + \frac{A^*z}{(z-re^{-j\beta})^2}\leftrightarrow 2\lvert A \rvert {\color{red}k\uparrow} \cos(\beta (k-1) + \theta) \varepsilon(k),\; r<1 (z−rejβ)2Az+(z−re−jβ)2A∗z↔2∣A∣k↑cos(β(k−1)+θ)ε(k),r<1
- 一阶极点:
-
结论: 一阶极点对应 h ( k ) h(k) h(k)为稳态分量;二阶及二阶以上极点对应 h ( k ) h(k) h(k)增长。
-
-
极点在单位圆外
-
在实轴上:
- 一阶极点:
A z z − a ↔ A a k ↑ ε ( k ) , ∣ a ∣ > 1 \frac{Az}{z-a}\leftrightarrow A {\color{blue}a^k\uparrow} \varepsilon(k),\; \lvert a \rvert >1 z−aAz↔Aak↑ε(k),∣a∣>1 - 二阶极点:
A z ( z − a ) 2 ↔ A k a k ↑ ε ( k ) , ∣ a ∣ > 1 \frac{Az}{(z-a)^2}\leftrightarrow A{\color{blue}ka^k\uparrow} \varepsilon(k),\; \lvert a \rvert >1 (z−a)2Az↔Akak↑ε(k),∣a∣>1
- 一阶极点:
-
不在实轴上:
- 一阶极点:
A z z − r e j β + A ∗ z z − r e − j β ↔ 2 ∣ A ∣ r k ↑ cos ( β k + θ ) ε ( k ) , r > 1 \frac{Az}{z-r e^{j\beta}} + \frac{A^*z}{z-re^{-j\beta}}\leftrightarrow 2\lvert A \rvert {\color{blue}r^k\uparrow} \cos(\beta k + \theta) \varepsilon(k),\; r>1 z−rejβAz+z−re−jβA∗z↔2∣A∣rk↑cos(βk+θ)ε(k),r>1 - 二阶极点:
A z ( z − r e j β ) 2 + A ∗ z ( z − r e − j β ) 2 ↔ 2 ∣ A ∣ r k − 1 ↑ cos ( β ( k − 1 ) + θ ) ε ( k ) , r > 1 \frac{Az}{(z-r e^{j\beta})^2} + \frac{A^*z}{(z-re^{-j\beta})^2}\leftrightarrow 2\lvert A \rvert {\color{blue}r^{k-1}\uparrow} \cos(\beta (k-1) + \theta) \varepsilon(k),\; r>1 (z−rejβ)2Az+(z−re−jβ)2A∗z↔2∣A∣rk−1↑cos(β(k−1)+θ)ε(k),r>1
- 一阶极点:
-
结论: 对应 h ( k ) h(k) h(k)按指数规律增长。
-
8.2.3. 离散系统稳定性判据(因果系统)
- 离散系统稳定的时域充要条件:(绝对可和)
∑ k = − ∞ ∞ ∣ h ( k ) ∣ < ∞ \sum^{\infty}_{k=-\infty} \lvert h(k)\rvert < \infty k=−∞∑∞∣h(k)∣<∞
-
离散系统稳定性的Z域充要条件:
- 若LTI离散系统的系统函数 H ( z ) H(z) H(z) 的收敛域包含单位圆,则系统为稳定系统。
- 若LTI离散因果系统稳定,要求其系统函数 H ( z ) H(z) H(z) 的极点全部在单位圆内。
- ∣ p j ∣ < 1 \lvert p_j \rvert <1 ∣pj∣<1
-
离散因果系统稳定性判定--朱里准则(Jury stability criterion)
H ( z ) = B ( z ) A ( z ) = b m z m + b m − 1 z m − 1 + ⋯ + b 1 z + b 0 a n z n + a n − 1 z n − 1 + ⋯ + a 1 z + a 0 \begin{aligned}H(z) = \displaystyle \frac{B(z)}{A(z)} = \displaystyle \frac{b_m z^m + b_{m-1}z^{m-1} + \cdots + b_1 z + b_0}{a_nz^n + a_{n-1}z^{n-1} + \cdots +a_1z+a_0}\end{aligned} H(z)=A(z)B(z)=anzn+an−1zn−1+⋯+a1z+a0bmzm+bm−1zm−1+⋯+b1z+b0
-
要判断 A ( z ) = 0 A(z)=0 A(z)=0 的所有根的绝对值是否都小于 1 1 1 。
-
朱里准则指出: A ( z ) = 0 A(z)=0 A(z)=0 的所有根都在单位圆内的充要条件是:
- A ( 1 ) > 0 A(1)>0 A(1)>0
- ( − 1 ) n A ( − 1 ) > 0 (-1)^nA(-1)>0 (−1)nA(−1)>0
-
a
n
>
∣
a
0
∣
c
n
−
1
>
∣
c
0
∣
⋯
r
2
>
∣
r
0
∣
a_n>\lvert a_0\rvert \; c_{n-1}>\lvert c_0 \rvert \cdots r_2>\lvert r_0\rvert
an>∣a0∣cn−1>∣c0∣⋯r2>∣r0∣
对奇数行,其第1个元素必大于最后一个元素的绝对值。
- 特例: 对二阶系统:
A ( z ) = a 2 z 2 + a 1 z + a 0 A(z) = a_2z^2 + a_1z + a_0 A(z)=a2z2+a1z+a0
易得 A ( 1 ) > 0 , A ( − 1 ) , a 2 > ∣ a 0 ∣ A(1)>0, \; A(-1),\; a_2>\vert a_0\vert A(1)>0,A(−1),a2>∣a0∣
-