信号分析与处理 - 杨西侠 - 课后答案二三五章(1)汇总

更新时间:2024-05-15 12:57:01 阅读量: 综合文库 文档下载

说明:文章内容仅供预览,部分内容可能不全。下载后的文档,内容与下面显示的完全一致。下载之前请确认下面内容是否您想要的,是否完整无缺。

2-1 画出下列各时间函数的波形图,注意它们的区别

1)x1(t) = sin ? t·u(t)

1 0 -

xπ 234t

2)x2(t) = sin[ ? ( t – t0 ) ]·u(t)

1 0 -xt3)x3(t) = sin ? t·u ( t – t0 )

t

x31 0 t4)x2(t) = sin[ ? ( t – t0 ) ]·u ( t – t0 )

t

1 0 -xtt

1

2-2 已知波形图如图2-76所示,试画出经下列各种运算后的波形图

1 x(t) 1 2 3 -1

0 t

图 2-76 (1)x ( t-2 )

1 x t -0 1 2 3 4

(2)x ( t+2 )

x 1 t ----0 1

(3)x (2t)

1 x(2t) t -1 0 1 2 3

(4)x ( t/2 )

1 x t --0 1 2 3 4

(5)x (-t)

2

x (-t) 1 t -3 -2 -1 0 1 2

(6)x (-t-2)

1 x (-t-2) t 1

-----0

(7)x ( -t/2-2 )

1 x ( -t/2-2 )

t -8 -7 -6 -5 -4 -3 -2 -1 0 1

(8)dx/dt

1 dx/dt t -2 -1 0 1 2 3 -δ (t-2) 2-3 应用脉冲函数的抽样特性,求下列表达式的函数值

(1)

????????x(t?t0)δ(t) dt = x(-t)

0

(2)

??x(t0?t)δ(t) dt = x(t)

0

3

(3)

?t0?(t?t0) u(t -??2??t0) dt = u(

20

)

(4)

??????(t0?t) u(t – 2t) dt = u(-t)

0

(5)

??e?????t?t?δ(t+2) dt = e2-2

(6)

??t?sint??????δ(t-

?) dt =

66+

12

(7)

?????e?j?t???t????t?t0??dt

=

?????e?j?t??t?dt–?e?j?t?(t?t0)dt

?? = 1 – cosΩt0 + jsinΩt0

??= 1-

e?j?t02-4 求下列各函数x1(t)与x2(t) 之卷积,x1(t)* x2(t)

(1) x1(t) = u(t), x2(t) = e-at · u(t) ( a>0 )

x1(t)* x2(t) =

?????u(?)eu(t??)d??a? =

1?a??ated?(1?e) = ?0a) · u(t) t?(2) x1(t) =δ(t+1) -δ(t-1) , x2(t) = cos(Ωt + 4?[cos(?t?)u(?)][?(t???1)??(t???1)]d?

x(t)* x(t) =???4??1

2

?= cos[Ω(t+1)+

?]u(t+1) – cos[Ω(t-1)+

44]u(t-1)

(3) x1(t) = u(t) – u(t-1) , x2(t) = u(t) – u(t-2)

x1(t)* x2(t) =

?????[u(?)?u(??2)][u(t??)?u(t???1)]d?

t当 t <0时,x1(t)* x2(t) = 0

当 0

?0d? = t

4

当 1

?211d? = 1 d?=3-t

当 2

?t?2x1(t)* x2(t) 1 t 0 1 2 3

(4) x1(t) = u(t-1) , x2(t) = sin t · u(t)

x1(t)* x2(t) =

?????sin(?) u(?) u(t???1)d?

=

??0t-1sin ? u(t-?-1)d? ??sin ? d? ? -cos ?|0

0t-1= 1- cos(t-1)

2-5 已知周期函数x(t)前1/4周期的波形如图2-77所示,根据下列各种情况的要求画出x(t)在一个周期( 0

(1) x(t)是偶函数,只含有偶次谐波分量 f(t) = f(-t), f(t) = f(t±T/2)

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

(2) x(t)是偶函数,只含有奇次谐波分量 f(t) = f(-t), f(t) = -f(t±T/2)

5

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

(3) x(t)是偶函数,含有偶次和奇次谐波分量 f(t) = f(-t)

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

(4) x(t)是奇函数,只含有奇次谐波分量 f(t) = -f(-t), f(t) = -f(t±T/2)

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

(5) x(t)是奇函数,只含有偶次谐波分量 f(t) = -f(-t), f(t) = f(t±T/2)

6

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

(6) x(t)是奇函数,含有偶次和奇次谐波分量 f(t) = -f(-t)

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

f(t) t -T/2 -T/4 0 T/4 T/2 3T/4 T

2-6 利用信号x(t)的对称性,定性判断图2-78所示各周期信号的傅里叶级数中所含有的频率分量

(a)

7

x(t) t -2T -T 0 T 2T

(b)

这是一个非奇、非偶、非奇偶谐波函数,且正负半波不对称,所以含有直流、正弦等所有谐波分量,

因为去除直流后为奇函数。

x (t) t -T 0 T

(c)

