IIR滤波器scalar版本的传递函数
上图是CCES2.9.1关于SHARC架构下IIR滤波器 Scalar版本的滤波器图示，为了方便使用，下面手动推导一下该滤波器的传递函数。
对应上图 Output可以描述为：
O
u
t
p
u
t
=
S
a
m
p
l
e
∗
b
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
O
u
t
p
u
t
∗
a
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
Output=Sample*bcoeffs[TAPS-1]+Output*acoeffs[TAPS-1]+
Output=Sample∗bcoeffs[TAPS−1]+Output∗acoeffs[TAPS−1]+
S a m p l e ∗ b c o e f f s [ T A P S − 2 ] + O u t p u t ∗ a c o e f f s [ T A P S − 2 ] + . . . Sample*bcoeffs[TAPS-2]+Output*acoeffs[TAPS-2]+... Sample∗bcoeffs[TAPS−2]+Output∗acoeffs[TAPS−2]+...
+ S a m p l e ∗ b c o e f f s [ 0 ] + O u t p u t ∗ a c o e f f s [ 0 ] ( 公式一 ) +Sample*bcoeffs[0]+Output*acoeffs[0](公式一) +Sample∗bcoeffs[0]+Output∗acoeffs[0](公式一)
公式一中一共2*TAPS项。公式一进行化简：左右同时除以Sample
O
u
t
p
u
t
/
S
a
m
p
l
e
=
(
b
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
b
c
o
e
f
f
s
[
T
A
P
S
−
2
]
+
.
.
.
+
b
c
o
e
f
f
s
[
0
]
)
+
(
O
u
t
p
u
t
/
S
a
m
p
l
e
)
(
a
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
a
c
o
e
f
f
s
[
T
A
P
S
−
2
]
+
.
.
.
+
a
c
o
e
f
f
s
[
0
]
)
Output/Sample = (bcoeffs[TAPS-1]+bcoeffs[TAPS-2]+...+bcoeffs[0])+(Output/Sample)(acoeffs[TAPS-1]+acoeffs[TAPS-2]+...+acoeffs[0])
Output/Sample=(bcoeffs[TAPS−1]+bcoeffs[TAPS−2]+...+bcoeffs[0])+(Output/Sample)(acoeffs[TAPS−1]+acoeffs[TAPS−2]+...+acoeffs[0])
接着，把Output/Sample全部移到左侧进行化简
O
u
t
p
u
t
/
S
a
m
p
l
e
=
(
b
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
b
c
o
e
f
f
s
[
T
A
P
S
−
2
]
+
.
.
.
+
b
c
o
e
f
f
s
[
0
]
)
/
[
1
−
(
a
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
.
.
.
.
.
a
c
o
e
f
f
s
[
0
)
]
Output/Sample = (bcoeffs[TAPS-1]+bcoeffs[TAPS-2]+...+bcoeffs[0])/[1-(acoeffs[TAPS-1]+.....acoeffs[0)]
Output/Sample=(bcoeffs[TAPS−1]+bcoeffs[TAPS−2]+...+bcoeffs[0])/[1−(acoeffs[TAPS−1]+.....acoeffs[0)]
又因为acoeffs[0]=1，所以IIR滤波器scalar版本的传递函数为
H
(
Z
)
=
(
b
c
o
e
f
f
s
[
T
A
P
S
−
1
]
+
b
c
o
e
f
f
s
[
T
A
P
S
−
2
]
+
.
.
.
+
b
c
o
e
f
f
s
[
0
]
)
/
(
−
a
c
o
e
f
f
s
[
T
A
P
S
−
1
]
−
.
.
.
−
a
c
o
e
f
f
s
[
1
]
)
H(Z)=(bcoeffs[TAPS-1]+bcoeffs[TAPS-2]+...+bcoeffs[0])/(-acoeffs[TAPS-1]-...-acoeffs[1])
H(Z)=(bcoeffs[TAPS−1]+bcoeffs[TAPS−2]+...+bcoeffs[0])/(−acoeffs[TAPS−1]−...−acoeffs[1])
如有错误欢迎指出！！！万分感谢！！！
扫一扫
949
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
