导航
模糊交集
模糊交集(t-norms)是一种具有两个参数的函数,定义为
t
:
[
0
,
1
]
×
[
0
,
1
]
→
[
0
,
1
]
t:[0, 1]\times [0, 1]\to [0, 1]
t:[0,1]×[0,1]→[0,1]
使得
μ
A
∩
B
(
x
)
=
t
[
μ
a
(
x
)
,
μ
B
(
x
)
]
\mu_{A\cap B}(x)=t[\mu_a(x), \mu_B(x)]
μA∩B(x)=t[μa(x),μB(x)]
其中,模糊交集函数必须符合以下条件
- 边界条件 t ( 0 , 0 ) = 0 t(0, 0)=0 t(0,0)=0以及 t ( μ A ( x ) , 1 ) = t ( 1 , μ A ( x ) ) = μ A ( x ) t(\mu_A(x), 1)=t(1, \mu_A(x))=\mu_A(x) t(μA(x),1)=t(1,μA(x))=μA(x)
- 单调性,如果 μ A ( x ) < μ C ( x ) \mu_A(x)<\mu_C(x) μA(x)<μC(x)以及 μ B ( x ) < μ D ( x ) \mu_B(x)<\mu_D(x) μB(x)<μD(x),则 t ( μ A ( x ) , μ B ( x ) ) < t ( μ C ( x ) , μ D ( x ) ) t(\mu_A(x), \mu_B(x))<t(\mu_C(x), \mu_D(x)) t(μA(x),μB(x))<t(μC(x),μD(x))
- 交换性, t ( μ A ( x ) , μ B ( x ) ) = t ( μ B ( x ) , μ A ( x ) ) t(\mu_A(x), \mu_B(x))=t(\mu_B(x), \mu_A(x)) t(μA(x),μB(x))=t(μB(x),μA(x))
- 结合性, t ( μ A ( x ) , t ( μ B ( x ) , μ C ( x ) ) ) = t ( t ( μ A ( x ) , μ B ( x ) ) , μ C ( x ) ) t(\mu_A(x), t(\mu_B(x), \mu_C(x)))=t(t(\mu_A(x), \mu_B(x)), \mu_C(x)) t(μA(x),t(μB(x),μC(x)))=t(t(μA(x),μB(x)),μC(x))
非参数模糊交集
令 a ≡ μ A ( x ) , b ≡ μ B ( x ) a\equiv \mu_A(x), b\equiv \mu_B(x) a≡μA(x),b≡μB(x)
- 最小值(
minimum
): t min = a ∧ b = min ( a , b ) t_{\min}=a\wedge b=\min(a, b) tmin=a∧b=min(a,b) - 代数积(
algebraic product
): t a p ( a , b ) = a ⋅ b = a b t_{ap}(a, b)=a\cdot b=ab tap(a,b)=a⋅b=ab - 边界积(
bounded product
): t b p = a Θ b = max ( a , b ) t_{bp}=a\Theta b=\max(a, b) tbp=aΘb=max(a,b) - 激烈积(
drastic product
):
t d p ( a , b ) = a ⌢ b = { a , b = 1 b , a = 1 0 , a , b < 1 t_{dp}(a, b)=a\frown b= \begin{cases} a, b=1\\ b, a=1\\ 0, a,b<1 \end{cases} tdp(a,b)=a⌢b=⎩⎪⎨⎪⎧a,b=1b,a=10,a,b<1
关系: a ⌢ b ≤ a Θ b ≤ a ⋅ b ≤ a ∧ b a\frown b\leq a\Theta b\leq a\cdot b\leq a\wedge b a⌢b≤aΘb≤a⋅b≤a∧b
- 最大值(
maximum
): S max ( a , b ) = a ∨ b = max ( a , b ) S_{\max}(a, b)=a\vee b=\max(a, b) Smax(a,b)=a∨b=max(a,b) - 代数和(
algebraic sum
): S a s ( a , b ) = a ∔ b = a + b − a b S_{as}(a, b)=a\dotplus b=a+b-ab Sas(a,b)=a∔b=a+b−ab - 边界和(
bounded sum