这是一个奇函数。也是一个奇谐波函数,所以只含有基波、奇次正弦谐波分量。

x(t) t -T -T/2 0 T/2 T

(d)

除去直流分量后是奇函数,又f(t) = f(t±T/2),是偶谐波函数,所以含有直流、偶次正弦谐波。

x (t) t -T -T/2 0 T/2 T

8

(e)

正负半波对称,偶函数,奇谐波函数,所以只含有基波、奇次余弦分量。

x (t) t -T/2 0 T/2 T

(f)

奇函数、正负半波对称,所以只含有正弦分量(基、谐)

x(t) t -T -T/2 0 T/2 T

正负半波对称、奇函数、奇谐波函数,所以只含有基波和奇次正弦谐波。

2-7 试画出x(t) = 3cosΩ1t + 5sin2Ω1t的复数谱图(幅度谱和相位谱)

解:a0 = 0, a1 = 3, b2 = 5, c1 = 3, c2 = 5 |x1| = |

12(a1-jb1)| =

32, |x2| =

12c2 =

52

035φ2 = arctan (-0φ1 = arctan (-

) = 0, φ-1= 0 ) = -

?2, φ-2=

?2

9

|xn| 3 2 1 nΩ1

-2Ω1 -Ω1 0 Ω1 2Ω1

π/2 nΩ1

-2Ω1 -Ω1 0 Ω1 2Ω1 -π/2

2-8 求图2-8所示对称周期矩形信号的傅里叶级数

E/2 x (t) t -T -T/2 0 T/2 T -E/2

解:这是一个正负半波对称的奇函数,奇谐函数,所以只含有基波和奇次正弦谐波。

bn =

2T?T0x(t) sin n?t dt

T =

2T?T202E sin n?t dt–

T2?T2E sin n?t dt 2 10

=

ET?T20[sin n?t - sin n?(t-TT) ]dt 2T

EET22?cos n?t| ? cos [n?(t?)]|0 = 02n?2n?2 =

EE?(cos n? -1) ? cos (1-cos n?) 2n?2n?2En? ,n为奇数,n = 1,3,5 ……

=

?E(cos n? -1) ? n? 0 ,n 为偶数,n = 2,4,6 ……

2E∴ x(t) =

11 [ sin ?t ? sin 3?t ? sin 5?t ? ??? ] ?35指数形式的傅里叶级数

0 , n = 0, ±2, ±4 ……

Xn=

12(an-jbn) =

??jEn? , n = ±1, ±3, ±5 ……

∴ x(t) = a0 +

jn?t?jn?t(Xe?Xe) ?nnn?02-9 求图2-9所示周期信号的傅里叶级数

E x (t) t -T/2 0 T/4 T/2 3T/4 T

解:此函数是一个偶函数 x(t) = x(-t) ∴ 其傅里叶级数含有直流分量和余弦分量

T ao =

144E ? t dt= ET0T8 +

3T1Tt14 ?T E dt+ ?3T4E(1-) dt

T4TT4 11

= =

EE2E292T) + + E–2(T?8216T6E3E3E– = 444 an =

2T ? x(t) cos n?t dtT0

=

1T ? x(t) (ejn?t ? e-jn?t) dtT0 =

=

1T?4En?(1?cos), n = 1, 2, … 22(n?) ∴ x(t) =

3E4–

11[ cos ?t ? cos2?t ? cos 3?t ? ...] 249?4E2-10 若已知F[x(t)] = X(Ω)利用傅里叶变换的性质确定下列信号的傅里叶变换

(1) x(2t–5) (2) x(1–t) (3) x(t) · cos t

解:(1) 由时移特性和尺度变换特性可得

1?-j2? X () eF [x( 2t - 5)] = 22(2) 由时移特性和尺度变换特性

5 F [x(at)] =

1? X () |a|a F [x(t-t0)] =

X (?) e-j?t0 X (-?) e-j?

F [x(1–t)] =

(3) 由欧拉公式和频移特性

cos t =

1 ( ejt? e-jt) 2

F [

x (t) e?j?0t] = X(Ω?Ω0)

12Ω0 = 1

F [x(t) · cos t] = [ X(Ω–1) + X(Ω+1)]

12

2-11已知升余弦脉冲x(t) =

E?t( 1 ? cos ) (???t??)求其傅里叶变换 22)–u( t–τ)]

解:x(t) = 求微分

E?t( 1 ? cos )[ u( t +τ22

E??t?x(t) = ? sin [ u(t ? ?) - u(t-?)]

2??

E?2?t??x(t) = ?2 cos [ u(t ? ?) - u(t-?)]

2??E?3?tE?2x???(t)=3 sin [ u(t ? ?) - u(t-?)] +2 [ ?(t ? ?) - ?(t-?)]

2??2?

?2E?2 x?(t) + [ ?(t ? ?) - ?(t-?)] = 22?2?由微分特性可得:

2Ej????j??3[-(j?) X(?) ? (e?e)]2( jΩ) X(Ω) =

2?

