Information Technology Reference
In-Depth Information
∂ω
r
j
x q ω
w i
N j
M i
x q w i (
b ji )
b l ji
s
j
c kj )
c kj
(
n
m
l , p = 1
r , s = 1
+
c kj +
b ji
x k .
M i
2
N j
2
l = 1 w i
r = 1 ω
k
=
0
j
=
1
r
j
(A.24)
The derivative of the degree of activation of the rules of the plant with respect to
each of the state variables,
w i /∂
x q , and the derivative of the degree of activation
of the rules of the plant with respect to each of the state variables,
r j /∂
∂ω
x q , will be
determined below.
A.3 Derivative of the Degree of Activation of the Rules
of the Plant
n
m
w i (
l
l
l
l
x
,
u
) =
1 μ
ki (
x k , σ
ki )
1 μ
ji (
u j , α
ji )
,
(A.25)
k
=
j
=
w i (
x
)
w i (
u
)
so that:
w i (
w i
w i (
w i (
)
x
)
)
u
)
w i (
w i (
w i (
x q =
x
)
u
)
=
u
+
x
.
(A.26)
x q
x q
x q
Given that:
n
w i (
x
)
l
l
=
1 μ
ki (
x k , σ
ki )
,
(A.27)
x q
x q
k
=
and
n
n
w i (
l
l
l
l
x
)
= ∂μ
1 i (
x 1 , σ
1 i )
ki ) + ∂μ
2 i (
x 2 , σ
2 i )
l
l
l
l
1 μ
ki (
x k , σ
2 μ
ki (
x k , σ
ki )
x q
x q
x q
k
=
1
,
k
=
k
=
1
,
k
=
n
l
l
+ ···+ ∂μ
ni (
x n , σ
ni )
l
l
n μ
ki (
x k , σ
ki ),
(A.28)
x q
k
=
1
,
k
=
which can be represented more compactly by:
 
Search WWH ::




Custom Search