): S b s ( a , b ) = a ⊕ b = m i n ( 1 , a + b ) S_{bs}(a, b)=a\oplus b=min(1, a+b) Sbs(a,b)=a⊕b=min(1,a+b) - 激烈和(
drastic sum
):
S d s ( a , b ) = a ∨ b ≤ a ∔ b ≤ a ⊕ b ≤ a ⌣ b S_{ds}(a, b)=a\vee b\leq a\dotplus b\leq a\oplus b\leq a\smile b Sds(a,b)=a∨b≤a∔b≤a⊕b≤a⌣b
关系
模糊关系也是一种模糊集合
并集:
μ
R
∪
S
(
x
,
y
)
=
max
(
μ
R
(
x
,
y
)
,
μ
S
(
x
,
y
)
)
\mu_{R\cup S}(x, y)=\max(\mu_R(x, y), \mu_S(x, y))
μR∪S(x,y)=max(μR(x,y),μS(x,y))
交集:
μ
R
∩
S
(
x
,
y
)
=
min
(
μ
R
(
x
,
y
)
,
μ
S
(
x
,
y
)
)
\mu_{R\cap S}(x, y)=\min(\mu_R(x, y), \mu_S(x, y))
μR∩S(x,y)=min(μR(x,y),μS(x,y))
补集:
μ
R
ˉ
(
x
,
y
)
=
1
−
μ
R
(
x
,
y
)
\mu_{\bar{R}}(x, y)=1-\mu_R(x, y)
μRˉ(x,y)=1−μR(x,y)
包含:
R
⊆
S
⇔
μ
R
(
x
,
y
)
≤
μ
S
(
x
,
y
)
,
∀
(
x
,
y
)
R\subseteq S \Leftrightarrow \mu_R(x, y)\leq \mu_S(x, y), \forall (x, y)
R⊆S⇔μR(x,y)≤μS(x,y),∀(x,y)
投影与柱状扩充
投影
R
R
R表示在
X
×
Y
X\times Y
X×Y上的一个模糊关系,如果
R
R
R对
X
X
X以及
Y
Y
Y的投影分别定义为
R
X
=
R
↓
X
=
∫
X
max
y
μ
R
(
x
,
y
)
/
x
R_X=R\downarrow X=\int_X\max_y\mu_R(x, y)/x
RX=R↓X=∫XymaxμR(x,y)/x
和
R
Y
=
R
↓
Y
=
∫
Y
max
x
μ
R
(
x
,
y
)
/
x
R_Y=R\downarrow Y=\int_Y\max_x\mu_R(x, y)/x
RY=R↓Y=∫YxmaxμR(x,y)/x
隶属函数定义如下
μ
R
X
=
μ
R
↓
X
(
x
)
=
max
y
μ
R
(
x
,
y
)
\mu_{R_X}=\mu_{R\downarrow X}(x)=\max_y\mu_R(x, y)
μRX=μR↓X(x)=ymaxμR(x,y)
和
μ
R
Y
=
μ
R
↓
Y
(
y
)
=
max
x
μ
R
(
x
,
y
)
\mu_{R_Y}=\mu_{R\downarrow Y}(y)=\max_x\mu_R(x, y)
μRY=μR↓Y(y)=xmaxμR(x,y)
柱状扩充
R
R
R表示
X
X
X或者
Y
Y
Y上的一个模糊关系或者集合,那么其在
X
×
Y
X\times Y
X×Y上的柱状扩充
C
(
A
)
C(A)
C(A)的定义分别如下
C
(
A
)
=
R
↑
X
×
Y
=
∫
X
×
Y
μ
R
(
x
)
/
(
x
,
y
)
C
(
A
)
=
R
↑
X
×
Y
=
∫
X
×
Y
μ
R
(
y
)
/
(
x
,
y
)
C(A)=R\uparrow X\times Y=\int_{X\times Y}\mu_R(x)/(x, y)\\ C(A)=R\uparrow X\times Y=\int_{X\times Y}\mu_R(y)/(x, y)
C(A)=R↑X×Y=∫X×YμR(x)/(x,y)C(A)=R↑X×Y=∫X×YμR(y)/(x,y)
隶属函数如下
μ
C
(
A
)
μ
R
(
x
)
,
∀
x
∈
X
,
∀
y
∈
Y
μ
C
(
A
)
μ
R
(
y
)
,
∀
x
∈
X
,
∀
y
∈
Y
\mu_{C(A)}\mu_R(x), \forall x\in X, \forall y\in Y\\ \mu_{C(A)}\mu_R(y), \forall x\in X, \forall y\in Y
μC(A)μR(x),∀x∈X,∀y∈YμC(A)μR(y),∀x∈X,∀y∈Y