∴ X(Ω) =

?2E2sin2?? 2 ?2?(2??)?2-12已知一信号如图2-81所示,求其傅里叶变换

x(t) t -τ/2 0 τ/2

解:(1) 由卷积定理求

x(t) =

G?(t) * G?(t)

22 13

G?(t) =

22E[u(t?)?u(t?)] ?442E???Sa()

?24E?2??Sa() 24??

G?(?) =

2 由时域卷积定理

X(Ω) =

G?(?) G?(?) =

22(2) 由微分特性求

2E ,–?< t < 0 ?2

x?(t) = –2E ,0 < t < ???2

2

0 ,| t | >

x??(t) = 2E [δ( t +??由微分特性

?) +δ( t–22??j?2)–2δ(t)]

( jΩ) X(Ω) =

2

2E?(e?j?2?e???2)?(2cos?2)

?22E X(Ω) =

E?2??Sa() 242-13已知矩形脉冲的傅里叶变换,利用时移特性求图2-82所示信号的傅里叶变换,并大致画出幅度谱

解:

G?(t) = E [ u( t +

?2)–u( t–

?)] 2

G?(?) =

??E? Sa()

2?)–G?2?( t–)

2

x(t) = G?( t +

由时移特性和线性性

14

X(Ω) =

??j??E? Sa()e22E? Sa(??e)2???j??2E? Sa()e–

2?j?2 =

?e2j?j?2·2j = 2jE? Sa(??)sin ??22

2Eτ -2? ?-? ?0 ? ? 2?Ω ?

2-14已知三角脉冲x1(t)的傅里叶变换为

X1(Ω) =

E?2??Sa() 24) cosΩ0t的傅里叶变换

?试利用有关性质和定理求x2(t) = x1(t–

2

由时移性质

-j??2F [x1 (t–)] = X1 (?) e

2解:由时移性质和频域卷积定理可解得此题

?

由频移特性和频域卷积定理可知:

F [x(t )cosΩ0t]=

12[X(Ω–Ω0)+ X(Ω+Ω0)]

?X2 (Ω) = F [x1 (t–

2

=

)cosΩ0t]

???0?2???0?2

12[ X1 (Ω–Ω0)

e?j + X(Ω+Ω0)

e?j]

15

=

E?4(???0)??j2e[Sa

4???0?2(???0)??j2e+ Sa

4???0?2]

2-15求图2-82所示X(Ω)的傅里叶逆变换x(t)

|X(Ω)| A A |X(Ω)| Ω -Ω0 0 Ω0 -Ω0 0 Ω0 Ω φ(Ω) π/2 π/2 φ(Ω) Ω -Ω0 0 -π/2 Ω0 -Ω0 0 -π/2 Ω0 Ω a) b)

?j?(?)e解:a) X(Ω) = | X(Ω)|

j?t0G(?)e = 2?0由定义:

x(t) =

12??????X(?)ej?td?

?0 =

12?A2????0Aej?t0ej?td? ej?(t?t0)d?

=

??0??0 =

A0ej?(t?t0)|???02?j(t?t0)

=

Asin[?0(t?t0)]

?(t?t0) 16

=

A?0?Sa[?0(t?t0)]

b)

1x(t)?2?

?????X(?)ej?td?

?j

1=

2?=

?0?2??00Aeeej?t1d?+

2???00Ae2ej?td? ej(?t?j?

A2??j(?t??2)??0Ad?+

2?+

??0?2)0d?

=

j(?t?)A2e|0??02?j??j(?t?)A02e|?02?j??j(?0t??

A

=

j?2?A2?j(?0t??2?2e))

A

j?2?A2?j(?0t?sin[(?0t?)?2e))]=

j(?0t??2)

A

=

?2A?(?0t??2?Sa[?0t??2]

2-16确定下列信号的最低抽样频率与抽样间隔

(1) Sa(100t) (2) Sa2(100t)

(3) Sa(100t)+ Sa2(100t) 解:(1)由对偶性质可知:

Sa(100t)的频谱是个矩形脉冲,其脉宽为[-100,100] 即Ωm = 100 =2πfm

∴ fm =

50?

由抽样定理 fs ≥ 2fm ∴ fs ≥ 2×50? =

100?

Ts≤

?100

17

(2) 由对偶性质可知

Sa(100t)的频谱是个矩形脉冲,其脉宽为[-100,100] 又由频域卷积定理可知

Sa2(100t)的频谱是脉宽为[–200,–200]的三角形脉冲 即Ωm = 200 =2πfm

∴ fm =

100?

由抽样定理 fs ≥ 2fm ∴ fs ≥ 2×Ts≤

100? =

200?

?200

(3) 由线性性质可知

Sa(100t)+ Sa2(100t) 的频谱是Sa(100t)和Sa2(100t)之和 ∴其Ωm =2πfm= 200 即 fm =

100?

则fs ≥ 2fm = Ts≤

200?

?200

2-17已知人的脑电波频率范围为0~45Hz,对其作数字处理时,可以使用的最大抽样周期T是多少?若以T = 5ms抽样,要使抽样信号通过一理想低通滤波器后,能不是真的回复原信号,问理想低通滤波器的截至频率fc应满足什么条件?

解:由已知条件,可知fm = 45Hz 由抽样定理fs ≥ 2fm = 90Hz ∴ T ≤

1 901T =

x(f) T = 0.005 ∴ fs =

10005 = 200

f -45 0 45 由抽样定理和低通滤波可知 45 ≤ fc ≤ 200-45 = 155 即45 ≤ fc ≤ 155

x(f) 2-18若F[a(t)] = X(Ω), 如图2-85所示,当抽样脉冲p(t)为下列信号时,试分别求抽样后的抽样信号的频谱X s (Ω), 并画出相应的频谱图

(1) p(t) = cos t

f -45 0 45 1 X(Ω) 200 Ω

18

-1 0 1 图 2-85

(2) p(t) = cos2 t (3) p(t) =

n???????(t?2?n) ??(t??n)

??(4) p(t) =

n???解:由抽样特性可知 x s = x(t) p(t) 由频域卷积定理可知 X s (Ω) =

1X(?)*P(?) 2?1 1/2 X s (Ω) (1) P(Ω) = [δ(Ω+1)+δ(Ω-1)]

1X(?)*P(?) ∴ X s (Ω) = 2?1 = [X(??1)?X(??1)]

2(2) P(Ω) = [δ(Ω+2)+δ(Ω-2)] ∴ X s (Ω) =

Ω -2 -1 0 1 1/2 1 2 18 (1) X s (Ω) 1X(?)*P(?) 2?1 = [X(??2)?X(??2)]

2Ω -3 -2 -1 0 1 2 3 18 (2) (3) P(Ω) =

2?2?n???????(??n)

??1X s (Ω) 2? =

n?????(??n)

-3 -2 -1 0 1 2 3 Ω 18 (3) ∴ X s (Ω) =

1X(?)*P(?) 2? =

12???n????X(??n)

1X (Ω) s???(4) P(Ω) =

2??n???????(??2n)

=

2??(??2n)

n???Ω -3 -2 -1 0 1 2 3 ∴ X s (Ω) =

1X(?)*P(?) 2?18 (3) 19

=

1?n????X(??2n)

??Xp (1) = 2, Xp (2) = 0, Xp (3) = 2

3-1 解:序列频谱的定义为

X(e) =

(1)

j?n?-??jn?x(n)e?????

X(ej?) = X(ej?) = X(e) =

=

j?n?-??jn??(n)e?= 1

(2)

n?-?????(n?3)e???jn?=

e-j3?

(3)

n?-??jn?[0.5?(n?1)??(n)?0.5?(n?1)]e?

0.5ej?+ 1 +0.5e-j?= 1 +

n?jn?au(n)e?????ej??e?j? = 1 +cos ?

2(4)

X(e) =

=

j?

n?-?

?aen?0n?jn? =

?j?n(ae) (∵0 < a < 1, ∴收敛) ?n?0??

1 = ?j?

1?ae(5)

X(ej?) =

n?-??j?Re??j2??N(n)ej?jn?=

?en?0?jN?2N?1?jn?1?e?jN? = ?j?

1?e-jN-1?2N?2 =

·eN?2?e?esinee?j2??j2 = eN?2 ? sin23-2 (1) DTFT[x(n-n0)] =

n?-??x(n?n)e0???jn?

m?n?n0m?-??jm??jn0??jn0?j?x(m)ee?= X(e)e

?? 20

(2) DTFT[x(n)] =

??*

n?-??x(n)e*???jn?jn?*[x(n)e] = ?n?-????jn(??)*[x(n)e]= X*(e-j?) = ?n?-????? (3) DTFT[x(-n)] =

n?-??x(-n)e??n?-??jn?m??n?x(m)e?jm(??)= X(e-j?)

m???(4) DTFT[x(n)* y(n)] =

?jn?[x(n) * y(n)]e???

??=

n?-?m?-???????x(m)y(n?m)ej??jm????jn?=

j?m????jn?x(m)y(n?m)e??n?????

=

m????x(m)Y(e)e

?jm?Y(e)x(m)e?=

m??? =

??X(ej?)Y(ej?)

(5) DTFT[x(n) y(n)] =

n????jn?x(n)y(n)e?

=

1[?n???2?12???????X(ej?)ejn?d?]y(n)e?jn?

j???

? =

????X(e)[?y(n)e?jn(???)]d?

n??? =

12????X(ej?)Y(ej(???))d??jn?nx(n)e? = j[????=

1X(ej?)*Y(ej?) 2? (6) DTFT[nx(n)] =

n???dX(ej?)] d? (7) DTFT[x(2n)] =

n????jn?x(2n)e?

m?2n

m????x(m)e???jm?2

21

m取整数?=

??m????jm?jm1m22[x(m)e?(?1)x(m)e] 2??

??jm1??2x(m)e?2m??????j1??2mx(m)(?e) +?2m???? =

jj112X(e)+X(?e2) 222

(8) DTFT[x(n)] =

1X(ej?)*X(ej?) 2? (9) DTFT[xa(n)] =

??n????xa(n)e???jn?=

n????j2n?x(2n)e?a

?? =

n????j2n?j2?x(n)e) ?= X(e3-3 解:x(n) =

12?????X(e)ej?jn?d? =

12?????00ejn?d?

=

10ejn?|???02?jnsinn?0n? =

=

1ejn?0?e?jn?0?n?2j =

=

?0sinn?0??n?0?0Sa(n?0) ?3-4解: 由DFS的定义

∴ Xp (0) = Xp (k) =

nkx(n)W?pN n?0N?1?xn?03N?1p(n)e?jn?02?2

1 n p

3=

?xn?0(n) = 4

?-2 -1 0 1 2 3 4 5 图3-44 ?jn2 Xp (1) =

?xn?0p(n)e= 2 + (–j )

+ 0 + j = 2

22

Xp (2) =

?jn?x(n)e?p= 2 + (–1 ) + 0 + (–1 ) = 0 n?03 Xp (3) =

?xn?03p(n)e?jn3?2= 2 + j + 0 + (–j ) = 2∵ Xp (k)是周期函数,其周期长度N=4

?∴ Xp (k) = Z[1+cos( k)]或 Xp (0) = 4, Xp (1) = 2, Xp (2) = 0, Xp (3) = 2

2?j2?nkxp(n)eN ?j2?nkxp(n)e2N

3-5 解: 由DFS的定义

N?1 Xp1 (k) =

?n?02N?1 Xp2 (k) =

?n?0?j2?nk2N?1?j2?nk=xp(n)e2N+xp(n)e2N n?0n?NN?1??N?1 m?n?Nxp?n?0?j2?n(k)N?1(n)eN2+xm?0??j2?(m?N)kN2

p(m?N)eN?1?j2?mk?jk?k = Xp1 () +xp(m)eN2?e

2m?0? =Xp1()?Xp1() ? ek2k2?jk?=Xp1()(1? ek2?jk?)

?0 , k为奇数? =? k2Xp1() , k为偶数??23-6。解:与3-4答案相同,可由定义求出。

只不过此时的x(k)非周期的。

x(n) 2 1 ?Xp (k) = Z[1+cos( k)]R4(k)

2或 Xp1 (0) = 4, Xp1 (1) = 2, Xp1 (2) = 0, Xp1 (3)

-2 = 2

3-8 解:(1)由定义得,

-1 n 0 1 2 3 4 图3-45 离散时间信号 5 X(k)??n?02?j2?nke3

23

∴X(0)?e0?3 ?n?022X(1)??2?j2?ne3?j4?ne3?1?e?j2?3?e?j4?3?0

n?0?j4??1?e3?j2?n?e3X(2)?? (2)∵N??0

n?02?m??4m

2 ∴只要m?1,N就取整数N?4

∴X(k)??n?033?j2?nk?cosneN

2 ∴X(0)??j0cosne?1?0?1?0?0 ?2n?03?

X(1)??n?03?j?n?cosne2?1?0?1?0?2

2

X(2)??cosne?j?n?1?0?1?0?0

n?03?2

X(3)???j3?n?cosne2?1?0?1?0?2n?02

X(k)?1?cosk?, k=0,1,2,3

(3)X(k)??n?033?j2?nkx(n)eN

∴X(0)?x(n)?5 ?n?03?jn2

X(1)??x(n)en?03??1?(?2j?)?1j3??2j

X(2)??x(n)e?j?n?1?(?2?)?(?1)?(?3?)

n?035

X(3)??x(n)en?0?j3?n2?1?2j?1??(j3?)?2j

24

DFT?x(n)??3-9

N?1N?1:

?j2?nk(n)eN?n?0?j2?nkx(n)eN?X(k)(

1

DFT??(n)??∴X(k)???n?0. . .X(N?1)?1 ?1?1?X(k)∴X(0)?1 X(1)1?R ?0,1,2N,?... ,N(k),kN?1(2)DFT??(n?3)??N?1??n?0n?j2?nk(n?3)eN?e?j6?kN

(3)DFT??a???na?n?0?j2?nkeN?1?aNe?j2?k1?ae?j2?kN?1?aN1?ae?j2?kN

(4)DFT??ejw0nN?1????n?02?jw0n?jNnkee?N?1j(w?2?k)ne0N n?0?

?1?ejw0Nej2?kj(w0?2?k)N1?e?1?ejw0Nj(w0?2?k)N1?e

?N?1j2?n?j2?nkN?1j(2??2?k)nn??j2NeNeN?eNN (5)DFT?e??n?0??n?0??

?1?ej2?(1?k)j2?(1?k)1?eN???N , k?1???N?(k?1)

?0 , k?1 ?N?1n?03-10 解:(1)Z?x(n)??n????x(n)zN?1n?0??n??z2?nkN?n1?z?N? (z?1) 1?z?12?kN (2)DFT?x(n)???x(n)ejw?j?1?e?j2?k1?e?j?N?(k)

N?1n?0 (3)DTFT?x(n)??X(e)?n?????x(n)eNw2?jnw??e?jnw

?jNw21?ee(e?e)??www?jj?j1?e?jw22e(e?e2) NsinwN?1?j()w2?e2wsin2?jNw?jNw2j 25

jw 当w?0时,X(e)?N

当w?2?k时,X(ejw)?0 N11(4)由(3)可得,当x(n)由4点通过补零扩为10点时,此时的圆卷积和线卷积的结果相同。由于线卷积的长度为4+4-1=7

∴可知x(n)由4点通过补零扩为最少7点时,圆卷积和线卷积相等。

?ln?X(k?l)R(k)?x(n)W3-12 证明:频移定理为 IDFT? pNN?? 由IDFT的定义可知,

IDFT??Xp(k?l)RN(k)??1?N

?Xk?0k??ljN?1p(k?l)ej2?nkN?1???NN?l?1?Xp(m)e?j2?nmN?ln?j2Ne??

?x(n)e2?lnN?x(n)WN?ln3-13 解:频移定理

?ln?X(k?l)R(k)?x(n)W IDFT? pNN???2?mn?jmn2?1j2N1?mnmnNmn)?(e?e)?(WN?WN) (1)∵cos(N22 ∴DFT?x(n)cos( 由频移特性:

??2?1?1?mnmn??mn)??DFT?x(n)W?DFT?x(n)WNN???? N22? DFT?x(n)cos(mn)???Xp(k?m)?Xp(k?m)?RN(k) ??N??2?2?mn?jmn2?1j2N1?mnmn (2)∵sin(mn)?(e?eN)?(WN?WN)

N2j2j?2??1 ∴DFT?x(n)sin( 由频移特性:

??2?1?1?mnmn??? mn)??DFT?x(n)W?DFTx(n)WNN????N2j?2j? DFT?x(n)sin(mn)??Xp(k?m)?Xp(k?m)???RN(k) N2j??3-14

解:由DFT的定义可知,

?2??1 26

DFT?y(n)??

rN?1n?0?y(n)e?j2?nkrN??x(n)en?0N?1?j2?nkrN??x(n)en?0N?12?k?jn()Nrk?X()r

3-15 证明:频域圆卷积定理,

若y(n)?x(n)h(n) 则

Y(k)?Xk(?)Hk()1N?1 ? ?XlH(p)k?l(RNl)

Nl?01N?1 ? ?HlX(p)k?l(RNl)Nl?0N?1n?0()()?nkY(k)?DFT?y(n)???x(n)h(n)WN ??x(n)??IDFT?H(k)???W

n?0N?1?nkN?1??x(n)?n?0?NN?1?H(k)Wl?0N?1?lnN?nk?WN? ?1N?H(l)?x(n)Wl?0n?0N?1N?1

(k?l)nN1 ?N?H(l)Xl?0N?1p(k?l)RN(l)1N?1 同理可证Y(k)??X(l)Hp(k?l)RN(l)

Nl?03-16 证明:由卷积的定义可知 (1)x(n)??(n)?m????x(m)?(n?m)?x(n)

?x(m)?(n?n?0?(2)x(n)??(n?n0)??m)?x(n?n0)

m???3-19解:(1)T1min (3)N??1111??0.02s(2)Tmax???0.5?10?3s

2fn2?1000??F??50T10.026??40 ∴N?2?64 min?3T0.5?10 (4)分辨力提高一倍,则T1min?0.04s,则N?80,取N?27?128

N)又DFT?RN(n)??N?(k) 2nn3-17解:18..DFT??(?1)???DFT??(?1)RN(n)???X(k?n?(?1) ∴DFT????N?(k?N) 227

x(0) x(8) x(4) x(12) x(2) x(10) x(6) x(14) x(1) x(9) x(5) x(13) x(3) x(11) x(7) x(15) x1(0) x1(1) x1(2) x1(3) x1(4) x1(5) x1(6) x1(7) x1(8) x1(9) x1(10) x1(11) x1(12) x1(13) x1(14) x1(15) x2(0) x2(1) x2(2) x2(3) x2(4) x2(5) x2(6) x2(7) x2(8) x2(9) x2(10) x2(11) x2(12) x2(13) x2(14) x2(15) x3(0) x3(1) x3(2) x3 (3) x3 (4) x3 (5) x3 (6) x3 (7) x3 (8) x3 (9) x3 (10) x3 (11) x3 (12) x3 (13) x3 (14) x3 (15) X(0) X(1) X(2) X (3) X (4) X (5) X(6) X (7) X(8) X (9) X (10) X (11) X (12) X (13) X (14) X(15)

3-18 解:DFT??X(N)???N?12r?0nk2rk(2r?1)k (?1)nWN??(?1)2rWN??(?1)2r?1WN?n?0r?0r?0N?1N?12N?12rkk??WN?WN?WNrk

2r?02N?12

nn??DFT?(?1)?DFT(?1)RN(n)??????X(k?N) 2 28

5-1 用冲击响应不变法求相应的数字滤波器系统函数H(z)

s?3 2s?3s?2s?12)Ha(s) = 2

s?2s?41)Ha(s) =

解:由Ha(s)分解成部分分式之和 1)Ha(s) =

2s?31s?3==–

s?2s2?3s?2(s?2)(s?1)s?1211?e?T(1?2e?T)z?1∴H(z) = –=

1?e?Tz?11?e?2Tz?11?e?T(1?e?T)z?1?e?3Tz?2s?1=2s?2s?42)Ha(s) =

12s?2ej?3+

12s?2e?j?3

∴H(z) =

121?e?2Tej?3z?1+

121?e?2Te?j?3z?1=

1?e?Tcos(3T)z?11?2e?Tcos(3T)z?1?e?2Tz?2

5-2 设ha(t)表示一个模拟滤波器的单位冲击响应 ha(t)=

0 , t<0

(1)用冲击响应不变法,将此模拟滤波器转换成数字滤波器,确定系统函数H(z)(以T作为参数)

(2)证明,T为任何值时,数字滤波器是稳定的,并说明数字滤波器近似为低通滤波器,还是高通滤波器

e?0.9t , t≥0

解:(1)∵ ha(t)= e?0.9tu(t) 29

∴ Ha(s) =

1 s?0.91 ∴ H(z) = ?0.9T?11?ez1(2)∵ H(z) = ?0.9T?11?ez

则其极点为z=e?0.9T

∵ T > 0 ∴ |z| < 1

j?eH(ej?) =H(z)|z?ej? = j?

e?e?0.9T可以看出当ω↑时,| H(ej?) |↓ ∴ 是低通滤波

5-3 图5-40是由RC组成的模拟滤波器,写出其系统函数Ha(s),并选用一种合适的转换方法,将Ha(s)转换成数字滤波器H(z) 解:由回路法可知(这是一个高通滤波器) ya(t)=RC∴

dUc(t)dx(t)dy(t)= RCa–RCa dtdtdtC xa(t)

R

ya(t)

RCsY(s)== Ha(s) X(s)1?RCs由于脉冲响应不变法只适宜于实现带通滤波器,所以最好用双线性变换法实现H(z)

∴H(z) =

Ha(s)|21?z?1s??T1?z?12RC1?z?1?2RC(1?z?1)T1?z?1==

2RC1?z?1(T?2RC)?(T?2RC)z?1?z?21??T1?z?1 30

?c5-4 设模拟滤波器的系统函数为Ha(s)= ,式中Ωc是模拟滤波

s??c器的3dB带宽,利用双线性变换,设计一个具有0.2π的3dB带宽的单极点低通数字滤波器

解:由预畸可知

210.65tan(?0.2?)?c==

TT20.65∴ Ha(s) =T

0.65s?T由双线性变换法可得

H(z) =

Ha(s)|21?z?1s??T1?z?10.650.65(1?z?1)T== 21?z?10.652.65?1.35z?1???1T1?zT5-5 要求通过模拟滤波器设计数字滤波器,给定指标:3dB截至角频率ωc=π/2,通带内ωp=0.4π处起伏不超过1dB,阻带内ωs=0.8π处衰减不小于20dB,用Butterworth滤波特性实现

(1)用冲击响应不变法 (2)用双线性变换法 解:(1)用冲击响应不变法

① 先将数字指标转换为低通原型模拟滤波器指标

0.4??p==

TT?p0.8??s==

TT

31

?s②设计模拟滤波器,求出Ha(s) Butterworth的频响函数为

1|Ha(j?)|2=?2n

1?()?c∴ Ha(j?p)=

1?(1?p?c)2n=

?p2n=101?()?c1?110

2011?10Ha(j?s)=10== ?s2n?s2n1?()1?()?c?clg(102?1110)∴ n =

10?1=2.14 ?2lg(s)?p∴ 取 n = 3 ③ 求?c

|Ha(j?)|=

21?s2n1?()?c=10

?20.8?∴ ωc = 62rad/s = 6= 0.372π

9910?1∴ ?c=

?s?cT 设T = 1, 则 ?c= 0.372π

④ 求Ha(s)查表可得

Ha(s?)?1

(s??1)(s?2?s??1)s?c∴ Ha(s) = Ha(s?)|s???1

ss2s(?1)(2??1)?c?c?c32

⑤ 由冲击响应不变法 先将Ha(s)分解成部分分式

AA2A31Ha(s) =++

s?s1s?s2s?s3 =

A3则H(z) =++

1?e?s1Tz?11?e?s2Tz?11?e?s3Tz?1

=

A1A2(2)用双线性变换法

①由预畸求模拟滤波器原型指标

?p1.4532?p=tan=

TT2?s0.1552?s=tan=

TT2 ②设计模拟滤波器,求出Ha(s)

Butterworth的频响函数为

1|Ha(j?)|=?2n

1?()?c2∴ Ha(j?p)=

1?(1?p?c)2n=10?110

201?10Ha(j?s)=10= ?s2n1?()?c 33

lg(102?1110) ∴ n =

10?1=1.51 ?2lg(s)?p 取n =2

③求?c

1?2|Ha(j?s)|=10= ?2n1?()?c2取T=1

6.155∴ ?c=62rad/s = 6= 2.862

9910?1?s ④求Ha(s)

查表可得:

1Ha(s?)=2

s??1.4142s??1Ha(s) = Ha(s?)|s??s=

?c1ss?1.4142?12?c?c2

=

⑤由双线性变换法求 H(z) =

Ha(s)|21?z?1s??T1?z?1=

5-6 已知图5-41h1(n)是偶对称序列N=8,h2(n)是h1(n)圆周位移后的序列。设H1(k)=DFT[h1(n)], H2(k)=DFT[h2(n)]

(1) 问|H1(k)| = |H2(k)|是否成立?θ1(k)与θ2(k)有什么关系? (2) h1(n),h2(n)各构成低通滤波器,试问它们是线性相位的?延时

34

是多少?

(3) 这两个滤波器的性能是否相同?为什么?若不同谁优谁劣? 解:(1) 由DFT的时移定理

mkDFT[xp(n-m)RN(n)]= WNX(k)可知

H1(k)和H2(k)只有相位差,幅值相等,即有 |H1(k)| = |H2(k)| θ1(k)和θ2(k)相差WN 即θ2(k)–θ1(k)= W=e4k8?j2?4k8mk=e?jk?

(2) ∵ 无论h1(n),h2(n)都是偶对称序列

∴ 所以他们构成的低通滤波器具有线性相位

N?18?1延时 α===3.5

22(3) 不相同,相位相差kπ

h1(n)要优于h2(n),因为其相位滞后时间少

5-7用矩形容器设计一个近似理想频率响应的FIR线性相位的数字滤

?j?? , 0? |?|??c 波器 e

Hd(ej?) =

, ?c? |?|??

0

(1) 求出相应于理想低通的单位脉冲响应hd(n)

(2) 求出矩形窗设计法的h(n)表达式确定τ与N之间的关系 (3) N取奇数或偶数对滤波特性有什么影响?

35

1解:(1) hd(n)=

2?????Hd(ej?)ej?nd?

1 =

2?????cce?j??ej?nd? =

sin[?c(n??)]

?(n??)(2) h(n)= hd(n) RN(n), h(n)只能取偶对称序列,由线性相位 τ=

N?1 2(3) 由于N无论取奇数还是偶数,都可实现低通滤波,而且只

N?1的偶对称函数,就能保证线性相关,2要N的取值使h(n)为关于

另外N的大小,只影响余振的多少和过滤带的窄宽,不会影响阻带良域。

5-8用矩形容器设计一个线性相位高通FIR数字滤波器

?j?? , ?c? |?|?? eHd(ej?) =

, 0? |?|??c

0

(1) 求出响应于理想高通的单位脉冲响应hd(n)

(2) 求出矩形窗口设计法的h(n)表达式,确定τ与N之间的关系 (3) N的取值有什么限制?为什么?

1解:(1) hd(n)=

2?????j???Hd(ej?)ej?nd? ejn?

1= 2?1= 2?1= 2??????ce1d?+

2????ce?j??ejn?d?

???ce?j?(n??)1d?+

2????cej?(n??)d?

???c[e?j?(n??)?ej?(n??)]d?

36

=

cos[?(n??)]d? ???c1?1?sin[?(n??)]|?c =

?(n??)sin[?c(n??)]1sin[(n??)?]–=

?(n??)?(n??)?cSa[?c(n??)] = Sa[?(n??)]–?∴ hd(n)仍然是偶函数

(2) h(n)= hd(n) RN(n)

∴ h(n)为偶对称序列,要保持滤波器具有线性相位,则须有 τ=

N?1 2(3) 这是一个高通滤波器,由于h(n)为偶对称,而当N取偶数时,

所得到的滤波器不能实现高通特性

∴ N只能取奇数

5-9考虑一个长度为M=15的线性相位FIR滤波器,设滤波器具有对称单位样值响应,并且它的幅度响应满足条件

1, k = 0, 1, 2, 3

2?kH() = 15 0, k = 4, 5, 6, 7 确定该滤波器的系数h(n)

j?H(e)|解:由于H(k) =a??2?kN

37

1N?1j2?nk∴ h(n) = IDFT[H(k)] =

NN?H(k)e

k?0∴ h(0) = 1142?15?H(k)ej150k= 1 k?0j2815?h(1) = 114j2?

15k11?e15?H(k)e=?2? k?0151?ej15j5615?

h(2) = 114j4?15?H(k)e15k=1?1?e4? k?0151?ej15j8415? h(3) = 114j6?15k15?H(k)e=1k?015?1?e61?ej? 15

h(4) = 0 h(5) = 0 h(6) = 0 h(7) = 0

由频率特性可知,这是一个低通滤波器

∴ 要取h(n)关于α=N?115?12=2=7这一点偶对称时,可实

现低通滤波(奇对称时,无法实现低通滤波) ∴ 取 h(8) = h(6) h(9) = h(5) h(10) = h(4) h(11) = h(3) h(12) = h(2)

h(13) = h(1) h(14)=h(0)

38

本文来源:https://www.bwwdw.com/article/l0s7.html

Top